[llvm] [RISCV] Support postRA vsetvl insertion pass (PR #70549)
Piyou Chen via llvm-commits
llvm-commits at lists.llvm.org
Tue May 14 22:17:13 PDT 2024
https://github.com/BeMg updated https://github.com/llvm/llvm-project/pull/70549
>From c92c6af71b5626578f9e5a357c8e98497451aa56 Mon Sep 17 00:00:00 2001
From: Piyou Chen <piyou.chen at sifive.com>
Date: Wed, 8 May 2024 04:48:14 -0700
Subject: [PATCH] [RISCV] Support post-regalloc vsetvli pass
---
llvm/lib/Target/RISCV/RISCVTargetMachine.cpp | 19 +-
llvm/test/CodeGen/RISCV/O0-pipeline.ll | 6 +-
llvm/test/CodeGen/RISCV/O3-pipeline.ll | 2 +-
.../early-clobber-tied-def-subreg-liveness.ll | 18 +-
.../RISCV/intrinsic-cttz-elts-vscale.ll | 26 +-
llvm/test/CodeGen/RISCV/pr69586.ll | 210 +-
...regalloc-last-chance-recoloring-failure.ll | 4 +-
llvm/test/CodeGen/RISCV/rvv/abs-vp.ll | 11 +-
.../RISCV/rvv/access-fixed-objects-by-rvv.ll | 4 +-
.../CodeGen/RISCV/rvv/active_lane_mask.ll | 84 +-
llvm/test/CodeGen/RISCV/rvv/bitreverse-vp.ll | 234 +-
llvm/test/CodeGen/RISCV/rvv/bswap-vp.ll | 258 +--
.../CodeGen/RISCV/rvv/calling-conv-fastcc.ll | 23 +-
llvm/test/CodeGen/RISCV/rvv/calling-conv.ll | 8 +-
llvm/test/CodeGen/RISCV/rvv/ceil-vp.ll | 92 +-
.../RISCV/rvv/combine-store-extract-crash.ll | 24 +-
llvm/test/CodeGen/RISCV/rvv/compressstore.ll | 53 +-
.../RISCV/rvv/concat-vector-insert-elt.ll | 10 +-
.../RISCV/rvv/constant-folding-crash.ll | 18 +-
llvm/test/CodeGen/RISCV/rvv/ctlz-sdnode.ll | 10 +-
llvm/test/CodeGen/RISCV/rvv/ctlz-vp.ll | 8 +-
llvm/test/CodeGen/RISCV/rvv/ctpop-vp.ll | 129 +-
llvm/test/CodeGen/RISCV/rvv/cttz-vp.ll | 182 +-
llvm/test/CodeGen/RISCV/rvv/extractelt-i1.ll | 32 +-
.../rvv/fixed-vector-i8-index-cornercase.ll | 36 +-
.../CodeGen/RISCV/rvv/fixed-vectors-abs-vp.ll | 4 +-
.../RISCV/rvv/fixed-vectors-bitreverse-vp.ll | 410 ++--
.../RISCV/rvv/fixed-vectors-bswap-vp.ll | 234 +-
.../rvv/fixed-vectors-buildvec-of-binop.ll | 7 +-
.../RISCV/rvv/fixed-vectors-ceil-vp.ll | 94 +-
.../RISCV/rvv/fixed-vectors-ctlz-vp.ll | 1150 +++++-----
.../RISCV/rvv/fixed-vectors-ctpop-vp.ll | 523 +++--
.../RISCV/rvv/fixed-vectors-cttz-vp.ll | 1068 +++++----
.../rvv/fixed-vectors-deinterleave-load.ll | 4 +-
.../RISCV/rvv/fixed-vectors-extract-i1.ll | 16 +-
.../RISCV/rvv/fixed-vectors-extract.ll | 8 +-
.../RISCV/rvv/fixed-vectors-floor-vp.ll | 94 +-
.../RISCV/rvv/fixed-vectors-fmaximum-vp.ll | 31 +-
.../RISCV/rvv/fixed-vectors-fminimum-vp.ll | 31 +-
...d-vectors-fnearbyint-constrained-sdnode.ll | 24 +-
.../RISCV/rvv/fixed-vectors-fp-buildvec.ll | 9 +-
.../RISCV/rvv/fixed-vectors-fp-interleave.ll | 10 +-
.../RISCV/rvv/fixed-vectors-fp-shuffles.ll | 21 +-
.../CodeGen/RISCV/rvv/fixed-vectors-fp.ll | 2 +-
.../RISCV/rvv/fixed-vectors-fp2i-sat.ll | 130 +-
.../RISCV/rvv/fixed-vectors-fpext-vp.ll | 4 +-
.../RISCV/rvv/fixed-vectors-fptosi-vp.ll | 4 +-
.../RISCV/rvv/fixed-vectors-fptoui-vp.ll | 4 +-
.../RISCV/rvv/fixed-vectors-fptrunc-vp.ll | 4 +-
.../CodeGen/RISCV/rvv/fixed-vectors-i2fp.ll | 80 +-
.../rvv/fixed-vectors-insert-subvector.ll | 19 +-
.../CodeGen/RISCV/rvv/fixed-vectors-insert.ll | 6 +-
.../RISCV/rvv/fixed-vectors-int-buildvec.ll | 177 +-
.../rvv/fixed-vectors-int-explodevector.ll | 181 +-
.../RISCV/rvv/fixed-vectors-int-interleave.ll | 14 +-
.../RISCV/rvv/fixed-vectors-int-shuffles.ll | 60 +-
.../CodeGen/RISCV/rvv/fixed-vectors-int.ll | 192 +-
...fixed-vectors-interleaved-access-zve32x.ll | 40 +-
.../rvv/fixed-vectors-interleaved-access.ll | 903 ++++----
.../CodeGen/RISCV/rvv/fixed-vectors-llrint.ll | 82 +-
.../CodeGen/RISCV/rvv/fixed-vectors-lrint.ll | 62 +-
.../RISCV/rvv/fixed-vectors-mask-buildvec.ll | 22 +-
.../RISCV/rvv/fixed-vectors-mask-splat.ll | 6 +-
.../RISCV/rvv/fixed-vectors-masked-gather.ll | 1583 ++++++-------
.../rvv/fixed-vectors-masked-load-int.ll | 8 +-
.../RISCV/rvv/fixed-vectors-masked-scatter.ll | 1979 ++++++++---------
.../rvv/fixed-vectors-masked-store-int.ll | 41 +-
.../RISCV/rvv/fixed-vectors-nearbyint-vp.ll | 127 +-
.../rvv/fixed-vectors-reduction-formation.ll | 140 +-
.../rvv/fixed-vectors-reduction-fp-vp.ll | 8 +-
.../RISCV/rvv/fixed-vectors-reduction-fp.ll | 20 +-
.../rvv/fixed-vectors-reduction-int-vp.ll | 8 +-
.../RISCV/rvv/fixed-vectors-reduction-int.ll | 52 +-
.../rvv/fixed-vectors-reduction-mask-vp.ll | 58 +-
.../RISCV/rvv/fixed-vectors-rint-vp.ll | 104 +-
.../RISCV/rvv/fixed-vectors-round-vp.ll | 94 +-
.../RISCV/rvv/fixed-vectors-roundeven-vp.ll | 94 +-
.../RISCV/rvv/fixed-vectors-roundtozero-vp.ll | 94 +-
.../CodeGen/RISCV/rvv/fixed-vectors-sad.ll | 12 +-
.../RISCV/rvv/fixed-vectors-setcc-fp-vp.ll | 55 +-
.../RISCV/rvv/fixed-vectors-setcc-int-vp.ll | 42 +-
.../RISCV/rvv/fixed-vectors-sext-vp.ll | 4 +-
.../RISCV/rvv/fixed-vectors-shuffle-concat.ll | 4 +-
.../rvv/fixed-vectors-shuffle-exact-vlen.ll | 3 +-
.../rvv/fixed-vectors-shuffle-reverse.ll | 18 +-
.../rvv/fixed-vectors-shuffle-transpose.ll | 37 +-
.../RISCV/rvv/fixed-vectors-sitofp-vp.ll | 4 +-
.../rvv/fixed-vectors-strided-load-combine.ll | 8 +-
.../RISCV/rvv/fixed-vectors-strided-vpload.ll | 14 +-
.../rvv/fixed-vectors-strided-vpstore.ll | 2 +-
.../RISCV/rvv/fixed-vectors-trunc-vp.ll | 156 +-
.../RISCV/rvv/fixed-vectors-uitofp-vp.ll | 4 +-
.../RISCV/rvv/fixed-vectors-vadd-vp.ll | 16 +-
.../RISCV/rvv/fixed-vectors-vand-vp.ll | 14 +-
.../RISCV/rvv/fixed-vectors-vcopysign-vp.ll | 29 +-
.../RISCV/rvv/fixed-vectors-vfabs-vp.ll | 4 +-
.../RISCV/rvv/fixed-vectors-vfma-vp.ll | 70 +-
.../RISCV/rvv/fixed-vectors-vfmax-vp.ll | 29 +-
.../RISCV/rvv/fixed-vectors-vfmin-vp.ll | 29 +-
.../RISCV/rvv/fixed-vectors-vfmuladd-vp.ll | 70 +-
.../RISCV/rvv/fixed-vectors-vfneg-vp.ll | 4 +-
.../RISCV/rvv/fixed-vectors-vfsqrt-vp.ll | 4 +-
.../CodeGen/RISCV/rvv/fixed-vectors-vfwadd.ll | 10 +-
.../CodeGen/RISCV/rvv/fixed-vectors-vfwmul.ll | 10 +-
.../CodeGen/RISCV/rvv/fixed-vectors-vfwsub.ll | 10 +-
.../RISCV/rvv/fixed-vectors-vmax-vp.ll | 12 +-
.../RISCV/rvv/fixed-vectors-vmaxu-vp.ll | 12 +-
.../RISCV/rvv/fixed-vectors-vmin-vp.ll | 12 +-
.../RISCV/rvv/fixed-vectors-vminu-vp.ll | 12 +-
.../RISCV/rvv/fixed-vectors-vpgather.ll | 80 +-
.../CodeGen/RISCV/rvv/fixed-vectors-vpload.ll | 8 +-
.../RISCV/rvv/fixed-vectors-vpmerge.ll | 18 +-
.../RISCV/rvv/fixed-vectors-vpscatter.ll | 87 +-
.../RISCV/rvv/fixed-vectors-vpstore.ll | 2 +-
.../RISCV/rvv/fixed-vectors-vsadd-vp.ll | 22 +-
.../RISCV/rvv/fixed-vectors-vsaddu-vp.ll | 22 +-
.../RISCV/rvv/fixed-vectors-vscale-range.ll | 22 +-
.../RISCV/rvv/fixed-vectors-vselect-vp.ll | 86 +-
.../RISCV/rvv/fixed-vectors-vselect.ll | 24 +-
.../RISCV/rvv/fixed-vectors-vssub-vp.ll | 22 +-
.../RISCV/rvv/fixed-vectors-vssubu-vp.ll | 22 +-
.../CodeGen/RISCV/rvv/fixed-vectors-vwadd.ll | 23 +-
.../CodeGen/RISCV/rvv/fixed-vectors-vwaddu.ll | 23 +-
.../CodeGen/RISCV/rvv/fixed-vectors-vwmul.ll | 17 +-
.../RISCV/rvv/fixed-vectors-vwmulsu.ll | 21 +-
.../CodeGen/RISCV/rvv/fixed-vectors-vwmulu.ll | 15 +-
.../CodeGen/RISCV/rvv/fixed-vectors-vwsub.ll | 29 +-
.../CodeGen/RISCV/rvv/fixed-vectors-vwsubu.ll | 29 +-
.../RISCV/rvv/fixed-vectors-zext-vp.ll | 4 +-
llvm/test/CodeGen/RISCV/rvv/floor-vp.ll | 92 +-
.../test/CodeGen/RISCV/rvv/fmaximum-sdnode.ll | 26 +-
llvm/test/CodeGen/RISCV/rvv/fmaximum-vp.ll | 239 +-
.../test/CodeGen/RISCV/rvv/fminimum-sdnode.ll | 26 +-
llvm/test/CodeGen/RISCV/rvv/fminimum-vp.ll | 239 +-
.../rvv/fnearbyint-constrained-sdnode.ll | 30 +-
.../CodeGen/RISCV/rvv/fnearbyint-sdnode.ll | 30 +-
.../CodeGen/RISCV/rvv/fpclamptosat_vec.ll | 132 +-
llvm/test/CodeGen/RISCV/rvv/fptosi-sat.ll | 10 +-
llvm/test/CodeGen/RISCV/rvv/fshr-fshl-vp.ll | 301 ++-
llvm/test/CodeGen/RISCV/rvv/llrint-vp.ll | 2 +-
llvm/test/CodeGen/RISCV/rvv/lrint-vp.ll | 2 +-
llvm/test/CodeGen/RISCV/rvv/masked-tama.ll | 6 +-
llvm/test/CodeGen/RISCV/rvv/mgather-sdnode.ll | 22 +-
.../test/CodeGen/RISCV/rvv/mscatter-sdnode.ll | 57 +-
llvm/test/CodeGen/RISCV/rvv/nearbyint-vp.ll | 243 +-
llvm/test/CodeGen/RISCV/rvv/pr63596.ll | 14 +-
llvm/test/CodeGen/RISCV/rvv/rint-vp.ll | 137 +-
llvm/test/CodeGen/RISCV/rvv/round-vp.ll | 143 +-
llvm/test/CodeGen/RISCV/rvv/roundeven-vp.ll | 143 +-
llvm/test/CodeGen/RISCV/rvv/roundtozero-vp.ll | 143 +-
.../RISCV/rvv/rv32-spill-vector-csr.ll | 2 +-
.../RISCV/rvv/rv64-spill-vector-csr.ll | 4 +-
.../RISCV/rvv/rvv-peephole-vmerge-vops.ll | 2 +-
llvm/test/CodeGen/RISCV/rvv/setcc-fp-vp.ll | 176 +-
llvm/test/CodeGen/RISCV/rvv/setcc-int-vp.ll | 54 +-
.../test/CodeGen/RISCV/rvv/shuffle-reverse.ll | 40 +-
.../CodeGen/RISCV/rvv/sink-splat-operands.ll | 4 +-
llvm/test/CodeGen/RISCV/rvv/sshl_sat_vec.ll | 68 +-
llvm/test/CodeGen/RISCV/rvv/stepvector.ll | 6 +-
llvm/test/CodeGen/RISCV/rvv/strided-vpload.ll | 24 +-
.../test/CodeGen/RISCV/rvv/strided-vpstore.ll | 26 +-
llvm/test/CodeGen/RISCV/rvv/vadd-vp.ll | 8 +-
llvm/test/CodeGen/RISCV/rvv/vcopysign-vp.ll | 6 +-
llvm/test/CodeGen/RISCV/rvv/vcpop.ll | 14 +-
.../RISCV/rvv/vector-deinterleave-fixed.ll | 4 +-
.../RISCV/rvv/vector-deinterleave-load.ll | 50 +-
.../CodeGen/RISCV/rvv/vector-deinterleave.ll | 46 +-
.../RISCV/rvv/vector-interleave-store.ll | 12 +-
.../CodeGen/RISCV/rvv/vector-interleave.ll | 192 +-
llvm/test/CodeGen/RISCV/rvv/vector-splice.ll | 4 +-
llvm/test/CodeGen/RISCV/rvv/vfabs-vp.ll | 4 +-
llvm/test/CodeGen/RISCV/rvv/vfadd-vp.ll | 28 +-
llvm/test/CodeGen/RISCV/rvv/vfdiv-vp.ll | 28 +-
llvm/test/CodeGen/RISCV/rvv/vfirst.ll | 14 +-
llvm/test/CodeGen/RISCV/rvv/vfma-vp.ll | 64 +-
.../RISCV/rvv/vfmadd-constrained-sdnode.ll | 156 +-
llvm/test/CodeGen/RISCV/rvv/vfmadd-sdnode.ll | 144 +-
llvm/test/CodeGen/RISCV/rvv/vfmax-vp.ll | 6 +-
llvm/test/CodeGen/RISCV/rvv/vfmin-vp.ll | 6 +-
.../RISCV/rvv/vfmsub-constrained-sdnode.ll | 149 +-
llvm/test/CodeGen/RISCV/rvv/vfmul-vp.ll | 28 +-
llvm/test/CodeGen/RISCV/rvv/vfmuladd-vp.ll | 64 +-
llvm/test/CodeGen/RISCV/rvv/vfneg-vp.ll | 4 +-
.../RISCV/rvv/vfnmadd-constrained-sdnode.ll | 165 +-
.../RISCV/rvv/vfnmsub-constrained-sdnode.ll | 141 +-
llvm/test/CodeGen/RISCV/rvv/vfpext-vp.ll | 2 +-
llvm/test/CodeGen/RISCV/rvv/vfptosi-vp.ll | 26 +-
llvm/test/CodeGen/RISCV/rvv/vfptoui-vp.ll | 26 +-
llvm/test/CodeGen/RISCV/rvv/vfptrunc-vp.ll | 22 +-
llvm/test/CodeGen/RISCV/rvv/vfsqrt-vp.ll | 4 +-
llvm/test/CodeGen/RISCV/rvv/vfsub-vp.ll | 28 +-
llvm/test/CodeGen/RISCV/rvv/vfwmacc-vp.ll | 2 +-
llvm/test/CodeGen/RISCV/rvv/vfwnmacc-vp.ll | 14 +-
llvm/test/CodeGen/RISCV/rvv/vfwnmsac-vp.ll | 14 +-
llvm/test/CodeGen/RISCV/rvv/vitofp-sdnode.ll | 4 +-
llvm/test/CodeGen/RISCV/rvv/vmax-vp.ll | 8 +-
llvm/test/CodeGen/RISCV/rvv/vmaxu-vp.ll | 8 +-
llvm/test/CodeGen/RISCV/rvv/vmfeq.ll | 96 +-
llvm/test/CodeGen/RISCV/rvv/vmfge.ll | 96 +-
llvm/test/CodeGen/RISCV/rvv/vmfgt.ll | 96 +-
llvm/test/CodeGen/RISCV/rvv/vmfle.ll | 96 +-
llvm/test/CodeGen/RISCV/rvv/vmflt.ll | 96 +-
llvm/test/CodeGen/RISCV/rvv/vmfne.ll | 96 +-
llvm/test/CodeGen/RISCV/rvv/vmin-vp.ll | 8 +-
llvm/test/CodeGen/RISCV/rvv/vminu-vp.ll | 8 +-
llvm/test/CodeGen/RISCV/rvv/vmsbf.ll | 14 +-
llvm/test/CodeGen/RISCV/rvv/vmseq.ll | 178 +-
llvm/test/CodeGen/RISCV/rvv/vmsge.ll | 178 +-
llvm/test/CodeGen/RISCV/rvv/vmsgeu.ll | 176 +-
llvm/test/CodeGen/RISCV/rvv/vmsgt.ll | 178 +-
llvm/test/CodeGen/RISCV/rvv/vmsgtu.ll | 178 +-
llvm/test/CodeGen/RISCV/rvv/vmsif.ll | 14 +-
llvm/test/CodeGen/RISCV/rvv/vmsle.ll | 178 +-
llvm/test/CodeGen/RISCV/rvv/vmsleu.ll | 178 +-
llvm/test/CodeGen/RISCV/rvv/vmslt.ll | 178 +-
llvm/test/CodeGen/RISCV/rvv/vmsltu.ll | 178 +-
llvm/test/CodeGen/RISCV/rvv/vmsne.ll | 178 +-
llvm/test/CodeGen/RISCV/rvv/vmsof.ll | 14 +-
llvm/test/CodeGen/RISCV/rvv/vp-cttz-elts.ll | 8 +-
llvm/test/CodeGen/RISCV/rvv/vp-reverse-int.ll | 8 +-
.../rvv/vp-reverse-mask-fixed-vectors.ll | 8 +-
.../test/CodeGen/RISCV/rvv/vp-reverse-mask.ll | 22 +-
.../RISCV/rvv/vp-splice-mask-fixed-vectors.ll | 8 +-
.../RISCV/rvv/vp-splice-mask-vectors.ll | 14 +-
.../test/CodeGen/RISCV/rvv/vpgather-sdnode.ll | 80 +-
llvm/test/CodeGen/RISCV/rvv/vpload.ll | 10 +-
llvm/test/CodeGen/RISCV/rvv/vpmerge-sdnode.ll | 20 +-
.../CodeGen/RISCV/rvv/vpscatter-sdnode.ll | 44 +-
llvm/test/CodeGen/RISCV/rvv/vpstore.ll | 28 +-
.../RISCV/rvv/vreductions-fp-sdnode.ll | 8 +-
.../CodeGen/RISCV/rvv/vreductions-fp-vp.ll | 12 +-
.../CodeGen/RISCV/rvv/vreductions-int-vp.ll | 2 +-
.../CodeGen/RISCV/rvv/vreductions-mask-vp.ll | 76 +-
.../RISCV/rvv/vrgatherei16-subreg-liveness.ll | 16 +-
llvm/test/CodeGen/RISCV/rvv/vsadd-vp.ll | 4 +-
llvm/test/CodeGen/RISCV/rvv/vsaddu-vp.ll | 4 +-
llvm/test/CodeGen/RISCV/rvv/vselect-fp.ll | 8 +-
llvm/test/CodeGen/RISCV/rvv/vselect-vp.ll | 106 +-
.../RISCV/rvv/vsetvli-insert-crossbb.ll | 2 +-
.../CodeGen/RISCV/rvv/vsetvli-regression.ll | 5 +-
llvm/test/CodeGen/RISCV/rvv/vsext-vp.ll | 2 +-
llvm/test/CodeGen/RISCV/rvv/vsitofp-vp.ll | 28 +-
llvm/test/CodeGen/RISCV/rvv/vssub-vp.ll | 48 +-
llvm/test/CodeGen/RISCV/rvv/vssubu-vp.ll | 48 +-
llvm/test/CodeGen/RISCV/rvv/vtrunc-vp.ll | 22 +-
llvm/test/CodeGen/RISCV/rvv/vuitofp-vp.ll | 28 +-
llvm/test/CodeGen/RISCV/rvv/vxrm-insert.ll | 4 +-
llvm/test/CodeGen/RISCV/rvv/vzext-vp.ll | 2 +-
.../CodeGen/RISCV/srem-seteq-illegal-types.ll | 4 +-
249 files changed, 10385 insertions(+), 9571 deletions(-)
diff --git a/llvm/lib/Target/RISCV/RISCVTargetMachine.cpp b/llvm/lib/Target/RISCV/RISCVTargetMachine.cpp
index 5d598a275a008..f2af3cd7e74bd 100644
--- a/llvm/lib/Target/RISCV/RISCVTargetMachine.cpp
+++ b/llvm/lib/Target/RISCV/RISCVTargetMachine.cpp
@@ -96,6 +96,10 @@ static cl::opt<bool> EnableMISchedLoadClustering(
cl::desc("Enable load clustering in the machine scheduler"),
cl::init(false));
+static cl::opt<bool> EnableVSETVLIAfterRVVRegAlloc(
+ "riscv-vsetvli-after-rvv-regalloc", cl::Hidden,
+ cl::desc("vsetvl insertion after rvv regalloc"), cl::init(true));
+
extern "C" LLVM_EXTERNAL_VISIBILITY void LLVMInitializeRISCVTarget() {
RegisterTargetMachine<RISCVTargetMachine> X(getTheRISCV32Target());
RegisterTargetMachine<RISCVTargetMachine> Y(getTheRISCV64Target());
@@ -389,6 +393,8 @@ FunctionPass *RISCVPassConfig::createRVVRegAllocPass(bool Optimized) {
bool RISCVPassConfig::addRegAssignAndRewriteFast() {
addPass(createRVVRegAllocPass(false));
+ if (EnableVSETVLIAfterRVVRegAlloc)
+ addPass(createRISCVInsertVSETVLIPass());
addPass(createRISCVCoalesceVSETVLIPass());
if (TM->getOptLevel() != CodeGenOptLevel::None &&
EnableRISCVDeadRegisterElimination)
@@ -399,6 +405,8 @@ bool RISCVPassConfig::addRegAssignAndRewriteFast() {
bool RISCVPassConfig::addRegAssignAndRewriteOptimized() {
addPass(createRVVRegAllocPass(true));
addPass(createVirtRegRewriter(false));
+ if (EnableVSETVLIAfterRVVRegAlloc)
+ addPass(createRISCVInsertVSETVLIPass());
addPass(createRISCVCoalesceVSETVLIPass());
if (TM->getOptLevel() != CodeGenOptLevel::None &&
EnableRISCVDeadRegisterElimination)
@@ -547,10 +555,13 @@ void RISCVPassConfig::addPreRegAlloc() {
// Run RISCVInsertVSETVLI after PHI elimination. On O1 and above do it after
// register coalescing so needVSETVLIPHI doesn't need to look through COPYs.
- if (TM->getOptLevel() == CodeGenOptLevel::None)
- insertPass(&PHIEliminationID, createRISCVInsertVSETVLIPass());
- else
- insertPass(&RegisterCoalescerID, createRISCVInsertVSETVLIPass());
+ if (!EnableVSETVLIAfterRVVRegAlloc) {
+ if (TM->getOptLevel() == CodeGenOptLevel::None)
+ insertPass(&PHIEliminationID, createRISCVInsertVSETVLIPass());
+ else
+ insertPass(&RegisterCoalescerID, createRISCVInsertVSETVLIPass());
+ }
+
}
void RISCVPassConfig::addFastRegAlloc() {
diff --git a/llvm/test/CodeGen/RISCV/O0-pipeline.ll b/llvm/test/CodeGen/RISCV/O0-pipeline.ll
index 3aaa5dc03a7dc..e4abc93d1a8a1 100644
--- a/llvm/test/CodeGen/RISCV/O0-pipeline.ll
+++ b/llvm/test/CodeGen/RISCV/O0-pipeline.ll
@@ -44,14 +44,12 @@
; CHECK-NEXT: RISC-V Insert Write VXRM Pass
; CHECK-NEXT: Init Undef Pass
; CHECK-NEXT: Eliminate PHI nodes for register allocation
-; CHECK-NEXT: MachineDominator Tree Construction
-; CHECK-NEXT: Slot index numbering
-; CHECK-NEXT: Live Interval Analysis
-; CHECK-NEXT: RISC-V Insert VSETVLI pass
; CHECK-NEXT: Two-Address instruction pass
; CHECK-NEXT: Fast Register Allocator
+; CHECK-NEXT: MachineDominator Tree Construction
; CHECK-NEXT: Slot index numbering
; CHECK-NEXT: Live Interval Analysis
+; CHECK-NEXT: RISC-V Insert VSETVLI pass
; CHECK-NEXT: RISC-V Coalesce VSETVLI pass
; CHECK-NEXT: Fast Register Allocator
; CHECK-NEXT: Remove Redundant DEBUG_VALUE analysis
diff --git a/llvm/test/CodeGen/RISCV/O3-pipeline.ll b/llvm/test/CodeGen/RISCV/O3-pipeline.ll
index 52634b2a81629..0528b00d408b2 100644
--- a/llvm/test/CodeGen/RISCV/O3-pipeline.ll
+++ b/llvm/test/CodeGen/RISCV/O3-pipeline.ll
@@ -128,7 +128,6 @@
; CHECK-NEXT: Slot index numbering
; CHECK-NEXT: Live Interval Analysis
; CHECK-NEXT: Register Coalescer
-; CHECK-NEXT: RISC-V Insert VSETVLI pass
; CHECK-NEXT: Rename Disconnected Subregister Components
; CHECK-NEXT: Machine Instruction Scheduler
; CHECK-NEXT: Machine Block Frequency Analysis
@@ -142,6 +141,7 @@
; CHECK-NEXT: Machine Optimization Remark Emitter
; CHECK-NEXT: Greedy Register Allocator
; CHECK-NEXT: Virtual Register Rewriter
+; CHECK-NEXT: RISC-V Insert VSETVLI pass
; CHECK-NEXT: RISC-V Coalesce VSETVLI pass
; CHECK-NEXT: RISC-V Dead register definitions
; CHECK-NEXT: Virtual Register Map
diff --git a/llvm/test/CodeGen/RISCV/early-clobber-tied-def-subreg-liveness.ll b/llvm/test/CodeGen/RISCV/early-clobber-tied-def-subreg-liveness.ll
index 83a4f63add337..eb6ac985287a1 100644
--- a/llvm/test/CodeGen/RISCV/early-clobber-tied-def-subreg-liveness.ll
+++ b/llvm/test/CodeGen/RISCV/early-clobber-tied-def-subreg-liveness.ll
@@ -24,36 +24,36 @@ define void @_Z3foov() {
; CHECK-NEXT: lui a0, %hi(.L__const._Z3foov.var_49)
; CHECK-NEXT: addi a0, a0, %lo(.L__const._Z3foov.var_49)
; CHECK-NEXT: vsetivli zero, 2, e16, m2, ta, ma
-; CHECK-NEXT: vle16.v v8, (a0)
+; CHECK-NEXT: vle16.v v10, (a0)
; CHECK-NEXT: lui a0, %hi(.L__const._Z3foov.var_48)
; CHECK-NEXT: addi a0, a0, %lo(.L__const._Z3foov.var_48)
-; CHECK-NEXT: vle8.v v10, (a0)
+; CHECK-NEXT: vle8.v v8, (a0)
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs1r.v v10, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vs1r.v v8, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: lui a0, %hi(.L__const._Z3foov.var_46)
; CHECK-NEXT: addi a0, a0, %lo(.L__const._Z3foov.var_46)
-; CHECK-NEXT: vle16.v v10, (a0)
+; CHECK-NEXT: vle16.v v12, (a0)
; CHECK-NEXT: lui a0, %hi(.L__const._Z3foov.var_45)
; CHECK-NEXT: addi a0, a0, %lo(.L__const._Z3foov.var_45)
-; CHECK-NEXT: vle16.v v12, (a0)
+; CHECK-NEXT: vle16.v v14, (a0)
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: csrr a1, vlenb
; CHECK-NEXT: slli a1, a1, 1
-; CHECK-NEXT: vs2r.v v8, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: add a0, a0, a1
; CHECK-NEXT: vs2r.v v10, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: add a0, a0, a1
; CHECK-NEXT: vs2r.v v12, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: add a0, a0, a1
; CHECK-NEXT: vs2r.v v14, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: add a0, a0, a1
+; CHECK-NEXT: vs2r.v v16, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: #APP
; CHECK-NEXT: #NO_APP
-; CHECK-NEXT: vsetivli zero, 2, e16, m2, ta, ma
; CHECK-NEXT: lui a0, %hi(.L__const._Z3foov.var_40)
; CHECK-NEXT: addi a0, a0, %lo(.L__const._Z3foov.var_40)
+; CHECK-NEXT: vsetivli zero, 2, e16, m2, ta, ma
; CHECK-NEXT: vle16.v v8, (a0)
; CHECK-NEXT: lui a0, %hi(.L__const._Z3foov.var_44)
; CHECK-NEXT: addi a0, a0, %lo(.L__const._Z3foov.var_44)
@@ -71,12 +71,12 @@ define void @_Z3foov() {
; CHECK-NEXT: lui a0, 1048572
; CHECK-NEXT: addi a0, a0, 928
; CHECK-NEXT: vmsbc.vx v0, v8, a0
-; CHECK-NEXT: vsetvli zero, zero, e16, m2, tu, mu
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vl1r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, zero, e16, m2, tu, mu
; CHECK-NEXT: vsext.vf2 v10, v8, v0.t
; CHECK-NEXT: lui a0, %hi(var_47)
; CHECK-NEXT: addi a0, a0, %lo(var_47)
diff --git a/llvm/test/CodeGen/RISCV/intrinsic-cttz-elts-vscale.ll b/llvm/test/CodeGen/RISCV/intrinsic-cttz-elts-vscale.ll
index ea8feef332984..6009a6c7e138a 100644
--- a/llvm/test/CodeGen/RISCV/intrinsic-cttz-elts-vscale.ll
+++ b/llvm/test/CodeGen/RISCV/intrinsic-cttz-elts-vscale.ll
@@ -75,18 +75,18 @@ define i64 @ctz_nxv8i1_no_range(<vscale x 8 x i16> %a) {
; RV32-NEXT: sw a0, 16(sp)
; RV32-NEXT: addi a2, sp, 16
; RV32-NEXT: vsetvli a3, zero, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a2), zero
-; RV32-NEXT: vid.v v8
+; RV32-NEXT: vlse64.v v8, (a2), zero
+; RV32-NEXT: vid.v v16
; RV32-NEXT: li a2, -1
-; RV32-NEXT: vmadd.vx v8, a2, v16
-; RV32-NEXT: vsetvli zero, zero, e16, m2, ta, ma
+; RV32-NEXT: vmadd.vx v16, a2, v8
; RV32-NEXT: addi a2, sp, 32
-; RV32-NEXT: vl2r.v v16, (a2) # Unknown-size Folded Reload
-; RV32-NEXT: vmsne.vi v0, v16, 0
+; RV32-NEXT: vl2r.v v8, (a2) # Unknown-size Folded Reload
+; RV32-NEXT: vsetvli zero, zero, e16, m2, ta, ma
+; RV32-NEXT: vmsne.vi v0, v8, 0
; RV32-NEXT: vsetvli zero, zero, e64, m8, ta, ma
-; RV32-NEXT: vmv.v.i v16, 0
-; RV32-NEXT: vmerge.vim v16, v16, -1, v0
-; RV32-NEXT: vand.vv v8, v8, v16
+; RV32-NEXT: vmv.v.i v8, 0
+; RV32-NEXT: vmerge.vim v8, v8, -1, v0
+; RV32-NEXT: vand.vv v8, v16, v8
; RV32-NEXT: vredmaxu.vs v8, v8, v8
; RV32-NEXT: vmv.x.s a2, v8
; RV32-NEXT: sltu a3, a0, a2
@@ -108,15 +108,15 @@ define i64 @ctz_nxv8i1_no_range(<vscale x 8 x i16> %a) {
; RV64: # %bb.0:
; RV64-NEXT: csrr a0, vlenb
; RV64-NEXT: vsetvli a1, zero, e64, m8, ta, ma
-; RV64-NEXT: vmv.v.x v24, a0
-; RV64-NEXT: vid.v v16
+; RV64-NEXT: vmv.v.x v16, a0
+; RV64-NEXT: vid.v v24
; RV64-NEXT: li a1, -1
-; RV64-NEXT: vmadd.vx v16, a1, v24
+; RV64-NEXT: vmadd.vx v24, a1, v16
; RV64-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; RV64-NEXT: vmsne.vi v0, v8, 0
; RV64-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; RV64-NEXT: vmv.v.i v8, 0
-; RV64-NEXT: vmerge.vvm v8, v8, v16, v0
+; RV64-NEXT: vmerge.vvm v8, v8, v24, v0
; RV64-NEXT: vredmaxu.vs v8, v8, v8
; RV64-NEXT: vmv.x.s a1, v8
; RV64-NEXT: sub a0, a0, a1
diff --git a/llvm/test/CodeGen/RISCV/pr69586.ll b/llvm/test/CodeGen/RISCV/pr69586.ll
index 15daf2c577906..7084c04805be7 100644
--- a/llvm/test/CodeGen/RISCV/pr69586.ll
+++ b/llvm/test/CodeGen/RISCV/pr69586.ll
@@ -927,258 +927,258 @@ define void @test(ptr %0, ptr %1, i64 %2) {
; REMAT-NEXT: add a2, a0, a2
; REMAT-NEXT: vle32.v v14, (a2)
; REMAT-NEXT: sf.vc.vv 3, 0, v10, v12
-; REMAT-NEXT: vle32.v v12, (a2)
+; REMAT-NEXT: vle32.v v10, (a2)
; REMAT-NEXT: li a2, 11
; REMAT-NEXT: slli a2, a2, 10
; REMAT-NEXT: add a2, a0, a2
-; REMAT-NEXT: vle32.v v16, (a2)
+; REMAT-NEXT: vle32.v v26, (a2)
; REMAT-NEXT: sf.vc.vv 3, 0, v8, v14
-; REMAT-NEXT: vle32.v v10, (a2)
+; REMAT-NEXT: vle32.v v12, (a2)
; REMAT-NEXT: li a2, 23
; REMAT-NEXT: slli a2, a2, 9
; REMAT-NEXT: add a2, a0, a2
-; REMAT-NEXT: vle32.v v26, (a2)
-; REMAT-NEXT: sf.vc.vv 3, 0, v12, v16
-; REMAT-NEXT: vle32.v v12, (a2)
-; REMAT-NEXT: lui a2, 3
-; REMAT-NEXT: add a2, a0, a2
; REMAT-NEXT: vle32.v v28, (a2)
; REMAT-NEXT: vle32.v v14, (a2)
-; REMAT-NEXT: li a2, 25
-; REMAT-NEXT: slli a2, a2, 9
+; REMAT-NEXT: lui a2, 3
; REMAT-NEXT: add a2, a0, a2
; REMAT-NEXT: vle32.v v30, (a2)
; REMAT-NEXT: vle32.v v16, (a2)
-; REMAT-NEXT: li a2, 13
-; REMAT-NEXT: slli a2, a2, 10
+; REMAT-NEXT: li a2, 25
+; REMAT-NEXT: slli a2, a2, 9
; REMAT-NEXT: add a2, a0, a2
; REMAT-NEXT: vle32.v v6, (a2)
; REMAT-NEXT: vle32.v v18, (a2)
-; REMAT-NEXT: li a2, 27
-; REMAT-NEXT: slli a2, a2, 9
+; REMAT-NEXT: li a2, 13
+; REMAT-NEXT: slli a2, a2, 10
; REMAT-NEXT: add a2, a0, a2
; REMAT-NEXT: vle32.v v4, (a2)
; REMAT-NEXT: vle32.v v20, (a2)
-; REMAT-NEXT: li a2, 7
-; REMAT-NEXT: slli a2, a2, 11
+; REMAT-NEXT: li a2, 27
+; REMAT-NEXT: slli a2, a2, 9
; REMAT-NEXT: add a2, a0, a2
; REMAT-NEXT: vle32.v v2, (a2)
; REMAT-NEXT: vle32.v v22, (a2)
-; REMAT-NEXT: li a2, 29
-; REMAT-NEXT: slli a2, a2, 9
+; REMAT-NEXT: li a2, 7
+; REMAT-NEXT: slli a2, a2, 11
; REMAT-NEXT: add a2, a0, a2
; REMAT-NEXT: vle32.v v24, (a2)
; REMAT-NEXT: vle32.v v8, (a2)
; REMAT-NEXT: sf.vc.vv 3, 0, v10, v26
-; REMAT-NEXT: li a2, 15
-; REMAT-NEXT: slli a2, a2, 10
+; REMAT-NEXT: li a2, 29
+; REMAT-NEXT: slli a2, a2, 9
; REMAT-NEXT: add a2, a0, a2
; REMAT-NEXT: vle32.v v26, (a2)
; REMAT-NEXT: vle32.v v10, (a2)
; REMAT-NEXT: sf.vc.vv 3, 0, v12, v28
-; REMAT-NEXT: li a2, 31
-; REMAT-NEXT: slli a2, a2, 9
+; REMAT-NEXT: li a2, 15
+; REMAT-NEXT: slli a2, a2, 10
; REMAT-NEXT: add a2, a0, a2
; REMAT-NEXT: vle32.v v28, (a2)
; REMAT-NEXT: vle32.v v12, (a2)
; REMAT-NEXT: sf.vc.vv 3, 0, v14, v30
-; REMAT-NEXT: lui a2, 4
+; REMAT-NEXT: li a2, 31
+; REMAT-NEXT: slli a2, a2, 9
; REMAT-NEXT: add a2, a0, a2
; REMAT-NEXT: vle32.v v30, (a2)
; REMAT-NEXT: vle32.v v14, (a2)
; REMAT-NEXT: sf.vc.vv 3, 0, v16, v6
; REMAT-NEXT: lui a2, 4
-; REMAT-NEXT: addiw a2, a2, 512
; REMAT-NEXT: add a2, a0, a2
; REMAT-NEXT: vle32.v v6, (a2)
; REMAT-NEXT: vle32.v v16, (a2)
; REMAT-NEXT: sf.vc.vv 3, 0, v18, v4
-; REMAT-NEXT: li a2, 17
-; REMAT-NEXT: slli a2, a2, 10
+; REMAT-NEXT: lui a2, 4
+; REMAT-NEXT: addiw a2, a2, 512
; REMAT-NEXT: add a2, a0, a2
; REMAT-NEXT: vle32.v v4, (a2)
; REMAT-NEXT: vle32.v v18, (a2)
; REMAT-NEXT: sf.vc.vv 3, 0, v20, v2
-; REMAT-NEXT: lui a2, 4
-; REMAT-NEXT: addiw a2, a2, 1536
+; REMAT-NEXT: li a2, 17
+; REMAT-NEXT: slli a2, a2, 10
; REMAT-NEXT: add a2, a0, a2
; REMAT-NEXT: vle32.v v2, (a2)
; REMAT-NEXT: vle32.v v20, (a2)
; REMAT-NEXT: sf.vc.vv 3, 0, v22, v24
-; REMAT-NEXT: li a2, 9
-; REMAT-NEXT: slli a2, a2, 11
+; REMAT-NEXT: lui a2, 4
+; REMAT-NEXT: addiw a2, a2, 1536
; REMAT-NEXT: add a2, a0, a2
; REMAT-NEXT: vle32.v v24, (a2)
; REMAT-NEXT: vle32.v v22, (a2)
; REMAT-NEXT: sf.vc.vv 3, 0, v8, v26
-; REMAT-NEXT: lui a2, 5
-; REMAT-NEXT: addiw a2, a2, -1536
+; REMAT-NEXT: li a2, 9
+; REMAT-NEXT: slli a2, a2, 11
; REMAT-NEXT: add a2, a0, a2
; REMAT-NEXT: vle32.v v26, (a2)
; REMAT-NEXT: vle32.v v8, (a2)
; REMAT-NEXT: sf.vc.vv 3, 0, v10, v28
-; REMAT-NEXT: li a2, 19
-; REMAT-NEXT: slli a2, a2, 10
+; REMAT-NEXT: lui a2, 5
+; REMAT-NEXT: addiw a2, a2, -1536
; REMAT-NEXT: add a2, a0, a2
; REMAT-NEXT: vle32.v v28, (a2)
; REMAT-NEXT: vle32.v v10, (a2)
; REMAT-NEXT: sf.vc.vv 3, 0, v12, v30
-; REMAT-NEXT: lui ra, 5
-; REMAT-NEXT: addiw ra, ra, -512
-; REMAT-NEXT: add a2, a0, ra
+; REMAT-NEXT: li a2, 19
+; REMAT-NEXT: slli a2, a2, 10
+; REMAT-NEXT: add a2, a0, a2
; REMAT-NEXT: vle32.v v30, (a2)
; REMAT-NEXT: vle32.v v12, (a2)
; REMAT-NEXT: sf.vc.vv 3, 0, v14, v6
-; REMAT-NEXT: lui s11, 5
-; REMAT-NEXT: add a2, a0, s11
+; REMAT-NEXT: lui ra, 5
+; REMAT-NEXT: addiw ra, ra, -512
+; REMAT-NEXT: add a2, a0, ra
; REMAT-NEXT: vle32.v v6, (a2)
; REMAT-NEXT: vle32.v v14, (a2)
; REMAT-NEXT: sf.vc.vv 3, 0, v16, v4
-; REMAT-NEXT: lui s10, 5
-; REMAT-NEXT: addiw s10, s10, 512
-; REMAT-NEXT: add a2, a0, s10
+; REMAT-NEXT: lui s11, 5
+; REMAT-NEXT: add a2, a0, s11
; REMAT-NEXT: vle32.v v4, (a2)
; REMAT-NEXT: vle32.v v16, (a2)
; REMAT-NEXT: sf.vc.vv 3, 0, v18, v2
-; REMAT-NEXT: li s9, 21
-; REMAT-NEXT: slli s9, s9, 10
-; REMAT-NEXT: add a2, a0, s9
+; REMAT-NEXT: lui s10, 5
+; REMAT-NEXT: addiw s10, s10, 512
+; REMAT-NEXT: add a2, a0, s10
; REMAT-NEXT: vle32.v v2, (a2)
; REMAT-NEXT: vle32.v v18, (a2)
; REMAT-NEXT: sf.vc.vv 3, 0, v20, v24
-; REMAT-NEXT: lui s8, 5
-; REMAT-NEXT: addiw s8, s8, 1536
-; REMAT-NEXT: add a2, a0, s8
+; REMAT-NEXT: li s9, 21
+; REMAT-NEXT: slli s9, s9, 10
+; REMAT-NEXT: add a2, a0, s9
; REMAT-NEXT: vle32.v v24, (a2)
; REMAT-NEXT: vle32.v v20, (a2)
; REMAT-NEXT: sf.vc.vv 3, 0, v22, v26
-; REMAT-NEXT: li s7, 11
-; REMAT-NEXT: slli s7, s7, 11
-; REMAT-NEXT: add a2, a0, s7
+; REMAT-NEXT: lui s8, 5
+; REMAT-NEXT: addiw s8, s8, 1536
+; REMAT-NEXT: add a2, a0, s8
; REMAT-NEXT: vle32.v v26, (a2)
; REMAT-NEXT: vle32.v v22, (a2)
; REMAT-NEXT: sf.vc.vv 3, 0, v8, v28
-; REMAT-NEXT: lui s6, 6
-; REMAT-NEXT: addiw s6, s6, -1536
-; REMAT-NEXT: add a2, a0, s6
+; REMAT-NEXT: li s7, 11
+; REMAT-NEXT: slli s7, s7, 11
+; REMAT-NEXT: add a2, a0, s7
; REMAT-NEXT: vle32.v v28, (a2)
; REMAT-NEXT: vle32.v v8, (a2)
; REMAT-NEXT: sf.vc.vv 3, 0, v10, v30
-; REMAT-NEXT: li s5, 23
-; REMAT-NEXT: slli s5, s5, 10
-; REMAT-NEXT: add a2, a0, s5
+; REMAT-NEXT: lui s6, 6
+; REMAT-NEXT: addiw s6, s6, -1536
+; REMAT-NEXT: add a2, a0, s6
; REMAT-NEXT: vle32.v v30, (a2)
; REMAT-NEXT: vle32.v v10, (a2)
; REMAT-NEXT: sf.vc.vv 3, 0, v12, v6
-; REMAT-NEXT: lui s4, 6
-; REMAT-NEXT: addiw s4, s4, -512
-; REMAT-NEXT: add a2, a0, s4
+; REMAT-NEXT: li s5, 23
+; REMAT-NEXT: slli s5, s5, 10
+; REMAT-NEXT: add a2, a0, s5
; REMAT-NEXT: vle32.v v6, (a2)
; REMAT-NEXT: vle32.v v12, (a2)
; REMAT-NEXT: sf.vc.vv 3, 0, v14, v4
-; REMAT-NEXT: lui s3, 6
-; REMAT-NEXT: add a2, a0, s3
+; REMAT-NEXT: lui s4, 6
+; REMAT-NEXT: addiw s4, s4, -512
+; REMAT-NEXT: add a2, a0, s4
; REMAT-NEXT: vle32.v v4, (a2)
; REMAT-NEXT: vle32.v v14, (a2)
; REMAT-NEXT: sf.vc.vv 3, 0, v16, v2
-; REMAT-NEXT: lui s2, 6
-; REMAT-NEXT: addiw s2, s2, 512
-; REMAT-NEXT: add a2, a0, s2
+; REMAT-NEXT: lui s3, 6
+; REMAT-NEXT: add a2, a0, s3
; REMAT-NEXT: vle32.v v2, (a2)
; REMAT-NEXT: vle32.v v16, (a2)
; REMAT-NEXT: sf.vc.vv 3, 0, v18, v24
-; REMAT-NEXT: li s1, 25
-; REMAT-NEXT: slli s1, s1, 10
-; REMAT-NEXT: add a2, a0, s1
+; REMAT-NEXT: lui s2, 6
+; REMAT-NEXT: addiw s2, s2, 512
+; REMAT-NEXT: add a2, a0, s2
; REMAT-NEXT: vle32.v v0, (a2)
; REMAT-NEXT: vle32.v v18, (a2)
; REMAT-NEXT: sf.vc.vv 3, 0, v20, v26
-; REMAT-NEXT: lui s0, 6
-; REMAT-NEXT: addiw s0, s0, 1536
-; REMAT-NEXT: add a2, a0, s0
+; REMAT-NEXT: li s1, 25
+; REMAT-NEXT: slli s1, s1, 10
+; REMAT-NEXT: add a2, a0, s1
; REMAT-NEXT: vle32.v v26, (a2)
; REMAT-NEXT: vle32.v v20, (a2)
; REMAT-NEXT: sf.vc.vv 3, 0, v22, v28
-; REMAT-NEXT: li t6, 13
-; REMAT-NEXT: slli t6, t6, 11
-; REMAT-NEXT: add a2, a0, t6
+; REMAT-NEXT: lui s0, 6
+; REMAT-NEXT: addiw s0, s0, 1536
+; REMAT-NEXT: add a2, a0, s0
; REMAT-NEXT: vle32.v v28, (a2)
; REMAT-NEXT: vle32.v v22, (a2)
; REMAT-NEXT: sf.vc.vv 3, 0, v8, v30
-; REMAT-NEXT: lui t5, 7
-; REMAT-NEXT: addiw t5, t5, -1536
-; REMAT-NEXT: add a2, a0, t5
+; REMAT-NEXT: li t6, 13
+; REMAT-NEXT: slli t6, t6, 11
+; REMAT-NEXT: add a2, a0, t6
; REMAT-NEXT: vle32.v v30, (a2)
; REMAT-NEXT: vle32.v v24, (a2)
; REMAT-NEXT: sf.vc.vv 3, 0, v10, v6
-; REMAT-NEXT: li t4, 27
-; REMAT-NEXT: slli t4, t4, 10
-; REMAT-NEXT: add a2, a0, t4
+; REMAT-NEXT: lui t5, 7
+; REMAT-NEXT: addiw t5, t5, -1536
+; REMAT-NEXT: add a2, a0, t5
; REMAT-NEXT: vle32.v v6, (a2)
; REMAT-NEXT: vle32.v v10, (a2)
; REMAT-NEXT: sf.vc.vv 3, 0, v12, v4
-; REMAT-NEXT: lui t3, 7
-; REMAT-NEXT: addiw t3, t3, -512
-; REMAT-NEXT: add a2, a0, t3
+; REMAT-NEXT: li t4, 27
+; REMAT-NEXT: slli t4, t4, 10
+; REMAT-NEXT: add a2, a0, t4
; REMAT-NEXT: vle32.v v4, (a2)
; REMAT-NEXT: vle32.v v12, (a2)
; REMAT-NEXT: sf.vc.vv 3, 0, v14, v2
+; REMAT-NEXT: lui t3, 7
+; REMAT-NEXT: addiw t3, t3, -512
+; REMAT-NEXT: add a2, a0, t3
+; REMAT-NEXT: vle32.v v2, (a2)
+; REMAT-NEXT: vle32.v v14, (a2)
+; REMAT-NEXT: sf.vc.vv 3, 0, v16, v0
; REMAT-NEXT: lui t2, 7
; REMAT-NEXT: add a2, a0, t2
-; REMAT-NEXT: vle32.v v2, (a2)
+; REMAT-NEXT: vle32.v v0, (a2)
; REMAT-NEXT: vle32.v v8, (a2)
-; REMAT-NEXT: sf.vc.vv 3, 0, v16, v0
+; REMAT-NEXT: sf.vc.vv 3, 0, v18, v26
; REMAT-NEXT: lui t1, 7
; REMAT-NEXT: addiw t1, t1, 512
; REMAT-NEXT: add a2, a0, t1
-; REMAT-NEXT: vle32.v v14, (a2)
; REMAT-NEXT: vle32.v v16, (a2)
-; REMAT-NEXT: sf.vc.vv 3, 0, v18, v26
+; REMAT-NEXT: vle32.v v18, (a2)
+; REMAT-NEXT: sf.vc.vv 3, 0, v20, v28
; REMAT-NEXT: li t0, 29
; REMAT-NEXT: slli t0, t0, 10
; REMAT-NEXT: add a2, a0, t0
-; REMAT-NEXT: vle32.v v18, (a2)
+; REMAT-NEXT: vle32.v v20, (a2)
; REMAT-NEXT: vle32.v v26, (a2)
-; REMAT-NEXT: sf.vc.vv 3, 0, v20, v28
+; REMAT-NEXT: sf.vc.vv 3, 0, v22, v30
; REMAT-NEXT: lui a7, 7
; REMAT-NEXT: addiw a7, a7, 1536
; REMAT-NEXT: add a2, a0, a7
-; REMAT-NEXT: vle32.v v20, (a2)
+; REMAT-NEXT: vle32.v v22, (a2)
; REMAT-NEXT: vle32.v v28, (a2)
-; REMAT-NEXT: sf.vc.vv 3, 0, v22, v30
+; REMAT-NEXT: sf.vc.vv 3, 0, v24, v6
; REMAT-NEXT: li a6, 15
; REMAT-NEXT: slli a6, a6, 11
; REMAT-NEXT: add a2, a0, a6
-; REMAT-NEXT: vle32.v v22, (a2)
+; REMAT-NEXT: vle32.v v24, (a2)
; REMAT-NEXT: vle32.v v30, (a2)
-; REMAT-NEXT: sf.vc.vv 3, 0, v24, v6
+; REMAT-NEXT: sf.vc.vv 3, 0, v10, v4
; REMAT-NEXT: lui a5, 8
; REMAT-NEXT: addiw a5, a5, -1536
; REMAT-NEXT: add a2, a0, a5
-; REMAT-NEXT: vle32.v v24, (a2)
+; REMAT-NEXT: vle32.v v10, (a2)
; REMAT-NEXT: vle32.v v6, (a2)
-; REMAT-NEXT: sf.vc.vv 3, 0, v10, v4
+; REMAT-NEXT: sf.vc.vv 3, 0, v12, v2
; REMAT-NEXT: li a4, 31
; REMAT-NEXT: slli a4, a4, 10
; REMAT-NEXT: add a2, a0, a4
-; REMAT-NEXT: vle32.v v10, (a2)
+; REMAT-NEXT: vle32.v v12, (a2)
; REMAT-NEXT: vle32.v v4, (a2)
-; REMAT-NEXT: sf.vc.vv 3, 0, v12, v2
+; REMAT-NEXT: sf.vc.vv 3, 0, v14, v0
; REMAT-NEXT: lui a3, 8
; REMAT-NEXT: addiw a3, a3, -512
; REMAT-NEXT: add a2, a0, a3
-; REMAT-NEXT: vle32.v v12, (a2)
+; REMAT-NEXT: vle32.v v14, (a2)
; REMAT-NEXT: vle32.v v2, (a2)
; REMAT-NEXT: lui a2, 8
; REMAT-NEXT: add a0, a0, a2
; REMAT-NEXT: vle32.v v0, (a0)
-; REMAT-NEXT: sf.vc.vv 3, 0, v8, v14
-; REMAT-NEXT: sf.vc.vv 3, 0, v16, v18
-; REMAT-NEXT: sf.vc.vv 3, 0, v26, v20
-; REMAT-NEXT: sf.vc.vv 3, 0, v28, v22
-; REMAT-NEXT: sf.vc.vv 3, 0, v30, v24
-; REMAT-NEXT: sf.vc.vv 3, 0, v6, v10
-; REMAT-NEXT: sf.vc.vv 3, 0, v4, v12
+; REMAT-NEXT: sf.vc.vv 3, 0, v8, v16
+; REMAT-NEXT: sf.vc.vv 3, 0, v18, v20
+; REMAT-NEXT: sf.vc.vv 3, 0, v26, v22
+; REMAT-NEXT: sf.vc.vv 3, 0, v28, v24
+; REMAT-NEXT: sf.vc.vv 3, 0, v30, v10
+; REMAT-NEXT: sf.vc.vv 3, 0, v6, v12
+; REMAT-NEXT: sf.vc.vv 3, 0, v4, v14
; REMAT-NEXT: sf.vc.vv 3, 0, v2, v0
; REMAT-NEXT: sf.vc.v.i 2, 0, v8, 0
; REMAT-NEXT: addi a0, a1, 1024
diff --git a/llvm/test/CodeGen/RISCV/regalloc-last-chance-recoloring-failure.ll b/llvm/test/CodeGen/RISCV/regalloc-last-chance-recoloring-failure.ll
index 81ef6072449e8..c92ba98dcc338 100644
--- a/llvm/test/CodeGen/RISCV/regalloc-last-chance-recoloring-failure.ll
+++ b/llvm/test/CodeGen/RISCV/regalloc-last-chance-recoloring-failure.ll
@@ -43,7 +43,6 @@ define void @last_chance_recoloring_failure() {
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: call func
-; CHECK-NEXT: vsetvli zero, s0, e16, m4, ta, ma
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
@@ -55,6 +54,7 @@ define void @last_chance_recoloring_failure() {
; CHECK-NEXT: vl4r.v v20, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, s0, e16, m4, ta, ma
; CHECK-NEXT: vfwsub.wv v8, v24, v16
; CHECK-NEXT: vsetvli zero, zero, e32, m8, tu, mu
; CHECK-NEXT: vfdiv.vv v8, v24, v8, v0.t
@@ -99,7 +99,6 @@ define void @last_chance_recoloring_failure() {
; SUBREGLIVENESS-NEXT: addi a0, sp, 16
; SUBREGLIVENESS-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
; SUBREGLIVENESS-NEXT: call func
-; SUBREGLIVENESS-NEXT: vsetvli zero, s0, e16, m4, ta, ma
; SUBREGLIVENESS-NEXT: csrr a0, vlenb
; SUBREGLIVENESS-NEXT: slli a0, a0, 3
; SUBREGLIVENESS-NEXT: add a0, sp, a0
@@ -111,6 +110,7 @@ define void @last_chance_recoloring_failure() {
; SUBREGLIVENESS-NEXT: vl4r.v v20, (a0) # Unknown-size Folded Reload
; SUBREGLIVENESS-NEXT: addi a0, sp, 16
; SUBREGLIVENESS-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; SUBREGLIVENESS-NEXT: vsetvli zero, s0, e16, m4, ta, ma
; SUBREGLIVENESS-NEXT: vfwsub.wv v8, v24, v16
; SUBREGLIVENESS-NEXT: vsetvli zero, zero, e32, m8, tu, mu
; SUBREGLIVENESS-NEXT: vfdiv.vv v8, v24, v8, v0.t
diff --git a/llvm/test/CodeGen/RISCV/rvv/abs-vp.ll b/llvm/test/CodeGen/RISCV/rvv/abs-vp.ll
index eb74e2d302f1a..05d6716e47192 100644
--- a/llvm/test/CodeGen/RISCV/rvv/abs-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/abs-vp.ll
@@ -590,13 +590,12 @@ define <vscale x 16 x i64> @vp_abs_nxv16i64(<vscale x 16 x i64> %va, <vscale x 1
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB46_2:
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: slli a1, a1, 3
+; CHECK-NEXT: add a1, sp, a1
+; CHECK-NEXT: addi a1, a1, 16
+; CHECK-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vrsub.vi v16, v8, 0, v0.t
; CHECK-NEXT: vmax.vv v8, v8, v16, v0.t
; CHECK-NEXT: addi a0, sp, 16
diff --git a/llvm/test/CodeGen/RISCV/rvv/access-fixed-objects-by-rvv.ll b/llvm/test/CodeGen/RISCV/rvv/access-fixed-objects-by-rvv.ll
index e578aada5a9cf..91f700ef96800 100644
--- a/llvm/test/CodeGen/RISCV/rvv/access-fixed-objects-by-rvv.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/access-fixed-objects-by-rvv.ll
@@ -39,9 +39,9 @@ define <vscale x 1 x i64> @access_fixed_and_vector_objects(ptr %val) {
; RV64IV-NEXT: addi a0, sp, 8
; RV64IV-NEXT: vl1re64.v v8, (a0)
; RV64IV-NEXT: addi a0, sp, 528
-; RV64IV-NEXT: ld a1, 520(sp)
; RV64IV-NEXT: vl1re64.v v9, (a0)
-; RV64IV-NEXT: vsetvli zero, a1, e64, m1, ta, ma
+; RV64IV-NEXT: ld a0, 520(sp)
+; RV64IV-NEXT: vsetvli zero, a0, e64, m1, ta, ma
; RV64IV-NEXT: vadd.vv v8, v8, v9
; RV64IV-NEXT: csrr a0, vlenb
; RV64IV-NEXT: slli a0, a0, 1
diff --git a/llvm/test/CodeGen/RISCV/rvv/active_lane_mask.ll b/llvm/test/CodeGen/RISCV/rvv/active_lane_mask.ll
index 139579b3d2a36..9cb3991f31f94 100644
--- a/llvm/test/CodeGen/RISCV/rvv/active_lane_mask.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/active_lane_mask.ll
@@ -103,9 +103,9 @@ define <8 x i1> @fv8(ptr %p, i64 %index, i64 %tc) {
define <32 x i1> @fv32(ptr %p, i64 %index, i64 %tc) {
; CHECK-LABEL: fv32:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; CHECK-NEXT: lui a0, %hi(.LCPI8_0)
; CHECK-NEXT: addi a0, a0, %lo(.LCPI8_0)
+; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; CHECK-NEXT: vle8.v v8, (a0)
; CHECK-NEXT: vid.v v16
; CHECK-NEXT: vsaddu.vx v16, v16, a1
@@ -124,31 +124,30 @@ define <64 x i1> @fv64(ptr %p, i64 %index, i64 %tc) {
; CHECK-LABEL: fv64:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; CHECK-NEXT: vid.v v8
+; CHECK-NEXT: vsaddu.vx v8, v8, a1
+; CHECK-NEXT: vmsltu.vx v0, v8, a2
; CHECK-NEXT: lui a0, %hi(.LCPI9_0)
; CHECK-NEXT: addi a0, a0, %lo(.LCPI9_0)
; CHECK-NEXT: vle8.v v8, (a0)
-; CHECK-NEXT: vid.v v16
-; CHECK-NEXT: vsaddu.vx v16, v16, a1
-; CHECK-NEXT: vmsltu.vx v0, v16, a2
-; CHECK-NEXT: vsext.vf8 v16, v8
-; CHECK-NEXT: vsaddu.vx v8, v16, a1
-; CHECK-NEXT: vmsltu.vx v16, v8, a2
-; CHECK-NEXT: vsetivli zero, 4, e8, mf2, tu, ma
-; CHECK-NEXT: vslideup.vi v0, v16, 2
; CHECK-NEXT: lui a0, %hi(.LCPI9_1)
; CHECK-NEXT: addi a0, a0, %lo(.LCPI9_1)
-; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; CHECK-NEXT: vle8.v v8, (a0)
+; CHECK-NEXT: vle8.v v9, (a0)
; CHECK-NEXT: vsext.vf8 v16, v8
-; CHECK-NEXT: vsaddu.vx v8, v16, a1
-; CHECK-NEXT: vmsltu.vx v16, v8, a2
-; CHECK-NEXT: vsetivli zero, 6, e8, mf2, tu, ma
-; CHECK-NEXT: vslideup.vi v0, v16, 4
+; CHECK-NEXT: vsaddu.vx v16, v16, a1
+; CHECK-NEXT: vmsltu.vx v8, v16, a2
+; CHECK-NEXT: vsext.vf8 v16, v9
+; CHECK-NEXT: vsaddu.vx v16, v16, a1
; CHECK-NEXT: lui a0, %hi(.LCPI9_2)
; CHECK-NEXT: addi a0, a0, %lo(.LCPI9_2)
+; CHECK-NEXT: vle8.v v9, (a0)
+; CHECK-NEXT: vmsltu.vx v10, v16, a2
+; CHECK-NEXT: vsetivli zero, 4, e8, mf2, tu, ma
+; CHECK-NEXT: vslideup.vi v0, v8, 2
+; CHECK-NEXT: vsetivli zero, 6, e8, mf2, tu, ma
+; CHECK-NEXT: vslideup.vi v0, v10, 4
; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; CHECK-NEXT: vle8.v v8, (a0)
-; CHECK-NEXT: vsext.vf8 v16, v8
+; CHECK-NEXT: vsext.vf8 v16, v9
; CHECK-NEXT: vsaddu.vx v8, v16, a1
; CHECK-NEXT: vmsltu.vx v16, v8, a2
; CHECK-NEXT: vsetivli zero, 8, e8, mf2, ta, ma
@@ -174,51 +173,48 @@ define <128 x i1> @fv128(ptr %p, i64 %index, i64 %tc) {
; CHECK-NEXT: vsext.vf8 v16, v9
; CHECK-NEXT: vsaddu.vx v16, v16, a1
; CHECK-NEXT: vmsltu.vx v8, v16, a2
-; CHECK-NEXT: vsetivli zero, 4, e8, mf2, tu, ma
-; CHECK-NEXT: vslideup.vi v8, v10, 2
; CHECK-NEXT: lui a0, %hi(.LCPI10_2)
; CHECK-NEXT: addi a0, a0, %lo(.LCPI10_2)
-; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; CHECK-NEXT: vle8.v v9, (a0)
-; CHECK-NEXT: vsext.vf8 v16, v9
-; CHECK-NEXT: vsaddu.vx v16, v16, a1
-; CHECK-NEXT: vmsltu.vx v9, v16, a2
-; CHECK-NEXT: vsetivli zero, 6, e8, mf2, tu, ma
-; CHECK-NEXT: vslideup.vi v8, v9, 4
; CHECK-NEXT: lui a0, %hi(.LCPI10_3)
; CHECK-NEXT: addi a0, a0, %lo(.LCPI10_3)
-; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; CHECK-NEXT: vle8.v v9, (a0)
+; CHECK-NEXT: vle8.v v11, (a0)
; CHECK-NEXT: vsext.vf8 v16, v9
; CHECK-NEXT: vsaddu.vx v16, v16, a1
; CHECK-NEXT: vmsltu.vx v9, v16, a2
-; CHECK-NEXT: vsetivli zero, 8, e8, mf2, ta, ma
-; CHECK-NEXT: vslideup.vi v8, v9, 6
-; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; CHECK-NEXT: lui a0, %hi(.LCPI10_4)
-; CHECK-NEXT: addi a0, a0, %lo(.LCPI10_4)
-; CHECK-NEXT: vle8.v v9, (a0)
+; CHECK-NEXT: vsext.vf8 v16, v11
+; CHECK-NEXT: vsaddu.vx v16, v16, a1
+; CHECK-NEXT: vmsltu.vx v11, v16, a2
; CHECK-NEXT: vid.v v16
; CHECK-NEXT: vsaddu.vx v16, v16, a1
; CHECK-NEXT: vmsltu.vx v0, v16, a2
-; CHECK-NEXT: vsext.vf8 v16, v9
-; CHECK-NEXT: vsaddu.vx v16, v16, a1
-; CHECK-NEXT: vmsltu.vx v9, v16, a2
-; CHECK-NEXT: vsetivli zero, 4, e8, mf2, tu, ma
-; CHECK-NEXT: vslideup.vi v0, v9, 2
+; CHECK-NEXT: lui a0, %hi(.LCPI10_4)
+; CHECK-NEXT: addi a0, a0, %lo(.LCPI10_4)
+; CHECK-NEXT: vle8.v v12, (a0)
; CHECK-NEXT: lui a0, %hi(.LCPI10_5)
; CHECK-NEXT: addi a0, a0, %lo(.LCPI10_5)
-; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; CHECK-NEXT: vle8.v v9, (a0)
-; CHECK-NEXT: vsext.vf8 v16, v9
+; CHECK-NEXT: vle8.v v13, (a0)
+; CHECK-NEXT: vsext.vf8 v16, v12
; CHECK-NEXT: vsaddu.vx v16, v16, a1
-; CHECK-NEXT: vmsltu.vx v9, v16, a2
+; CHECK-NEXT: vmsltu.vx v12, v16, a2
+; CHECK-NEXT: vsext.vf8 v16, v13
+; CHECK-NEXT: vsaddu.vx v16, v16, a1
+; CHECK-NEXT: vmsltu.vx v13, v16, a2
+; CHECK-NEXT: vsetivli zero, 4, e8, mf2, tu, ma
+; CHECK-NEXT: vslideup.vi v8, v10, 2
; CHECK-NEXT: vsetivli zero, 6, e8, mf2, tu, ma
-; CHECK-NEXT: vslideup.vi v0, v9, 4
+; CHECK-NEXT: vslideup.vi v8, v9, 4
; CHECK-NEXT: lui a0, %hi(.LCPI10_6)
; CHECK-NEXT: addi a0, a0, %lo(.LCPI10_6)
-; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; CHECK-NEXT: vsetivli zero, 16, e8, m1, ta, ma
; CHECK-NEXT: vle8.v v9, (a0)
+; CHECK-NEXT: vsetivli zero, 8, e8, mf2, ta, ma
+; CHECK-NEXT: vslideup.vi v8, v11, 6
+; CHECK-NEXT: vsetivli zero, 4, e8, mf2, tu, ma
+; CHECK-NEXT: vslideup.vi v0, v12, 2
+; CHECK-NEXT: vsetivli zero, 6, e8, mf2, tu, ma
+; CHECK-NEXT: vslideup.vi v0, v13, 4
+; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; CHECK-NEXT: vsext.vf8 v16, v9
; CHECK-NEXT: vsaddu.vx v16, v16, a1
; CHECK-NEXT: vmsltu.vx v9, v16, a2
diff --git a/llvm/test/CodeGen/RISCV/rvv/bitreverse-vp.ll b/llvm/test/CodeGen/RISCV/rvv/bitreverse-vp.ll
index 879dff4a6e490..5217148ba4f4e 100644
--- a/llvm/test/CodeGen/RISCV/rvv/bitreverse-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/bitreverse-vp.ll
@@ -1449,27 +1449,27 @@ define <vscale x 1 x i64> @vp_bitreverse_nxv1i64(<vscale x 1 x i64> %va, <vscale
; RV32-NEXT: li a3, 40
; RV32-NEXT: vsll.vx v10, v10, a3, v0.t
; RV32-NEXT: vor.vv v9, v9, v10, v0.t
+; RV32-NEXT: addi a4, sp, 8
+; RV32-NEXT: vsetvli a5, zero, e64, m1, ta, ma
+; RV32-NEXT: vlse64.v v10, (a4), zero
; RV32-NEXT: lui a4, 4080
-; RV32-NEXT: vand.vx v10, v8, a4, v0.t
-; RV32-NEXT: vsll.vi v10, v10, 24, v0.t
-; RV32-NEXT: addi a5, sp, 8
-; RV32-NEXT: vsetvli a6, zero, e64, m1, ta, ma
-; RV32-NEXT: vlse64.v v11, (a5), zero
; RV32-NEXT: vsetvli zero, a0, e64, m1, ta, ma
-; RV32-NEXT: vand.vv v12, v8, v11, v0.t
+; RV32-NEXT: vand.vx v11, v8, a4, v0.t
+; RV32-NEXT: vsll.vi v11, v11, 24, v0.t
+; RV32-NEXT: vand.vv v12, v8, v10, v0.t
; RV32-NEXT: vsll.vi v12, v12, 8, v0.t
-; RV32-NEXT: vor.vv v10, v10, v12, v0.t
-; RV32-NEXT: vor.vv v9, v9, v10, v0.t
-; RV32-NEXT: vsrl.vx v10, v8, a1, v0.t
+; RV32-NEXT: vor.vv v11, v11, v12, v0.t
+; RV32-NEXT: vor.vv v9, v9, v11, v0.t
+; RV32-NEXT: vsrl.vx v11, v8, a1, v0.t
; RV32-NEXT: vsrl.vx v12, v8, a3, v0.t
; RV32-NEXT: vand.vx v12, v12, a2, v0.t
-; RV32-NEXT: vor.vv v10, v12, v10, v0.t
+; RV32-NEXT: vor.vv v11, v12, v11, v0.t
; RV32-NEXT: vsrl.vi v12, v8, 24, v0.t
; RV32-NEXT: vand.vx v12, v12, a4, v0.t
; RV32-NEXT: vsrl.vi v8, v8, 8, v0.t
-; RV32-NEXT: vand.vv v8, v8, v11, v0.t
+; RV32-NEXT: vand.vv v8, v8, v10, v0.t
; RV32-NEXT: vor.vv v8, v8, v12, v0.t
-; RV32-NEXT: vor.vv v8, v8, v10, v0.t
+; RV32-NEXT: vor.vv v8, v8, v11, v0.t
; RV32-NEXT: vor.vv v8, v9, v8, v0.t
; RV32-NEXT: vsrl.vi v9, v8, 4, v0.t
; RV32-NEXT: lui a1, 61681
@@ -1590,27 +1590,27 @@ define <vscale x 1 x i64> @vp_bitreverse_nxv1i64_unmasked(<vscale x 1 x i64> %va
; RV32-NEXT: li a3, 40
; RV32-NEXT: vsll.vx v10, v10, a3
; RV32-NEXT: vor.vv v9, v9, v10
+; RV32-NEXT: addi a4, sp, 8
+; RV32-NEXT: vsetvli a5, zero, e64, m1, ta, ma
+; RV32-NEXT: vlse64.v v10, (a4), zero
; RV32-NEXT: lui a4, 4080
-; RV32-NEXT: vand.vx v10, v8, a4
-; RV32-NEXT: vsll.vi v10, v10, 24
-; RV32-NEXT: addi a5, sp, 8
-; RV32-NEXT: vsetvli a6, zero, e64, m1, ta, ma
-; RV32-NEXT: vlse64.v v11, (a5), zero
; RV32-NEXT: vsetvli zero, a0, e64, m1, ta, ma
-; RV32-NEXT: vand.vv v12, v8, v11
+; RV32-NEXT: vand.vx v11, v8, a4
+; RV32-NEXT: vsll.vi v11, v11, 24
+; RV32-NEXT: vand.vv v12, v8, v10
; RV32-NEXT: vsll.vi v12, v12, 8
-; RV32-NEXT: vor.vv v10, v10, v12
-; RV32-NEXT: vor.vv v9, v9, v10
-; RV32-NEXT: vsrl.vx v10, v8, a1
+; RV32-NEXT: vor.vv v11, v11, v12
+; RV32-NEXT: vor.vv v9, v9, v11
+; RV32-NEXT: vsrl.vx v11, v8, a1
; RV32-NEXT: vsrl.vx v12, v8, a3
; RV32-NEXT: vand.vx v12, v12, a2
-; RV32-NEXT: vor.vv v10, v12, v10
+; RV32-NEXT: vor.vv v11, v12, v11
; RV32-NEXT: vsrl.vi v12, v8, 24
; RV32-NEXT: vand.vx v12, v12, a4
; RV32-NEXT: vsrl.vi v8, v8, 8
-; RV32-NEXT: vand.vv v8, v8, v11
+; RV32-NEXT: vand.vv v8, v8, v10
; RV32-NEXT: vor.vv v8, v8, v12
-; RV32-NEXT: vor.vv v8, v8, v10
+; RV32-NEXT: vor.vv v8, v8, v11
; RV32-NEXT: vor.vv v8, v9, v8
; RV32-NEXT: vsrl.vi v9, v8, 4
; RV32-NEXT: lui a1, 61681
@@ -1733,27 +1733,27 @@ define <vscale x 2 x i64> @vp_bitreverse_nxv2i64(<vscale x 2 x i64> %va, <vscale
; RV32-NEXT: li a3, 40
; RV32-NEXT: vsll.vx v12, v12, a3, v0.t
; RV32-NEXT: vor.vv v10, v10, v12, v0.t
+; RV32-NEXT: addi a4, sp, 8
+; RV32-NEXT: vsetvli a5, zero, e64, m2, ta, ma
+; RV32-NEXT: vlse64.v v12, (a4), zero
; RV32-NEXT: lui a4, 4080
-; RV32-NEXT: vand.vx v12, v8, a4, v0.t
-; RV32-NEXT: vsll.vi v12, v12, 24, v0.t
-; RV32-NEXT: addi a5, sp, 8
-; RV32-NEXT: vsetvli a6, zero, e64, m2, ta, ma
-; RV32-NEXT: vlse64.v v14, (a5), zero
; RV32-NEXT: vsetvli zero, a0, e64, m2, ta, ma
-; RV32-NEXT: vand.vv v16, v8, v14, v0.t
+; RV32-NEXT: vand.vx v14, v8, a4, v0.t
+; RV32-NEXT: vsll.vi v14, v14, 24, v0.t
+; RV32-NEXT: vand.vv v16, v8, v12, v0.t
; RV32-NEXT: vsll.vi v16, v16, 8, v0.t
-; RV32-NEXT: vor.vv v12, v12, v16, v0.t
-; RV32-NEXT: vor.vv v10, v10, v12, v0.t
-; RV32-NEXT: vsrl.vx v12, v8, a1, v0.t
+; RV32-NEXT: vor.vv v14, v14, v16, v0.t
+; RV32-NEXT: vor.vv v10, v10, v14, v0.t
+; RV32-NEXT: vsrl.vx v14, v8, a1, v0.t
; RV32-NEXT: vsrl.vx v16, v8, a3, v0.t
; RV32-NEXT: vand.vx v16, v16, a2, v0.t
-; RV32-NEXT: vor.vv v12, v16, v12, v0.t
+; RV32-NEXT: vor.vv v14, v16, v14, v0.t
; RV32-NEXT: vsrl.vi v16, v8, 24, v0.t
; RV32-NEXT: vand.vx v16, v16, a4, v0.t
; RV32-NEXT: vsrl.vi v8, v8, 8, v0.t
-; RV32-NEXT: vand.vv v8, v8, v14, v0.t
+; RV32-NEXT: vand.vv v8, v8, v12, v0.t
; RV32-NEXT: vor.vv v8, v8, v16, v0.t
-; RV32-NEXT: vor.vv v8, v8, v12, v0.t
+; RV32-NEXT: vor.vv v8, v8, v14, v0.t
; RV32-NEXT: vor.vv v8, v10, v8, v0.t
; RV32-NEXT: vsrl.vi v10, v8, 4, v0.t
; RV32-NEXT: lui a1, 61681
@@ -1874,27 +1874,27 @@ define <vscale x 2 x i64> @vp_bitreverse_nxv2i64_unmasked(<vscale x 2 x i64> %va
; RV32-NEXT: li a3, 40
; RV32-NEXT: vsll.vx v12, v12, a3
; RV32-NEXT: vor.vv v10, v10, v12
+; RV32-NEXT: addi a4, sp, 8
+; RV32-NEXT: vsetvli a5, zero, e64, m2, ta, ma
+; RV32-NEXT: vlse64.v v12, (a4), zero
; RV32-NEXT: lui a4, 4080
-; RV32-NEXT: vand.vx v12, v8, a4
-; RV32-NEXT: vsll.vi v12, v12, 24
-; RV32-NEXT: addi a5, sp, 8
-; RV32-NEXT: vsetvli a6, zero, e64, m2, ta, ma
-; RV32-NEXT: vlse64.v v14, (a5), zero
; RV32-NEXT: vsetvli zero, a0, e64, m2, ta, ma
-; RV32-NEXT: vand.vv v16, v8, v14
+; RV32-NEXT: vand.vx v14, v8, a4
+; RV32-NEXT: vsll.vi v14, v14, 24
+; RV32-NEXT: vand.vv v16, v8, v12
; RV32-NEXT: vsll.vi v16, v16, 8
-; RV32-NEXT: vor.vv v12, v12, v16
-; RV32-NEXT: vor.vv v10, v10, v12
-; RV32-NEXT: vsrl.vx v12, v8, a1
+; RV32-NEXT: vor.vv v14, v14, v16
+; RV32-NEXT: vor.vv v10, v10, v14
+; RV32-NEXT: vsrl.vx v14, v8, a1
; RV32-NEXT: vsrl.vx v16, v8, a3
; RV32-NEXT: vand.vx v16, v16, a2
-; RV32-NEXT: vor.vv v12, v16, v12
+; RV32-NEXT: vor.vv v14, v16, v14
; RV32-NEXT: vsrl.vi v16, v8, 24
; RV32-NEXT: vand.vx v16, v16, a4
; RV32-NEXT: vsrl.vi v8, v8, 8
-; RV32-NEXT: vand.vv v8, v8, v14
+; RV32-NEXT: vand.vv v8, v8, v12
; RV32-NEXT: vor.vv v8, v8, v16
-; RV32-NEXT: vor.vv v8, v8, v12
+; RV32-NEXT: vor.vv v8, v8, v14
; RV32-NEXT: vor.vv v8, v10, v8
; RV32-NEXT: vsrl.vi v10, v8, 4
; RV32-NEXT: lui a1, 61681
@@ -2017,13 +2017,13 @@ define <vscale x 4 x i64> @vp_bitreverse_nxv4i64(<vscale x 4 x i64> %va, <vscale
; RV32-NEXT: li a3, 40
; RV32-NEXT: vsll.vx v16, v16, a3, v0.t
; RV32-NEXT: vor.vv v16, v12, v16, v0.t
+; RV32-NEXT: addi a4, sp, 8
+; RV32-NEXT: vsetvli a5, zero, e64, m4, ta, ma
+; RV32-NEXT: vlse64.v v12, (a4), zero
; RV32-NEXT: lui a4, 4080
-; RV32-NEXT: vand.vx v12, v8, a4, v0.t
-; RV32-NEXT: vsll.vi v20, v12, 24, v0.t
-; RV32-NEXT: addi a5, sp, 8
-; RV32-NEXT: vsetvli a6, zero, e64, m4, ta, ma
-; RV32-NEXT: vlse64.v v12, (a5), zero
; RV32-NEXT: vsetvli zero, a0, e64, m4, ta, ma
+; RV32-NEXT: vand.vx v20, v8, a4, v0.t
+; RV32-NEXT: vsll.vi v20, v20, 24, v0.t
; RV32-NEXT: vand.vv v24, v8, v12, v0.t
; RV32-NEXT: vsll.vi v24, v24, 8, v0.t
; RV32-NEXT: vor.vv v20, v20, v24, v0.t
@@ -2158,27 +2158,27 @@ define <vscale x 4 x i64> @vp_bitreverse_nxv4i64_unmasked(<vscale x 4 x i64> %va
; RV32-NEXT: li a3, 40
; RV32-NEXT: vsll.vx v16, v16, a3
; RV32-NEXT: vor.vv v12, v12, v16
+; RV32-NEXT: addi a4, sp, 8
+; RV32-NEXT: vsetvli a5, zero, e64, m4, ta, ma
+; RV32-NEXT: vlse64.v v16, (a4), zero
; RV32-NEXT: lui a4, 4080
-; RV32-NEXT: vand.vx v16, v8, a4
-; RV32-NEXT: vsll.vi v16, v16, 24
-; RV32-NEXT: addi a5, sp, 8
-; RV32-NEXT: vsetvli a6, zero, e64, m4, ta, ma
-; RV32-NEXT: vlse64.v v20, (a5), zero
; RV32-NEXT: vsetvli zero, a0, e64, m4, ta, ma
-; RV32-NEXT: vand.vv v24, v8, v20
+; RV32-NEXT: vand.vx v20, v8, a4
+; RV32-NEXT: vsll.vi v20, v20, 24
+; RV32-NEXT: vand.vv v24, v8, v16
; RV32-NEXT: vsll.vi v24, v24, 8
-; RV32-NEXT: vor.vv v16, v16, v24
-; RV32-NEXT: vor.vv v12, v12, v16
-; RV32-NEXT: vsrl.vx v16, v8, a1
+; RV32-NEXT: vor.vv v20, v20, v24
+; RV32-NEXT: vor.vv v12, v12, v20
+; RV32-NEXT: vsrl.vx v20, v8, a1
; RV32-NEXT: vsrl.vx v24, v8, a3
; RV32-NEXT: vand.vx v24, v24, a2
-; RV32-NEXT: vor.vv v16, v24, v16
+; RV32-NEXT: vor.vv v20, v24, v20
; RV32-NEXT: vsrl.vi v24, v8, 24
; RV32-NEXT: vand.vx v24, v24, a4
; RV32-NEXT: vsrl.vi v8, v8, 8
-; RV32-NEXT: vand.vv v8, v8, v20
+; RV32-NEXT: vand.vv v8, v8, v16
; RV32-NEXT: vor.vv v8, v8, v24
-; RV32-NEXT: vor.vv v8, v8, v16
+; RV32-NEXT: vor.vv v8, v8, v20
; RV32-NEXT: vor.vv v8, v12, v8
; RV32-NEXT: vsrl.vi v12, v8, 4
; RV32-NEXT: lui a1, 61681
@@ -2311,20 +2311,23 @@ define <vscale x 7 x i64> @vp_bitreverse_nxv7i64(<vscale x 7 x i64> %va, <vscale
; RV32-NEXT: add a4, sp, a4
; RV32-NEXT: addi a4, a4, 16
; RV32-NEXT: vs8r.v v16, (a4) # Unknown-size Folded Spill
+; RV32-NEXT: addi a4, sp, 8
+; RV32-NEXT: vsetvli a5, zero, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v16, (a4), zero
+; RV32-NEXT: csrr a4, vlenb
+; RV32-NEXT: slli a4, a4, 3
+; RV32-NEXT: add a4, sp, a4
+; RV32-NEXT: addi a4, a4, 16
+; RV32-NEXT: vs8r.v v16, (a4) # Unknown-size Folded Spill
; RV32-NEXT: lui a4, 4080
-; RV32-NEXT: vand.vx v16, v8, a4, v0.t
-; RV32-NEXT: vsll.vi v24, v16, 24, v0.t
-; RV32-NEXT: addi a5, sp, 8
-; RV32-NEXT: vsetvli a6, zero, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a5), zero
-; RV32-NEXT: csrr a5, vlenb
-; RV32-NEXT: slli a5, a5, 3
-; RV32-NEXT: add a5, sp, a5
-; RV32-NEXT: addi a5, a5, 16
-; RV32-NEXT: vs8r.v v16, (a5) # Unknown-size Folded Spill
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v8, v16, v0.t
-; RV32-NEXT: vsll.vi v16, v16, 8, v0.t
+; RV32-NEXT: vand.vx v24, v8, a4, v0.t
+; RV32-NEXT: vsll.vi v24, v24, 24, v0.t
+; RV32-NEXT: addi a5, sp, 16
+; RV32-NEXT: vs8r.v v24, (a5) # Unknown-size Folded Spill
+; RV32-NEXT: vand.vv v24, v8, v16, v0.t
+; RV32-NEXT: vsll.vi v16, v24, 8, v0.t
+; RV32-NEXT: vl8r.v v24, (a5) # Unknown-size Folded Reload
; RV32-NEXT: vor.vv v16, v24, v16, v0.t
; RV32-NEXT: csrr a5, vlenb
; RV32-NEXT: slli a5, a5, 4
@@ -2337,10 +2340,10 @@ define <vscale x 7 x i64> @vp_bitreverse_nxv7i64(<vscale x 7 x i64> %va, <vscale
; RV32-NEXT: add a5, sp, a5
; RV32-NEXT: addi a5, a5, 16
; RV32-NEXT: vs8r.v v16, (a5) # Unknown-size Folded Spill
-; RV32-NEXT: vsrl.vx v24, v8, a1, v0.t
-; RV32-NEXT: vsrl.vx v16, v8, a3, v0.t
-; RV32-NEXT: vand.vx v16, v16, a2, v0.t
-; RV32-NEXT: vor.vv v16, v16, v24, v0.t
+; RV32-NEXT: vsrl.vx v16, v8, a1, v0.t
+; RV32-NEXT: vsrl.vx v24, v8, a3, v0.t
+; RV32-NEXT: vand.vx v24, v24, a2, v0.t
+; RV32-NEXT: vor.vv v16, v24, v16, v0.t
; RV32-NEXT: addi a1, sp, 16
; RV32-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
; RV32-NEXT: vsrl.vi v24, v8, 24, v0.t
@@ -2508,13 +2511,13 @@ define <vscale x 7 x i64> @vp_bitreverse_nxv7i64_unmasked(<vscale x 7 x i64> %va
; RV32-NEXT: vor.vv v16, v16, v24
; RV32-NEXT: addi a4, sp, 16
; RV32-NEXT: vs8r.v v16, (a4) # Unknown-size Folded Spill
+; RV32-NEXT: addi a4, sp, 8
+; RV32-NEXT: vsetvli a5, zero, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v16, (a4), zero
; RV32-NEXT: lui a4, 4080
-; RV32-NEXT: vand.vx v16, v8, a4
-; RV32-NEXT: vsll.vi v0, v16, 24
-; RV32-NEXT: addi a5, sp, 8
-; RV32-NEXT: vsetvli a6, zero, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a5), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vand.vx v0, v8, a4
+; RV32-NEXT: vsll.vi v0, v0, 24
; RV32-NEXT: vand.vv v24, v8, v16
; RV32-NEXT: vsll.vi v24, v24, 8
; RV32-NEXT: vor.vv v24, v0, v24
@@ -2669,20 +2672,23 @@ define <vscale x 8 x i64> @vp_bitreverse_nxv8i64(<vscale x 8 x i64> %va, <vscale
; RV32-NEXT: add a4, sp, a4
; RV32-NEXT: addi a4, a4, 16
; RV32-NEXT: vs8r.v v16, (a4) # Unknown-size Folded Spill
+; RV32-NEXT: addi a4, sp, 8
+; RV32-NEXT: vsetvli a5, zero, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v16, (a4), zero
+; RV32-NEXT: csrr a4, vlenb
+; RV32-NEXT: slli a4, a4, 3
+; RV32-NEXT: add a4, sp, a4
+; RV32-NEXT: addi a4, a4, 16
+; RV32-NEXT: vs8r.v v16, (a4) # Unknown-size Folded Spill
; RV32-NEXT: lui a4, 4080
-; RV32-NEXT: vand.vx v16, v8, a4, v0.t
-; RV32-NEXT: vsll.vi v24, v16, 24, v0.t
-; RV32-NEXT: addi a5, sp, 8
-; RV32-NEXT: vsetvli a6, zero, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a5), zero
-; RV32-NEXT: csrr a5, vlenb
-; RV32-NEXT: slli a5, a5, 3
-; RV32-NEXT: add a5, sp, a5
-; RV32-NEXT: addi a5, a5, 16
-; RV32-NEXT: vs8r.v v16, (a5) # Unknown-size Folded Spill
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v8, v16, v0.t
-; RV32-NEXT: vsll.vi v16, v16, 8, v0.t
+; RV32-NEXT: vand.vx v24, v8, a4, v0.t
+; RV32-NEXT: vsll.vi v24, v24, 24, v0.t
+; RV32-NEXT: addi a5, sp, 16
+; RV32-NEXT: vs8r.v v24, (a5) # Unknown-size Folded Spill
+; RV32-NEXT: vand.vv v24, v8, v16, v0.t
+; RV32-NEXT: vsll.vi v16, v24, 8, v0.t
+; RV32-NEXT: vl8r.v v24, (a5) # Unknown-size Folded Reload
; RV32-NEXT: vor.vv v16, v24, v16, v0.t
; RV32-NEXT: csrr a5, vlenb
; RV32-NEXT: slli a5, a5, 4
@@ -2695,10 +2701,10 @@ define <vscale x 8 x i64> @vp_bitreverse_nxv8i64(<vscale x 8 x i64> %va, <vscale
; RV32-NEXT: add a5, sp, a5
; RV32-NEXT: addi a5, a5, 16
; RV32-NEXT: vs8r.v v16, (a5) # Unknown-size Folded Spill
-; RV32-NEXT: vsrl.vx v24, v8, a1, v0.t
-; RV32-NEXT: vsrl.vx v16, v8, a3, v0.t
-; RV32-NEXT: vand.vx v16, v16, a2, v0.t
-; RV32-NEXT: vor.vv v16, v16, v24, v0.t
+; RV32-NEXT: vsrl.vx v16, v8, a1, v0.t
+; RV32-NEXT: vsrl.vx v24, v8, a3, v0.t
+; RV32-NEXT: vand.vx v24, v24, a2, v0.t
+; RV32-NEXT: vor.vv v16, v24, v16, v0.t
; RV32-NEXT: addi a1, sp, 16
; RV32-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
; RV32-NEXT: vsrl.vi v24, v8, 24, v0.t
@@ -2866,13 +2872,13 @@ define <vscale x 8 x i64> @vp_bitreverse_nxv8i64_unmasked(<vscale x 8 x i64> %va
; RV32-NEXT: vor.vv v16, v16, v24
; RV32-NEXT: addi a4, sp, 16
; RV32-NEXT: vs8r.v v16, (a4) # Unknown-size Folded Spill
+; RV32-NEXT: addi a4, sp, 8
+; RV32-NEXT: vsetvli a5, zero, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v16, (a4), zero
; RV32-NEXT: lui a4, 4080
-; RV32-NEXT: vand.vx v16, v8, a4
-; RV32-NEXT: vsll.vi v0, v16, 24
-; RV32-NEXT: addi a5, sp, 8
-; RV32-NEXT: vsetvli a6, zero, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a5), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vand.vx v0, v8, a4
+; RV32-NEXT: vsll.vi v0, v0, 24
; RV32-NEXT: vand.vv v24, v8, v16
; RV32-NEXT: vsll.vi v24, v24, 8
; RV32-NEXT: vor.vv v24, v0, v24
@@ -3056,13 +3062,13 @@ define <vscale x 64 x i16> @vp_bitreverse_nxv64i16(<vscale x 64 x i16> %va, <vsc
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a2
; CHECK-NEXT: .LBB46_2:
-; CHECK-NEXT: vsetvli zero, a0, e16, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: csrr a2, vlenb
+; CHECK-NEXT: slli a2, a2, 3
+; CHECK-NEXT: add a2, sp, a2
+; CHECK-NEXT: addi a2, a2, 16
+; CHECK-NEXT: vl8r.v v8, (a2) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, a0, e16, m8, ta, ma
; CHECK-NEXT: vsrl.vi v16, v8, 8, v0.t
; CHECK-NEXT: vsll.vi v8, v8, 8, v0.t
; CHECK-NEXT: vor.vv v8, v8, v16, v0.t
@@ -3107,8 +3113,8 @@ define <vscale x 64 x i16> @vp_bitreverse_nxv64i16(<vscale x 64 x i16> %va, <vsc
; CHECK-ZVBB-NEXT: # %bb.1:
; CHECK-ZVBB-NEXT: mv a0, a1
; CHECK-ZVBB-NEXT: .LBB46_2:
-; CHECK-ZVBB-NEXT: vsetvli zero, a0, e16, m8, ta, ma
; CHECK-ZVBB-NEXT: vmv1r.v v0, v24
+; CHECK-ZVBB-NEXT: vsetvli zero, a0, e16, m8, ta, ma
; CHECK-ZVBB-NEXT: vbrev.v v8, v8, v0.t
; CHECK-ZVBB-NEXT: ret
%v = call <vscale x 64 x i16> @llvm.vp.bitreverse.nxv64i16(<vscale x 64 x i16> %va, <vscale x 64 x i1> %m, i32 %evl)
diff --git a/llvm/test/CodeGen/RISCV/rvv/bswap-vp.ll b/llvm/test/CodeGen/RISCV/rvv/bswap-vp.ll
index 800d06c5a78f5..aadd9852af11e 100644
--- a/llvm/test/CodeGen/RISCV/rvv/bswap-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/bswap-vp.ll
@@ -525,27 +525,27 @@ define <vscale x 1 x i64> @vp_bswap_nxv1i64(<vscale x 1 x i64> %va, <vscale x 1
; RV32-NEXT: li a3, 40
; RV32-NEXT: vsll.vx v10, v10, a3, v0.t
; RV32-NEXT: vor.vv v9, v9, v10, v0.t
+; RV32-NEXT: addi a4, sp, 8
+; RV32-NEXT: vsetvli a5, zero, e64, m1, ta, ma
+; RV32-NEXT: vlse64.v v10, (a4), zero
; RV32-NEXT: lui a4, 4080
-; RV32-NEXT: vand.vx v10, v8, a4, v0.t
-; RV32-NEXT: vsll.vi v10, v10, 24, v0.t
-; RV32-NEXT: addi a5, sp, 8
-; RV32-NEXT: vsetvli a6, zero, e64, m1, ta, ma
-; RV32-NEXT: vlse64.v v11, (a5), zero
; RV32-NEXT: vsetvli zero, a0, e64, m1, ta, ma
-; RV32-NEXT: vand.vv v12, v8, v11, v0.t
+; RV32-NEXT: vand.vx v11, v8, a4, v0.t
+; RV32-NEXT: vsll.vi v11, v11, 24, v0.t
+; RV32-NEXT: vand.vv v12, v8, v10, v0.t
; RV32-NEXT: vsll.vi v12, v12, 8, v0.t
-; RV32-NEXT: vor.vv v10, v10, v12, v0.t
-; RV32-NEXT: vor.vv v9, v9, v10, v0.t
-; RV32-NEXT: vsrl.vx v10, v8, a1, v0.t
+; RV32-NEXT: vor.vv v11, v11, v12, v0.t
+; RV32-NEXT: vor.vv v9, v9, v11, v0.t
+; RV32-NEXT: vsrl.vx v11, v8, a1, v0.t
; RV32-NEXT: vsrl.vx v12, v8, a3, v0.t
; RV32-NEXT: vand.vx v12, v12, a2, v0.t
-; RV32-NEXT: vor.vv v10, v12, v10, v0.t
+; RV32-NEXT: vor.vv v11, v12, v11, v0.t
; RV32-NEXT: vsrl.vi v12, v8, 24, v0.t
; RV32-NEXT: vand.vx v12, v12, a4, v0.t
; RV32-NEXT: vsrl.vi v8, v8, 8, v0.t
-; RV32-NEXT: vand.vv v8, v8, v11, v0.t
+; RV32-NEXT: vand.vv v8, v8, v10, v0.t
; RV32-NEXT: vor.vv v8, v8, v12, v0.t
-; RV32-NEXT: vor.vv v8, v8, v10, v0.t
+; RV32-NEXT: vor.vv v8, v8, v11, v0.t
; RV32-NEXT: vor.vv v8, v9, v8, v0.t
; RV32-NEXT: addi sp, sp, 16
; RV32-NEXT: ret
@@ -609,27 +609,27 @@ define <vscale x 1 x i64> @vp_bswap_nxv1i64_unmasked(<vscale x 1 x i64> %va, i32
; RV32-NEXT: li a3, 40
; RV32-NEXT: vsll.vx v10, v10, a3
; RV32-NEXT: vor.vv v9, v9, v10
+; RV32-NEXT: addi a4, sp, 8
+; RV32-NEXT: vsetvli a5, zero, e64, m1, ta, ma
+; RV32-NEXT: vlse64.v v10, (a4), zero
; RV32-NEXT: lui a4, 4080
-; RV32-NEXT: vand.vx v10, v8, a4
-; RV32-NEXT: vsll.vi v10, v10, 24
-; RV32-NEXT: addi a5, sp, 8
-; RV32-NEXT: vsetvli a6, zero, e64, m1, ta, ma
-; RV32-NEXT: vlse64.v v11, (a5), zero
; RV32-NEXT: vsetvli zero, a0, e64, m1, ta, ma
-; RV32-NEXT: vand.vv v12, v8, v11
+; RV32-NEXT: vand.vx v11, v8, a4
+; RV32-NEXT: vsll.vi v11, v11, 24
+; RV32-NEXT: vand.vv v12, v8, v10
; RV32-NEXT: vsll.vi v12, v12, 8
-; RV32-NEXT: vor.vv v10, v10, v12
-; RV32-NEXT: vor.vv v9, v9, v10
-; RV32-NEXT: vsrl.vx v10, v8, a1
+; RV32-NEXT: vor.vv v11, v11, v12
+; RV32-NEXT: vor.vv v9, v9, v11
+; RV32-NEXT: vsrl.vx v11, v8, a1
; RV32-NEXT: vsrl.vx v12, v8, a3
; RV32-NEXT: vand.vx v12, v12, a2
-; RV32-NEXT: vor.vv v10, v12, v10
+; RV32-NEXT: vor.vv v11, v12, v11
; RV32-NEXT: vsrl.vi v12, v8, 24
; RV32-NEXT: vand.vx v12, v12, a4
; RV32-NEXT: vsrl.vi v8, v8, 8
-; RV32-NEXT: vand.vv v8, v8, v11
+; RV32-NEXT: vand.vv v8, v8, v10
; RV32-NEXT: vor.vv v8, v8, v12
-; RV32-NEXT: vor.vv v8, v8, v10
+; RV32-NEXT: vor.vv v8, v8, v11
; RV32-NEXT: vor.vv v8, v9, v8
; RV32-NEXT: addi sp, sp, 16
; RV32-NEXT: ret
@@ -695,27 +695,27 @@ define <vscale x 2 x i64> @vp_bswap_nxv2i64(<vscale x 2 x i64> %va, <vscale x 2
; RV32-NEXT: li a3, 40
; RV32-NEXT: vsll.vx v12, v12, a3, v0.t
; RV32-NEXT: vor.vv v10, v10, v12, v0.t
+; RV32-NEXT: addi a4, sp, 8
+; RV32-NEXT: vsetvli a5, zero, e64, m2, ta, ma
+; RV32-NEXT: vlse64.v v12, (a4), zero
; RV32-NEXT: lui a4, 4080
-; RV32-NEXT: vand.vx v12, v8, a4, v0.t
-; RV32-NEXT: vsll.vi v12, v12, 24, v0.t
-; RV32-NEXT: addi a5, sp, 8
-; RV32-NEXT: vsetvli a6, zero, e64, m2, ta, ma
-; RV32-NEXT: vlse64.v v14, (a5), zero
; RV32-NEXT: vsetvli zero, a0, e64, m2, ta, ma
-; RV32-NEXT: vand.vv v16, v8, v14, v0.t
+; RV32-NEXT: vand.vx v14, v8, a4, v0.t
+; RV32-NEXT: vsll.vi v14, v14, 24, v0.t
+; RV32-NEXT: vand.vv v16, v8, v12, v0.t
; RV32-NEXT: vsll.vi v16, v16, 8, v0.t
-; RV32-NEXT: vor.vv v12, v12, v16, v0.t
-; RV32-NEXT: vor.vv v10, v10, v12, v0.t
-; RV32-NEXT: vsrl.vx v12, v8, a1, v0.t
+; RV32-NEXT: vor.vv v14, v14, v16, v0.t
+; RV32-NEXT: vor.vv v10, v10, v14, v0.t
+; RV32-NEXT: vsrl.vx v14, v8, a1, v0.t
; RV32-NEXT: vsrl.vx v16, v8, a3, v0.t
; RV32-NEXT: vand.vx v16, v16, a2, v0.t
-; RV32-NEXT: vor.vv v12, v16, v12, v0.t
+; RV32-NEXT: vor.vv v14, v16, v14, v0.t
; RV32-NEXT: vsrl.vi v16, v8, 24, v0.t
; RV32-NEXT: vand.vx v16, v16, a4, v0.t
; RV32-NEXT: vsrl.vi v8, v8, 8, v0.t
-; RV32-NEXT: vand.vv v8, v8, v14, v0.t
+; RV32-NEXT: vand.vv v8, v8, v12, v0.t
; RV32-NEXT: vor.vv v8, v8, v16, v0.t
-; RV32-NEXT: vor.vv v8, v8, v12, v0.t
+; RV32-NEXT: vor.vv v8, v8, v14, v0.t
; RV32-NEXT: vor.vv v8, v10, v8, v0.t
; RV32-NEXT: addi sp, sp, 16
; RV32-NEXT: ret
@@ -779,27 +779,27 @@ define <vscale x 2 x i64> @vp_bswap_nxv2i64_unmasked(<vscale x 2 x i64> %va, i32
; RV32-NEXT: li a3, 40
; RV32-NEXT: vsll.vx v12, v12, a3
; RV32-NEXT: vor.vv v10, v10, v12
+; RV32-NEXT: addi a4, sp, 8
+; RV32-NEXT: vsetvli a5, zero, e64, m2, ta, ma
+; RV32-NEXT: vlse64.v v12, (a4), zero
; RV32-NEXT: lui a4, 4080
-; RV32-NEXT: vand.vx v12, v8, a4
-; RV32-NEXT: vsll.vi v12, v12, 24
-; RV32-NEXT: addi a5, sp, 8
-; RV32-NEXT: vsetvli a6, zero, e64, m2, ta, ma
-; RV32-NEXT: vlse64.v v14, (a5), zero
; RV32-NEXT: vsetvli zero, a0, e64, m2, ta, ma
-; RV32-NEXT: vand.vv v16, v8, v14
+; RV32-NEXT: vand.vx v14, v8, a4
+; RV32-NEXT: vsll.vi v14, v14, 24
+; RV32-NEXT: vand.vv v16, v8, v12
; RV32-NEXT: vsll.vi v16, v16, 8
-; RV32-NEXT: vor.vv v12, v12, v16
-; RV32-NEXT: vor.vv v10, v10, v12
-; RV32-NEXT: vsrl.vx v12, v8, a1
+; RV32-NEXT: vor.vv v14, v14, v16
+; RV32-NEXT: vor.vv v10, v10, v14
+; RV32-NEXT: vsrl.vx v14, v8, a1
; RV32-NEXT: vsrl.vx v16, v8, a3
; RV32-NEXT: vand.vx v16, v16, a2
-; RV32-NEXT: vor.vv v12, v16, v12
+; RV32-NEXT: vor.vv v14, v16, v14
; RV32-NEXT: vsrl.vi v16, v8, 24
; RV32-NEXT: vand.vx v16, v16, a4
; RV32-NEXT: vsrl.vi v8, v8, 8
-; RV32-NEXT: vand.vv v8, v8, v14
+; RV32-NEXT: vand.vv v8, v8, v12
; RV32-NEXT: vor.vv v8, v8, v16
-; RV32-NEXT: vor.vv v8, v8, v12
+; RV32-NEXT: vor.vv v8, v8, v14
; RV32-NEXT: vor.vv v8, v10, v8
; RV32-NEXT: addi sp, sp, 16
; RV32-NEXT: ret
@@ -865,13 +865,13 @@ define <vscale x 4 x i64> @vp_bswap_nxv4i64(<vscale x 4 x i64> %va, <vscale x 4
; RV32-NEXT: li a3, 40
; RV32-NEXT: vsll.vx v16, v16, a3, v0.t
; RV32-NEXT: vor.vv v16, v12, v16, v0.t
+; RV32-NEXT: addi a4, sp, 8
+; RV32-NEXT: vsetvli a5, zero, e64, m4, ta, ma
+; RV32-NEXT: vlse64.v v12, (a4), zero
; RV32-NEXT: lui a4, 4080
-; RV32-NEXT: vand.vx v12, v8, a4, v0.t
-; RV32-NEXT: vsll.vi v20, v12, 24, v0.t
-; RV32-NEXT: addi a5, sp, 8
-; RV32-NEXT: vsetvli a6, zero, e64, m4, ta, ma
-; RV32-NEXT: vlse64.v v12, (a5), zero
; RV32-NEXT: vsetvli zero, a0, e64, m4, ta, ma
+; RV32-NEXT: vand.vx v20, v8, a4, v0.t
+; RV32-NEXT: vsll.vi v20, v20, 24, v0.t
; RV32-NEXT: vand.vv v24, v8, v12, v0.t
; RV32-NEXT: vsll.vi v24, v24, 8, v0.t
; RV32-NEXT: vor.vv v20, v20, v24, v0.t
@@ -949,27 +949,27 @@ define <vscale x 4 x i64> @vp_bswap_nxv4i64_unmasked(<vscale x 4 x i64> %va, i32
; RV32-NEXT: li a3, 40
; RV32-NEXT: vsll.vx v16, v16, a3
; RV32-NEXT: vor.vv v12, v12, v16
+; RV32-NEXT: addi a4, sp, 8
+; RV32-NEXT: vsetvli a5, zero, e64, m4, ta, ma
+; RV32-NEXT: vlse64.v v16, (a4), zero
; RV32-NEXT: lui a4, 4080
-; RV32-NEXT: vand.vx v16, v8, a4
-; RV32-NEXT: vsll.vi v16, v16, 24
-; RV32-NEXT: addi a5, sp, 8
-; RV32-NEXT: vsetvli a6, zero, e64, m4, ta, ma
-; RV32-NEXT: vlse64.v v20, (a5), zero
; RV32-NEXT: vsetvli zero, a0, e64, m4, ta, ma
-; RV32-NEXT: vand.vv v24, v8, v20
+; RV32-NEXT: vand.vx v20, v8, a4
+; RV32-NEXT: vsll.vi v20, v20, 24
+; RV32-NEXT: vand.vv v24, v8, v16
; RV32-NEXT: vsll.vi v24, v24, 8
-; RV32-NEXT: vor.vv v16, v16, v24
-; RV32-NEXT: vor.vv v12, v12, v16
-; RV32-NEXT: vsrl.vx v16, v8, a1
+; RV32-NEXT: vor.vv v20, v20, v24
+; RV32-NEXT: vor.vv v12, v12, v20
+; RV32-NEXT: vsrl.vx v20, v8, a1
; RV32-NEXT: vsrl.vx v24, v8, a3
; RV32-NEXT: vand.vx v24, v24, a2
-; RV32-NEXT: vor.vv v16, v24, v16
+; RV32-NEXT: vor.vv v20, v24, v20
; RV32-NEXT: vsrl.vi v24, v8, 24
; RV32-NEXT: vand.vx v24, v24, a4
; RV32-NEXT: vsrl.vi v8, v8, 8
-; RV32-NEXT: vand.vv v8, v8, v20
+; RV32-NEXT: vand.vv v8, v8, v16
; RV32-NEXT: vor.vv v8, v8, v24
-; RV32-NEXT: vor.vv v8, v8, v16
+; RV32-NEXT: vor.vv v8, v8, v20
; RV32-NEXT: vor.vv v8, v12, v8
; RV32-NEXT: addi sp, sp, 16
; RV32-NEXT: ret
@@ -1045,20 +1045,23 @@ define <vscale x 7 x i64> @vp_bswap_nxv7i64(<vscale x 7 x i64> %va, <vscale x 7
; RV32-NEXT: add a4, sp, a4
; RV32-NEXT: addi a4, a4, 16
; RV32-NEXT: vs8r.v v16, (a4) # Unknown-size Folded Spill
+; RV32-NEXT: addi a4, sp, 8
+; RV32-NEXT: vsetvli a5, zero, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v16, (a4), zero
+; RV32-NEXT: csrr a4, vlenb
+; RV32-NEXT: slli a4, a4, 3
+; RV32-NEXT: add a4, sp, a4
+; RV32-NEXT: addi a4, a4, 16
+; RV32-NEXT: vs8r.v v16, (a4) # Unknown-size Folded Spill
; RV32-NEXT: lui a4, 4080
-; RV32-NEXT: vand.vx v16, v8, a4, v0.t
-; RV32-NEXT: vsll.vi v24, v16, 24, v0.t
-; RV32-NEXT: addi a5, sp, 8
-; RV32-NEXT: vsetvli a6, zero, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a5), zero
-; RV32-NEXT: csrr a5, vlenb
-; RV32-NEXT: slli a5, a5, 3
-; RV32-NEXT: add a5, sp, a5
-; RV32-NEXT: addi a5, a5, 16
-; RV32-NEXT: vs8r.v v16, (a5) # Unknown-size Folded Spill
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v8, v16, v0.t
-; RV32-NEXT: vsll.vi v16, v16, 8, v0.t
+; RV32-NEXT: vand.vx v24, v8, a4, v0.t
+; RV32-NEXT: vsll.vi v24, v24, 24, v0.t
+; RV32-NEXT: addi a0, sp, 16
+; RV32-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
+; RV32-NEXT: vand.vv v24, v8, v16, v0.t
+; RV32-NEXT: vsll.vi v16, v24, 8, v0.t
+; RV32-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
; RV32-NEXT: vor.vv v16, v24, v16, v0.t
; RV32-NEXT: csrr a0, vlenb
; RV32-NEXT: slli a0, a0, 4
@@ -1071,10 +1074,10 @@ define <vscale x 7 x i64> @vp_bswap_nxv7i64(<vscale x 7 x i64> %va, <vscale x 7
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 16
; RV32-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
-; RV32-NEXT: vsrl.vx v24, v8, a1, v0.t
-; RV32-NEXT: vsrl.vx v16, v8, a3, v0.t
-; RV32-NEXT: vand.vx v16, v16, a2, v0.t
-; RV32-NEXT: vor.vv v16, v16, v24, v0.t
+; RV32-NEXT: vsrl.vx v16, v8, a1, v0.t
+; RV32-NEXT: vsrl.vx v24, v8, a3, v0.t
+; RV32-NEXT: vand.vx v24, v24, a2, v0.t
+; RV32-NEXT: vor.vv v16, v24, v16, v0.t
; RV32-NEXT: addi a0, sp, 16
; RV32-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
; RV32-NEXT: vsrl.vi v24, v8, 24, v0.t
@@ -1185,13 +1188,13 @@ define <vscale x 7 x i64> @vp_bswap_nxv7i64_unmasked(<vscale x 7 x i64> %va, i32
; RV32-NEXT: vor.vv v16, v16, v24
; RV32-NEXT: addi a4, sp, 16
; RV32-NEXT: vs8r.v v16, (a4) # Unknown-size Folded Spill
+; RV32-NEXT: addi a4, sp, 8
+; RV32-NEXT: vsetvli a5, zero, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v16, (a4), zero
; RV32-NEXT: lui a4, 4080
-; RV32-NEXT: vand.vx v16, v8, a4
-; RV32-NEXT: vsll.vi v0, v16, 24
-; RV32-NEXT: addi a5, sp, 8
-; RV32-NEXT: vsetvli a6, zero, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a5), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vand.vx v0, v8, a4
+; RV32-NEXT: vsll.vi v0, v0, 24
; RV32-NEXT: vand.vv v24, v8, v16
; RV32-NEXT: vsll.vi v24, v24, 8
; RV32-NEXT: vor.vv v24, v0, v24
@@ -1288,20 +1291,23 @@ define <vscale x 8 x i64> @vp_bswap_nxv8i64(<vscale x 8 x i64> %va, <vscale x 8
; RV32-NEXT: add a4, sp, a4
; RV32-NEXT: addi a4, a4, 16
; RV32-NEXT: vs8r.v v16, (a4) # Unknown-size Folded Spill
+; RV32-NEXT: addi a4, sp, 8
+; RV32-NEXT: vsetvli a5, zero, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v16, (a4), zero
+; RV32-NEXT: csrr a4, vlenb
+; RV32-NEXT: slli a4, a4, 3
+; RV32-NEXT: add a4, sp, a4
+; RV32-NEXT: addi a4, a4, 16
+; RV32-NEXT: vs8r.v v16, (a4) # Unknown-size Folded Spill
; RV32-NEXT: lui a4, 4080
-; RV32-NEXT: vand.vx v16, v8, a4, v0.t
-; RV32-NEXT: vsll.vi v24, v16, 24, v0.t
-; RV32-NEXT: addi a5, sp, 8
-; RV32-NEXT: vsetvli a6, zero, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a5), zero
-; RV32-NEXT: csrr a5, vlenb
-; RV32-NEXT: slli a5, a5, 3
-; RV32-NEXT: add a5, sp, a5
-; RV32-NEXT: addi a5, a5, 16
-; RV32-NEXT: vs8r.v v16, (a5) # Unknown-size Folded Spill
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v8, v16, v0.t
-; RV32-NEXT: vsll.vi v16, v16, 8, v0.t
+; RV32-NEXT: vand.vx v24, v8, a4, v0.t
+; RV32-NEXT: vsll.vi v24, v24, 24, v0.t
+; RV32-NEXT: addi a0, sp, 16
+; RV32-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
+; RV32-NEXT: vand.vv v24, v8, v16, v0.t
+; RV32-NEXT: vsll.vi v16, v24, 8, v0.t
+; RV32-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
; RV32-NEXT: vor.vv v16, v24, v16, v0.t
; RV32-NEXT: csrr a0, vlenb
; RV32-NEXT: slli a0, a0, 4
@@ -1314,10 +1320,10 @@ define <vscale x 8 x i64> @vp_bswap_nxv8i64(<vscale x 8 x i64> %va, <vscale x 8
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 16
; RV32-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
-; RV32-NEXT: vsrl.vx v24, v8, a1, v0.t
-; RV32-NEXT: vsrl.vx v16, v8, a3, v0.t
-; RV32-NEXT: vand.vx v16, v16, a2, v0.t
-; RV32-NEXT: vor.vv v16, v16, v24, v0.t
+; RV32-NEXT: vsrl.vx v16, v8, a1, v0.t
+; RV32-NEXT: vsrl.vx v24, v8, a3, v0.t
+; RV32-NEXT: vand.vx v24, v24, a2, v0.t
+; RV32-NEXT: vor.vv v16, v24, v16, v0.t
; RV32-NEXT: addi a0, sp, 16
; RV32-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
; RV32-NEXT: vsrl.vi v24, v8, 24, v0.t
@@ -1428,13 +1434,13 @@ define <vscale x 8 x i64> @vp_bswap_nxv8i64_unmasked(<vscale x 8 x i64> %va, i32
; RV32-NEXT: vor.vv v16, v16, v24
; RV32-NEXT: addi a4, sp, 16
; RV32-NEXT: vs8r.v v16, (a4) # Unknown-size Folded Spill
+; RV32-NEXT: addi a4, sp, 8
+; RV32-NEXT: vsetvli a5, zero, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v16, (a4), zero
; RV32-NEXT: lui a4, 4080
-; RV32-NEXT: vand.vx v16, v8, a4
-; RV32-NEXT: vsll.vi v0, v16, 24
-; RV32-NEXT: addi a5, sp, 8
-; RV32-NEXT: vsetvli a6, zero, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a5), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vand.vx v0, v8, a4
+; RV32-NEXT: vsll.vi v0, v0, 24
; RV32-NEXT: vand.vv v24, v8, v16
; RV32-NEXT: vsll.vi v24, v24, 8
; RV32-NEXT: vor.vv v24, v0, v24
@@ -1539,13 +1545,13 @@ define <vscale x 64 x i16> @vp_bswap_nxv64i16(<vscale x 64 x i16> %va, <vscale x
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB32_2:
-; CHECK-NEXT: vsetvli zero, a0, e16, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: csrr a1, vlenb
+; CHECK-NEXT: slli a1, a1, 3
+; CHECK-NEXT: add a1, sp, a1
+; CHECK-NEXT: addi a1, a1, 16
+; CHECK-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, a0, e16, m8, ta, ma
; CHECK-NEXT: vsrl.vi v16, v8, 8, v0.t
; CHECK-NEXT: vsll.vi v8, v8, 8, v0.t
; CHECK-NEXT: vor.vv v8, v8, v16, v0.t
@@ -1575,8 +1581,8 @@ define <vscale x 64 x i16> @vp_bswap_nxv64i16(<vscale x 64 x i16> %va, <vscale x
; CHECK-ZVKB-NEXT: # %bb.1:
; CHECK-ZVKB-NEXT: mv a0, a1
; CHECK-ZVKB-NEXT: .LBB32_2:
-; CHECK-ZVKB-NEXT: vsetvli zero, a0, e16, m8, ta, ma
; CHECK-ZVKB-NEXT: vmv1r.v v0, v24
+; CHECK-ZVKB-NEXT: vsetvli zero, a0, e16, m8, ta, ma
; CHECK-ZVKB-NEXT: vrev8.v v8, v8, v0.t
; CHECK-ZVKB-NEXT: ret
%v = call <vscale x 64 x i16> @llvm.vp.bswap.nxv64i16(<vscale x 64 x i16> %va, <vscale x 64 x i1> %m, i32 %evl)
@@ -1646,27 +1652,27 @@ define <vscale x 1 x i48> @vp_bswap_nxv1i48(<vscale x 1 x i48> %va, <vscale x 1
; RV32-NEXT: li a3, 40
; RV32-NEXT: vsll.vx v10, v10, a3, v0.t
; RV32-NEXT: vor.vv v9, v9, v10, v0.t
+; RV32-NEXT: addi a4, sp, 8
+; RV32-NEXT: vsetvli a5, zero, e64, m1, ta, ma
+; RV32-NEXT: vlse64.v v10, (a4), zero
; RV32-NEXT: lui a4, 4080
-; RV32-NEXT: vand.vx v10, v8, a4, v0.t
-; RV32-NEXT: vsll.vi v10, v10, 24, v0.t
-; RV32-NEXT: addi a5, sp, 8
-; RV32-NEXT: vsetvli a6, zero, e64, m1, ta, ma
-; RV32-NEXT: vlse64.v v11, (a5), zero
; RV32-NEXT: vsetvli zero, a0, e64, m1, ta, ma
-; RV32-NEXT: vand.vv v12, v8, v11, v0.t
+; RV32-NEXT: vand.vx v11, v8, a4, v0.t
+; RV32-NEXT: vsll.vi v11, v11, 24, v0.t
+; RV32-NEXT: vand.vv v12, v8, v10, v0.t
; RV32-NEXT: vsll.vi v12, v12, 8, v0.t
-; RV32-NEXT: vor.vv v10, v10, v12, v0.t
-; RV32-NEXT: vor.vv v9, v9, v10, v0.t
-; RV32-NEXT: vsrl.vx v10, v8, a1, v0.t
+; RV32-NEXT: vor.vv v11, v11, v12, v0.t
+; RV32-NEXT: vor.vv v9, v9, v11, v0.t
+; RV32-NEXT: vsrl.vx v11, v8, a1, v0.t
; RV32-NEXT: vsrl.vx v12, v8, a3, v0.t
; RV32-NEXT: vand.vx v12, v12, a2, v0.t
-; RV32-NEXT: vor.vv v10, v12, v10, v0.t
+; RV32-NEXT: vor.vv v11, v12, v11, v0.t
; RV32-NEXT: vsrl.vi v12, v8, 24, v0.t
; RV32-NEXT: vand.vx v12, v12, a4, v0.t
; RV32-NEXT: vsrl.vi v8, v8, 8, v0.t
-; RV32-NEXT: vand.vv v8, v8, v11, v0.t
+; RV32-NEXT: vand.vv v8, v8, v10, v0.t
; RV32-NEXT: vor.vv v8, v8, v12, v0.t
-; RV32-NEXT: vor.vv v8, v8, v10, v0.t
+; RV32-NEXT: vor.vv v8, v8, v11, v0.t
; RV32-NEXT: vor.vv v8, v9, v8, v0.t
; RV32-NEXT: vsrl.vi v8, v8, 16, v0.t
; RV32-NEXT: addi sp, sp, 16
diff --git a/llvm/test/CodeGen/RISCV/rvv/calling-conv-fastcc.ll b/llvm/test/CodeGen/RISCV/rvv/calling-conv-fastcc.ll
index 187f758b78020..2f0d5bb6e19c4 100644
--- a/llvm/test/CodeGen/RISCV/rvv/calling-conv-fastcc.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/calling-conv-fastcc.ll
@@ -244,29 +244,28 @@ define fastcc <vscale x 32 x i32> @ret_nxv32i32_param_nxv32i32_nxv32i32_nxv32i32
; CHECK-NEXT: add a1, sp, a1
; CHECK-NEXT: addi a1, a1, 16
; CHECK-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
-; CHECK-NEXT: vmv8r.v v24, v8
; CHECK-NEXT: csrr a1, vlenb
; CHECK-NEXT: slli a1, a1, 3
; CHECK-NEXT: add a3, a2, a1
-; CHECK-NEXT: vl8re32.v v8, (a3)
-; CHECK-NEXT: addi a3, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
; CHECK-NEXT: add a1, a0, a1
-; CHECK-NEXT: vl8re32.v v0, (a0)
-; CHECK-NEXT: vl8re32.v v8, (a1)
+; CHECK-NEXT: vl8re32.v v24, (a0)
+; CHECK-NEXT: vl8re32.v v0, (a1)
+; CHECK-NEXT: vl8re32.v v16, (a3)
+; CHECK-NEXT: addi a0, sp, 16
+; CHECK-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: vl8re32.v v16, (a2)
; CHECK-NEXT: vsetvli a0, zero, e32, m8, ta, ma
-; CHECK-NEXT: vadd.vv v0, v24, v0
+; CHECK-NEXT: vadd.vv v24, v8, v24
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vadd.vv v8, v24, v8
+; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vadd.vv v0, v8, v0
; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vadd.vv v8, v8, v24
-; CHECK-NEXT: vadd.vv v24, v0, v16
+; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vadd.vv v8, v0, v8
+; CHECK-NEXT: vadd.vv v24, v24, v16
; CHECK-NEXT: vadd.vx v16, v8, a4
; CHECK-NEXT: vadd.vx v8, v24, a4
; CHECK-NEXT: csrr a0, vlenb
diff --git a/llvm/test/CodeGen/RISCV/rvv/calling-conv.ll b/llvm/test/CodeGen/RISCV/rvv/calling-conv.ll
index 647d3158b6167..fa62143546df6 100644
--- a/llvm/test/CodeGen/RISCV/rvv/calling-conv.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/calling-conv.ll
@@ -39,11 +39,11 @@ define <vscale x 32 x i32> @caller_scalable_vector_split_indirect(<vscale x 32 x
; RV32-NEXT: vs8r.v v8, (a0)
; RV32-NEXT: csrr a1, vlenb
; RV32-NEXT: slli a1, a1, 3
-; RV32-NEXT: add a0, a0, a1
-; RV32-NEXT: vs8r.v v16, (a0)
+; RV32-NEXT: add a1, a0, a1
; RV32-NEXT: vsetvli a0, zero, e32, m8, ta, ma
; RV32-NEXT: vmv.v.i v8, 0
; RV32-NEXT: addi a0, sp, 128
+; RV32-NEXT: vs8r.v v16, (a1)
; RV32-NEXT: vmv.v.i v16, 0
; RV32-NEXT: call callee_scalable_vector_split_indirect
; RV32-NEXT: addi sp, s0, -144
@@ -70,11 +70,11 @@ define <vscale x 32 x i32> @caller_scalable_vector_split_indirect(<vscale x 32 x
; RV64-NEXT: vs8r.v v8, (a0)
; RV64-NEXT: csrr a1, vlenb
; RV64-NEXT: slli a1, a1, 3
-; RV64-NEXT: add a0, a0, a1
-; RV64-NEXT: vs8r.v v16, (a0)
+; RV64-NEXT: add a1, a0, a1
; RV64-NEXT: vsetvli a0, zero, e32, m8, ta, ma
; RV64-NEXT: vmv.v.i v8, 0
; RV64-NEXT: addi a0, sp, 128
+; RV64-NEXT: vs8r.v v16, (a1)
; RV64-NEXT: vmv.v.i v16, 0
; RV64-NEXT: call callee_scalable_vector_split_indirect
; RV64-NEXT: addi sp, s0, -144
diff --git a/llvm/test/CodeGen/RISCV/rvv/ceil-vp.ll b/llvm/test/CodeGen/RISCV/rvv/ceil-vp.ll
index aa11e012af201..dec67721514de 100644
--- a/llvm/test/CodeGen/RISCV/rvv/ceil-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/ceil-vp.ll
@@ -135,16 +135,16 @@ declare <vscale x 8 x half> @llvm.vp.ceil.nxv8f16(<vscale x 8 x half>, <vscale x
define <vscale x 8 x half> @vp_ceil_vv_nxv8f16(<vscale x 8 x half> %va, <vscale x 8 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_ceil_vv_nxv8f16:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v10, v0
; CHECK-NEXT: lui a1, %hi(.LCPI6_0)
; CHECK-NEXT: flh fa5, %lo(.LCPI6_0)(a1)
+; CHECK-NEXT: vmv1r.v v10, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, ma
; CHECK-NEXT: vfabs.v v12, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, mu
; CHECK-NEXT: vmflt.vf v10, v12, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 3
-; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
@@ -179,16 +179,16 @@ declare <vscale x 16 x half> @llvm.vp.ceil.nxv16f16(<vscale x 16 x half>, <vscal
define <vscale x 16 x half> @vp_ceil_vv_nxv16f16(<vscale x 16 x half> %va, <vscale x 16 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_ceil_vv_nxv16f16:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v12, v0
; CHECK-NEXT: lui a1, %hi(.LCPI8_0)
; CHECK-NEXT: flh fa5, %lo(.LCPI8_0)(a1)
+; CHECK-NEXT: vmv1r.v v12, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, ma
; CHECK-NEXT: vfabs.v v16, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e16, m4, ta, mu
; CHECK-NEXT: vmflt.vf v12, v16, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 3
-; CHECK-NEXT: vsetvli zero, zero, e16, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, zero, e16, m4, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
@@ -223,16 +223,16 @@ declare <vscale x 32 x half> @llvm.vp.ceil.nxv32f16(<vscale x 32 x half>, <vscal
define <vscale x 32 x half> @vp_ceil_vv_nxv32f16(<vscale x 32 x half> %va, <vscale x 32 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_ceil_vv_nxv32f16:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: lui a1, %hi(.LCPI10_0)
; CHECK-NEXT: flh fa5, %lo(.LCPI10_0)(a1)
+; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m8, ta, ma
; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e16, m8, ta, mu
; CHECK-NEXT: vmflt.vf v16, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 3
-; CHECK-NEXT: vsetvli zero, zero, e16, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vsetvli zero, zero, e16, m8, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
@@ -359,8 +359,8 @@ define <vscale x 4 x float> @vp_ceil_vv_nxv4f32(<vscale x 4 x float> %va, <vscal
; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, mu
; CHECK-NEXT: vmflt.vf v10, v12, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 3
-; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
@@ -403,8 +403,8 @@ define <vscale x 8 x float> @vp_ceil_vv_nxv8f32(<vscale x 8 x float> %va, <vscal
; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, mu
; CHECK-NEXT: vmflt.vf v12, v16, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 3
-; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
@@ -447,8 +447,8 @@ define <vscale x 16 x float> @vp_ceil_vv_nxv16f32(<vscale x 16 x float> %va, <vs
; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, mu
; CHECK-NEXT: vmflt.vf v16, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 3
-; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
@@ -525,16 +525,16 @@ declare <vscale x 2 x double> @llvm.vp.ceil.nxv2f64(<vscale x 2 x double>, <vsca
define <vscale x 2 x double> @vp_ceil_vv_nxv2f64(<vscale x 2 x double> %va, <vscale x 2 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_ceil_vv_nxv2f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v10, v0
; CHECK-NEXT: lui a1, %hi(.LCPI24_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI24_0)(a1)
+; CHECK-NEXT: vmv1r.v v10, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, ma
; CHECK-NEXT: vfabs.v v12, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, mu
; CHECK-NEXT: vmflt.vf v10, v12, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 3
-; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
@@ -569,16 +569,16 @@ declare <vscale x 4 x double> @llvm.vp.ceil.nxv4f64(<vscale x 4 x double>, <vsca
define <vscale x 4 x double> @vp_ceil_vv_nxv4f64(<vscale x 4 x double> %va, <vscale x 4 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_ceil_vv_nxv4f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v12, v0
; CHECK-NEXT: lui a1, %hi(.LCPI26_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI26_0)(a1)
+; CHECK-NEXT: vmv1r.v v12, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, ma
; CHECK-NEXT: vfabs.v v16, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, mu
; CHECK-NEXT: vmflt.vf v12, v16, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 3
-; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
@@ -613,16 +613,16 @@ declare <vscale x 7 x double> @llvm.vp.ceil.nxv7f64(<vscale x 7 x double>, <vsca
define <vscale x 7 x double> @vp_ceil_vv_nxv7f64(<vscale x 7 x double> %va, <vscale x 7 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_ceil_vv_nxv7f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: lui a1, %hi(.LCPI28_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI28_0)(a1)
+; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
; CHECK-NEXT: vmflt.vf v16, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 3
-; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
@@ -657,16 +657,16 @@ declare <vscale x 8 x double> @llvm.vp.ceil.nxv8f64(<vscale x 8 x double>, <vsca
define <vscale x 8 x double> @vp_ceil_vv_nxv8f64(<vscale x 8 x double> %va, <vscale x 8 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_ceil_vv_nxv8f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: lui a1, %hi(.LCPI30_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI30_0)(a1)
+; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
; CHECK-NEXT: vmflt.vf v16, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 3
-; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
@@ -705,66 +705,56 @@ define <vscale x 16 x double> @vp_ceil_vv_nxv16f64(<vscale x 16 x double> %va, <
; CHECK-NEXT: addi sp, sp, -16
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a1, a1, 4
+; CHECK-NEXT: slli a1, a1, 3
; CHECK-NEXT: sub sp, sp, a1
-; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
-; CHECK-NEXT: vmv1r.v v24, v0
-; CHECK-NEXT: addi a1, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
+; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; CHECK-NEXT: vmv1r.v v7, v0
; CHECK-NEXT: csrr a1, vlenb
; CHECK-NEXT: srli a2, a1, 3
; CHECK-NEXT: vsetvli a3, zero, e8, mf4, ta, ma
-; CHECK-NEXT: vslidedown.vx v25, v0, a2
+; CHECK-NEXT: vslidedown.vx v6, v0, a2
; CHECK-NEXT: sub a2, a0, a1
+; CHECK-NEXT: lui a3, %hi(.LCPI32_0)
+; CHECK-NEXT: fld fa5, %lo(.LCPI32_0)(a3)
; CHECK-NEXT: sltu a3, a0, a2
; CHECK-NEXT: addi a3, a3, -1
; CHECK-NEXT: and a2, a3, a2
-; CHECK-NEXT: lui a3, %hi(.LCPI32_0)
-; CHECK-NEXT: fld fa5, %lo(.LCPI32_0)(a3)
+; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: vsetvli zero, a2, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v25
-; CHECK-NEXT: vfabs.v v8, v16, v0.t
+; CHECK-NEXT: vfabs.v v24, v16, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vmflt.vf v25, v8, fa5, v0.t
+; CHECK-NEXT: vmflt.vf v6, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a2, 3
+; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v25
-; CHECK-NEXT: vfcvt.x.f.v v8, v16, v0.t
+; CHECK-NEXT: vfcvt.x.f.v v24, v16, v0.t
+; CHECK-NEXT: addi a3, sp, 16
+; CHECK-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
; CHECK-NEXT: fsrm a2
-; CHECK-NEXT: vfcvt.f.x.v v8, v8, v0.t
+; CHECK-NEXT: addi a2, sp, 16
+; CHECK-NEXT: vl8r.v v24, (a2) # Unknown-size Folded Reload
+; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vfsgnj.vv v16, v8, v16, v0.t
-; CHECK-NEXT: csrr a2, vlenb
-; CHECK-NEXT: slli a2, a2, 3
-; CHECK-NEXT: add a2, sp, a2
-; CHECK-NEXT: addi a2, a2, 16
-; CHECK-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
+; CHECK-NEXT: vfsgnj.vv v16, v24, v16, v0.t
; CHECK-NEXT: bltu a0, a1, .LBB32_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB32_2:
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v24
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vfabs.v v16, v8, v0.t
+; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vmflt.vf v24, v16, fa5, v0.t
+; CHECK-NEXT: vmflt.vf v7, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 3
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v24
-; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
+; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: fsrm a0
-; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
+; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vfsgnj.vv v8, v16, v8, v0.t
+; CHECK-NEXT: vfsgnj.vv v8, v24, v8, v0.t
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 4
; CHECK-NEXT: add sp, sp, a0
; CHECK-NEXT: addi sp, sp, 16
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/combine-store-extract-crash.ll b/llvm/test/CodeGen/RISCV/rvv/combine-store-extract-crash.ll
index ed434deea1a83..482cf83d540c4 100644
--- a/llvm/test/CodeGen/RISCV/rvv/combine-store-extract-crash.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/combine-store-extract-crash.ll
@@ -10,19 +10,19 @@ define void @test(ptr %ref_array, ptr %sad_array) {
; RV32-NEXT: th.lwd a2, a3, (a0), 0, 3
; RV32-NEXT: vsetivli zero, 4, e8, mf4, ta, ma
; RV32-NEXT: vle8.v v8, (a2)
-; RV32-NEXT: vmv.v.i v9, 0
; RV32-NEXT: vsetivli zero, 16, e32, m4, ta, ma
; RV32-NEXT: vzext.vf4 v12, v8
; RV32-NEXT: vmv.s.x v8, zero
-; RV32-NEXT: vredsum.vs v10, v12, v8
-; RV32-NEXT: vmv.x.s a0, v10
+; RV32-NEXT: vredsum.vs v9, v12, v8
+; RV32-NEXT: vmv.x.s a0, v9
; RV32-NEXT: th.swia a0, (a1), 4, 0
; RV32-NEXT: vsetivli zero, 4, e8, mf4, ta, ma
-; RV32-NEXT: vle8.v v10, (a3)
+; RV32-NEXT: vle8.v v9, (a3)
+; RV32-NEXT: vmv.v.i v10, 0
; RV32-NEXT: vsetivli zero, 8, e8, mf2, ta, ma
-; RV32-NEXT: vslideup.vi v10, v9, 4
+; RV32-NEXT: vslideup.vi v9, v10, 4
; RV32-NEXT: vsetivli zero, 16, e32, m4, ta, ma
-; RV32-NEXT: vzext.vf4 v12, v10
+; RV32-NEXT: vzext.vf4 v12, v9
; RV32-NEXT: vredsum.vs v8, v12, v8
; RV32-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32-NEXT: vse32.v v8, (a1)
@@ -33,19 +33,19 @@ define void @test(ptr %ref_array, ptr %sad_array) {
; RV64-NEXT: th.ldd a2, a3, (a0), 0, 4
; RV64-NEXT: vsetivli zero, 4, e8, mf4, ta, ma
; RV64-NEXT: vle8.v v8, (a2)
-; RV64-NEXT: vmv.v.i v9, 0
; RV64-NEXT: vsetivli zero, 16, e32, m4, ta, ma
; RV64-NEXT: vzext.vf4 v12, v8
; RV64-NEXT: vmv.s.x v8, zero
-; RV64-NEXT: vredsum.vs v10, v12, v8
-; RV64-NEXT: vmv.x.s a0, v10
+; RV64-NEXT: vredsum.vs v9, v12, v8
+; RV64-NEXT: vmv.x.s a0, v9
; RV64-NEXT: th.swia a0, (a1), 4, 0
; RV64-NEXT: vsetivli zero, 4, e8, mf4, ta, ma
-; RV64-NEXT: vle8.v v10, (a3)
+; RV64-NEXT: vle8.v v9, (a3)
+; RV64-NEXT: vmv.v.i v10, 0
; RV64-NEXT: vsetivli zero, 8, e8, mf2, ta, ma
-; RV64-NEXT: vslideup.vi v10, v9, 4
+; RV64-NEXT: vslideup.vi v9, v10, 4
; RV64-NEXT: vsetivli zero, 16, e32, m4, ta, ma
-; RV64-NEXT: vzext.vf4 v12, v10
+; RV64-NEXT: vzext.vf4 v12, v9
; RV64-NEXT: vredsum.vs v8, v12, v8
; RV64-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV64-NEXT: vse32.v v8, (a1)
diff --git a/llvm/test/CodeGen/RISCV/rvv/compressstore.ll b/llvm/test/CodeGen/RISCV/rvv/compressstore.ll
index 673008d9c0b3d..52811133c53f3 100644
--- a/llvm/test/CodeGen/RISCV/rvv/compressstore.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/compressstore.ll
@@ -197,28 +197,51 @@ entry:
define void @test_compresstore_v256i8(ptr %p, <256 x i1> %mask, <256 x i8> %data) {
; RV64-LABEL: test_compresstore_v256i8:
; RV64: # %bb.0: # %entry
-; RV64-NEXT: vmv1r.v v7, v8
+; RV64-NEXT: addi sp, sp, -16
+; RV64-NEXT: .cfi_def_cfa_offset 16
+; RV64-NEXT: csrr a2, vlenb
+; RV64-NEXT: slli a2, a2, 4
+; RV64-NEXT: sub sp, sp, a2
+; RV64-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
+; RV64-NEXT: csrr a2, vlenb
+; RV64-NEXT: slli a2, a2, 3
+; RV64-NEXT: add a2, sp, a2
+; RV64-NEXT: addi a2, a2, 16
+; RV64-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
; RV64-NEXT: li a2, 128
; RV64-NEXT: vsetvli zero, a2, e8, m8, ta, ma
-; RV64-NEXT: vle8.v v24, (a1)
+; RV64-NEXT: vle8.v v16, (a1)
+; RV64-NEXT: addi a1, sp, 16
+; RV64-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
; RV64-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; RV64-NEXT: vslidedown.vi v9, v0, 1
; RV64-NEXT: vmv.x.s a1, v9
; RV64-NEXT: vmv.x.s a3, v0
+; RV64-NEXT: csrr a4, vlenb
+; RV64-NEXT: slli a4, a4, 3
+; RV64-NEXT: add a4, sp, a4
+; RV64-NEXT: addi a4, a4, 16
+; RV64-NEXT: vl8r.v v24, (a4) # Unknown-size Folded Reload
; RV64-NEXT: vsetvli zero, a2, e8, m8, ta, ma
-; RV64-NEXT: vcompress.vm v8, v16, v0
+; RV64-NEXT: vcompress.vm v16, v24, v0
; RV64-NEXT: vcpop.m a4, v0
; RV64-NEXT: vsetvli zero, a4, e8, m8, ta, ma
-; RV64-NEXT: vse8.v v8, (a0)
+; RV64-NEXT: vse8.v v16, (a0)
+; RV64-NEXT: addi a4, sp, 16
+; RV64-NEXT: vl8r.v v24, (a4) # Unknown-size Folded Reload
; RV64-NEXT: vsetvli zero, a2, e8, m8, ta, ma
-; RV64-NEXT: vcompress.vm v8, v24, v7
-; RV64-NEXT: vcpop.m a2, v7
+; RV64-NEXT: vcompress.vm v16, v24, v8
+; RV64-NEXT: vcpop.m a2, v8
; RV64-NEXT: cpop a3, a3
; RV64-NEXT: cpop a1, a1
; RV64-NEXT: add a0, a0, a3
; RV64-NEXT: add a0, a0, a1
; RV64-NEXT: vsetvli zero, a2, e8, m8, ta, ma
-; RV64-NEXT: vse8.v v8, (a0)
+; RV64-NEXT: vse8.v v16, (a0)
+; RV64-NEXT: csrr a0, vlenb
+; RV64-NEXT: slli a0, a0, 4
+; RV64-NEXT: add sp, sp, a0
+; RV64-NEXT: addi sp, sp, 16
; RV64-NEXT: ret
;
; RV32-LABEL: test_compresstore_v256i8:
@@ -796,18 +819,18 @@ define void @test_compresstore_v32i64(ptr %p, <32 x i1> %mask, <32 x i64> %data)
; RV64-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; RV64-NEXT: vse64.v v24, (a0)
; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
-; RV64-NEXT: vslidedown.vi v24, v0, 2
+; RV64-NEXT: vslidedown.vi v8, v0, 2
; RV64-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV64-NEXT: vcompress.vm v8, v16, v24
+; RV64-NEXT: vcompress.vm v24, v16, v8
; RV64-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; RV64-NEXT: vmv.x.s a1, v0
; RV64-NEXT: zext.h a1, a1
; RV64-NEXT: cpopw a1, a1
; RV64-NEXT: slli a1, a1, 3
; RV64-NEXT: add a0, a0, a1
-; RV64-NEXT: vcpop.m a1, v24
+; RV64-NEXT: vcpop.m a1, v8
; RV64-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; RV64-NEXT: vse64.v v8, (a0)
+; RV64-NEXT: vse64.v v24, (a0)
; RV64-NEXT: ret
;
; RV32-LABEL: test_compresstore_v32i64:
@@ -818,18 +841,18 @@ define void @test_compresstore_v32i64(ptr %p, <32 x i1> %mask, <32 x i64> %data)
; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; RV32-NEXT: vse64.v v24, (a0)
; RV32-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
-; RV32-NEXT: vslidedown.vi v24, v0, 2
+; RV32-NEXT: vslidedown.vi v8, v0, 2
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vcompress.vm v8, v16, v24
+; RV32-NEXT: vcompress.vm v24, v16, v8
; RV32-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; RV32-NEXT: vmv.x.s a1, v0
; RV32-NEXT: zext.h a1, a1
; RV32-NEXT: cpop a1, a1
; RV32-NEXT: slli a1, a1, 3
; RV32-NEXT: add a0, a0, a1
-; RV32-NEXT: vcpop.m a1, v24
+; RV32-NEXT: vcpop.m a1, v8
; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; RV32-NEXT: vse64.v v8, (a0)
+; RV32-NEXT: vse64.v v24, (a0)
; RV32-NEXT: ret
entry:
tail call void @llvm.masked.compressstore.v32i64(<32 x i64> %data, ptr align 8 %p, <32 x i1> %mask)
diff --git a/llvm/test/CodeGen/RISCV/rvv/concat-vector-insert-elt.ll b/llvm/test/CodeGen/RISCV/rvv/concat-vector-insert-elt.ll
index bd65ed52be680..1343b64b876dc 100644
--- a/llvm/test/CodeGen/RISCV/rvv/concat-vector-insert-elt.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/concat-vector-insert-elt.ll
@@ -189,16 +189,16 @@ define void @v4xi64_concat_vector_insert_idx3(ptr %a, ptr %b, i64 %x) {
; RV32-LABEL: v4xi64_concat_vector_insert_idx3:
; RV32: # %bb.0:
; RV32-NEXT: vsetivli zero, 2, e64, m1, ta, ma
-; RV32-NEXT: vle64.v v8, (a0)
-; RV32-NEXT: vle64.v v10, (a1)
+; RV32-NEXT: vle64.v v8, (a1)
+; RV32-NEXT: vle64.v v10, (a0)
; RV32-NEXT: vsetivli zero, 2, e32, m1, ta, ma
; RV32-NEXT: vslide1down.vx v9, v8, a2
; RV32-NEXT: vslide1down.vx v9, v9, a3
; RV32-NEXT: vsetivli zero, 2, e64, m1, ta, ma
-; RV32-NEXT: vslideup.vi v10, v9, 1
+; RV32-NEXT: vslideup.vi v8, v9, 1
; RV32-NEXT: vsetivli zero, 4, e64, m2, ta, ma
-; RV32-NEXT: vslideup.vi v8, v10, 2
-; RV32-NEXT: vse64.v v8, (a0)
+; RV32-NEXT: vslideup.vi v10, v8, 2
+; RV32-NEXT: vse64.v v10, (a0)
; RV32-NEXT: ret
;
; RV64-LABEL: v4xi64_concat_vector_insert_idx3:
diff --git a/llvm/test/CodeGen/RISCV/rvv/constant-folding-crash.ll b/llvm/test/CodeGen/RISCV/rvv/constant-folding-crash.ll
index 113154c0f9855..7839b602706db 100644
--- a/llvm/test/CodeGen/RISCV/rvv/constant-folding-crash.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/constant-folding-crash.ll
@@ -19,18 +19,19 @@ define void @constant_folding_crash(ptr %v54, <4 x ptr> %lanes.a, <4 x ptr> %lan
; RV32-LABEL: constant_folding_crash:
; RV32: # %bb.0: # %entry
; RV32-NEXT: lw a0, 8(a0)
-; RV32-NEXT: vmv1r.v v10, v0
; RV32-NEXT: andi a0, a0, 1
; RV32-NEXT: seqz a0, a0
; RV32-NEXT: vsetivli zero, 4, e8, mf4, ta, ma
-; RV32-NEXT: vmv.v.x v11, a0
-; RV32-NEXT: vmsne.vi v0, v11, 0
+; RV32-NEXT: vmv.v.x v10, a0
+; RV32-NEXT: vmsne.vi v10, v10, 0
+; RV32-NEXT: vmv1r.v v11, v0
+; RV32-NEXT: vmv1r.v v0, v10
; RV32-NEXT: vsetvli zero, zero, e32, m1, ta, ma
; RV32-NEXT: vmerge.vvm v8, v9, v8, v0
; RV32-NEXT: vmv.x.s a0, v8
; RV32-NEXT: vsetvli zero, zero, e8, mf4, ta, ma
; RV32-NEXT: vmv.v.i v8, 0
-; RV32-NEXT: vmv1r.v v0, v10
+; RV32-NEXT: vmv1r.v v0, v11
; RV32-NEXT: vmerge.vim v8, v8, 1, v0
; RV32-NEXT: vrgather.vi v9, v8, 0
; RV32-NEXT: vmsne.vi v0, v9, 0
@@ -42,18 +43,19 @@ define void @constant_folding_crash(ptr %v54, <4 x ptr> %lanes.a, <4 x ptr> %lan
; RV64-LABEL: constant_folding_crash:
; RV64: # %bb.0: # %entry
; RV64-NEXT: ld a0, 8(a0)
-; RV64-NEXT: vmv1r.v v12, v0
; RV64-NEXT: andi a0, a0, 1
; RV64-NEXT: seqz a0, a0
; RV64-NEXT: vsetivli zero, 4, e8, mf4, ta, ma
-; RV64-NEXT: vmv.v.x v13, a0
-; RV64-NEXT: vmsne.vi v0, v13, 0
+; RV64-NEXT: vmv.v.x v12, a0
+; RV64-NEXT: vmsne.vi v12, v12, 0
+; RV64-NEXT: vmv1r.v v13, v0
+; RV64-NEXT: vmv1r.v v0, v12
; RV64-NEXT: vsetvli zero, zero, e64, m2, ta, ma
; RV64-NEXT: vmerge.vvm v8, v10, v8, v0
; RV64-NEXT: vmv.x.s a0, v8
; RV64-NEXT: vsetvli zero, zero, e8, mf4, ta, ma
; RV64-NEXT: vmv.v.i v8, 0
-; RV64-NEXT: vmv1r.v v0, v12
+; RV64-NEXT: vmv1r.v v0, v13
; RV64-NEXT: vmerge.vim v8, v8, 1, v0
; RV64-NEXT: vrgather.vi v9, v8, 0
; RV64-NEXT: vmsne.vi v0, v9, 0
diff --git a/llvm/test/CodeGen/RISCV/rvv/ctlz-sdnode.ll b/llvm/test/CodeGen/RISCV/rvv/ctlz-sdnode.ll
index 41ec102c34efb..6e538f3dfb38e 100644
--- a/llvm/test/CodeGen/RISCV/rvv/ctlz-sdnode.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/ctlz-sdnode.ll
@@ -3341,16 +3341,16 @@ define <vscale x 8 x i64> @ctlz_zero_undef_nxv8i64(<vscale x 8 x i64> %va) {
;
; RV32F-LABEL: ctlz_zero_undef_nxv8i64:
; RV32F: # %bb.0:
-; RV32F-NEXT: vmv8r.v v16, v8
; RV32F-NEXT: li a0, 190
; RV32F-NEXT: vsetvli a1, zero, e64, m8, ta, ma
-; RV32F-NEXT: vmv.v.x v8, a0
+; RV32F-NEXT: vmv.v.x v16, a0
; RV32F-NEXT: fsrmi a0, 1
; RV32F-NEXT: vsetvli zero, zero, e32, m4, ta, ma
-; RV32F-NEXT: vfncvt.f.xu.w v24, v16
-; RV32F-NEXT: vsrl.vi v16, v24, 23
-; RV32F-NEXT: vwsubu.wv v8, v8, v16
+; RV32F-NEXT: vfncvt.f.xu.w v24, v8
+; RV32F-NEXT: vsrl.vi v8, v24, 23
+; RV32F-NEXT: vwsubu.wv v16, v16, v8
; RV32F-NEXT: fsrm a0
+; RV32F-NEXT: vmv8r.v v8, v16
; RV32F-NEXT: ret
;
; RV64F-LABEL: ctlz_zero_undef_nxv8i64:
diff --git a/llvm/test/CodeGen/RISCV/rvv/ctlz-vp.ll b/llvm/test/CodeGen/RISCV/rvv/ctlz-vp.ll
index 86086f5dc88f7..fff280c005b54 100644
--- a/llvm/test/CodeGen/RISCV/rvv/ctlz-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/ctlz-vp.ll
@@ -1259,8 +1259,8 @@ define <vscale x 16 x i64> @vp_ctlz_nxv16i64(<vscale x 16 x i64> %va, <vscale x
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB46_2:
; CHECK-NEXT: fsrmi a1, 1
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfcvt.f.xu.v v8, v8, v0.t
; CHECK-NEXT: vsrl.vx v8, v8, a2, v0.t
; CHECK-NEXT: vrsub.vx v8, v8, a3, v0.t
@@ -1285,8 +1285,8 @@ define <vscale x 16 x i64> @vp_ctlz_nxv16i64(<vscale x 16 x i64> %va, <vscale x
; CHECK-ZVBB-NEXT: # %bb.1:
; CHECK-ZVBB-NEXT: mv a0, a1
; CHECK-ZVBB-NEXT: .LBB46_2:
-; CHECK-ZVBB-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-ZVBB-NEXT: vmv1r.v v0, v24
+; CHECK-ZVBB-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-ZVBB-NEXT: vclz.v v8, v8, v0.t
; CHECK-ZVBB-NEXT: ret
%v = call <vscale x 16 x i64> @llvm.vp.ctlz.nxv16i64(<vscale x 16 x i64> %va, i1 false, <vscale x 16 x i1> %m, i32 %evl)
@@ -2487,8 +2487,8 @@ define <vscale x 16 x i64> @vp_ctlz_zero_undef_nxv16i64(<vscale x 16 x i64> %va,
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB94_2:
; CHECK-NEXT: fsrmi a1, 1
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfcvt.f.xu.v v8, v8, v0.t
; CHECK-NEXT: vsrl.vx v8, v8, a2, v0.t
; CHECK-NEXT: vrsub.vx v8, v8, a3, v0.t
@@ -2512,8 +2512,8 @@ define <vscale x 16 x i64> @vp_ctlz_zero_undef_nxv16i64(<vscale x 16 x i64> %va,
; CHECK-ZVBB-NEXT: # %bb.1:
; CHECK-ZVBB-NEXT: mv a0, a1
; CHECK-ZVBB-NEXT: .LBB94_2:
-; CHECK-ZVBB-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-ZVBB-NEXT: vmv1r.v v0, v24
+; CHECK-ZVBB-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-ZVBB-NEXT: vclz.v v8, v8, v0.t
; CHECK-ZVBB-NEXT: ret
%v = call <vscale x 16 x i64> @llvm.vp.ctlz.nxv16i64(<vscale x 16 x i64> %va, i1 true, <vscale x 16 x i1> %m, i32 %evl)
diff --git a/llvm/test/CodeGen/RISCV/rvv/ctpop-vp.ll b/llvm/test/CodeGen/RISCV/rvv/ctpop-vp.ll
index 883f68aec1f42..e3c53212e91b7 100644
--- a/llvm/test/CodeGen/RISCV/rvv/ctpop-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/ctpop-vp.ll
@@ -2024,8 +2024,7 @@ define <vscale x 16 x i64> @vp_ctpop_nxv16i64(<vscale x 16 x i64> %va, <vscale x
; RV32-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x38, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 56 * vlenb
; RV32-NEXT: vmv1r.v v24, v0
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a2, 40
-; RV32-NEXT: mul a1, a1, a2
+; RV32-NEXT: slli a1, a1, 5
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
; RV32-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
@@ -2044,30 +2043,35 @@ define <vscale x 16 x i64> @vp_ctpop_nxv16i64(<vscale x 16 x i64> %va, <vscale x
; RV32-NEXT: addi a3, a3, -1
; RV32-NEXT: and a2, a3, a2
; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vi v8, v16, 1, v0.t
; RV32-NEXT: csrr a3, vlenb
; RV32-NEXT: li a4, 40
; RV32-NEXT: mul a3, a3, a4
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 16
-; RV32-NEXT: vl8r.v v8, (a3) # Unknown-size Folded Reload
-; RV32-NEXT: vsrl.vi v16, v8, 1, v0.t
+; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
; RV32-NEXT: lui a3, 349525
; RV32-NEXT: addi a3, a3, 1365
; RV32-NEXT: vsetvli a4, zero, e32, m8, ta, ma
-; RV32-NEXT: vmv.v.x v8, a3
-; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
+; RV32-NEXT: vmv.v.x v16, a3
; RV32-NEXT: csrr a3, vlenb
; RV32-NEXT: li a4, 24
; RV32-NEXT: mul a3, a3, a4
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 16
-; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: vand.vv v8, v16, v8, v0.t
+; RV32-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
; RV32-NEXT: csrr a3, vlenb
; RV32-NEXT: li a4, 40
; RV32-NEXT: mul a3, a3, a4
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 16
+; RV32-NEXT: vl8r.v v8, (a3) # Unknown-size Folded Reload
+; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
+; RV32-NEXT: vand.vv v8, v8, v16, v0.t
+; RV32-NEXT: csrr a3, vlenb
+; RV32-NEXT: slli a3, a3, 5
+; RV32-NEXT: add a3, sp, a3
+; RV32-NEXT: addi a3, a3, 16
; RV32-NEXT: vl8r.v v16, (a3) # Unknown-size Folded Reload
; RV32-NEXT: vsub.vv v8, v16, v8, v0.t
; RV32-NEXT: csrr a3, vlenb
@@ -2078,64 +2082,64 @@ define <vscale x 16 x i64> @vp_ctpop_nxv16i64(<vscale x 16 x i64> %va, <vscale x
; RV32-NEXT: lui a3, 209715
; RV32-NEXT: addi a3, a3, 819
; RV32-NEXT: vsetvli a4, zero, e32, m8, ta, ma
-; RV32-NEXT: vmv.v.x v8, a3
-; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
+; RV32-NEXT: vmv.v.x v16, a3
; RV32-NEXT: csrr a3, vlenb
; RV32-NEXT: slli a3, a3, 5
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 16
-; RV32-NEXT: vl8r.v v16, (a3) # Unknown-size Folded Reload
-; RV32-NEXT: vand.vv v16, v16, v8, v0.t
+; RV32-NEXT: vl8r.v v8, (a3) # Unknown-size Folded Reload
+; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
+; RV32-NEXT: vand.vv v8, v8, v16, v0.t
; RV32-NEXT: csrr a3, vlenb
; RV32-NEXT: slli a3, a3, 4
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 16
-; RV32-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
+; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
; RV32-NEXT: csrr a3, vlenb
; RV32-NEXT: slli a3, a3, 5
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 16
-; RV32-NEXT: vl8r.v v16, (a3) # Unknown-size Folded Reload
-; RV32-NEXT: vsrl.vi v16, v16, 2, v0.t
+; RV32-NEXT: vl8r.v v8, (a3) # Unknown-size Folded Reload
+; RV32-NEXT: vsrl.vi v8, v8, 2, v0.t
; RV32-NEXT: csrr a3, vlenb
; RV32-NEXT: li a4, 40
; RV32-NEXT: mul a3, a3, a4
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 16
-; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: vand.vv v16, v16, v8, v0.t
+; RV32-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
+; RV32-NEXT: vand.vv v8, v8, v16, v0.t
; RV32-NEXT: csrr a3, vlenb
; RV32-NEXT: slli a3, a3, 4
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 16
-; RV32-NEXT: vl8r.v v8, (a3) # Unknown-size Folded Reload
-; RV32-NEXT: vadd.vv v8, v8, v16, v0.t
+; RV32-NEXT: vl8r.v v16, (a3) # Unknown-size Folded Reload
+; RV32-NEXT: vadd.vv v8, v16, v8, v0.t
; RV32-NEXT: vsrl.vi v16, v8, 4, v0.t
-; RV32-NEXT: vadd.vv v16, v8, v16, v0.t
+; RV32-NEXT: vadd.vv v8, v8, v16, v0.t
; RV32-NEXT: lui a3, 61681
; RV32-NEXT: addi a3, a3, -241
; RV32-NEXT: vsetvli a4, zero, e32, m8, ta, ma
-; RV32-NEXT: vmv.v.x v8, a3
-; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
+; RV32-NEXT: vmv.v.x v16, a3
; RV32-NEXT: csrr a3, vlenb
; RV32-NEXT: slli a3, a3, 5
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 16
-; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: vand.vv v16, v16, v8, v0.t
+; RV32-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
+; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
+; RV32-NEXT: vand.vv v16, v8, v16, v0.t
; RV32-NEXT: lui a3, 4112
; RV32-NEXT: addi a3, a3, 257
; RV32-NEXT: vsetvli a4, zero, e32, m8, ta, ma
; RV32-NEXT: vmv.v.x v8, a3
+; RV32-NEXT: csrr a3, vlenb
+; RV32-NEXT: slli a3, a3, 4
+; RV32-NEXT: add a3, sp, a3
+; RV32-NEXT: addi a3, a3, 16
+; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
-; RV32-NEXT: csrr a2, vlenb
-; RV32-NEXT: slli a2, a2, 4
-; RV32-NEXT: add a2, sp, a2
-; RV32-NEXT: addi a2, a2, 16
-; RV32-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
-; RV32-NEXT: vmul.vv v8, v16, v8, v0.t
+; RV32-NEXT: vmul.vv v16, v16, v8, v0.t
; RV32-NEXT: li a2, 56
-; RV32-NEXT: vsrl.vx v8, v8, a2, v0.t
+; RV32-NEXT: vsrl.vx v8, v16, a2, v0.t
; RV32-NEXT: csrr a3, vlenb
; RV32-NEXT: slli a3, a3, 3
; RV32-NEXT: add a3, sp, a3
@@ -2145,14 +2149,13 @@ define <vscale x 16 x i64> @vp_ctpop_nxv16i64(<vscale x 16 x i64> %va, <vscale x
; RV32-NEXT: # %bb.1:
; RV32-NEXT: mv a0, a1
; RV32-NEXT: .LBB46_2:
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vmv1r.v v0, v24
-; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: li a1, 48
-; RV32-NEXT: mul a0, a0, a1
-; RV32-NEXT: add a0, sp, a0
-; RV32-NEXT: addi a0, a0, 16
-; RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: li a3, 48
+; RV32-NEXT: mul a1, a1, a3
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 16
+; RV32-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vsrl.vi v8, v8, 1, v0.t
; RV32-NEXT: addi a0, sp, 16
; RV32-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
@@ -2161,17 +2164,17 @@ define <vscale x 16 x i64> @vp_ctpop_nxv16i64(<vscale x 16 x i64> %va, <vscale x
; RV32-NEXT: mul a0, a0, a1
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 16
-; RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; RV32-NEXT: addi a0, sp, 16
; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; RV32-NEXT: vand.vv v8, v16, v8, v0.t
+; RV32-NEXT: addi a0, sp, 16
+; RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v16, v8, v16, v0.t
; RV32-NEXT: csrr a0, vlenb
; RV32-NEXT: li a1, 48
; RV32-NEXT: mul a0, a0, a1
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 16
-; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; RV32-NEXT: vsub.vv v8, v16, v8, v0.t
+; RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vsub.vv v8, v8, v16, v0.t
; RV32-NEXT: csrr a0, vlenb
; RV32-NEXT: li a1, 48
; RV32-NEXT: mul a0, a0, a1
@@ -2183,41 +2186,41 @@ define <vscale x 16 x i64> @vp_ctpop_nxv16i64(<vscale x 16 x i64> %va, <vscale x
; RV32-NEXT: mul a0, a0, a1
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 16
-; RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; RV32-NEXT: csrr a0, vlenb
; RV32-NEXT: li a1, 48
; RV32-NEXT: mul a0, a0, a1
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 16
-; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; RV32-NEXT: vand.vv v16, v16, v8, v0.t
+; RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v8, v8, v16, v0.t
; RV32-NEXT: csrr a0, vlenb
; RV32-NEXT: li a1, 24
; RV32-NEXT: mul a0, a0, a1
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 16
-; RV32-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
+; RV32-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; RV32-NEXT: csrr a0, vlenb
; RV32-NEXT: li a1, 48
; RV32-NEXT: mul a0, a0, a1
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 16
; RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; RV32-NEXT: vsrl.vi v8, v8, 2, v0.t
+; RV32-NEXT: vsrl.vi v16, v8, 2, v0.t
; RV32-NEXT: csrr a0, vlenb
; RV32-NEXT: li a1, 40
; RV32-NEXT: mul a0, a0, a1
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 16
-; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; RV32-NEXT: vand.vv v8, v8, v16, v0.t
+; RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v16, v16, v8, v0.t
; RV32-NEXT: csrr a0, vlenb
; RV32-NEXT: li a1, 24
; RV32-NEXT: mul a0, a0, a1
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 16
-; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; RV32-NEXT: vadd.vv v8, v16, v8, v0.t
+; RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vadd.vv v8, v8, v16, v0.t
; RV32-NEXT: vsrl.vi v16, v8, 4, v0.t
; RV32-NEXT: vadd.vv v8, v8, v16, v0.t
; RV32-NEXT: csrr a0, vlenb
@@ -2303,13 +2306,13 @@ define <vscale x 16 x i64> @vp_ctpop_nxv16i64(<vscale x 16 x i64> %va, <vscale x
; RV64-NEXT: sltu a0, a0, a1
; RV64-NEXT: addi a0, a0, -1
; RV64-NEXT: and a0, a0, a1
-; RV64-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV64-NEXT: vmv1r.v v0, v24
-; RV64-NEXT: csrr a0, vlenb
-; RV64-NEXT: slli a0, a0, 3
-; RV64-NEXT: add a0, sp, a0
-; RV64-NEXT: addi a0, a0, 16
-; RV64-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; RV64-NEXT: csrr a1, vlenb
+; RV64-NEXT: slli a1, a1, 3
+; RV64-NEXT: add a1, sp, a1
+; RV64-NEXT: addi a1, a1, 16
+; RV64-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
+; RV64-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV64-NEXT: vsrl.vi v16, v8, 1, v0.t
; RV64-NEXT: vand.vx v16, v16, a2, v0.t
; RV64-NEXT: vsub.vv v16, v8, v16, v0.t
@@ -2347,8 +2350,8 @@ define <vscale x 16 x i64> @vp_ctpop_nxv16i64(<vscale x 16 x i64> %va, <vscale x
; CHECK-ZVBB-NEXT: # %bb.1:
; CHECK-ZVBB-NEXT: mv a0, a1
; CHECK-ZVBB-NEXT: .LBB46_2:
-; CHECK-ZVBB-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-ZVBB-NEXT: vmv1r.v v0, v24
+; CHECK-ZVBB-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-ZVBB-NEXT: vcpop.v v8, v8, v0.t
; CHECK-ZVBB-NEXT: ret
%v = call <vscale x 16 x i64> @llvm.vp.ctpop.nxv16i64(<vscale x 16 x i64> %va, <vscale x 16 x i1> %m, i32 %evl)
@@ -2375,13 +2378,13 @@ define <vscale x 16 x i64> @vp_ctpop_nxv16i64_unmasked(<vscale x 16 x i64> %va,
; RV32-NEXT: addi a3, a3, 1365
; RV32-NEXT: vsetvli a4, zero, e32, m8, ta, ma
; RV32-NEXT: vmv.v.x v0, a3
-; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
; RV32-NEXT: csrr a3, vlenb
; RV32-NEXT: li a4, 24
; RV32-NEXT: mul a3, a3, a4
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 16
; RV32-NEXT: vs8r.v v0, (a3) # Unknown-size Folded Spill
+; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
; RV32-NEXT: vand.vv v24, v24, v0
; RV32-NEXT: vsub.vv v24, v16, v24
; RV32-NEXT: lui a3, 209715
@@ -2404,20 +2407,20 @@ define <vscale x 16 x i64> @vp_ctpop_nxv16i64_unmasked(<vscale x 16 x i64> %va,
; RV32-NEXT: addi a3, a3, -241
; RV32-NEXT: vsetvli a4, zero, e32, m8, ta, ma
; RV32-NEXT: vmv.v.x v24, a3
-; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
; RV32-NEXT: csrr a3, vlenb
; RV32-NEXT: slli a3, a3, 3
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 16
; RV32-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
+; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
; RV32-NEXT: vand.vv v16, v16, v24
; RV32-NEXT: lui a3, 4112
; RV32-NEXT: addi a3, a3, 257
; RV32-NEXT: vsetvli a4, zero, e32, m8, ta, ma
; RV32-NEXT: vmv.v.x v24, a3
+; RV32-NEXT: addi a3, sp, 16
+; RV32-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
-; RV32-NEXT: addi a2, sp, 16
-; RV32-NEXT: vs8r.v v24, (a2) # Unknown-size Folded Spill
; RV32-NEXT: vmul.vv v16, v16, v24
; RV32-NEXT: li a2, 56
; RV32-NEXT: vsrl.vx v16, v16, a2
diff --git a/llvm/test/CodeGen/RISCV/rvv/cttz-vp.ll b/llvm/test/CodeGen/RISCV/rvv/cttz-vp.ll
index ef8a6c704a44b..d7fcb94bbf497 100644
--- a/llvm/test/CodeGen/RISCV/rvv/cttz-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/cttz-vp.ll
@@ -2282,7 +2282,6 @@ define <vscale x 16 x i64> @vp_cttz_nxv16i64(<vscale x 16 x i64> %va, <vscale x
; RV32-NEXT: addi a4, a4, 1365
; RV32-NEXT: vsetvli a5, zero, e32, m8, ta, ma
; RV32-NEXT: vmv.v.x v8, a4
-; RV32-NEXT: vsetvli zero, a3, e64, m8, ta, ma
; RV32-NEXT: csrr a4, vlenb
; RV32-NEXT: li a5, 24
; RV32-NEXT: mul a4, a4, a5
@@ -2295,6 +2294,7 @@ define <vscale x 16 x i64> @vp_cttz_nxv16i64(<vscale x 16 x i64> %va, <vscale x
; RV32-NEXT: add a4, sp, a4
; RV32-NEXT: addi a4, a4, 16
; RV32-NEXT: vl8r.v v16, (a4) # Unknown-size Folded Reload
+; RV32-NEXT: vsetvli zero, a3, e64, m8, ta, ma
; RV32-NEXT: vand.vv v8, v16, v8, v0.t
; RV32-NEXT: csrr a4, vlenb
; RV32-NEXT: li a5, 40
@@ -2312,82 +2312,81 @@ define <vscale x 16 x i64> @vp_cttz_nxv16i64(<vscale x 16 x i64> %va, <vscale x
; RV32-NEXT: lui a4, 209715
; RV32-NEXT: addi a4, a4, 819
; RV32-NEXT: vsetvli a5, zero, e32, m8, ta, ma
-; RV32-NEXT: vmv.v.x v8, a4
-; RV32-NEXT: vsetvli zero, a3, e64, m8, ta, ma
+; RV32-NEXT: vmv.v.x v16, a4
; RV32-NEXT: csrr a4, vlenb
; RV32-NEXT: li a5, 40
; RV32-NEXT: mul a4, a4, a5
; RV32-NEXT: add a4, sp, a4
; RV32-NEXT: addi a4, a4, 16
-; RV32-NEXT: vl8r.v v16, (a4) # Unknown-size Folded Reload
-; RV32-NEXT: vand.vv v16, v16, v8, v0.t
+; RV32-NEXT: vl8r.v v8, (a4) # Unknown-size Folded Reload
+; RV32-NEXT: vsetvli zero, a3, e64, m8, ta, ma
+; RV32-NEXT: vand.vv v8, v8, v16, v0.t
; RV32-NEXT: csrr a4, vlenb
; RV32-NEXT: slli a4, a4, 4
; RV32-NEXT: add a4, sp, a4
; RV32-NEXT: addi a4, a4, 16
-; RV32-NEXT: vs8r.v v16, (a4) # Unknown-size Folded Spill
+; RV32-NEXT: vs8r.v v8, (a4) # Unknown-size Folded Spill
; RV32-NEXT: csrr a4, vlenb
; RV32-NEXT: li a5, 40
; RV32-NEXT: mul a4, a4, a5
; RV32-NEXT: add a4, sp, a4
; RV32-NEXT: addi a4, a4, 16
-; RV32-NEXT: vl8r.v v16, (a4) # Unknown-size Folded Reload
-; RV32-NEXT: vsrl.vi v16, v16, 2, v0.t
+; RV32-NEXT: vl8r.v v8, (a4) # Unknown-size Folded Reload
+; RV32-NEXT: vsrl.vi v8, v8, 2, v0.t
; RV32-NEXT: csrr a4, vlenb
; RV32-NEXT: li a5, 48
; RV32-NEXT: mul a4, a4, a5
; RV32-NEXT: add a4, sp, a4
; RV32-NEXT: addi a4, a4, 16
-; RV32-NEXT: vs8r.v v8, (a4) # Unknown-size Folded Spill
-; RV32-NEXT: vand.vv v16, v16, v8, v0.t
+; RV32-NEXT: vs8r.v v16, (a4) # Unknown-size Folded Spill
+; RV32-NEXT: vand.vv v8, v8, v16, v0.t
; RV32-NEXT: csrr a4, vlenb
; RV32-NEXT: slli a4, a4, 4
; RV32-NEXT: add a4, sp, a4
; RV32-NEXT: addi a4, a4, 16
-; RV32-NEXT: vl8r.v v8, (a4) # Unknown-size Folded Reload
-; RV32-NEXT: vadd.vv v8, v8, v16, v0.t
+; RV32-NEXT: vl8r.v v16, (a4) # Unknown-size Folded Reload
+; RV32-NEXT: vadd.vv v8, v16, v8, v0.t
; RV32-NEXT: vsrl.vi v16, v8, 4, v0.t
-; RV32-NEXT: vadd.vv v16, v8, v16, v0.t
+; RV32-NEXT: vadd.vv v8, v8, v16, v0.t
; RV32-NEXT: lui a4, 61681
; RV32-NEXT: addi a4, a4, -241
; RV32-NEXT: vsetvli a5, zero, e32, m8, ta, ma
-; RV32-NEXT: vmv.v.x v8, a4
-; RV32-NEXT: vsetvli zero, a3, e64, m8, ta, ma
+; RV32-NEXT: vmv.v.x v16, a4
; RV32-NEXT: csrr a4, vlenb
; RV32-NEXT: slli a4, a4, 4
; RV32-NEXT: add a4, sp, a4
; RV32-NEXT: addi a4, a4, 16
-; RV32-NEXT: vs8r.v v8, (a4) # Unknown-size Folded Spill
-; RV32-NEXT: vand.vv v16, v16, v8, v0.t
+; RV32-NEXT: vs8r.v v16, (a4) # Unknown-size Folded Spill
+; RV32-NEXT: vsetvli zero, a3, e64, m8, ta, ma
+; RV32-NEXT: vand.vv v16, v8, v16, v0.t
; RV32-NEXT: lui a4, 4112
; RV32-NEXT: addi a4, a4, 257
; RV32-NEXT: vsetvli a5, zero, e32, m8, ta, ma
; RV32-NEXT: vmv.v.x v8, a4
+; RV32-NEXT: csrr a4, vlenb
+; RV32-NEXT: slli a4, a4, 3
+; RV32-NEXT: add a4, sp, a4
+; RV32-NEXT: addi a4, a4, 16
+; RV32-NEXT: vs8r.v v8, (a4) # Unknown-size Folded Spill
; RV32-NEXT: vsetvli zero, a3, e64, m8, ta, ma
-; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: slli a3, a3, 3
-; RV32-NEXT: add a3, sp, a3
-; RV32-NEXT: addi a3, a3, 16
-; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: vmul.vv v8, v16, v8, v0.t
+; RV32-NEXT: vmul.vv v16, v16, v8, v0.t
; RV32-NEXT: li a3, 56
-; RV32-NEXT: vsrl.vx v8, v8, a3, v0.t
+; RV32-NEXT: vsrl.vx v8, v16, a3, v0.t
; RV32-NEXT: addi a4, sp, 16
; RV32-NEXT: vs8r.v v8, (a4) # Unknown-size Folded Spill
; RV32-NEXT: bltu a0, a1, .LBB46_2
; RV32-NEXT: # %bb.1:
; RV32-NEXT: mv a0, a1
; RV32-NEXT: .LBB46_2:
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vmv1r.v v0, v24
-; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 5
-; RV32-NEXT: add a0, sp, a0
-; RV32-NEXT: addi a0, a0, 16
-; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; RV32-NEXT: vsub.vx v8, v16, a2, v0.t
-; RV32-NEXT: vnot.v v16, v16, v0.t
-; RV32-NEXT: vand.vv v8, v16, v8, v0.t
+; RV32-NEXT: slli a1, a1, 5
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 16
+; RV32-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vsub.vx v16, v8, a2, v0.t
+; RV32-NEXT: vnot.v v8, v8, v0.t
+; RV32-NEXT: vand.vv v8, v8, v16, v0.t
; RV32-NEXT: csrr a0, vlenb
; RV32-NEXT: slli a0, a0, 5
; RV32-NEXT: add a0, sp, a0
@@ -2549,13 +2548,12 @@ define <vscale x 16 x i64> @vp_cttz_nxv16i64(<vscale x 16 x i64> %va, <vscale x
; RV64-NEXT: # %bb.1:
; RV64-NEXT: mv a0, a1
; RV64-NEXT: .LBB46_2:
-; RV64-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV64-NEXT: vmv1r.v v0, v24
-; RV64-NEXT: csrr a0, vlenb
-; RV64-NEXT: slli a0, a0, 3
-; RV64-NEXT: add a0, sp, a0
-; RV64-NEXT: addi a0, a0, 16
-; RV64-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; RV64-NEXT: slli a1, a1, 3
+; RV64-NEXT: add a1, sp, a1
+; RV64-NEXT: addi a1, a1, 16
+; RV64-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
+; RV64-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV64-NEXT: vsub.vx v16, v8, a2, v0.t
; RV64-NEXT: vnot.v v8, v8, v0.t
; RV64-NEXT: vand.vv v8, v8, v16, v0.t
@@ -2596,8 +2594,8 @@ define <vscale x 16 x i64> @vp_cttz_nxv16i64(<vscale x 16 x i64> %va, <vscale x
; CHECK-ZVBB-NEXT: # %bb.1:
; CHECK-ZVBB-NEXT: mv a0, a1
; CHECK-ZVBB-NEXT: .LBB46_2:
-; CHECK-ZVBB-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-ZVBB-NEXT: vmv1r.v v0, v24
+; CHECK-ZVBB-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-ZVBB-NEXT: vctz.v v8, v8, v0.t
; CHECK-ZVBB-NEXT: ret
%v = call <vscale x 16 x i64> @llvm.vp.cttz.nxv16i64(<vscale x 16 x i64> %va, i1 false, <vscale x 16 x i1> %m, i32 %evl)
@@ -2628,98 +2626,97 @@ define <vscale x 16 x i64> @vp_cttz_nxv16i64_unmasked(<vscale x 16 x i64> %va, i
; RV32-NEXT: vsetvli zero, a3, e64, m8, ta, ma
; RV32-NEXT: vsub.vx v8, v16, a2
; RV32-NEXT: vnot.v v16, v16
-; RV32-NEXT: vand.vv v8, v16, v8
-; RV32-NEXT: vsrl.vi v24, v8, 1
+; RV32-NEXT: vand.vv v16, v16, v8
+; RV32-NEXT: vsrl.vi v24, v16, 1
; RV32-NEXT: lui a4, 349525
; RV32-NEXT: addi a4, a4, 1365
; RV32-NEXT: vsetvli a5, zero, e32, m8, ta, ma
-; RV32-NEXT: vmv.v.x v16, a4
-; RV32-NEXT: vsetvli zero, a3, e64, m8, ta, ma
+; RV32-NEXT: vmv.v.x v8, a4
; RV32-NEXT: csrr a4, vlenb
; RV32-NEXT: li a5, 24
; RV32-NEXT: mul a4, a4, a5
; RV32-NEXT: add a4, sp, a4
; RV32-NEXT: addi a4, a4, 16
-; RV32-NEXT: vs8r.v v16, (a4) # Unknown-size Folded Spill
-; RV32-NEXT: vand.vv v24, v24, v16
-; RV32-NEXT: vsub.vv v8, v8, v24
+; RV32-NEXT: vs8r.v v8, (a4) # Unknown-size Folded Spill
+; RV32-NEXT: vsetvli zero, a3, e64, m8, ta, ma
+; RV32-NEXT: vand.vv v24, v24, v8
+; RV32-NEXT: vsub.vv v16, v16, v24
; RV32-NEXT: lui a4, 209715
; RV32-NEXT: addi a4, a4, 819
; RV32-NEXT: vsetvli a5, zero, e32, m8, ta, ma
; RV32-NEXT: vmv.v.x v0, a4
; RV32-NEXT: vsetvli zero, a3, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v24, v8, v0
-; RV32-NEXT: vsrl.vi v8, v8, 2
-; RV32-NEXT: vand.vv v8, v8, v0
-; RV32-NEXT: vadd.vv v8, v24, v8
-; RV32-NEXT: vsrl.vi v24, v8, 4
-; RV32-NEXT: vadd.vv v8, v8, v24
+; RV32-NEXT: vand.vv v24, v16, v0
+; RV32-NEXT: vsrl.vi v16, v16, 2
+; RV32-NEXT: vand.vv v16, v16, v0
+; RV32-NEXT: vadd.vv v16, v24, v16
+; RV32-NEXT: vsrl.vi v24, v16, 4
+; RV32-NEXT: vadd.vv v16, v16, v24
; RV32-NEXT: lui a4, 61681
; RV32-NEXT: addi a4, a4, -241
; RV32-NEXT: vsetvli a5, zero, e32, m8, ta, ma
-; RV32-NEXT: vmv.v.x v16, a4
-; RV32-NEXT: vsetvli zero, a3, e64, m8, ta, ma
+; RV32-NEXT: vmv.v.x v8, a4
; RV32-NEXT: csrr a4, vlenb
; RV32-NEXT: slli a4, a4, 4
; RV32-NEXT: add a4, sp, a4
; RV32-NEXT: addi a4, a4, 16
-; RV32-NEXT: vs8r.v v16, (a4) # Unknown-size Folded Spill
-; RV32-NEXT: vand.vv v16, v8, v16
+; RV32-NEXT: vs8r.v v8, (a4) # Unknown-size Folded Spill
+; RV32-NEXT: vsetvli zero, a3, e64, m8, ta, ma
+; RV32-NEXT: vand.vv v8, v16, v8
; RV32-NEXT: lui a4, 4112
; RV32-NEXT: addi a4, a4, 257
; RV32-NEXT: vsetvli a5, zero, e32, m8, ta, ma
-; RV32-NEXT: vmv.v.x v8, a4
+; RV32-NEXT: vmv.v.x v16, a4
+; RV32-NEXT: csrr a4, vlenb
+; RV32-NEXT: slli a4, a4, 3
+; RV32-NEXT: add a4, sp, a4
+; RV32-NEXT: addi a4, a4, 16
+; RV32-NEXT: vs8r.v v16, (a4) # Unknown-size Folded Spill
; RV32-NEXT: vsetvli zero, a3, e64, m8, ta, ma
-; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: slli a3, a3, 3
-; RV32-NEXT: add a3, sp, a3
-; RV32-NEXT: addi a3, a3, 16
-; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: vmul.vv v16, v16, v8
+; RV32-NEXT: vmul.vv v8, v8, v16
; RV32-NEXT: li a3, 56
-; RV32-NEXT: vsrl.vx v8, v16, a3
+; RV32-NEXT: vsrl.vx v8, v8, a3
; RV32-NEXT: addi a4, sp, 16
; RV32-NEXT: vs8r.v v8, (a4) # Unknown-size Folded Spill
; RV32-NEXT: bltu a0, a1, .LBB47_2
; RV32-NEXT: # %bb.1:
; RV32-NEXT: mv a0, a1
; RV32-NEXT: .LBB47_2:
+; RV32-NEXT: slli a1, a1, 5
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 16
+; RV32-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 5
-; RV32-NEXT: add a0, sp, a0
-; RV32-NEXT: addi a0, a0, 16
-; RV32-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
-; RV32-NEXT: vsub.vx v16, v24, a2
+; RV32-NEXT: vsub.vx v8, v24, a2
; RV32-NEXT: vnot.v v24, v24
-; RV32-NEXT: vand.vv v16, v24, v16
-; RV32-NEXT: vsrl.vi v24, v16, 1
+; RV32-NEXT: vand.vv v8, v24, v8
+; RV32-NEXT: vsrl.vi v24, v8, 1
; RV32-NEXT: csrr a0, vlenb
; RV32-NEXT: li a1, 24
; RV32-NEXT: mul a0, a0, a1
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 16
-; RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; RV32-NEXT: vand.vv v24, v24, v8
-; RV32-NEXT: vsub.vv v16, v16, v24
-; RV32-NEXT: vand.vv v24, v16, v0
-; RV32-NEXT: vsrl.vi v16, v16, 2
-; RV32-NEXT: vand.vv v16, v16, v0
-; RV32-NEXT: vadd.vv v16, v24, v16
-; RV32-NEXT: vsrl.vi v24, v16, 4
-; RV32-NEXT: vadd.vv v16, v16, v24
+; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v24, v24, v16
+; RV32-NEXT: vsub.vv v8, v8, v24
+; RV32-NEXT: vand.vv v24, v8, v0
+; RV32-NEXT: vsrl.vi v8, v8, 2
+; RV32-NEXT: vand.vv v8, v8, v0
+; RV32-NEXT: vadd.vv v8, v24, v8
+; RV32-NEXT: vsrl.vi v24, v8, 4
+; RV32-NEXT: vadd.vv v8, v8, v24
; RV32-NEXT: csrr a0, vlenb
; RV32-NEXT: slli a0, a0, 4
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 16
-; RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; RV32-NEXT: vand.vv v16, v16, v8
+; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v8, v8, v16
; RV32-NEXT: csrr a0, vlenb
; RV32-NEXT: slli a0, a0, 3
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 16
-; RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; RV32-NEXT: vmul.vv v8, v16, v8
+; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vmul.vv v8, v8, v16
; RV32-NEXT: vsrl.vx v8, v8, a3
; RV32-NEXT: addi a0, sp, 16
; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
@@ -4038,13 +4035,12 @@ define <vscale x 16 x i64> @vp_cttz_zero_undef_nxv16i64(<vscale x 16 x i64> %va,
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB94_2:
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: slli a1, a1, 3
+; CHECK-NEXT: add a1, sp, a1
+; CHECK-NEXT: addi a1, a1, 16
+; CHECK-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vrsub.vi v16, v8, 0, v0.t
; CHECK-NEXT: vand.vv v8, v8, v16, v0.t
; CHECK-NEXT: fsrmi a0, 1
@@ -4077,8 +4073,8 @@ define <vscale x 16 x i64> @vp_cttz_zero_undef_nxv16i64(<vscale x 16 x i64> %va,
; CHECK-ZVBB-NEXT: # %bb.1:
; CHECK-ZVBB-NEXT: mv a0, a1
; CHECK-ZVBB-NEXT: .LBB94_2:
-; CHECK-ZVBB-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-ZVBB-NEXT: vmv1r.v v0, v24
+; CHECK-ZVBB-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-ZVBB-NEXT: vctz.v v8, v8, v0.t
; CHECK-ZVBB-NEXT: ret
%v = call <vscale x 16 x i64> @llvm.vp.cttz.nxv16i64(<vscale x 16 x i64> %va, i1 true, <vscale x 16 x i1> %m, i32 %evl)
diff --git a/llvm/test/CodeGen/RISCV/rvv/extractelt-i1.ll b/llvm/test/CodeGen/RISCV/rvv/extractelt-i1.ll
index e69b4789a09af..43bb8059f8352 100644
--- a/llvm/test/CodeGen/RISCV/rvv/extractelt-i1.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/extractelt-i1.ll
@@ -142,22 +142,22 @@ define i1 @extractelt_nxv128i1(ptr %x, i64 %idx) nounwind {
; RV32-NEXT: slli a3, a3, 4
; RV32-NEXT: sub sp, sp, a3
; RV32-NEXT: andi sp, sp, -64
-; RV32-NEXT: addi a3, sp, 64
; RV32-NEXT: slli a2, a2, 3
-; RV32-NEXT: add a4, a0, a2
-; RV32-NEXT: vl8r.v v16, (a4)
+; RV32-NEXT: add a3, a0, a2
+; RV32-NEXT: vl8r.v v16, (a3)
; RV32-NEXT: vl8r.v v24, (a0)
-; RV32-NEXT: add a1, a3, a1
-; RV32-NEXT: vsetvli a0, zero, e8, m8, ta, ma
+; RV32-NEXT: addi a0, sp, 64
+; RV32-NEXT: add a1, a0, a1
+; RV32-NEXT: vsetvli a3, zero, e8, m8, ta, ma
; RV32-NEXT: vmseq.vi v8, v16, 0
; RV32-NEXT: vmseq.vi v0, v24, 0
; RV32-NEXT: vmv.v.i v16, 0
; RV32-NEXT: vmerge.vim v24, v16, 1, v0
-; RV32-NEXT: vs8r.v v24, (a3)
-; RV32-NEXT: add a2, a3, a2
+; RV32-NEXT: vs8r.v v24, (a0)
+; RV32-NEXT: add a0, a0, a2
; RV32-NEXT: vmv1r.v v0, v8
; RV32-NEXT: vmerge.vim v8, v16, 1, v0
-; RV32-NEXT: vs8r.v v8, (a2)
+; RV32-NEXT: vs8r.v v8, (a0)
; RV32-NEXT: lbu a0, 0(a1)
; RV32-NEXT: addi sp, s0, -80
; RV32-NEXT: lw ra, 76(sp) # 4-byte Folded Reload
@@ -182,22 +182,22 @@ define i1 @extractelt_nxv128i1(ptr %x, i64 %idx) nounwind {
; RV64-NEXT: slli a3, a3, 4
; RV64-NEXT: sub sp, sp, a3
; RV64-NEXT: andi sp, sp, -64
-; RV64-NEXT: addi a3, sp, 64
; RV64-NEXT: slli a2, a2, 3
-; RV64-NEXT: add a4, a0, a2
-; RV64-NEXT: vl8r.v v16, (a4)
+; RV64-NEXT: add a3, a0, a2
+; RV64-NEXT: vl8r.v v16, (a3)
; RV64-NEXT: vl8r.v v24, (a0)
-; RV64-NEXT: add a1, a3, a1
-; RV64-NEXT: vsetvli a0, zero, e8, m8, ta, ma
+; RV64-NEXT: addi a0, sp, 64
+; RV64-NEXT: add a1, a0, a1
+; RV64-NEXT: vsetvli a3, zero, e8, m8, ta, ma
; RV64-NEXT: vmseq.vi v8, v16, 0
; RV64-NEXT: vmseq.vi v0, v24, 0
; RV64-NEXT: vmv.v.i v16, 0
; RV64-NEXT: vmerge.vim v24, v16, 1, v0
-; RV64-NEXT: vs8r.v v24, (a3)
-; RV64-NEXT: add a2, a3, a2
+; RV64-NEXT: vs8r.v v24, (a0)
+; RV64-NEXT: add a0, a0, a2
; RV64-NEXT: vmv1r.v v0, v8
; RV64-NEXT: vmerge.vim v8, v16, 1, v0
-; RV64-NEXT: vs8r.v v8, (a2)
+; RV64-NEXT: vs8r.v v8, (a0)
; RV64-NEXT: lbu a0, 0(a1)
; RV64-NEXT: addi sp, s0, -80
; RV64-NEXT: ld ra, 72(sp) # 8-byte Folded Reload
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vector-i8-index-cornercase.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vector-i8-index-cornercase.ll
index 875f4f239028b..6b8d778bc3242 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vector-i8-index-cornercase.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vector-i8-index-cornercase.ll
@@ -35,10 +35,10 @@ define <512 x i8> @single_source(<512 x i8> %a) {
; CHECK-NEXT: vslidedown.vi v16, v16, 4
; CHECK-NEXT: li a0, 466
; CHECK-NEXT: li a1, 465
+; CHECK-NEXT: lbu a2, 1012(sp)
; CHECK-NEXT: vsetvli zero, a0, e8, m8, tu, ma
-; CHECK-NEXT: lbu a0, 1012(sp)
; CHECK-NEXT: vslideup.vx v8, v16, a1
-; CHECK-NEXT: vmv.s.x v16, a0
+; CHECK-NEXT: vmv.s.x v16, a2
; CHECK-NEXT: li a0, 501
; CHECK-NEXT: li a1, 500
; CHECK-NEXT: vsetvli zero, a0, e8, m8, tu, ma
@@ -118,16 +118,16 @@ define <512 x i8> @two_source(<512 x i8> %a, <512 x i8> %b) {
; CHECK-NEXT: vslidedown.vi v24, v24, 4
; CHECK-NEXT: li a1, 466
; CHECK-NEXT: li a2, 465
+; CHECK-NEXT: lbu a3, 985(sp)
; CHECK-NEXT: vsetvli zero, a1, e8, m8, tu, ma
-; CHECK-NEXT: lbu a1, 985(sp)
; CHECK-NEXT: vslideup.vx v8, v24, a2
-; CHECK-NEXT: vmv.s.x v24, a1
+; CHECK-NEXT: vmv.s.x v24, a3
; CHECK-NEXT: li a1, 478
; CHECK-NEXT: li a2, 477
+; CHECK-NEXT: lbu a3, 1012(sp)
; CHECK-NEXT: vsetvli zero, a1, e8, m8, tu, ma
-; CHECK-NEXT: lbu a1, 1012(sp)
; CHECK-NEXT: vslideup.vx v8, v24, a2
-; CHECK-NEXT: vmv.s.x v24, a1
+; CHECK-NEXT: vmv.s.x v24, a3
; CHECK-NEXT: li a1, 501
; CHECK-NEXT: li a2, 500
; CHECK-NEXT: vsetvli zero, a1, e8, m8, tu, ma
@@ -137,21 +137,21 @@ define <512 x i8> @two_source(<512 x i8> %a, <512 x i8> %b) {
; CHECK-NEXT: addi a1, a1, 501
; CHECK-NEXT: slli a1, a1, 13
; CHECK-NEXT: addi a1, a1, 512
+; CHECK-NEXT: vsetivli zero, 8, e64, m1, ta, ma
+; CHECK-NEXT: vmv.v.i v24, 0
+; CHECK-NEXT: lui a2, 1047552
+; CHECK-NEXT: addiw a2, a2, 1
+; CHECK-NEXT: slli a2, a2, 23
+; CHECK-NEXT: addi a2, a2, 1
+; CHECK-NEXT: slli a2, a2, 18
+; CHECK-NEXT: vslide1down.vx v0, v24, a2
+; CHECK-NEXT: lui a2, 4
+; CHECK-NEXT: vmv.s.x v24, a2
; CHECK-NEXT: li a2, 64
+; CHECK-NEXT: vsetivli zero, 7, e64, m1, tu, ma
+; CHECK-NEXT: vslideup.vi v0, v24, 6
; CHECK-NEXT: vsetvli zero, a2, e64, m8, ta, ma
; CHECK-NEXT: vmv.v.x v24, a1
-; CHECK-NEXT: vsetivli zero, 8, e64, m1, ta, ma
-; CHECK-NEXT: vmv.v.i v7, 0
-; CHECK-NEXT: lui a1, 1047552
-; CHECK-NEXT: addiw a1, a1, 1
-; CHECK-NEXT: slli a1, a1, 23
-; CHECK-NEXT: addi a1, a1, 1
-; CHECK-NEXT: slli a1, a1, 18
-; CHECK-NEXT: vslide1down.vx v0, v7, a1
-; CHECK-NEXT: lui a1, 4
-; CHECK-NEXT: vmv.s.x v7, a1
-; CHECK-NEXT: vsetivli zero, 7, e64, m1, tu, ma
-; CHECK-NEXT: vslideup.vi v0, v7, 6
; CHECK-NEXT: vsetvli zero, a0, e8, m8, ta, mu
; CHECK-NEXT: vrgather.vv v8, v16, v24, v0.t
; CHECK-NEXT: addi sp, s0, -1536
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-abs-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-abs-vp.ll
index c0d366760d079..f3e823562888f 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-abs-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-abs-vp.ll
@@ -417,8 +417,8 @@ declare <32 x i64> @llvm.vp.abs.v32i64(<32 x i64>, i1 immarg, <32 x i1>, i32)
define <32 x i64> @vp_abs_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_abs_v32i64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; CHECK-NEXT: li a2, 16
+; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; CHECK-NEXT: vslidedown.vi v7, v0, 2
; CHECK-NEXT: mv a1, a0
; CHECK-NEXT: bltu a0, a2, .LBB34_2
@@ -432,8 +432,8 @@ define <32 x i64> @vp_abs_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %evl)
; CHECK-NEXT: sltu a0, a0, a1
; CHECK-NEXT: addi a0, a0, -1
; CHECK-NEXT: and a0, a0, a1
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v7
+; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vrsub.vi v24, v16, 0, v0.t
; CHECK-NEXT: vmax.vv v16, v16, v24, v0.t
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-bitreverse-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-bitreverse-vp.ll
index 943fc58d637a0..068c25b821002 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-bitreverse-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-bitreverse-vp.ll
@@ -847,27 +847,27 @@ define <2 x i64> @vp_bitreverse_v2i64(<2 x i64> %va, <2 x i1> %m, i32 zeroext %e
; RV32-NEXT: li a3, 40
; RV32-NEXT: vsll.vx v10, v10, a3, v0.t
; RV32-NEXT: vor.vv v9, v9, v10, v0.t
-; RV32-NEXT: lui a4, 4080
-; RV32-NEXT: vand.vx v10, v8, a4, v0.t
-; RV32-NEXT: vsll.vi v10, v10, 24, v0.t
-; RV32-NEXT: addi a5, sp, 8
+; RV32-NEXT: addi a4, sp, 8
; RV32-NEXT: vsetivli zero, 2, e64, m1, ta, ma
-; RV32-NEXT: vlse64.v v11, (a5), zero
+; RV32-NEXT: vlse64.v v10, (a4), zero
+; RV32-NEXT: lui a4, 4080
; RV32-NEXT: vsetvli zero, a0, e64, m1, ta, ma
-; RV32-NEXT: vand.vv v12, v8, v11, v0.t
+; RV32-NEXT: vand.vx v11, v8, a4, v0.t
+; RV32-NEXT: vsll.vi v11, v11, 24, v0.t
+; RV32-NEXT: vand.vv v12, v8, v10, v0.t
; RV32-NEXT: vsll.vi v12, v12, 8, v0.t
-; RV32-NEXT: vor.vv v10, v10, v12, v0.t
-; RV32-NEXT: vor.vv v9, v9, v10, v0.t
-; RV32-NEXT: vsrl.vx v10, v8, a1, v0.t
+; RV32-NEXT: vor.vv v11, v11, v12, v0.t
+; RV32-NEXT: vor.vv v9, v9, v11, v0.t
+; RV32-NEXT: vsrl.vx v11, v8, a1, v0.t
; RV32-NEXT: vsrl.vx v12, v8, a3, v0.t
; RV32-NEXT: vand.vx v12, v12, a2, v0.t
-; RV32-NEXT: vor.vv v10, v12, v10, v0.t
+; RV32-NEXT: vor.vv v11, v12, v11, v0.t
; RV32-NEXT: vsrl.vi v12, v8, 24, v0.t
; RV32-NEXT: vand.vx v12, v12, a4, v0.t
; RV32-NEXT: vsrl.vi v8, v8, 8, v0.t
-; RV32-NEXT: vand.vv v8, v8, v11, v0.t
+; RV32-NEXT: vand.vv v8, v8, v10, v0.t
; RV32-NEXT: vor.vv v8, v8, v12, v0.t
-; RV32-NEXT: vor.vv v8, v8, v10, v0.t
+; RV32-NEXT: vor.vv v8, v8, v11, v0.t
; RV32-NEXT: vor.vv v8, v9, v8, v0.t
; RV32-NEXT: vsrl.vi v9, v8, 4, v0.t
; RV32-NEXT: lui a1, 61681
@@ -982,27 +982,27 @@ define <2 x i64> @vp_bitreverse_v2i64_unmasked(<2 x i64> %va, i32 zeroext %evl)
; RV32-NEXT: li a3, 40
; RV32-NEXT: vsll.vx v10, v10, a3
; RV32-NEXT: vor.vv v9, v9, v10
-; RV32-NEXT: lui a4, 4080
-; RV32-NEXT: vand.vx v10, v8, a4
-; RV32-NEXT: vsll.vi v10, v10, 24
-; RV32-NEXT: addi a5, sp, 8
+; RV32-NEXT: addi a4, sp, 8
; RV32-NEXT: vsetivli zero, 2, e64, m1, ta, ma
-; RV32-NEXT: vlse64.v v11, (a5), zero
+; RV32-NEXT: vlse64.v v10, (a4), zero
+; RV32-NEXT: lui a4, 4080
; RV32-NEXT: vsetvli zero, a0, e64, m1, ta, ma
-; RV32-NEXT: vand.vv v12, v8, v11
+; RV32-NEXT: vand.vx v11, v8, a4
+; RV32-NEXT: vsll.vi v11, v11, 24
+; RV32-NEXT: vand.vv v12, v8, v10
; RV32-NEXT: vsll.vi v12, v12, 8
-; RV32-NEXT: vor.vv v10, v10, v12
-; RV32-NEXT: vor.vv v9, v9, v10
-; RV32-NEXT: vsrl.vx v10, v8, a1
+; RV32-NEXT: vor.vv v11, v11, v12
+; RV32-NEXT: vor.vv v9, v9, v11
+; RV32-NEXT: vsrl.vx v11, v8, a1
; RV32-NEXT: vsrl.vx v12, v8, a3
; RV32-NEXT: vand.vx v12, v12, a2
-; RV32-NEXT: vor.vv v10, v12, v10
+; RV32-NEXT: vor.vv v11, v12, v11
; RV32-NEXT: vsrl.vi v12, v8, 24
; RV32-NEXT: vand.vx v12, v12, a4
; RV32-NEXT: vsrl.vi v8, v8, 8
-; RV32-NEXT: vand.vv v8, v8, v11
+; RV32-NEXT: vand.vv v8, v8, v10
; RV32-NEXT: vor.vv v8, v8, v12
-; RV32-NEXT: vor.vv v8, v8, v10
+; RV32-NEXT: vor.vv v8, v8, v11
; RV32-NEXT: vor.vv v8, v9, v8
; RV32-NEXT: vsrl.vi v9, v8, 4
; RV32-NEXT: lui a1, 61681
@@ -1119,27 +1119,27 @@ define <4 x i64> @vp_bitreverse_v4i64(<4 x i64> %va, <4 x i1> %m, i32 zeroext %e
; RV32-NEXT: li a3, 40
; RV32-NEXT: vsll.vx v12, v12, a3, v0.t
; RV32-NEXT: vor.vv v10, v10, v12, v0.t
-; RV32-NEXT: lui a4, 4080
-; RV32-NEXT: vand.vx v12, v8, a4, v0.t
-; RV32-NEXT: vsll.vi v12, v12, 24, v0.t
-; RV32-NEXT: addi a5, sp, 8
+; RV32-NEXT: addi a4, sp, 8
; RV32-NEXT: vsetivli zero, 4, e64, m2, ta, ma
-; RV32-NEXT: vlse64.v v14, (a5), zero
+; RV32-NEXT: vlse64.v v12, (a4), zero
+; RV32-NEXT: lui a4, 4080
; RV32-NEXT: vsetvli zero, a0, e64, m2, ta, ma
-; RV32-NEXT: vand.vv v16, v8, v14, v0.t
+; RV32-NEXT: vand.vx v14, v8, a4, v0.t
+; RV32-NEXT: vsll.vi v14, v14, 24, v0.t
+; RV32-NEXT: vand.vv v16, v8, v12, v0.t
; RV32-NEXT: vsll.vi v16, v16, 8, v0.t
-; RV32-NEXT: vor.vv v12, v12, v16, v0.t
-; RV32-NEXT: vor.vv v10, v10, v12, v0.t
-; RV32-NEXT: vsrl.vx v12, v8, a1, v0.t
+; RV32-NEXT: vor.vv v14, v14, v16, v0.t
+; RV32-NEXT: vor.vv v10, v10, v14, v0.t
+; RV32-NEXT: vsrl.vx v14, v8, a1, v0.t
; RV32-NEXT: vsrl.vx v16, v8, a3, v0.t
; RV32-NEXT: vand.vx v16, v16, a2, v0.t
-; RV32-NEXT: vor.vv v12, v16, v12, v0.t
+; RV32-NEXT: vor.vv v14, v16, v14, v0.t
; RV32-NEXT: vsrl.vi v16, v8, 24, v0.t
; RV32-NEXT: vand.vx v16, v16, a4, v0.t
; RV32-NEXT: vsrl.vi v8, v8, 8, v0.t
-; RV32-NEXT: vand.vv v8, v8, v14, v0.t
+; RV32-NEXT: vand.vv v8, v8, v12, v0.t
; RV32-NEXT: vor.vv v8, v8, v16, v0.t
-; RV32-NEXT: vor.vv v8, v8, v12, v0.t
+; RV32-NEXT: vor.vv v8, v8, v14, v0.t
; RV32-NEXT: vor.vv v8, v10, v8, v0.t
; RV32-NEXT: vsrl.vi v10, v8, 4, v0.t
; RV32-NEXT: lui a1, 61681
@@ -1254,27 +1254,27 @@ define <4 x i64> @vp_bitreverse_v4i64_unmasked(<4 x i64> %va, i32 zeroext %evl)
; RV32-NEXT: li a3, 40
; RV32-NEXT: vsll.vx v12, v12, a3
; RV32-NEXT: vor.vv v10, v10, v12
-; RV32-NEXT: lui a4, 4080
-; RV32-NEXT: vand.vx v12, v8, a4
-; RV32-NEXT: vsll.vi v12, v12, 24
-; RV32-NEXT: addi a5, sp, 8
+; RV32-NEXT: addi a4, sp, 8
; RV32-NEXT: vsetivli zero, 4, e64, m2, ta, ma
-; RV32-NEXT: vlse64.v v14, (a5), zero
+; RV32-NEXT: vlse64.v v12, (a4), zero
+; RV32-NEXT: lui a4, 4080
; RV32-NEXT: vsetvli zero, a0, e64, m2, ta, ma
-; RV32-NEXT: vand.vv v16, v8, v14
+; RV32-NEXT: vand.vx v14, v8, a4
+; RV32-NEXT: vsll.vi v14, v14, 24
+; RV32-NEXT: vand.vv v16, v8, v12
; RV32-NEXT: vsll.vi v16, v16, 8
-; RV32-NEXT: vor.vv v12, v12, v16
-; RV32-NEXT: vor.vv v10, v10, v12
-; RV32-NEXT: vsrl.vx v12, v8, a1
+; RV32-NEXT: vor.vv v14, v14, v16
+; RV32-NEXT: vor.vv v10, v10, v14
+; RV32-NEXT: vsrl.vx v14, v8, a1
; RV32-NEXT: vsrl.vx v16, v8, a3
; RV32-NEXT: vand.vx v16, v16, a2
-; RV32-NEXT: vor.vv v12, v16, v12
+; RV32-NEXT: vor.vv v14, v16, v14
; RV32-NEXT: vsrl.vi v16, v8, 24
; RV32-NEXT: vand.vx v16, v16, a4
; RV32-NEXT: vsrl.vi v8, v8, 8
-; RV32-NEXT: vand.vv v8, v8, v14
+; RV32-NEXT: vand.vv v8, v8, v12
; RV32-NEXT: vor.vv v8, v8, v16
-; RV32-NEXT: vor.vv v8, v8, v12
+; RV32-NEXT: vor.vv v8, v8, v14
; RV32-NEXT: vor.vv v8, v10, v8
; RV32-NEXT: vsrl.vi v10, v8, 4
; RV32-NEXT: lui a1, 61681
@@ -1391,13 +1391,13 @@ define <8 x i64> @vp_bitreverse_v8i64(<8 x i64> %va, <8 x i1> %m, i32 zeroext %e
; RV32-NEXT: li a3, 40
; RV32-NEXT: vsll.vx v16, v16, a3, v0.t
; RV32-NEXT: vor.vv v16, v12, v16, v0.t
-; RV32-NEXT: lui a4, 4080
-; RV32-NEXT: vand.vx v12, v8, a4, v0.t
-; RV32-NEXT: vsll.vi v20, v12, 24, v0.t
-; RV32-NEXT: addi a5, sp, 8
+; RV32-NEXT: addi a4, sp, 8
; RV32-NEXT: vsetivli zero, 8, e64, m4, ta, ma
-; RV32-NEXT: vlse64.v v12, (a5), zero
+; RV32-NEXT: vlse64.v v12, (a4), zero
+; RV32-NEXT: lui a4, 4080
; RV32-NEXT: vsetvli zero, a0, e64, m4, ta, ma
+; RV32-NEXT: vand.vx v20, v8, a4, v0.t
+; RV32-NEXT: vsll.vi v20, v20, 24, v0.t
; RV32-NEXT: vand.vv v24, v8, v12, v0.t
; RV32-NEXT: vsll.vi v24, v24, 8, v0.t
; RV32-NEXT: vor.vv v20, v20, v24, v0.t
@@ -1526,27 +1526,27 @@ define <8 x i64> @vp_bitreverse_v8i64_unmasked(<8 x i64> %va, i32 zeroext %evl)
; RV32-NEXT: li a3, 40
; RV32-NEXT: vsll.vx v16, v16, a3
; RV32-NEXT: vor.vv v12, v12, v16
-; RV32-NEXT: lui a4, 4080
-; RV32-NEXT: vand.vx v16, v8, a4
-; RV32-NEXT: vsll.vi v16, v16, 24
-; RV32-NEXT: addi a5, sp, 8
+; RV32-NEXT: addi a4, sp, 8
; RV32-NEXT: vsetivli zero, 8, e64, m4, ta, ma
-; RV32-NEXT: vlse64.v v20, (a5), zero
+; RV32-NEXT: vlse64.v v16, (a4), zero
+; RV32-NEXT: lui a4, 4080
; RV32-NEXT: vsetvli zero, a0, e64, m4, ta, ma
-; RV32-NEXT: vand.vv v24, v8, v20
+; RV32-NEXT: vand.vx v20, v8, a4
+; RV32-NEXT: vsll.vi v20, v20, 24
+; RV32-NEXT: vand.vv v24, v8, v16
; RV32-NEXT: vsll.vi v24, v24, 8
-; RV32-NEXT: vor.vv v16, v16, v24
-; RV32-NEXT: vor.vv v12, v12, v16
-; RV32-NEXT: vsrl.vx v16, v8, a1
+; RV32-NEXT: vor.vv v20, v20, v24
+; RV32-NEXT: vor.vv v12, v12, v20
+; RV32-NEXT: vsrl.vx v20, v8, a1
; RV32-NEXT: vsrl.vx v24, v8, a3
; RV32-NEXT: vand.vx v24, v24, a2
-; RV32-NEXT: vor.vv v16, v24, v16
+; RV32-NEXT: vor.vv v20, v24, v20
; RV32-NEXT: vsrl.vi v24, v8, 24
; RV32-NEXT: vand.vx v24, v24, a4
; RV32-NEXT: vsrl.vi v8, v8, 8
-; RV32-NEXT: vand.vv v8, v8, v20
+; RV32-NEXT: vand.vv v8, v8, v16
; RV32-NEXT: vor.vv v8, v8, v24
-; RV32-NEXT: vor.vv v8, v8, v16
+; RV32-NEXT: vor.vv v8, v8, v20
; RV32-NEXT: vor.vv v8, v12, v8
; RV32-NEXT: vsrl.vi v12, v8, 4
; RV32-NEXT: lui a1, 61681
@@ -1685,20 +1685,23 @@ define <15 x i64> @vp_bitreverse_v15i64(<15 x i64> %va, <15 x i1> %m, i32 zeroex
; RV32-NEXT: add a4, sp, a4
; RV32-NEXT: addi a4, a4, 48
; RV32-NEXT: vs8r.v v16, (a4) # Unknown-size Folded Spill
-; RV32-NEXT: lui a4, 4080
-; RV32-NEXT: vand.vx v16, v8, a4, v0.t
-; RV32-NEXT: vsll.vi v24, v16, 24, v0.t
-; RV32-NEXT: addi a5, sp, 16
+; RV32-NEXT: addi a4, sp, 16
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a5), zero
-; RV32-NEXT: csrr a5, vlenb
-; RV32-NEXT: slli a5, a5, 3
-; RV32-NEXT: add a5, sp, a5
-; RV32-NEXT: addi a5, a5, 48
-; RV32-NEXT: vs8r.v v16, (a5) # Unknown-size Folded Spill
+; RV32-NEXT: vlse64.v v16, (a4), zero
+; RV32-NEXT: csrr a4, vlenb
+; RV32-NEXT: slli a4, a4, 3
+; RV32-NEXT: add a4, sp, a4
+; RV32-NEXT: addi a4, a4, 48
+; RV32-NEXT: vs8r.v v16, (a4) # Unknown-size Folded Spill
+; RV32-NEXT: lui a4, 4080
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v8, v16, v0.t
-; RV32-NEXT: vsll.vi v16, v16, 8, v0.t
+; RV32-NEXT: vand.vx v24, v8, a4, v0.t
+; RV32-NEXT: vsll.vi v24, v24, 24, v0.t
+; RV32-NEXT: addi a5, sp, 48
+; RV32-NEXT: vs8r.v v24, (a5) # Unknown-size Folded Spill
+; RV32-NEXT: vand.vv v24, v8, v16, v0.t
+; RV32-NEXT: vsll.vi v16, v24, 8, v0.t
+; RV32-NEXT: vl8r.v v24, (a5) # Unknown-size Folded Reload
; RV32-NEXT: vor.vv v16, v24, v16, v0.t
; RV32-NEXT: csrr a5, vlenb
; RV32-NEXT: slli a5, a5, 4
@@ -1711,10 +1714,10 @@ define <15 x i64> @vp_bitreverse_v15i64(<15 x i64> %va, <15 x i1> %m, i32 zeroex
; RV32-NEXT: add a5, sp, a5
; RV32-NEXT: addi a5, a5, 48
; RV32-NEXT: vs8r.v v16, (a5) # Unknown-size Folded Spill
-; RV32-NEXT: vsrl.vx v24, v8, a1, v0.t
-; RV32-NEXT: vsrl.vx v16, v8, a3, v0.t
-; RV32-NEXT: vand.vx v16, v16, a2, v0.t
-; RV32-NEXT: vor.vv v16, v16, v24, v0.t
+; RV32-NEXT: vsrl.vx v16, v8, a1, v0.t
+; RV32-NEXT: vsrl.vx v24, v8, a3, v0.t
+; RV32-NEXT: vand.vx v24, v24, a2, v0.t
+; RV32-NEXT: vor.vv v16, v24, v16, v0.t
; RV32-NEXT: addi a1, sp, 48
; RV32-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
; RV32-NEXT: vsrl.vi v24, v8, 24, v0.t
@@ -1727,38 +1730,38 @@ define <15 x i64> @vp_bitreverse_v15i64(<15 x i64> %va, <15 x i1> %m, i32 zeroex
; RV32-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
; RV32-NEXT: vand.vv v8, v8, v16, v0.t
; RV32-NEXT: vor.vv v8, v8, v24, v0.t
+; RV32-NEXT: addi a1, sp, 40
+; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v24, (a1), zero
; RV32-NEXT: addi a1, sp, 48
; RV32-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vor.vv v8, v8, v16, v0.t
; RV32-NEXT: csrr a1, vlenb
; RV32-NEXT: slli a1, a1, 4
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 48
; RV32-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
-; RV32-NEXT: vor.vv v8, v16, v8, v0.t
-; RV32-NEXT: vsrl.vi v16, v8, 4, v0.t
-; RV32-NEXT: addi a1, sp, 40
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v24, (a1), zero
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v16, v24, v0.t
+; RV32-NEXT: vor.vv v16, v16, v8, v0.t
+; RV32-NEXT: vsrl.vi v8, v16, 4, v0.t
; RV32-NEXT: vand.vv v8, v8, v24, v0.t
-; RV32-NEXT: vsll.vi v8, v8, 4, v0.t
-; RV32-NEXT: vor.vv v8, v16, v8, v0.t
-; RV32-NEXT: vsrl.vi v16, v8, 2, v0.t
+; RV32-NEXT: vand.vv v16, v16, v24, v0.t
; RV32-NEXT: addi a1, sp, 32
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV32-NEXT: vlse64.v v24, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v16, v24, v0.t
+; RV32-NEXT: vsll.vi v16, v16, 4, v0.t
+; RV32-NEXT: vor.vv v16, v8, v16, v0.t
+; RV32-NEXT: vsrl.vi v8, v16, 2, v0.t
; RV32-NEXT: vand.vv v8, v8, v24, v0.t
-; RV32-NEXT: vsll.vi v8, v8, 2, v0.t
-; RV32-NEXT: vor.vv v8, v16, v8, v0.t
-; RV32-NEXT: vsrl.vi v16, v8, 1, v0.t
+; RV32-NEXT: vand.vv v16, v16, v24, v0.t
; RV32-NEXT: addi a1, sp, 24
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV32-NEXT: vlse64.v v24, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vsll.vi v16, v16, 2, v0.t
+; RV32-NEXT: vor.vv v8, v8, v16, v0.t
+; RV32-NEXT: vsrl.vi v16, v8, 1, v0.t
; RV32-NEXT: vand.vv v16, v16, v24, v0.t
; RV32-NEXT: vand.vv v8, v8, v24, v0.t
; RV32-NEXT: vsll.vi v8, v8, 1, v0.t
@@ -1885,60 +1888,60 @@ define <15 x i64> @vp_bitreverse_v15i64_unmasked(<15 x i64> %va, i32 zeroext %ev
; RV32-NEXT: vor.vv v16, v16, v24
; RV32-NEXT: addi a4, sp, 48
; RV32-NEXT: vs8r.v v16, (a4) # Unknown-size Folded Spill
-; RV32-NEXT: lui a4, 4080
-; RV32-NEXT: vand.vx v16, v8, a4
-; RV32-NEXT: vsll.vi v0, v16, 24
-; RV32-NEXT: addi a5, sp, 16
+; RV32-NEXT: addi a4, sp, 16
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a5), zero
+; RV32-NEXT: vlse64.v v24, (a4), zero
+; RV32-NEXT: lui a4, 4080
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v24, v8, v16
-; RV32-NEXT: vsll.vi v24, v24, 8
-; RV32-NEXT: vor.vv v24, v0, v24
+; RV32-NEXT: vand.vx v0, v8, a4
+; RV32-NEXT: vsll.vi v0, v0, 24
+; RV32-NEXT: vand.vv v16, v8, v24
+; RV32-NEXT: vsll.vi v16, v16, 8
+; RV32-NEXT: vor.vv v16, v0, v16
; RV32-NEXT: addi a5, sp, 48
; RV32-NEXT: vl8r.v v0, (a5) # Unknown-size Folded Reload
-; RV32-NEXT: vor.vv v24, v0, v24
-; RV32-NEXT: vs8r.v v24, (a5) # Unknown-size Folded Spill
+; RV32-NEXT: vor.vv v16, v0, v16
+; RV32-NEXT: vs8r.v v16, (a5) # Unknown-size Folded Spill
; RV32-NEXT: vsrl.vx v0, v8, a3
; RV32-NEXT: vand.vx v0, v0, a2
-; RV32-NEXT: vsrl.vx v24, v8, a1
-; RV32-NEXT: vor.vv v24, v0, v24
-; RV32-NEXT: vsrl.vi v0, v8, 8
-; RV32-NEXT: vand.vv v16, v0, v16
+; RV32-NEXT: vsrl.vx v16, v8, a1
+; RV32-NEXT: vor.vv v0, v0, v16
+; RV32-NEXT: vsrl.vi v16, v8, 8
+; RV32-NEXT: vand.vv v16, v16, v24
; RV32-NEXT: vsrl.vi v8, v8, 24
; RV32-NEXT: vand.vx v8, v8, a4
; RV32-NEXT: vor.vv v8, v16, v8
-; RV32-NEXT: vor.vv v8, v8, v24
-; RV32-NEXT: addi a1, sp, 48
-; RV32-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
-; RV32-NEXT: vor.vv v8, v16, v8
-; RV32-NEXT: vsrl.vi v16, v8, 4
; RV32-NEXT: addi a1, sp, 40
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v24, (a1), zero
+; RV32-NEXT: vlse64.v v16, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v16, v24
-; RV32-NEXT: vand.vv v8, v8, v24
-; RV32-NEXT: vsll.vi v8, v8, 4
-; RV32-NEXT: vor.vv v8, v16, v8
-; RV32-NEXT: vsrl.vi v16, v8, 2
+; RV32-NEXT: vor.vv v8, v8, v0
+; RV32-NEXT: addi a1, sp, 48
+; RV32-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vor.vv v8, v24, v8
+; RV32-NEXT: vsrl.vi v24, v8, 4
+; RV32-NEXT: vand.vv v24, v24, v16
+; RV32-NEXT: vand.vv v8, v8, v16
; RV32-NEXT: addi a1, sp, 32
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v24, (a1), zero
+; RV32-NEXT: vlse64.v v16, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v16, v24
-; RV32-NEXT: vand.vv v8, v8, v24
-; RV32-NEXT: vsll.vi v8, v8, 2
-; RV32-NEXT: vor.vv v8, v16, v8
-; RV32-NEXT: vsrl.vi v16, v8, 1
+; RV32-NEXT: vsll.vi v8, v8, 4
+; RV32-NEXT: vor.vv v8, v24, v8
+; RV32-NEXT: vsrl.vi v24, v8, 2
+; RV32-NEXT: vand.vv v24, v24, v16
+; RV32-NEXT: vand.vv v8, v8, v16
; RV32-NEXT: addi a1, sp, 24
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v24, (a1), zero
+; RV32-NEXT: vlse64.v v16, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v16, v24
-; RV32-NEXT: vand.vv v8, v8, v24
+; RV32-NEXT: vsll.vi v8, v8, 2
+; RV32-NEXT: vor.vv v8, v24, v8
+; RV32-NEXT: vsrl.vi v24, v8, 1
+; RV32-NEXT: vand.vv v24, v24, v16
+; RV32-NEXT: vand.vv v8, v8, v16
; RV32-NEXT: vadd.vv v8, v8, v8
-; RV32-NEXT: vor.vv v8, v16, v8
+; RV32-NEXT: vor.vv v8, v24, v8
; RV32-NEXT: csrr a0, vlenb
; RV32-NEXT: slli a0, a0, 3
; RV32-NEXT: add sp, sp, a0
@@ -2049,20 +2052,23 @@ define <16 x i64> @vp_bitreverse_v16i64(<16 x i64> %va, <16 x i1> %m, i32 zeroex
; RV32-NEXT: add a4, sp, a4
; RV32-NEXT: addi a4, a4, 48
; RV32-NEXT: vs8r.v v16, (a4) # Unknown-size Folded Spill
-; RV32-NEXT: lui a4, 4080
-; RV32-NEXT: vand.vx v16, v8, a4, v0.t
-; RV32-NEXT: vsll.vi v24, v16, 24, v0.t
-; RV32-NEXT: addi a5, sp, 16
+; RV32-NEXT: addi a4, sp, 16
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a5), zero
-; RV32-NEXT: csrr a5, vlenb
-; RV32-NEXT: slli a5, a5, 3
-; RV32-NEXT: add a5, sp, a5
-; RV32-NEXT: addi a5, a5, 48
-; RV32-NEXT: vs8r.v v16, (a5) # Unknown-size Folded Spill
+; RV32-NEXT: vlse64.v v16, (a4), zero
+; RV32-NEXT: csrr a4, vlenb
+; RV32-NEXT: slli a4, a4, 3
+; RV32-NEXT: add a4, sp, a4
+; RV32-NEXT: addi a4, a4, 48
+; RV32-NEXT: vs8r.v v16, (a4) # Unknown-size Folded Spill
+; RV32-NEXT: lui a4, 4080
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v8, v16, v0.t
-; RV32-NEXT: vsll.vi v16, v16, 8, v0.t
+; RV32-NEXT: vand.vx v24, v8, a4, v0.t
+; RV32-NEXT: vsll.vi v24, v24, 24, v0.t
+; RV32-NEXT: addi a5, sp, 48
+; RV32-NEXT: vs8r.v v24, (a5) # Unknown-size Folded Spill
+; RV32-NEXT: vand.vv v24, v8, v16, v0.t
+; RV32-NEXT: vsll.vi v16, v24, 8, v0.t
+; RV32-NEXT: vl8r.v v24, (a5) # Unknown-size Folded Reload
; RV32-NEXT: vor.vv v16, v24, v16, v0.t
; RV32-NEXT: csrr a5, vlenb
; RV32-NEXT: slli a5, a5, 4
@@ -2075,10 +2081,10 @@ define <16 x i64> @vp_bitreverse_v16i64(<16 x i64> %va, <16 x i1> %m, i32 zeroex
; RV32-NEXT: add a5, sp, a5
; RV32-NEXT: addi a5, a5, 48
; RV32-NEXT: vs8r.v v16, (a5) # Unknown-size Folded Spill
-; RV32-NEXT: vsrl.vx v24, v8, a1, v0.t
-; RV32-NEXT: vsrl.vx v16, v8, a3, v0.t
-; RV32-NEXT: vand.vx v16, v16, a2, v0.t
-; RV32-NEXT: vor.vv v16, v16, v24, v0.t
+; RV32-NEXT: vsrl.vx v16, v8, a1, v0.t
+; RV32-NEXT: vsrl.vx v24, v8, a3, v0.t
+; RV32-NEXT: vand.vx v24, v24, a2, v0.t
+; RV32-NEXT: vor.vv v16, v24, v16, v0.t
; RV32-NEXT: addi a1, sp, 48
; RV32-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
; RV32-NEXT: vsrl.vi v24, v8, 24, v0.t
@@ -2091,38 +2097,38 @@ define <16 x i64> @vp_bitreverse_v16i64(<16 x i64> %va, <16 x i1> %m, i32 zeroex
; RV32-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
; RV32-NEXT: vand.vv v8, v8, v16, v0.t
; RV32-NEXT: vor.vv v8, v8, v24, v0.t
+; RV32-NEXT: addi a1, sp, 40
+; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v24, (a1), zero
; RV32-NEXT: addi a1, sp, 48
; RV32-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vor.vv v8, v8, v16, v0.t
; RV32-NEXT: csrr a1, vlenb
; RV32-NEXT: slli a1, a1, 4
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 48
; RV32-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
-; RV32-NEXT: vor.vv v8, v16, v8, v0.t
-; RV32-NEXT: vsrl.vi v16, v8, 4, v0.t
-; RV32-NEXT: addi a1, sp, 40
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v24, (a1), zero
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v16, v24, v0.t
+; RV32-NEXT: vor.vv v16, v16, v8, v0.t
+; RV32-NEXT: vsrl.vi v8, v16, 4, v0.t
; RV32-NEXT: vand.vv v8, v8, v24, v0.t
-; RV32-NEXT: vsll.vi v8, v8, 4, v0.t
-; RV32-NEXT: vor.vv v8, v16, v8, v0.t
-; RV32-NEXT: vsrl.vi v16, v8, 2, v0.t
+; RV32-NEXT: vand.vv v16, v16, v24, v0.t
; RV32-NEXT: addi a1, sp, 32
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV32-NEXT: vlse64.v v24, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v16, v24, v0.t
+; RV32-NEXT: vsll.vi v16, v16, 4, v0.t
+; RV32-NEXT: vor.vv v16, v8, v16, v0.t
+; RV32-NEXT: vsrl.vi v8, v16, 2, v0.t
; RV32-NEXT: vand.vv v8, v8, v24, v0.t
-; RV32-NEXT: vsll.vi v8, v8, 2, v0.t
-; RV32-NEXT: vor.vv v8, v16, v8, v0.t
-; RV32-NEXT: vsrl.vi v16, v8, 1, v0.t
+; RV32-NEXT: vand.vv v16, v16, v24, v0.t
; RV32-NEXT: addi a1, sp, 24
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV32-NEXT: vlse64.v v24, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vsll.vi v16, v16, 2, v0.t
+; RV32-NEXT: vor.vv v8, v8, v16, v0.t
+; RV32-NEXT: vsrl.vi v16, v8, 1, v0.t
; RV32-NEXT: vand.vv v16, v16, v24, v0.t
; RV32-NEXT: vand.vv v8, v8, v24, v0.t
; RV32-NEXT: vsll.vi v8, v8, 1, v0.t
@@ -2249,60 +2255,60 @@ define <16 x i64> @vp_bitreverse_v16i64_unmasked(<16 x i64> %va, i32 zeroext %ev
; RV32-NEXT: vor.vv v16, v16, v24
; RV32-NEXT: addi a4, sp, 48
; RV32-NEXT: vs8r.v v16, (a4) # Unknown-size Folded Spill
-; RV32-NEXT: lui a4, 4080
-; RV32-NEXT: vand.vx v16, v8, a4
-; RV32-NEXT: vsll.vi v0, v16, 24
-; RV32-NEXT: addi a5, sp, 16
+; RV32-NEXT: addi a4, sp, 16
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a5), zero
+; RV32-NEXT: vlse64.v v24, (a4), zero
+; RV32-NEXT: lui a4, 4080
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v24, v8, v16
-; RV32-NEXT: vsll.vi v24, v24, 8
-; RV32-NEXT: vor.vv v24, v0, v24
+; RV32-NEXT: vand.vx v0, v8, a4
+; RV32-NEXT: vsll.vi v0, v0, 24
+; RV32-NEXT: vand.vv v16, v8, v24
+; RV32-NEXT: vsll.vi v16, v16, 8
+; RV32-NEXT: vor.vv v16, v0, v16
; RV32-NEXT: addi a5, sp, 48
; RV32-NEXT: vl8r.v v0, (a5) # Unknown-size Folded Reload
-; RV32-NEXT: vor.vv v24, v0, v24
-; RV32-NEXT: vs8r.v v24, (a5) # Unknown-size Folded Spill
+; RV32-NEXT: vor.vv v16, v0, v16
+; RV32-NEXT: vs8r.v v16, (a5) # Unknown-size Folded Spill
; RV32-NEXT: vsrl.vx v0, v8, a3
; RV32-NEXT: vand.vx v0, v0, a2
-; RV32-NEXT: vsrl.vx v24, v8, a1
-; RV32-NEXT: vor.vv v24, v0, v24
-; RV32-NEXT: vsrl.vi v0, v8, 8
-; RV32-NEXT: vand.vv v16, v0, v16
+; RV32-NEXT: vsrl.vx v16, v8, a1
+; RV32-NEXT: vor.vv v0, v0, v16
+; RV32-NEXT: vsrl.vi v16, v8, 8
+; RV32-NEXT: vand.vv v16, v16, v24
; RV32-NEXT: vsrl.vi v8, v8, 24
; RV32-NEXT: vand.vx v8, v8, a4
; RV32-NEXT: vor.vv v8, v16, v8
-; RV32-NEXT: vor.vv v8, v8, v24
-; RV32-NEXT: addi a1, sp, 48
-; RV32-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
-; RV32-NEXT: vor.vv v8, v16, v8
-; RV32-NEXT: vsrl.vi v16, v8, 4
; RV32-NEXT: addi a1, sp, 40
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v24, (a1), zero
+; RV32-NEXT: vlse64.v v16, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v16, v24
-; RV32-NEXT: vand.vv v8, v8, v24
-; RV32-NEXT: vsll.vi v8, v8, 4
-; RV32-NEXT: vor.vv v8, v16, v8
-; RV32-NEXT: vsrl.vi v16, v8, 2
+; RV32-NEXT: vor.vv v8, v8, v0
+; RV32-NEXT: addi a1, sp, 48
+; RV32-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vor.vv v8, v24, v8
+; RV32-NEXT: vsrl.vi v24, v8, 4
+; RV32-NEXT: vand.vv v24, v24, v16
+; RV32-NEXT: vand.vv v8, v8, v16
; RV32-NEXT: addi a1, sp, 32
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v24, (a1), zero
+; RV32-NEXT: vlse64.v v16, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v16, v24
-; RV32-NEXT: vand.vv v8, v8, v24
-; RV32-NEXT: vsll.vi v8, v8, 2
-; RV32-NEXT: vor.vv v8, v16, v8
-; RV32-NEXT: vsrl.vi v16, v8, 1
+; RV32-NEXT: vsll.vi v8, v8, 4
+; RV32-NEXT: vor.vv v8, v24, v8
+; RV32-NEXT: vsrl.vi v24, v8, 2
+; RV32-NEXT: vand.vv v24, v24, v16
+; RV32-NEXT: vand.vv v8, v8, v16
; RV32-NEXT: addi a1, sp, 24
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v24, (a1), zero
+; RV32-NEXT: vlse64.v v16, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v16, v24
-; RV32-NEXT: vand.vv v8, v8, v24
+; RV32-NEXT: vsll.vi v8, v8, 2
+; RV32-NEXT: vor.vv v8, v24, v8
+; RV32-NEXT: vsrl.vi v24, v8, 1
+; RV32-NEXT: vand.vv v24, v24, v16
+; RV32-NEXT: vand.vv v8, v8, v16
; RV32-NEXT: vadd.vv v8, v8, v8
-; RV32-NEXT: vor.vv v8, v16, v8
+; RV32-NEXT: vor.vv v8, v24, v8
; RV32-NEXT: csrr a0, vlenb
; RV32-NEXT: slli a0, a0, 3
; RV32-NEXT: add sp, sp, a0
@@ -2388,8 +2394,8 @@ define <128 x i16> @vp_bitreverse_v128i16(<128 x i16> %va, <128 x i1> %m, i32 ze
; CHECK-NEXT: add a1, sp, a1
; CHECK-NEXT: addi a1, a1, 16
; CHECK-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
-; CHECK-NEXT: vsetivli zero, 8, e8, m1, ta, ma
; CHECK-NEXT: li a2, 64
+; CHECK-NEXT: vsetivli zero, 8, e8, m1, ta, ma
; CHECK-NEXT: vslidedown.vi v24, v0, 8
; CHECK-NEXT: mv a1, a0
; CHECK-NEXT: bltu a0, a2, .LBB34_2
@@ -2427,13 +2433,13 @@ define <128 x i16> @vp_bitreverse_v128i16(<128 x i16> %va, <128 x i1> %m, i32 ze
; CHECK-NEXT: sltu a0, a0, a4
; CHECK-NEXT: addi a0, a0, -1
; CHECK-NEXT: and a0, a0, a4
-; CHECK-NEXT: vsetvli zero, a0, e16, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: csrr a4, vlenb
+; CHECK-NEXT: slli a4, a4, 3
+; CHECK-NEXT: add a4, sp, a4
+; CHECK-NEXT: addi a4, a4, 16
+; CHECK-NEXT: vl8r.v v8, (a4) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, a0, e16, m8, ta, ma
; CHECK-NEXT: vsrl.vi v16, v8, 8, v0.t
; CHECK-NEXT: vsll.vi v8, v8, 8, v0.t
; CHECK-NEXT: vor.vv v8, v8, v16, v0.t
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-bswap-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-bswap-vp.ll
index f80d4e5c0d7c3..1490738687322 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-bswap-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-bswap-vp.ll
@@ -295,27 +295,27 @@ define <2 x i64> @vp_bswap_v2i64(<2 x i64> %va, <2 x i1> %m, i32 zeroext %evl) {
; RV32-NEXT: li a3, 40
; RV32-NEXT: vsll.vx v10, v10, a3, v0.t
; RV32-NEXT: vor.vv v9, v9, v10, v0.t
-; RV32-NEXT: lui a4, 4080
-; RV32-NEXT: vand.vx v10, v8, a4, v0.t
-; RV32-NEXT: vsll.vi v10, v10, 24, v0.t
-; RV32-NEXT: addi a5, sp, 8
+; RV32-NEXT: addi a4, sp, 8
; RV32-NEXT: vsetivli zero, 2, e64, m1, ta, ma
-; RV32-NEXT: vlse64.v v11, (a5), zero
+; RV32-NEXT: vlse64.v v10, (a4), zero
+; RV32-NEXT: lui a4, 4080
; RV32-NEXT: vsetvli zero, a0, e64, m1, ta, ma
-; RV32-NEXT: vand.vv v12, v8, v11, v0.t
+; RV32-NEXT: vand.vx v11, v8, a4, v0.t
+; RV32-NEXT: vsll.vi v11, v11, 24, v0.t
+; RV32-NEXT: vand.vv v12, v8, v10, v0.t
; RV32-NEXT: vsll.vi v12, v12, 8, v0.t
-; RV32-NEXT: vor.vv v10, v10, v12, v0.t
-; RV32-NEXT: vor.vv v9, v9, v10, v0.t
-; RV32-NEXT: vsrl.vx v10, v8, a1, v0.t
+; RV32-NEXT: vor.vv v11, v11, v12, v0.t
+; RV32-NEXT: vor.vv v9, v9, v11, v0.t
+; RV32-NEXT: vsrl.vx v11, v8, a1, v0.t
; RV32-NEXT: vsrl.vx v12, v8, a3, v0.t
; RV32-NEXT: vand.vx v12, v12, a2, v0.t
-; RV32-NEXT: vor.vv v10, v12, v10, v0.t
+; RV32-NEXT: vor.vv v11, v12, v11, v0.t
; RV32-NEXT: vsrl.vi v12, v8, 24, v0.t
; RV32-NEXT: vand.vx v12, v12, a4, v0.t
; RV32-NEXT: vsrl.vi v8, v8, 8, v0.t
-; RV32-NEXT: vand.vv v8, v8, v11, v0.t
+; RV32-NEXT: vand.vv v8, v8, v10, v0.t
; RV32-NEXT: vor.vv v8, v8, v12, v0.t
-; RV32-NEXT: vor.vv v8, v8, v10, v0.t
+; RV32-NEXT: vor.vv v8, v8, v11, v0.t
; RV32-NEXT: vor.vv v8, v9, v8, v0.t
; RV32-NEXT: addi sp, sp, 16
; RV32-NEXT: ret
@@ -373,27 +373,27 @@ define <2 x i64> @vp_bswap_v2i64_unmasked(<2 x i64> %va, i32 zeroext %evl) {
; RV32-NEXT: li a3, 40
; RV32-NEXT: vsll.vx v10, v10, a3
; RV32-NEXT: vor.vv v9, v9, v10
-; RV32-NEXT: lui a4, 4080
-; RV32-NEXT: vand.vx v10, v8, a4
-; RV32-NEXT: vsll.vi v10, v10, 24
-; RV32-NEXT: addi a5, sp, 8
+; RV32-NEXT: addi a4, sp, 8
; RV32-NEXT: vsetivli zero, 2, e64, m1, ta, ma
-; RV32-NEXT: vlse64.v v11, (a5), zero
+; RV32-NEXT: vlse64.v v10, (a4), zero
+; RV32-NEXT: lui a4, 4080
; RV32-NEXT: vsetvli zero, a0, e64, m1, ta, ma
-; RV32-NEXT: vand.vv v12, v8, v11
+; RV32-NEXT: vand.vx v11, v8, a4
+; RV32-NEXT: vsll.vi v11, v11, 24
+; RV32-NEXT: vand.vv v12, v8, v10
; RV32-NEXT: vsll.vi v12, v12, 8
-; RV32-NEXT: vor.vv v10, v10, v12
-; RV32-NEXT: vor.vv v9, v9, v10
-; RV32-NEXT: vsrl.vx v10, v8, a1
+; RV32-NEXT: vor.vv v11, v11, v12
+; RV32-NEXT: vor.vv v9, v9, v11
+; RV32-NEXT: vsrl.vx v11, v8, a1
; RV32-NEXT: vsrl.vx v12, v8, a3
; RV32-NEXT: vand.vx v12, v12, a2
-; RV32-NEXT: vor.vv v10, v12, v10
+; RV32-NEXT: vor.vv v11, v12, v11
; RV32-NEXT: vsrl.vi v12, v8, 24
; RV32-NEXT: vand.vx v12, v12, a4
; RV32-NEXT: vsrl.vi v8, v8, 8
-; RV32-NEXT: vand.vv v8, v8, v11
+; RV32-NEXT: vand.vv v8, v8, v10
; RV32-NEXT: vor.vv v8, v8, v12
-; RV32-NEXT: vor.vv v8, v8, v10
+; RV32-NEXT: vor.vv v8, v8, v11
; RV32-NEXT: vor.vv v8, v9, v8
; RV32-NEXT: addi sp, sp, 16
; RV32-NEXT: ret
@@ -453,27 +453,27 @@ define <4 x i64> @vp_bswap_v4i64(<4 x i64> %va, <4 x i1> %m, i32 zeroext %evl) {
; RV32-NEXT: li a3, 40
; RV32-NEXT: vsll.vx v12, v12, a3, v0.t
; RV32-NEXT: vor.vv v10, v10, v12, v0.t
-; RV32-NEXT: lui a4, 4080
-; RV32-NEXT: vand.vx v12, v8, a4, v0.t
-; RV32-NEXT: vsll.vi v12, v12, 24, v0.t
-; RV32-NEXT: addi a5, sp, 8
+; RV32-NEXT: addi a4, sp, 8
; RV32-NEXT: vsetivli zero, 4, e64, m2, ta, ma
-; RV32-NEXT: vlse64.v v14, (a5), zero
+; RV32-NEXT: vlse64.v v12, (a4), zero
+; RV32-NEXT: lui a4, 4080
; RV32-NEXT: vsetvli zero, a0, e64, m2, ta, ma
-; RV32-NEXT: vand.vv v16, v8, v14, v0.t
+; RV32-NEXT: vand.vx v14, v8, a4, v0.t
+; RV32-NEXT: vsll.vi v14, v14, 24, v0.t
+; RV32-NEXT: vand.vv v16, v8, v12, v0.t
; RV32-NEXT: vsll.vi v16, v16, 8, v0.t
-; RV32-NEXT: vor.vv v12, v12, v16, v0.t
-; RV32-NEXT: vor.vv v10, v10, v12, v0.t
-; RV32-NEXT: vsrl.vx v12, v8, a1, v0.t
+; RV32-NEXT: vor.vv v14, v14, v16, v0.t
+; RV32-NEXT: vor.vv v10, v10, v14, v0.t
+; RV32-NEXT: vsrl.vx v14, v8, a1, v0.t
; RV32-NEXT: vsrl.vx v16, v8, a3, v0.t
; RV32-NEXT: vand.vx v16, v16, a2, v0.t
-; RV32-NEXT: vor.vv v12, v16, v12, v0.t
+; RV32-NEXT: vor.vv v14, v16, v14, v0.t
; RV32-NEXT: vsrl.vi v16, v8, 24, v0.t
; RV32-NEXT: vand.vx v16, v16, a4, v0.t
; RV32-NEXT: vsrl.vi v8, v8, 8, v0.t
-; RV32-NEXT: vand.vv v8, v8, v14, v0.t
+; RV32-NEXT: vand.vv v8, v8, v12, v0.t
; RV32-NEXT: vor.vv v8, v8, v16, v0.t
-; RV32-NEXT: vor.vv v8, v8, v12, v0.t
+; RV32-NEXT: vor.vv v8, v8, v14, v0.t
; RV32-NEXT: vor.vv v8, v10, v8, v0.t
; RV32-NEXT: addi sp, sp, 16
; RV32-NEXT: ret
@@ -531,27 +531,27 @@ define <4 x i64> @vp_bswap_v4i64_unmasked(<4 x i64> %va, i32 zeroext %evl) {
; RV32-NEXT: li a3, 40
; RV32-NEXT: vsll.vx v12, v12, a3
; RV32-NEXT: vor.vv v10, v10, v12
-; RV32-NEXT: lui a4, 4080
-; RV32-NEXT: vand.vx v12, v8, a4
-; RV32-NEXT: vsll.vi v12, v12, 24
-; RV32-NEXT: addi a5, sp, 8
+; RV32-NEXT: addi a4, sp, 8
; RV32-NEXT: vsetivli zero, 4, e64, m2, ta, ma
-; RV32-NEXT: vlse64.v v14, (a5), zero
+; RV32-NEXT: vlse64.v v12, (a4), zero
+; RV32-NEXT: lui a4, 4080
; RV32-NEXT: vsetvli zero, a0, e64, m2, ta, ma
-; RV32-NEXT: vand.vv v16, v8, v14
+; RV32-NEXT: vand.vx v14, v8, a4
+; RV32-NEXT: vsll.vi v14, v14, 24
+; RV32-NEXT: vand.vv v16, v8, v12
; RV32-NEXT: vsll.vi v16, v16, 8
-; RV32-NEXT: vor.vv v12, v12, v16
-; RV32-NEXT: vor.vv v10, v10, v12
-; RV32-NEXT: vsrl.vx v12, v8, a1
+; RV32-NEXT: vor.vv v14, v14, v16
+; RV32-NEXT: vor.vv v10, v10, v14
+; RV32-NEXT: vsrl.vx v14, v8, a1
; RV32-NEXT: vsrl.vx v16, v8, a3
; RV32-NEXT: vand.vx v16, v16, a2
-; RV32-NEXT: vor.vv v12, v16, v12
+; RV32-NEXT: vor.vv v14, v16, v14
; RV32-NEXT: vsrl.vi v16, v8, 24
; RV32-NEXT: vand.vx v16, v16, a4
; RV32-NEXT: vsrl.vi v8, v8, 8
-; RV32-NEXT: vand.vv v8, v8, v14
+; RV32-NEXT: vand.vv v8, v8, v12
; RV32-NEXT: vor.vv v8, v8, v16
-; RV32-NEXT: vor.vv v8, v8, v12
+; RV32-NEXT: vor.vv v8, v8, v14
; RV32-NEXT: vor.vv v8, v10, v8
; RV32-NEXT: addi sp, sp, 16
; RV32-NEXT: ret
@@ -611,13 +611,13 @@ define <8 x i64> @vp_bswap_v8i64(<8 x i64> %va, <8 x i1> %m, i32 zeroext %evl) {
; RV32-NEXT: li a3, 40
; RV32-NEXT: vsll.vx v16, v16, a3, v0.t
; RV32-NEXT: vor.vv v16, v12, v16, v0.t
-; RV32-NEXT: lui a4, 4080
-; RV32-NEXT: vand.vx v12, v8, a4, v0.t
-; RV32-NEXT: vsll.vi v20, v12, 24, v0.t
-; RV32-NEXT: addi a5, sp, 8
+; RV32-NEXT: addi a4, sp, 8
; RV32-NEXT: vsetivli zero, 8, e64, m4, ta, ma
-; RV32-NEXT: vlse64.v v12, (a5), zero
+; RV32-NEXT: vlse64.v v12, (a4), zero
+; RV32-NEXT: lui a4, 4080
; RV32-NEXT: vsetvli zero, a0, e64, m4, ta, ma
+; RV32-NEXT: vand.vx v20, v8, a4, v0.t
+; RV32-NEXT: vsll.vi v20, v20, 24, v0.t
; RV32-NEXT: vand.vv v24, v8, v12, v0.t
; RV32-NEXT: vsll.vi v24, v24, 8, v0.t
; RV32-NEXT: vor.vv v20, v20, v24, v0.t
@@ -689,27 +689,27 @@ define <8 x i64> @vp_bswap_v8i64_unmasked(<8 x i64> %va, i32 zeroext %evl) {
; RV32-NEXT: li a3, 40
; RV32-NEXT: vsll.vx v16, v16, a3
; RV32-NEXT: vor.vv v12, v12, v16
-; RV32-NEXT: lui a4, 4080
-; RV32-NEXT: vand.vx v16, v8, a4
-; RV32-NEXT: vsll.vi v16, v16, 24
-; RV32-NEXT: addi a5, sp, 8
+; RV32-NEXT: addi a4, sp, 8
; RV32-NEXT: vsetivli zero, 8, e64, m4, ta, ma
-; RV32-NEXT: vlse64.v v20, (a5), zero
+; RV32-NEXT: vlse64.v v16, (a4), zero
+; RV32-NEXT: lui a4, 4080
; RV32-NEXT: vsetvli zero, a0, e64, m4, ta, ma
-; RV32-NEXT: vand.vv v24, v8, v20
+; RV32-NEXT: vand.vx v20, v8, a4
+; RV32-NEXT: vsll.vi v20, v20, 24
+; RV32-NEXT: vand.vv v24, v8, v16
; RV32-NEXT: vsll.vi v24, v24, 8
-; RV32-NEXT: vor.vv v16, v16, v24
-; RV32-NEXT: vor.vv v12, v12, v16
-; RV32-NEXT: vsrl.vx v16, v8, a1
+; RV32-NEXT: vor.vv v20, v20, v24
+; RV32-NEXT: vor.vv v12, v12, v20
+; RV32-NEXT: vsrl.vx v20, v8, a1
; RV32-NEXT: vsrl.vx v24, v8, a3
; RV32-NEXT: vand.vx v24, v24, a2
-; RV32-NEXT: vor.vv v16, v24, v16
+; RV32-NEXT: vor.vv v20, v24, v20
; RV32-NEXT: vsrl.vi v24, v8, 24
; RV32-NEXT: vand.vx v24, v24, a4
; RV32-NEXT: vsrl.vi v8, v8, 8
-; RV32-NEXT: vand.vv v8, v8, v20
+; RV32-NEXT: vand.vv v8, v8, v16
; RV32-NEXT: vor.vv v8, v8, v24
-; RV32-NEXT: vor.vv v8, v8, v16
+; RV32-NEXT: vor.vv v8, v8, v20
; RV32-NEXT: vor.vv v8, v12, v8
; RV32-NEXT: addi sp, sp, 16
; RV32-NEXT: ret
@@ -779,20 +779,23 @@ define <15 x i64> @vp_bswap_v15i64(<15 x i64> %va, <15 x i1> %m, i32 zeroext %ev
; RV32-NEXT: add a4, sp, a4
; RV32-NEXT: addi a4, a4, 16
; RV32-NEXT: vs8r.v v16, (a4) # Unknown-size Folded Spill
-; RV32-NEXT: lui a4, 4080
-; RV32-NEXT: vand.vx v16, v8, a4, v0.t
-; RV32-NEXT: vsll.vi v24, v16, 24, v0.t
-; RV32-NEXT: addi a5, sp, 8
+; RV32-NEXT: addi a4, sp, 8
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a5), zero
-; RV32-NEXT: csrr a5, vlenb
-; RV32-NEXT: slli a5, a5, 3
-; RV32-NEXT: add a5, sp, a5
-; RV32-NEXT: addi a5, a5, 16
-; RV32-NEXT: vs8r.v v16, (a5) # Unknown-size Folded Spill
+; RV32-NEXT: vlse64.v v16, (a4), zero
+; RV32-NEXT: csrr a4, vlenb
+; RV32-NEXT: slli a4, a4, 3
+; RV32-NEXT: add a4, sp, a4
+; RV32-NEXT: addi a4, a4, 16
+; RV32-NEXT: vs8r.v v16, (a4) # Unknown-size Folded Spill
+; RV32-NEXT: lui a4, 4080
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v8, v16, v0.t
-; RV32-NEXT: vsll.vi v16, v16, 8, v0.t
+; RV32-NEXT: vand.vx v24, v8, a4, v0.t
+; RV32-NEXT: vsll.vi v24, v24, 24, v0.t
+; RV32-NEXT: addi a0, sp, 16
+; RV32-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
+; RV32-NEXT: vand.vv v24, v8, v16, v0.t
+; RV32-NEXT: vsll.vi v16, v24, 8, v0.t
+; RV32-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
; RV32-NEXT: vor.vv v16, v24, v16, v0.t
; RV32-NEXT: csrr a0, vlenb
; RV32-NEXT: slli a0, a0, 4
@@ -805,10 +808,10 @@ define <15 x i64> @vp_bswap_v15i64(<15 x i64> %va, <15 x i1> %m, i32 zeroext %ev
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 16
; RV32-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
-; RV32-NEXT: vsrl.vx v24, v8, a1, v0.t
-; RV32-NEXT: vsrl.vx v16, v8, a3, v0.t
-; RV32-NEXT: vand.vx v16, v16, a2, v0.t
-; RV32-NEXT: vor.vv v16, v16, v24, v0.t
+; RV32-NEXT: vsrl.vx v16, v8, a1, v0.t
+; RV32-NEXT: vsrl.vx v24, v8, a3, v0.t
+; RV32-NEXT: vand.vx v24, v24, a2, v0.t
+; RV32-NEXT: vor.vv v16, v24, v16, v0.t
; RV32-NEXT: addi a0, sp, 16
; RV32-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
; RV32-NEXT: vsrl.vi v24, v8, 24, v0.t
@@ -913,13 +916,13 @@ define <15 x i64> @vp_bswap_v15i64_unmasked(<15 x i64> %va, i32 zeroext %evl) {
; RV32-NEXT: vor.vv v16, v16, v24
; RV32-NEXT: addi a4, sp, 16
; RV32-NEXT: vs8r.v v16, (a4) # Unknown-size Folded Spill
-; RV32-NEXT: lui a4, 4080
-; RV32-NEXT: vand.vx v16, v8, a4
-; RV32-NEXT: vsll.vi v0, v16, 24
-; RV32-NEXT: addi a5, sp, 8
+; RV32-NEXT: addi a4, sp, 8
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a5), zero
+; RV32-NEXT: vlse64.v v16, (a4), zero
+; RV32-NEXT: lui a4, 4080
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vand.vx v0, v8, a4
+; RV32-NEXT: vsll.vi v0, v0, 24
; RV32-NEXT: vand.vv v24, v8, v16
; RV32-NEXT: vsll.vi v24, v24, 8
; RV32-NEXT: vor.vv v24, v0, v24
@@ -1010,20 +1013,23 @@ define <16 x i64> @vp_bswap_v16i64(<16 x i64> %va, <16 x i1> %m, i32 zeroext %ev
; RV32-NEXT: add a4, sp, a4
; RV32-NEXT: addi a4, a4, 16
; RV32-NEXT: vs8r.v v16, (a4) # Unknown-size Folded Spill
-; RV32-NEXT: lui a4, 4080
-; RV32-NEXT: vand.vx v16, v8, a4, v0.t
-; RV32-NEXT: vsll.vi v24, v16, 24, v0.t
-; RV32-NEXT: addi a5, sp, 8
+; RV32-NEXT: addi a4, sp, 8
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a5), zero
-; RV32-NEXT: csrr a5, vlenb
-; RV32-NEXT: slli a5, a5, 3
-; RV32-NEXT: add a5, sp, a5
-; RV32-NEXT: addi a5, a5, 16
-; RV32-NEXT: vs8r.v v16, (a5) # Unknown-size Folded Spill
+; RV32-NEXT: vlse64.v v16, (a4), zero
+; RV32-NEXT: csrr a4, vlenb
+; RV32-NEXT: slli a4, a4, 3
+; RV32-NEXT: add a4, sp, a4
+; RV32-NEXT: addi a4, a4, 16
+; RV32-NEXT: vs8r.v v16, (a4) # Unknown-size Folded Spill
+; RV32-NEXT: lui a4, 4080
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v8, v16, v0.t
-; RV32-NEXT: vsll.vi v16, v16, 8, v0.t
+; RV32-NEXT: vand.vx v24, v8, a4, v0.t
+; RV32-NEXT: vsll.vi v24, v24, 24, v0.t
+; RV32-NEXT: addi a0, sp, 16
+; RV32-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
+; RV32-NEXT: vand.vv v24, v8, v16, v0.t
+; RV32-NEXT: vsll.vi v16, v24, 8, v0.t
+; RV32-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
; RV32-NEXT: vor.vv v16, v24, v16, v0.t
; RV32-NEXT: csrr a0, vlenb
; RV32-NEXT: slli a0, a0, 4
@@ -1036,10 +1042,10 @@ define <16 x i64> @vp_bswap_v16i64(<16 x i64> %va, <16 x i1> %m, i32 zeroext %ev
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 16
; RV32-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
-; RV32-NEXT: vsrl.vx v24, v8, a1, v0.t
-; RV32-NEXT: vsrl.vx v16, v8, a3, v0.t
-; RV32-NEXT: vand.vx v16, v16, a2, v0.t
-; RV32-NEXT: vor.vv v16, v16, v24, v0.t
+; RV32-NEXT: vsrl.vx v16, v8, a1, v0.t
+; RV32-NEXT: vsrl.vx v24, v8, a3, v0.t
+; RV32-NEXT: vand.vx v24, v24, a2, v0.t
+; RV32-NEXT: vor.vv v16, v24, v16, v0.t
; RV32-NEXT: addi a0, sp, 16
; RV32-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
; RV32-NEXT: vsrl.vi v24, v8, 24, v0.t
@@ -1144,13 +1150,13 @@ define <16 x i64> @vp_bswap_v16i64_unmasked(<16 x i64> %va, i32 zeroext %evl) {
; RV32-NEXT: vor.vv v16, v16, v24
; RV32-NEXT: addi a4, sp, 16
; RV32-NEXT: vs8r.v v16, (a4) # Unknown-size Folded Spill
-; RV32-NEXT: lui a4, 4080
-; RV32-NEXT: vand.vx v16, v8, a4
-; RV32-NEXT: vsll.vi v0, v16, 24
-; RV32-NEXT: addi a5, sp, 8
+; RV32-NEXT: addi a4, sp, 8
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a5), zero
+; RV32-NEXT: vlse64.v v16, (a4), zero
+; RV32-NEXT: lui a4, 4080
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vand.vx v0, v8, a4
+; RV32-NEXT: vsll.vi v0, v0, 24
; RV32-NEXT: vand.vv v24, v8, v16
; RV32-NEXT: vsll.vi v24, v24, 8
; RV32-NEXT: vor.vv v24, v0, v24
@@ -1228,8 +1234,8 @@ define <128 x i16> @vp_bswap_v128i16(<128 x i16> %va, <128 x i1> %m, i32 zeroext
; CHECK-NEXT: add a1, sp, a1
; CHECK-NEXT: addi a1, a1, 16
; CHECK-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
-; CHECK-NEXT: vsetivli zero, 8, e8, m1, ta, ma
; CHECK-NEXT: li a2, 64
+; CHECK-NEXT: vsetivli zero, 8, e8, m1, ta, ma
; CHECK-NEXT: vslidedown.vi v24, v0, 8
; CHECK-NEXT: mv a1, a0
; CHECK-NEXT: bltu a0, a2, .LBB26_2
@@ -1246,13 +1252,13 @@ define <128 x i16> @vp_bswap_v128i16(<128 x i16> %va, <128 x i1> %m, i32 zeroext
; CHECK-NEXT: sltu a0, a0, a1
; CHECK-NEXT: addi a0, a0, -1
; CHECK-NEXT: and a0, a0, a1
-; CHECK-NEXT: vsetvli zero, a0, e16, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: csrr a1, vlenb
+; CHECK-NEXT: slli a1, a1, 3
+; CHECK-NEXT: add a1, sp, a1
+; CHECK-NEXT: addi a1, a1, 16
+; CHECK-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, a0, e16, m8, ta, ma
; CHECK-NEXT: vsrl.vi v16, v8, 8, v0.t
; CHECK-NEXT: vsll.vi v8, v8, 8, v0.t
; CHECK-NEXT: vor.vv v16, v8, v16, v0.t
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-buildvec-of-binop.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-buildvec-of-binop.ll
index af7d7f7ae755b..65a1035fd815c 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-buildvec-of-binop.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-buildvec-of-binop.ll
@@ -567,13 +567,14 @@ define <8 x i32> @add_constant_rhs_8xi32_partial(<8 x i32> %vin, i32 %a, i32 %b,
; CHECK-NEXT: vsetivli zero, 6, e32, m2, tu, ma
; CHECK-NEXT: vslideup.vi v8, v10, 5
; CHECK-NEXT: vmv.s.x v10, a2
+; CHECK-NEXT: lui a0, %hi(.LCPI19_0)
+; CHECK-NEXT: addi a0, a0, %lo(.LCPI19_0)
+; CHECK-NEXT: vsetivli zero, 8, e32, m2, ta, ma
+; CHECK-NEXT: vle32.v v12, (a0)
; CHECK-NEXT: vsetivli zero, 7, e32, m2, tu, ma
; CHECK-NEXT: vslideup.vi v8, v10, 6
; CHECK-NEXT: vmv.s.x v10, a3
; CHECK-NEXT: vsetivli zero, 8, e32, m2, ta, ma
-; CHECK-NEXT: lui a0, %hi(.LCPI19_0)
-; CHECK-NEXT: addi a0, a0, %lo(.LCPI19_0)
-; CHECK-NEXT: vle32.v v12, (a0)
; CHECK-NEXT: vslideup.vi v8, v10, 7
; CHECK-NEXT: vadd.vv v8, v8, v12
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-ceil-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-ceil-vp.ll
index 3e2af7e8267b9..befbfb88550ba 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-ceil-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-ceil-vp.ll
@@ -204,8 +204,8 @@ define <8 x half> @vp_ceil_v8f16(<8 x half> %va, <8 x i1> %m, i32 zeroext %evl)
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m2, ta, mu
; ZVFHMIN-NEXT: vmflt.vf v9, v12, fa5, v0.t
; ZVFHMIN-NEXT: fsrmi a0, 3
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v9
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; ZVFHMIN-NEXT: vfcvt.x.f.v v12, v10, v0.t
; ZVFHMIN-NEXT: fsrm a0
; ZVFHMIN-NEXT: vfcvt.f.x.v v12, v12, v0.t
@@ -261,16 +261,16 @@ declare <16 x half> @llvm.vp.ceil.v16f16(<16 x half>, <16 x i1>, i32)
define <16 x half> @vp_ceil_v16f16(<16 x half> %va, <16 x i1> %m, i32 zeroext %evl) {
; ZVFH-LABEL: vp_ceil_v16f16:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vmv1r.v v10, v0
; ZVFH-NEXT: lui a1, %hi(.LCPI6_0)
; ZVFH-NEXT: flh fa5, %lo(.LCPI6_0)(a1)
+; ZVFH-NEXT: vmv1r.v v10, v0
; ZVFH-NEXT: vsetvli zero, a0, e16, m2, ta, ma
; ZVFH-NEXT: vfabs.v v12, v8, v0.t
; ZVFH-NEXT: vsetvli zero, zero, e16, m2, ta, mu
; ZVFH-NEXT: vmflt.vf v10, v12, fa5, v0.t
; ZVFH-NEXT: fsrmi a0, 3
-; ZVFH-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; ZVFH-NEXT: vmv1r.v v0, v10
+; ZVFH-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; ZVFH-NEXT: vfcvt.x.f.v v12, v8, v0.t
; ZVFH-NEXT: fsrm a0
; ZVFH-NEXT: vfcvt.f.x.v v12, v12, v0.t
@@ -290,8 +290,8 @@ define <16 x half> @vp_ceil_v16f16(<16 x half> %va, <16 x i1> %m, i32 zeroext %e
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m4, ta, mu
; ZVFHMIN-NEXT: vmflt.vf v10, v16, fa5, v0.t
; ZVFHMIN-NEXT: fsrmi a0, 3
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v10
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; ZVFHMIN-NEXT: vfcvt.x.f.v v16, v12, v0.t
; ZVFHMIN-NEXT: fsrm a0
; ZVFHMIN-NEXT: vfcvt.f.x.v v16, v16, v0.t
@@ -439,8 +439,8 @@ define <8 x float> @vp_ceil_v8f32(<8 x float> %va, <8 x i1> %m, i32 zeroext %evl
; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, mu
; CHECK-NEXT: vmflt.vf v10, v12, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 3
-; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
@@ -483,8 +483,8 @@ define <16 x float> @vp_ceil_v16f32(<16 x float> %va, <16 x i1> %m, i32 zeroext
; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, mu
; CHECK-NEXT: vmflt.vf v12, v16, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 3
-; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
@@ -561,16 +561,16 @@ declare <4 x double> @llvm.vp.ceil.v4f64(<4 x double>, <4 x i1>, i32)
define <4 x double> @vp_ceil_v4f64(<4 x double> %va, <4 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_ceil_v4f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v10, v0
; CHECK-NEXT: lui a1, %hi(.LCPI18_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI18_0)(a1)
+; CHECK-NEXT: vmv1r.v v10, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, ma
; CHECK-NEXT: vfabs.v v12, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, mu
; CHECK-NEXT: vmflt.vf v10, v12, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 3
-; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
@@ -605,16 +605,16 @@ declare <8 x double> @llvm.vp.ceil.v8f64(<8 x double>, <8 x i1>, i32)
define <8 x double> @vp_ceil_v8f64(<8 x double> %va, <8 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_ceil_v8f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v12, v0
; CHECK-NEXT: lui a1, %hi(.LCPI20_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI20_0)(a1)
+; CHECK-NEXT: vmv1r.v v12, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, ma
; CHECK-NEXT: vfabs.v v16, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, mu
; CHECK-NEXT: vmflt.vf v12, v16, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 3
-; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
@@ -649,16 +649,16 @@ declare <15 x double> @llvm.vp.ceil.v15f64(<15 x double>, <15 x i1>, i32)
define <15 x double> @vp_ceil_v15f64(<15 x double> %va, <15 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_ceil_v15f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: lui a1, %hi(.LCPI22_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI22_0)(a1)
+; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
; CHECK-NEXT: vmflt.vf v16, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 3
-; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
@@ -693,16 +693,16 @@ declare <16 x double> @llvm.vp.ceil.v16f64(<16 x double>, <16 x i1>, i32)
define <16 x double> @vp_ceil_v16f64(<16 x double> %va, <16 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_ceil_v16f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: lui a1, %hi(.LCPI24_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI24_0)(a1)
+; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
; CHECK-NEXT: vmflt.vf v16, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 3
-; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
@@ -737,69 +737,59 @@ declare <32 x double> @llvm.vp.ceil.v32f64(<32 x double>, <32 x i1>, i32)
define <32 x double> @vp_ceil_v32f64(<32 x double> %va, <32 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_ceil_v32f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: addi sp, sp, -16
-; CHECK-NEXT: .cfi_def_cfa_offset 16
-; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a1, a1, 4
-; CHECK-NEXT: sub sp, sp, a1
-; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
-; CHECK-NEXT: vmv1r.v v25, v0
-; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a1, a1, 3
-; CHECK-NEXT: add a1, sp, a1
-; CHECK-NEXT: addi a1, a1, 16
-; CHECK-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
-; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
+; CHECK-NEXT: vmv1r.v v6, v0
; CHECK-NEXT: li a2, 16
-; CHECK-NEXT: vslidedown.vi v24, v0, 2
+; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
+; CHECK-NEXT: vslidedown.vi v7, v0, 2
; CHECK-NEXT: mv a1, a0
; CHECK-NEXT: bltu a0, a2, .LBB26_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: li a1, 16
; CHECK-NEXT: .LBB26_2:
+; CHECK-NEXT: addi sp, sp, -16
+; CHECK-NEXT: .cfi_def_cfa_offset 16
+; CHECK-NEXT: csrr a2, vlenb
+; CHECK-NEXT: slli a2, a2, 3
+; CHECK-NEXT: sub sp, sp, a2
+; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
; CHECK-NEXT: lui a2, %hi(.LCPI26_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI26_0)(a2)
+; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v25
-; CHECK-NEXT: vfabs.v v16, v8, v0.t
+; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vmflt.vf v25, v16, fa5, v0.t
+; CHECK-NEXT: vmflt.vf v6, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a1, 3
+; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v25
-; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
+; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
+; CHECK-NEXT: addi a2, sp, 16
+; CHECK-NEXT: vs8r.v v24, (a2) # Unknown-size Folded Spill
; CHECK-NEXT: fsrm a1
-; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
-; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vfsgnj.vv v8, v16, v8, v0.t
; CHECK-NEXT: addi a1, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
+; CHECK-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v24, v8, v0.t
; CHECK-NEXT: addi a1, a0, -16
; CHECK-NEXT: sltu a0, a0, a1
; CHECK-NEXT: addi a0, a0, -1
; CHECK-NEXT: and a0, a0, a1
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v24
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vfabs.v v8, v16, v0.t
+; CHECK-NEXT: vfabs.v v24, v16, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vmflt.vf v24, v8, fa5, v0.t
+; CHECK-NEXT: vmflt.vf v7, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 3
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v24
-; CHECK-NEXT: vfcvt.x.f.v v8, v16, v0.t
+; CHECK-NEXT: vfcvt.x.f.v v24, v16, v0.t
; CHECK-NEXT: fsrm a0
-; CHECK-NEXT: vfcvt.f.x.v v8, v8, v0.t
+; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vfsgnj.vv v16, v8, v16, v0.t
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vfsgnj.vv v16, v24, v16, v0.t
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 4
+; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add sp, sp, a0
; CHECK-NEXT: addi sp, sp, 16
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-ctlz-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-ctlz-vp.ll
index 2f4539d5038c2..b42fb8c686164 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-ctlz-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-ctlz-vp.ll
@@ -1503,24 +1503,28 @@ declare <15 x i64> @llvm.vp.ctlz.v15i64(<15 x i64>, i1 immarg, <15 x i1>, i32)
define <15 x i64> @vp_ctlz_v15i64(<15 x i64> %va, <15 x i1> %m, i32 zeroext %evl) {
; RV32-LABEL: vp_ctlz_v15i64:
; RV32: # %bb.0:
-; RV32-NEXT: addi sp, sp, -32
-; RV32-NEXT: .cfi_def_cfa_offset 32
+; RV32-NEXT: addi sp, sp, -48
+; RV32-NEXT: .cfi_def_cfa_offset 48
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 4
+; RV32-NEXT: sub sp, sp, a1
+; RV32-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x30, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 48 + 16 * vlenb
; RV32-NEXT: lui a1, 349525
; RV32-NEXT: addi a1, a1, 1365
-; RV32-NEXT: sw a1, 28(sp)
-; RV32-NEXT: sw a1, 24(sp)
+; RV32-NEXT: sw a1, 44(sp)
+; RV32-NEXT: sw a1, 40(sp)
; RV32-NEXT: lui a1, 209715
; RV32-NEXT: addi a1, a1, 819
-; RV32-NEXT: sw a1, 20(sp)
-; RV32-NEXT: sw a1, 16(sp)
+; RV32-NEXT: sw a1, 36(sp)
+; RV32-NEXT: sw a1, 32(sp)
; RV32-NEXT: lui a1, 61681
; RV32-NEXT: addi a1, a1, -241
-; RV32-NEXT: sw a1, 12(sp)
-; RV32-NEXT: sw a1, 8(sp)
+; RV32-NEXT: sw a1, 28(sp)
+; RV32-NEXT: sw a1, 24(sp)
; RV32-NEXT: lui a1, 4112
; RV32-NEXT: addi a1, a1, 257
-; RV32-NEXT: sw a1, 4(sp)
-; RV32-NEXT: sw a1, 0(sp)
+; RV32-NEXT: sw a1, 20(sp)
+; RV32-NEXT: sw a1, 16(sp)
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vsrl.vi v16, v8, 1, v0.t
; RV32-NEXT: vor.vv v8, v8, v16, v0.t
@@ -1535,37 +1539,60 @@ define <15 x i64> @vp_ctlz_v15i64(<15 x i64> %va, <15 x i1> %m, i32 zeroext %evl
; RV32-NEXT: li a1, 32
; RV32-NEXT: vsrl.vx v16, v8, a1, v0.t
; RV32-NEXT: vor.vv v8, v8, v16, v0.t
-; RV32-NEXT: vnot.v v8, v8, v0.t
-; RV32-NEXT: vsrl.vi v16, v8, 1, v0.t
-; RV32-NEXT: addi a1, sp, 24
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v24, (a1), zero
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v16, v24, v0.t
-; RV32-NEXT: vsub.vv v8, v8, v16, v0.t
-; RV32-NEXT: addi a1, sp, 16
+; RV32-NEXT: vnot.v v24, v8, v0.t
+; RV32-NEXT: addi a1, sp, 48
+; RV32-NEXT: vs8r.v v24, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: addi a1, sp, 40
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v8, (a1), zero
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 3
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 48
+; RV32-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: addi a1, sp, 32
; RV32-NEXT: vlse64.v v16, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vi v8, v24, 1, v0.t
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 3
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 48
+; RV32-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v8, v8, v24, v0.t
+; RV32-NEXT: addi a1, sp, 48
+; RV32-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vsub.vv v8, v24, v8, v0.t
; RV32-NEXT: vand.vv v24, v8, v16, v0.t
; RV32-NEXT: vsrl.vi v8, v8, 2, v0.t
; RV32-NEXT: vand.vv v8, v8, v16, v0.t
; RV32-NEXT: vadd.vv v8, v24, v8, v0.t
-; RV32-NEXT: vsrl.vi v16, v8, 4, v0.t
-; RV32-NEXT: vadd.vv v8, v8, v16, v0.t
-; RV32-NEXT: addi a1, sp, 8
+; RV32-NEXT: addi a1, sp, 24
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV32-NEXT: vlse64.v v16, (a1), zero
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 3
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 48
+; RV32-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: addi a1, sp, 16
+; RV32-NEXT: vlse64.v v24, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vi v16, v8, 4, v0.t
+; RV32-NEXT: vadd.vv v8, v8, v16, v0.t
+; RV32-NEXT: csrr a0, vlenb
+; RV32-NEXT: slli a0, a0, 3
+; RV32-NEXT: add a0, sp, a0
+; RV32-NEXT: addi a0, a0, 48
+; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; RV32-NEXT: vand.vv v8, v8, v16, v0.t
-; RV32-NEXT: mv a1, sp
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a1), zero
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vmul.vv v8, v8, v16, v0.t
+; RV32-NEXT: vmul.vv v8, v8, v24, v0.t
; RV32-NEXT: li a0, 56
; RV32-NEXT: vsrl.vx v8, v8, a0, v0.t
-; RV32-NEXT: addi sp, sp, 32
+; RV32-NEXT: csrr a0, vlenb
+; RV32-NEXT: slli a0, a0, 4
+; RV32-NEXT: add sp, sp, a0
+; RV32-NEXT: addi sp, sp, 48
; RV32-NEXT: ret
;
; RV64-LABEL: vp_ctlz_v15i64:
@@ -1655,33 +1682,29 @@ define <15 x i64> @vp_ctlz_v15i64_unmasked(<15 x i64> %va, i32 zeroext %evl) {
; RV32-NEXT: vsrl.vx v16, v8, a1
; RV32-NEXT: vor.vv v8, v8, v16
; RV32-NEXT: vnot.v v8, v8
-; RV32-NEXT: vsrl.vi v16, v8, 1
; RV32-NEXT: addi a1, sp, 24
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v16, (a1), zero
+; RV32-NEXT: addi a1, sp, 16
; RV32-NEXT: vlse64.v v24, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v16, v24
+; RV32-NEXT: vsrl.vi v0, v8, 1
+; RV32-NEXT: vand.vv v16, v0, v16
; RV32-NEXT: vsub.vv v8, v8, v16
-; RV32-NEXT: addi a1, sp, 16
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a1), zero
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v24, v8, v16
+; RV32-NEXT: vand.vv v16, v8, v24
; RV32-NEXT: vsrl.vi v8, v8, 2
-; RV32-NEXT: vand.vv v8, v8, v16
-; RV32-NEXT: vadd.vv v8, v24, v8
-; RV32-NEXT: vsrl.vi v16, v8, 4
-; RV32-NEXT: vadd.vv v8, v8, v16
+; RV32-NEXT: vand.vv v8, v8, v24
+; RV32-NEXT: vadd.vv v8, v16, v8
; RV32-NEXT: addi a1, sp, 8
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV32-NEXT: vlse64.v v16, (a1), zero
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v8, v8, v16
; RV32-NEXT: mv a1, sp
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a1), zero
+; RV32-NEXT: vlse64.v v24, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vmul.vv v8, v8, v16
+; RV32-NEXT: vsrl.vi v0, v8, 4
+; RV32-NEXT: vadd.vv v8, v8, v0
+; RV32-NEXT: vand.vv v8, v8, v16
+; RV32-NEXT: vmul.vv v8, v8, v24
; RV32-NEXT: li a0, 56
; RV32-NEXT: vsrl.vx v8, v8, a0
; RV32-NEXT: addi sp, sp, 32
@@ -1743,24 +1766,28 @@ declare <16 x i64> @llvm.vp.ctlz.v16i64(<16 x i64>, i1 immarg, <16 x i1>, i32)
define <16 x i64> @vp_ctlz_v16i64(<16 x i64> %va, <16 x i1> %m, i32 zeroext %evl) {
; RV32-LABEL: vp_ctlz_v16i64:
; RV32: # %bb.0:
-; RV32-NEXT: addi sp, sp, -32
-; RV32-NEXT: .cfi_def_cfa_offset 32
+; RV32-NEXT: addi sp, sp, -48
+; RV32-NEXT: .cfi_def_cfa_offset 48
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 4
+; RV32-NEXT: sub sp, sp, a1
+; RV32-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x30, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 48 + 16 * vlenb
; RV32-NEXT: lui a1, 349525
; RV32-NEXT: addi a1, a1, 1365
-; RV32-NEXT: sw a1, 28(sp)
-; RV32-NEXT: sw a1, 24(sp)
+; RV32-NEXT: sw a1, 44(sp)
+; RV32-NEXT: sw a1, 40(sp)
; RV32-NEXT: lui a1, 209715
; RV32-NEXT: addi a1, a1, 819
-; RV32-NEXT: sw a1, 20(sp)
-; RV32-NEXT: sw a1, 16(sp)
+; RV32-NEXT: sw a1, 36(sp)
+; RV32-NEXT: sw a1, 32(sp)
; RV32-NEXT: lui a1, 61681
; RV32-NEXT: addi a1, a1, -241
-; RV32-NEXT: sw a1, 12(sp)
-; RV32-NEXT: sw a1, 8(sp)
+; RV32-NEXT: sw a1, 28(sp)
+; RV32-NEXT: sw a1, 24(sp)
; RV32-NEXT: lui a1, 4112
; RV32-NEXT: addi a1, a1, 257
-; RV32-NEXT: sw a1, 4(sp)
-; RV32-NEXT: sw a1, 0(sp)
+; RV32-NEXT: sw a1, 20(sp)
+; RV32-NEXT: sw a1, 16(sp)
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vsrl.vi v16, v8, 1, v0.t
; RV32-NEXT: vor.vv v8, v8, v16, v0.t
@@ -1775,37 +1802,60 @@ define <16 x i64> @vp_ctlz_v16i64(<16 x i64> %va, <16 x i1> %m, i32 zeroext %evl
; RV32-NEXT: li a1, 32
; RV32-NEXT: vsrl.vx v16, v8, a1, v0.t
; RV32-NEXT: vor.vv v8, v8, v16, v0.t
-; RV32-NEXT: vnot.v v8, v8, v0.t
-; RV32-NEXT: vsrl.vi v16, v8, 1, v0.t
-; RV32-NEXT: addi a1, sp, 24
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v24, (a1), zero
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v16, v24, v0.t
-; RV32-NEXT: vsub.vv v8, v8, v16, v0.t
-; RV32-NEXT: addi a1, sp, 16
+; RV32-NEXT: vnot.v v24, v8, v0.t
+; RV32-NEXT: addi a1, sp, 48
+; RV32-NEXT: vs8r.v v24, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: addi a1, sp, 40
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v8, (a1), zero
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 3
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 48
+; RV32-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: addi a1, sp, 32
; RV32-NEXT: vlse64.v v16, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vi v8, v24, 1, v0.t
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 3
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 48
+; RV32-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v8, v8, v24, v0.t
+; RV32-NEXT: addi a1, sp, 48
+; RV32-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vsub.vv v8, v24, v8, v0.t
; RV32-NEXT: vand.vv v24, v8, v16, v0.t
; RV32-NEXT: vsrl.vi v8, v8, 2, v0.t
; RV32-NEXT: vand.vv v8, v8, v16, v0.t
; RV32-NEXT: vadd.vv v8, v24, v8, v0.t
-; RV32-NEXT: vsrl.vi v16, v8, 4, v0.t
-; RV32-NEXT: vadd.vv v8, v8, v16, v0.t
-; RV32-NEXT: addi a1, sp, 8
+; RV32-NEXT: addi a1, sp, 24
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV32-NEXT: vlse64.v v16, (a1), zero
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 3
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 48
+; RV32-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: addi a1, sp, 16
+; RV32-NEXT: vlse64.v v24, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vi v16, v8, 4, v0.t
+; RV32-NEXT: vadd.vv v8, v8, v16, v0.t
+; RV32-NEXT: csrr a0, vlenb
+; RV32-NEXT: slli a0, a0, 3
+; RV32-NEXT: add a0, sp, a0
+; RV32-NEXT: addi a0, a0, 48
+; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; RV32-NEXT: vand.vv v8, v8, v16, v0.t
-; RV32-NEXT: mv a1, sp
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a1), zero
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vmul.vv v8, v8, v16, v0.t
+; RV32-NEXT: vmul.vv v8, v8, v24, v0.t
; RV32-NEXT: li a0, 56
; RV32-NEXT: vsrl.vx v8, v8, a0, v0.t
-; RV32-NEXT: addi sp, sp, 32
+; RV32-NEXT: csrr a0, vlenb
+; RV32-NEXT: slli a0, a0, 4
+; RV32-NEXT: add sp, sp, a0
+; RV32-NEXT: addi sp, sp, 48
; RV32-NEXT: ret
;
; RV64-LABEL: vp_ctlz_v16i64:
@@ -1895,33 +1945,29 @@ define <16 x i64> @vp_ctlz_v16i64_unmasked(<16 x i64> %va, i32 zeroext %evl) {
; RV32-NEXT: vsrl.vx v16, v8, a1
; RV32-NEXT: vor.vv v8, v8, v16
; RV32-NEXT: vnot.v v8, v8
-; RV32-NEXT: vsrl.vi v16, v8, 1
; RV32-NEXT: addi a1, sp, 24
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v16, (a1), zero
+; RV32-NEXT: addi a1, sp, 16
; RV32-NEXT: vlse64.v v24, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v16, v24
+; RV32-NEXT: vsrl.vi v0, v8, 1
+; RV32-NEXT: vand.vv v16, v0, v16
; RV32-NEXT: vsub.vv v8, v8, v16
-; RV32-NEXT: addi a1, sp, 16
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a1), zero
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v24, v8, v16
+; RV32-NEXT: vand.vv v16, v8, v24
; RV32-NEXT: vsrl.vi v8, v8, 2
-; RV32-NEXT: vand.vv v8, v8, v16
-; RV32-NEXT: vadd.vv v8, v24, v8
-; RV32-NEXT: vsrl.vi v16, v8, 4
-; RV32-NEXT: vadd.vv v8, v8, v16
+; RV32-NEXT: vand.vv v8, v8, v24
+; RV32-NEXT: vadd.vv v8, v16, v8
; RV32-NEXT: addi a1, sp, 8
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV32-NEXT: vlse64.v v16, (a1), zero
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v8, v8, v16
; RV32-NEXT: mv a1, sp
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a1), zero
+; RV32-NEXT: vlse64.v v24, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vmul.vv v8, v8, v16
+; RV32-NEXT: vsrl.vi v0, v8, 4
+; RV32-NEXT: vadd.vv v8, v8, v0
+; RV32-NEXT: vand.vv v8, v8, v16
+; RV32-NEXT: vmul.vv v8, v8, v24
; RV32-NEXT: li a0, 56
; RV32-NEXT: vsrl.vx v8, v8, a0
; RV32-NEXT: addi sp, sp, 32
@@ -1991,7 +2037,7 @@ define <32 x i64> @vp_ctlz_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %evl
; RV32-NEXT: sub sp, sp, a1
; RV32-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x30, 0x22, 0x11, 0x38, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 48 + 56 * vlenb
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: slli a1, a1, 5
+; RV32-NEXT: slli a1, a1, 4
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 48
; RV32-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
@@ -2035,111 +2081,145 @@ define <32 x i64> @vp_ctlz_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %evl
; RV32-NEXT: vor.vv v8, v8, v16, v0.t
; RV32-NEXT: vnot.v v8, v8, v0.t
; RV32-NEXT: csrr a3, vlenb
+; RV32-NEXT: slli a3, a3, 5
+; RV32-NEXT: add a3, sp, a3
+; RV32-NEXT: addi a3, a3, 48
+; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
+; RV32-NEXT: addi a3, sp, 40
+; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v8, (a3), zero
+; RV32-NEXT: csrr a3, vlenb
; RV32-NEXT: li a4, 40
; RV32-NEXT: mul a3, a3, a4
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: vsrl.vi v8, v8, 1, v0.t
+; RV32-NEXT: addi a3, sp, 32
+; RV32-NEXT: vlse64.v v8, (a3), zero
; RV32-NEXT: csrr a3, vlenb
; RV32-NEXT: li a4, 48
; RV32-NEXT: mul a3, a3, a4
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: addi a3, sp, 40
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v8, (a3), zero
+; RV32-NEXT: csrr a3, vlenb
+; RV32-NEXT: slli a3, a3, 5
+; RV32-NEXT: add a3, sp, a3
+; RV32-NEXT: addi a3, a3, 48
+; RV32-NEXT: vl8r.v v16, (a3) # Unknown-size Folded Reload
+; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vi v8, v16, 1, v0.t
; RV32-NEXT: csrr a3, vlenb
; RV32-NEXT: li a4, 24
; RV32-NEXT: mul a3, a3, a4
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: li a4, 48
+; RV32-NEXT: li a4, 40
; RV32-NEXT: mul a3, a3, a4
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
; RV32-NEXT: vl8r.v v16, (a3) # Unknown-size Folded Reload
-; RV32-NEXT: vand.vv v16, v16, v8, v0.t
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: li a4, 40
+; RV32-NEXT: li a4, 24
; RV32-NEXT: mul a3, a3, a4
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
; RV32-NEXT: vl8r.v v8, (a3) # Unknown-size Folded Reload
-; RV32-NEXT: vsub.vv v8, v8, v16, v0.t
+; RV32-NEXT: vand.vv v8, v8, v16, v0.t
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: li a4, 40
-; RV32-NEXT: mul a3, a3, a4
+; RV32-NEXT: slli a3, a3, 5
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
-; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: addi a3, sp, 32
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v8, (a3), zero
+; RV32-NEXT: vl8r.v v16, (a3) # Unknown-size Folded Reload
+; RV32-NEXT: vsub.vv v16, v16, v8, v0.t
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: li a4, 48
-; RV32-NEXT: mul a3, a3, a4
+; RV32-NEXT: slli a3, a3, 5
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
-; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
+; RV32-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: li a4, 40
-; RV32-NEXT: mul a3, a3, a4
+; RV32-NEXT: slli a3, a3, 5
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
; RV32-NEXT: vl8r.v v16, (a3) # Unknown-size Folded Reload
+; RV32-NEXT: csrr a3, vlenb
+; RV32-NEXT: li a4, 48
+; RV32-NEXT: mul a3, a3, a4
+; RV32-NEXT: add a3, sp, a3
+; RV32-NEXT: addi a3, a3, 48
+; RV32-NEXT: vl8r.v v8, (a3) # Unknown-size Folded Reload
; RV32-NEXT: vand.vv v16, v16, v8, v0.t
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: slli a3, a3, 4
+; RV32-NEXT: li a4, 24
+; RV32-NEXT: mul a3, a3, a4
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
; RV32-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: li a4, 40
-; RV32-NEXT: mul a3, a3, a4
+; RV32-NEXT: slli a3, a3, 5
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
; RV32-NEXT: vl8r.v v16, (a3) # Unknown-size Folded Reload
; RV32-NEXT: vsrl.vi v16, v16, 2, v0.t
; RV32-NEXT: vand.vv v16, v16, v8, v0.t
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: slli a3, a3, 4
+; RV32-NEXT: li a4, 24
+; RV32-NEXT: mul a3, a3, a4
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
; RV32-NEXT: vl8r.v v8, (a3) # Unknown-size Folded Reload
; RV32-NEXT: vadd.vv v8, v8, v16, v0.t
-; RV32-NEXT: vsrl.vi v16, v8, 4, v0.t
-; RV32-NEXT: vadd.vv v16, v8, v16, v0.t
+; RV32-NEXT: csrr a3, vlenb
+; RV32-NEXT: slli a3, a3, 3
+; RV32-NEXT: add a3, sp, a3
+; RV32-NEXT: addi a3, a3, 48
+; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
; RV32-NEXT: addi a3, sp, 24
+; RV32-NEXT: addi a4, sp, 16
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV32-NEXT: vlse64.v v8, (a3), zero
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: li a4, 40
-; RV32-NEXT: mul a3, a3, a4
+; RV32-NEXT: li a5, 24
+; RV32-NEXT: mul a3, a3, a5
+; RV32-NEXT: add a3, sp, a3
+; RV32-NEXT: addi a3, a3, 48
+; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
+; RV32-NEXT: vlse64.v v8, (a4), zero
+; RV32-NEXT: csrr a3, vlenb
+; RV32-NEXT: slli a3, a3, 5
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v8, v16, v8, v0.t
-; RV32-NEXT: addi a3, sp, 16
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a3), zero
; RV32-NEXT: csrr a3, vlenb
; RV32-NEXT: slli a3, a3, 3
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
-; RV32-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
+; RV32-NEXT: vl8r.v v8, (a3) # Unknown-size Folded Reload
; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
-; RV32-NEXT: vmul.vv v8, v8, v16, v0.t
+; RV32-NEXT: vsrl.vi v16, v8, 4, v0.t
+; RV32-NEXT: addi a2, sp, 48
+; RV32-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
+; RV32-NEXT: vl8r.v v16, (a2) # Unknown-size Folded Reload
+; RV32-NEXT: vadd.vv v16, v8, v16, v0.t
+; RV32-NEXT: csrr a2, vlenb
+; RV32-NEXT: li a3, 24
+; RV32-NEXT: mul a2, a2, a3
+; RV32-NEXT: add a2, sp, a2
+; RV32-NEXT: addi a2, a2, 48
+; RV32-NEXT: vl8r.v v8, (a2) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v16, v16, v8, v0.t
+; RV32-NEXT: csrr a2, vlenb
+; RV32-NEXT: slli a2, a2, 5
+; RV32-NEXT: add a2, sp, a2
+; RV32-NEXT: addi a2, a2, 48
+; RV32-NEXT: vl8r.v v8, (a2) # Unknown-size Folded Reload
+; RV32-NEXT: vmul.vv v8, v16, v8, v0.t
; RV32-NEXT: li a2, 56
; RV32-NEXT: vsrl.vx v8, v8, a2, v0.t
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: slli a3, a3, 4
+; RV32-NEXT: slli a3, a3, 3
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
@@ -2147,13 +2227,13 @@ define <32 x i64> @vp_ctlz_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %evl
; RV32-NEXT: sltu a0, a0, a3
; RV32-NEXT: addi a0, a0, -1
; RV32-NEXT: and a0, a0, a3
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vmv1r.v v0, v24
-; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 5
-; RV32-NEXT: add a0, sp, a0
-; RV32-NEXT: addi a0, a0, 48
-; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: csrr a3, vlenb
+; RV32-NEXT: slli a3, a3, 4
+; RV32-NEXT: add a3, sp, a3
+; RV32-NEXT: addi a3, a3, 48
+; RV32-NEXT: vl8r.v v16, (a3) # Unknown-size Folded Reload
+; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vsrl.vi v8, v16, 1, v0.t
; RV32-NEXT: vor.vv v8, v16, v8, v0.t
; RV32-NEXT: vsrl.vi v16, v8, 2, v0.t
@@ -2171,18 +2251,18 @@ define <32 x i64> @vp_ctlz_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %evl
; RV32-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; RV32-NEXT: vsrl.vi v8, v8, 1, v0.t
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 5
+; RV32-NEXT: slli a0, a0, 4
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 48
; RV32-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: li a1, 24
+; RV32-NEXT: li a1, 40
; RV32-NEXT: mul a0, a0, a1
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 48
; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 5
+; RV32-NEXT: slli a0, a0, 4
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 48
; RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
@@ -2191,17 +2271,35 @@ define <32 x i64> @vp_ctlz_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %evl
; RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; RV32-NEXT: vsub.vv v8, v8, v16, v0.t
; RV32-NEXT: csrr a0, vlenb
+; RV32-NEXT: li a1, 40
+; RV32-NEXT: mul a0, a0, a1
+; RV32-NEXT: add a0, sp, a0
+; RV32-NEXT: addi a0, a0, 48
+; RV32-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; RV32-NEXT: csrr a0, vlenb
; RV32-NEXT: li a1, 48
; RV32-NEXT: mul a0, a0, a1
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 48
+; RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: csrr a0, vlenb
+; RV32-NEXT: li a1, 40
+; RV32-NEXT: mul a0, a0, a1
+; RV32-NEXT: add a0, sp, a0
+; RV32-NEXT: addi a0, a0, 48
; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; RV32-NEXT: vand.vv v16, v8, v16, v0.t
+; RV32-NEXT: vand.vv v16, v16, v8, v0.t
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 5
+; RV32-NEXT: slli a0, a0, 4
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 48
; RV32-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
+; RV32-NEXT: csrr a0, vlenb
+; RV32-NEXT: li a1, 40
+; RV32-NEXT: mul a0, a0, a1
+; RV32-NEXT: add a0, sp, a0
+; RV32-NEXT: addi a0, a0, 48
+; RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; RV32-NEXT: vsrl.vi v8, v8, 2, v0.t
; RV32-NEXT: csrr a0, vlenb
; RV32-NEXT: li a1, 48
@@ -2211,7 +2309,7 @@ define <32 x i64> @vp_ctlz_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %evl
; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; RV32-NEXT: vand.vv v8, v8, v16, v0.t
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 5
+; RV32-NEXT: slli a0, a0, 4
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 48
; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
@@ -2219,21 +2317,21 @@ define <32 x i64> @vp_ctlz_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %evl
; RV32-NEXT: vsrl.vi v16, v8, 4, v0.t
; RV32-NEXT: vadd.vv v8, v8, v16, v0.t
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: li a1, 40
+; RV32-NEXT: li a1, 24
; RV32-NEXT: mul a0, a0, a1
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 48
; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; RV32-NEXT: vand.vv v8, v8, v16, v0.t
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 3
+; RV32-NEXT: slli a0, a0, 5
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 48
; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; RV32-NEXT: vmul.vv v8, v8, v16, v0.t
; RV32-NEXT: vsrl.vx v16, v8, a2, v0.t
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 4
+; RV32-NEXT: slli a0, a0, 3
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 48
; RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
@@ -2257,8 +2355,8 @@ define <32 x i64> @vp_ctlz_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %evl
; RV64-NEXT: add a1, sp, a1
; RV64-NEXT: addi a1, a1, 16
; RV64-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
-; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64-NEXT: li a2, 16
+; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64-NEXT: vslidedown.vi v24, v0, 2
; RV64-NEXT: mv a1, a0
; RV64-NEXT: bltu a0, a2, .LBB34_2
@@ -2315,13 +2413,13 @@ define <32 x i64> @vp_ctlz_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %evl
; RV64-NEXT: sltu a0, a0, a7
; RV64-NEXT: addi a0, a0, -1
; RV64-NEXT: and a0, a0, a7
-; RV64-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV64-NEXT: vmv1r.v v0, v24
-; RV64-NEXT: csrr a0, vlenb
-; RV64-NEXT: slli a0, a0, 3
-; RV64-NEXT: add a0, sp, a0
-; RV64-NEXT: addi a0, a0, 16
-; RV64-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; RV64-NEXT: csrr a7, vlenb
+; RV64-NEXT: slli a7, a7, 3
+; RV64-NEXT: add a7, sp, a7
+; RV64-NEXT: addi a7, a7, 16
+; RV64-NEXT: vl8r.v v8, (a7) # Unknown-size Folded Reload
+; RV64-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV64-NEXT: vsrl.vi v16, v8, 1, v0.t
; RV64-NEXT: vor.vv v16, v8, v16, v0.t
; RV64-NEXT: vsrl.vi v8, v16, 2, v0.t
@@ -2364,10 +2462,14 @@ define <32 x i64> @vp_ctlz_v32i64_unmasked(<32 x i64> %va, i32 zeroext %evl) {
; RV32-NEXT: addi sp, sp, -48
; RV32-NEXT: .cfi_def_cfa_offset 48
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: slli a1, a1, 5
+; RV32-NEXT: slli a1, a1, 4
; RV32-NEXT: sub sp, sp, a1
-; RV32-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x30, 0x22, 0x11, 0x20, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 48 + 32 * vlenb
-; RV32-NEXT: vmv8r.v v24, v16
+; RV32-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x30, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 48 + 16 * vlenb
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 3
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 48
+; RV32-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
; RV32-NEXT: lui a1, 349525
; RV32-NEXT: addi a1, a1, 1365
; RV32-NEXT: sw a1, 44(sp)
@@ -2391,74 +2493,8 @@ define <32 x i64> @vp_ctlz_v32i64_unmasked(<32 x i64> %va, i32 zeroext %evl) {
; RV32-NEXT: li a1, 16
; RV32-NEXT: .LBB35_2:
; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; RV32-NEXT: vsrl.vi v16, v8, 1
-; RV32-NEXT: vor.vv v8, v8, v16
-; RV32-NEXT: vsrl.vi v16, v8, 2
-; RV32-NEXT: vor.vv v8, v8, v16
-; RV32-NEXT: vsrl.vi v16, v8, 4
-; RV32-NEXT: vor.vv v8, v8, v16
-; RV32-NEXT: vsrl.vi v16, v8, 8
-; RV32-NEXT: vor.vv v8, v8, v16
-; RV32-NEXT: vsrl.vi v16, v8, 16
-; RV32-NEXT: vor.vv v8, v8, v16
-; RV32-NEXT: li a2, 32
-; RV32-NEXT: vsrl.vx v16, v8, a2
-; RV32-NEXT: vor.vv v8, v8, v16
-; RV32-NEXT: vnot.v v8, v8
-; RV32-NEXT: vsrl.vi v16, v8, 1
-; RV32-NEXT: addi a3, sp, 40
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v0, (a3), zero
-; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: li a4, 24
-; RV32-NEXT: mul a3, a3, a4
-; RV32-NEXT: add a3, sp, a3
-; RV32-NEXT: addi a3, a3, 48
-; RV32-NEXT: vs8r.v v0, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v16, v0
-; RV32-NEXT: vsub.vv v8, v8, v16
-; RV32-NEXT: addi a3, sp, 32
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v0, (a3), zero
-; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v8, v0
-; RV32-NEXT: vsrl.vi v8, v8, 2
-; RV32-NEXT: vand.vv v8, v8, v0
-; RV32-NEXT: vadd.vv v8, v16, v8
-; RV32-NEXT: vsrl.vi v16, v8, 4
-; RV32-NEXT: vadd.vv v8, v8, v16
-; RV32-NEXT: addi a3, sp, 24
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a3), zero
-; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: slli a3, a3, 4
-; RV32-NEXT: add a3, sp, a3
-; RV32-NEXT: addi a3, a3, 48
-; RV32-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v8, v8, v16
-; RV32-NEXT: addi a3, sp, 16
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a3), zero
-; RV32-NEXT: addi a3, sp, 48
-; RV32-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; RV32-NEXT: vmul.vv v8, v8, v16
-; RV32-NEXT: li a1, 56
-; RV32-NEXT: vsrl.vx v8, v8, a1
-; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: slli a3, a3, 3
-; RV32-NEXT: add a3, sp, a3
-; RV32-NEXT: addi a3, a3, 48
-; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: addi a3, a0, -16
-; RV32-NEXT: sltu a0, a0, a3
-; RV32-NEXT: addi a0, a0, -1
-; RV32-NEXT: and a0, a0, a3
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vsrl.vi v8, v24, 1
-; RV32-NEXT: vor.vv v8, v24, v8
+; RV32-NEXT: vsrl.vi v24, v8, 1
+; RV32-NEXT: vor.vv v8, v8, v24
; RV32-NEXT: vsrl.vi v24, v8, 2
; RV32-NEXT: vor.vv v8, v8, v24
; RV32-NEXT: vsrl.vi v24, v8, 4
@@ -2467,41 +2503,84 @@ define <32 x i64> @vp_ctlz_v32i64_unmasked(<32 x i64> %va, i32 zeroext %evl) {
; RV32-NEXT: vor.vv v8, v8, v24
; RV32-NEXT: vsrl.vi v24, v8, 16
; RV32-NEXT: vor.vv v8, v8, v24
+; RV32-NEXT: li a2, 32
; RV32-NEXT: vsrl.vx v24, v8, a2
; RV32-NEXT: vor.vv v8, v8, v24
-; RV32-NEXT: vnot.v v8, v8
-; RV32-NEXT: vsrl.vi v24, v8, 1
-; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: li a2, 24
-; RV32-NEXT: mul a0, a0, a2
-; RV32-NEXT: add a0, sp, a0
-; RV32-NEXT: addi a0, a0, 48
-; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vnot.v v0, v8
+; RV32-NEXT: addi a3, sp, 40
+; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v16, (a3), zero
+; RV32-NEXT: addi a3, sp, 32
+; RV32-NEXT: vlse64.v v8, (a3), zero
+; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vi v24, v0, 1
; RV32-NEXT: vand.vv v24, v24, v16
-; RV32-NEXT: vsub.vv v8, v8, v24
-; RV32-NEXT: vand.vv v24, v8, v0
-; RV32-NEXT: vsrl.vi v8, v8, 2
-; RV32-NEXT: vand.vv v8, v8, v0
+; RV32-NEXT: vsub.vv v24, v0, v24
+; RV32-NEXT: vand.vv v0, v24, v8
+; RV32-NEXT: vsrl.vi v24, v24, 2
+; RV32-NEXT: vand.vv v24, v24, v8
+; RV32-NEXT: vadd.vv v24, v0, v24
+; RV32-NEXT: vsrl.vi v0, v24, 4
+; RV32-NEXT: vadd.vv v24, v24, v0
+; RV32-NEXT: addi a3, sp, 48
+; RV32-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
+; RV32-NEXT: addi a3, a0, -16
+; RV32-NEXT: sltu a0, a0, a3
+; RV32-NEXT: addi a0, a0, -1
+; RV32-NEXT: and a0, a0, a3
+; RV32-NEXT: csrr a3, vlenb
+; RV32-NEXT: slli a3, a3, 3
+; RV32-NEXT: add a3, sp, a3
+; RV32-NEXT: addi a3, a3, 48
+; RV32-NEXT: vl8r.v v0, (a3) # Unknown-size Folded Reload
+; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vi v24, v0, 1
+; RV32-NEXT: vor.vv v24, v0, v24
+; RV32-NEXT: vsrl.vi v0, v24, 2
+; RV32-NEXT: vor.vv v24, v24, v0
+; RV32-NEXT: vsrl.vi v0, v24, 4
+; RV32-NEXT: vor.vv v24, v24, v0
+; RV32-NEXT: vsrl.vi v0, v24, 8
+; RV32-NEXT: vor.vv v24, v24, v0
+; RV32-NEXT: vsrl.vi v0, v24, 16
+; RV32-NEXT: vor.vv v24, v24, v0
+; RV32-NEXT: vsrl.vx v0, v24, a2
+; RV32-NEXT: vor.vv v24, v24, v0
+; RV32-NEXT: vnot.v v24, v24
+; RV32-NEXT: vsrl.vi v0, v24, 1
+; RV32-NEXT: vand.vv v16, v0, v16
+; RV32-NEXT: addi a2, sp, 24
+; RV32-NEXT: vsub.vv v16, v24, v16
+; RV32-NEXT: vand.vv v24, v16, v8
+; RV32-NEXT: vsrl.vi v16, v16, 2
+; RV32-NEXT: vand.vv v8, v16, v8
+; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v16, (a2), zero
+; RV32-NEXT: addi a2, sp, 16
+; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vadd.vv v8, v24, v8
-; RV32-NEXT: vsrl.vi v24, v8, 4
-; RV32-NEXT: vadd.vv v8, v8, v24
-; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 4
-; RV32-NEXT: add a0, sp, a0
-; RV32-NEXT: addi a0, a0, 48
-; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v24, (a2), zero
+; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vi v0, v8, 4
+; RV32-NEXT: vadd.vv v8, v8, v0
+; RV32-NEXT: addi a2, sp, 48
+; RV32-NEXT: vl8r.v v0, (a2) # Unknown-size Folded Reload
+; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
+; RV32-NEXT: vand.vv v0, v0, v16
+; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vand.vv v8, v8, v16
-; RV32-NEXT: addi a0, sp, 48
-; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; RV32-NEXT: vmul.vv v8, v8, v16
-; RV32-NEXT: vsrl.vx v16, v8, a1
-; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 3
-; RV32-NEXT: add a0, sp, a0
-; RV32-NEXT: addi a0, a0, 48
-; RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
+; RV32-NEXT: vmul.vv v16, v0, v24
+; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vmul.vv v24, v8, v24
+; RV32-NEXT: li a2, 56
+; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vx v8, v16, a2
+; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vx v16, v24, a2
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 5
+; RV32-NEXT: slli a0, a0, 4
; RV32-NEXT: add sp, sp, a0
; RV32-NEXT: addi sp, sp, 48
; RV32-NEXT: ret
@@ -4060,24 +4139,28 @@ define <8 x i64> @vp_ctlz_zero_undef_v8i64_unmasked(<8 x i64> %va, i32 zeroext %
define <15 x i64> @vp_ctlz_zero_undef_v15i64(<15 x i64> %va, <15 x i1> %m, i32 zeroext %evl) {
; RV32-LABEL: vp_ctlz_zero_undef_v15i64:
; RV32: # %bb.0:
-; RV32-NEXT: addi sp, sp, -32
-; RV32-NEXT: .cfi_def_cfa_offset 32
+; RV32-NEXT: addi sp, sp, -48
+; RV32-NEXT: .cfi_def_cfa_offset 48
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 4
+; RV32-NEXT: sub sp, sp, a1
+; RV32-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x30, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 48 + 16 * vlenb
; RV32-NEXT: lui a1, 349525
; RV32-NEXT: addi a1, a1, 1365
-; RV32-NEXT: sw a1, 28(sp)
-; RV32-NEXT: sw a1, 24(sp)
+; RV32-NEXT: sw a1, 44(sp)
+; RV32-NEXT: sw a1, 40(sp)
; RV32-NEXT: lui a1, 209715
; RV32-NEXT: addi a1, a1, 819
-; RV32-NEXT: sw a1, 20(sp)
-; RV32-NEXT: sw a1, 16(sp)
+; RV32-NEXT: sw a1, 36(sp)
+; RV32-NEXT: sw a1, 32(sp)
; RV32-NEXT: lui a1, 61681
; RV32-NEXT: addi a1, a1, -241
-; RV32-NEXT: sw a1, 12(sp)
-; RV32-NEXT: sw a1, 8(sp)
+; RV32-NEXT: sw a1, 28(sp)
+; RV32-NEXT: sw a1, 24(sp)
; RV32-NEXT: lui a1, 4112
; RV32-NEXT: addi a1, a1, 257
-; RV32-NEXT: sw a1, 4(sp)
-; RV32-NEXT: sw a1, 0(sp)
+; RV32-NEXT: sw a1, 20(sp)
+; RV32-NEXT: sw a1, 16(sp)
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vsrl.vi v16, v8, 1, v0.t
; RV32-NEXT: vor.vv v8, v8, v16, v0.t
@@ -4092,37 +4175,60 @@ define <15 x i64> @vp_ctlz_zero_undef_v15i64(<15 x i64> %va, <15 x i1> %m, i32 z
; RV32-NEXT: li a1, 32
; RV32-NEXT: vsrl.vx v16, v8, a1, v0.t
; RV32-NEXT: vor.vv v8, v8, v16, v0.t
-; RV32-NEXT: vnot.v v8, v8, v0.t
-; RV32-NEXT: vsrl.vi v16, v8, 1, v0.t
-; RV32-NEXT: addi a1, sp, 24
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v24, (a1), zero
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v16, v24, v0.t
-; RV32-NEXT: vsub.vv v8, v8, v16, v0.t
-; RV32-NEXT: addi a1, sp, 16
+; RV32-NEXT: vnot.v v24, v8, v0.t
+; RV32-NEXT: addi a1, sp, 48
+; RV32-NEXT: vs8r.v v24, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: addi a1, sp, 40
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v8, (a1), zero
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 3
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 48
+; RV32-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: addi a1, sp, 32
; RV32-NEXT: vlse64.v v16, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vi v8, v24, 1, v0.t
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 3
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 48
+; RV32-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v8, v8, v24, v0.t
+; RV32-NEXT: addi a1, sp, 48
+; RV32-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vsub.vv v8, v24, v8, v0.t
; RV32-NEXT: vand.vv v24, v8, v16, v0.t
; RV32-NEXT: vsrl.vi v8, v8, 2, v0.t
; RV32-NEXT: vand.vv v8, v8, v16, v0.t
; RV32-NEXT: vadd.vv v8, v24, v8, v0.t
-; RV32-NEXT: vsrl.vi v16, v8, 4, v0.t
-; RV32-NEXT: vadd.vv v8, v8, v16, v0.t
-; RV32-NEXT: addi a1, sp, 8
+; RV32-NEXT: addi a1, sp, 24
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV32-NEXT: vlse64.v v16, (a1), zero
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 3
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 48
+; RV32-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: addi a1, sp, 16
+; RV32-NEXT: vlse64.v v24, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vi v16, v8, 4, v0.t
+; RV32-NEXT: vadd.vv v8, v8, v16, v0.t
+; RV32-NEXT: csrr a0, vlenb
+; RV32-NEXT: slli a0, a0, 3
+; RV32-NEXT: add a0, sp, a0
+; RV32-NEXT: addi a0, a0, 48
+; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; RV32-NEXT: vand.vv v8, v8, v16, v0.t
-; RV32-NEXT: mv a1, sp
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a1), zero
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vmul.vv v8, v8, v16, v0.t
+; RV32-NEXT: vmul.vv v8, v8, v24, v0.t
; RV32-NEXT: li a0, 56
; RV32-NEXT: vsrl.vx v8, v8, a0, v0.t
-; RV32-NEXT: addi sp, sp, 32
+; RV32-NEXT: csrr a0, vlenb
+; RV32-NEXT: slli a0, a0, 4
+; RV32-NEXT: add sp, sp, a0
+; RV32-NEXT: addi sp, sp, 48
; RV32-NEXT: ret
;
; RV64-LABEL: vp_ctlz_zero_undef_v15i64:
@@ -4212,33 +4318,29 @@ define <15 x i64> @vp_ctlz_zero_undef_v15i64_unmasked(<15 x i64> %va, i32 zeroex
; RV32-NEXT: vsrl.vx v16, v8, a1
; RV32-NEXT: vor.vv v8, v8, v16
; RV32-NEXT: vnot.v v8, v8
-; RV32-NEXT: vsrl.vi v16, v8, 1
; RV32-NEXT: addi a1, sp, 24
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v16, (a1), zero
+; RV32-NEXT: addi a1, sp, 16
; RV32-NEXT: vlse64.v v24, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v16, v24
+; RV32-NEXT: vsrl.vi v0, v8, 1
+; RV32-NEXT: vand.vv v16, v0, v16
; RV32-NEXT: vsub.vv v8, v8, v16
-; RV32-NEXT: addi a1, sp, 16
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a1), zero
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v24, v8, v16
+; RV32-NEXT: vand.vv v16, v8, v24
; RV32-NEXT: vsrl.vi v8, v8, 2
-; RV32-NEXT: vand.vv v8, v8, v16
-; RV32-NEXT: vadd.vv v8, v24, v8
-; RV32-NEXT: vsrl.vi v16, v8, 4
-; RV32-NEXT: vadd.vv v8, v8, v16
+; RV32-NEXT: vand.vv v8, v8, v24
+; RV32-NEXT: vadd.vv v8, v16, v8
; RV32-NEXT: addi a1, sp, 8
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV32-NEXT: vlse64.v v16, (a1), zero
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v8, v8, v16
; RV32-NEXT: mv a1, sp
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a1), zero
+; RV32-NEXT: vlse64.v v24, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vmul.vv v8, v8, v16
+; RV32-NEXT: vsrl.vi v0, v8, 4
+; RV32-NEXT: vadd.vv v8, v8, v0
+; RV32-NEXT: vand.vv v8, v8, v16
+; RV32-NEXT: vmul.vv v8, v8, v24
; RV32-NEXT: li a0, 56
; RV32-NEXT: vsrl.vx v8, v8, a0
; RV32-NEXT: addi sp, sp, 32
@@ -4298,24 +4400,28 @@ define <15 x i64> @vp_ctlz_zero_undef_v15i64_unmasked(<15 x i64> %va, i32 zeroex
define <16 x i64> @vp_ctlz_zero_undef_v16i64(<16 x i64> %va, <16 x i1> %m, i32 zeroext %evl) {
; RV32-LABEL: vp_ctlz_zero_undef_v16i64:
; RV32: # %bb.0:
-; RV32-NEXT: addi sp, sp, -32
-; RV32-NEXT: .cfi_def_cfa_offset 32
+; RV32-NEXT: addi sp, sp, -48
+; RV32-NEXT: .cfi_def_cfa_offset 48
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 4
+; RV32-NEXT: sub sp, sp, a1
+; RV32-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x30, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 48 + 16 * vlenb
; RV32-NEXT: lui a1, 349525
; RV32-NEXT: addi a1, a1, 1365
-; RV32-NEXT: sw a1, 28(sp)
-; RV32-NEXT: sw a1, 24(sp)
+; RV32-NEXT: sw a1, 44(sp)
+; RV32-NEXT: sw a1, 40(sp)
; RV32-NEXT: lui a1, 209715
; RV32-NEXT: addi a1, a1, 819
-; RV32-NEXT: sw a1, 20(sp)
-; RV32-NEXT: sw a1, 16(sp)
+; RV32-NEXT: sw a1, 36(sp)
+; RV32-NEXT: sw a1, 32(sp)
; RV32-NEXT: lui a1, 61681
; RV32-NEXT: addi a1, a1, -241
-; RV32-NEXT: sw a1, 12(sp)
-; RV32-NEXT: sw a1, 8(sp)
+; RV32-NEXT: sw a1, 28(sp)
+; RV32-NEXT: sw a1, 24(sp)
; RV32-NEXT: lui a1, 4112
; RV32-NEXT: addi a1, a1, 257
-; RV32-NEXT: sw a1, 4(sp)
-; RV32-NEXT: sw a1, 0(sp)
+; RV32-NEXT: sw a1, 20(sp)
+; RV32-NEXT: sw a1, 16(sp)
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vsrl.vi v16, v8, 1, v0.t
; RV32-NEXT: vor.vv v8, v8, v16, v0.t
@@ -4330,37 +4436,60 @@ define <16 x i64> @vp_ctlz_zero_undef_v16i64(<16 x i64> %va, <16 x i1> %m, i32 z
; RV32-NEXT: li a1, 32
; RV32-NEXT: vsrl.vx v16, v8, a1, v0.t
; RV32-NEXT: vor.vv v8, v8, v16, v0.t
-; RV32-NEXT: vnot.v v8, v8, v0.t
-; RV32-NEXT: vsrl.vi v16, v8, 1, v0.t
-; RV32-NEXT: addi a1, sp, 24
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v24, (a1), zero
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v16, v24, v0.t
-; RV32-NEXT: vsub.vv v8, v8, v16, v0.t
-; RV32-NEXT: addi a1, sp, 16
+; RV32-NEXT: vnot.v v24, v8, v0.t
+; RV32-NEXT: addi a1, sp, 48
+; RV32-NEXT: vs8r.v v24, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: addi a1, sp, 40
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v8, (a1), zero
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 3
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 48
+; RV32-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: addi a1, sp, 32
; RV32-NEXT: vlse64.v v16, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vi v8, v24, 1, v0.t
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 3
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 48
+; RV32-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v8, v8, v24, v0.t
+; RV32-NEXT: addi a1, sp, 48
+; RV32-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vsub.vv v8, v24, v8, v0.t
; RV32-NEXT: vand.vv v24, v8, v16, v0.t
; RV32-NEXT: vsrl.vi v8, v8, 2, v0.t
; RV32-NEXT: vand.vv v8, v8, v16, v0.t
; RV32-NEXT: vadd.vv v8, v24, v8, v0.t
-; RV32-NEXT: vsrl.vi v16, v8, 4, v0.t
-; RV32-NEXT: vadd.vv v8, v8, v16, v0.t
-; RV32-NEXT: addi a1, sp, 8
+; RV32-NEXT: addi a1, sp, 24
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV32-NEXT: vlse64.v v16, (a1), zero
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 3
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 48
+; RV32-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: addi a1, sp, 16
+; RV32-NEXT: vlse64.v v24, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vi v16, v8, 4, v0.t
+; RV32-NEXT: vadd.vv v8, v8, v16, v0.t
+; RV32-NEXT: csrr a0, vlenb
+; RV32-NEXT: slli a0, a0, 3
+; RV32-NEXT: add a0, sp, a0
+; RV32-NEXT: addi a0, a0, 48
+; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; RV32-NEXT: vand.vv v8, v8, v16, v0.t
-; RV32-NEXT: mv a1, sp
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a1), zero
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vmul.vv v8, v8, v16, v0.t
+; RV32-NEXT: vmul.vv v8, v8, v24, v0.t
; RV32-NEXT: li a0, 56
; RV32-NEXT: vsrl.vx v8, v8, a0, v0.t
-; RV32-NEXT: addi sp, sp, 32
+; RV32-NEXT: csrr a0, vlenb
+; RV32-NEXT: slli a0, a0, 4
+; RV32-NEXT: add sp, sp, a0
+; RV32-NEXT: addi sp, sp, 48
; RV32-NEXT: ret
;
; RV64-LABEL: vp_ctlz_zero_undef_v16i64:
@@ -4450,33 +4579,29 @@ define <16 x i64> @vp_ctlz_zero_undef_v16i64_unmasked(<16 x i64> %va, i32 zeroex
; RV32-NEXT: vsrl.vx v16, v8, a1
; RV32-NEXT: vor.vv v8, v8, v16
; RV32-NEXT: vnot.v v8, v8
-; RV32-NEXT: vsrl.vi v16, v8, 1
; RV32-NEXT: addi a1, sp, 24
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v16, (a1), zero
+; RV32-NEXT: addi a1, sp, 16
; RV32-NEXT: vlse64.v v24, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v16, v24
+; RV32-NEXT: vsrl.vi v0, v8, 1
+; RV32-NEXT: vand.vv v16, v0, v16
; RV32-NEXT: vsub.vv v8, v8, v16
-; RV32-NEXT: addi a1, sp, 16
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a1), zero
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v24, v8, v16
+; RV32-NEXT: vand.vv v16, v8, v24
; RV32-NEXT: vsrl.vi v8, v8, 2
-; RV32-NEXT: vand.vv v8, v8, v16
-; RV32-NEXT: vadd.vv v8, v24, v8
-; RV32-NEXT: vsrl.vi v16, v8, 4
-; RV32-NEXT: vadd.vv v8, v8, v16
+; RV32-NEXT: vand.vv v8, v8, v24
+; RV32-NEXT: vadd.vv v8, v16, v8
; RV32-NEXT: addi a1, sp, 8
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV32-NEXT: vlse64.v v16, (a1), zero
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v8, v8, v16
; RV32-NEXT: mv a1, sp
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a1), zero
+; RV32-NEXT: vlse64.v v24, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vmul.vv v8, v8, v16
+; RV32-NEXT: vsrl.vi v0, v8, 4
+; RV32-NEXT: vadd.vv v8, v8, v0
+; RV32-NEXT: vand.vv v8, v8, v16
+; RV32-NEXT: vmul.vv v8, v8, v24
; RV32-NEXT: li a0, 56
; RV32-NEXT: vsrl.vx v8, v8, a0
; RV32-NEXT: addi sp, sp, 32
@@ -4544,7 +4669,7 @@ define <32 x i64> @vp_ctlz_zero_undef_v32i64(<32 x i64> %va, <32 x i1> %m, i32 z
; RV32-NEXT: sub sp, sp, a1
; RV32-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x30, 0x22, 0x11, 0x38, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 48 + 56 * vlenb
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: slli a1, a1, 5
+; RV32-NEXT: slli a1, a1, 4
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 48
; RV32-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
@@ -4588,111 +4713,145 @@ define <32 x i64> @vp_ctlz_zero_undef_v32i64(<32 x i64> %va, <32 x i1> %m, i32 z
; RV32-NEXT: vor.vv v8, v8, v16, v0.t
; RV32-NEXT: vnot.v v8, v8, v0.t
; RV32-NEXT: csrr a3, vlenb
+; RV32-NEXT: slli a3, a3, 5
+; RV32-NEXT: add a3, sp, a3
+; RV32-NEXT: addi a3, a3, 48
+; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
+; RV32-NEXT: addi a3, sp, 40
+; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v8, (a3), zero
+; RV32-NEXT: csrr a3, vlenb
; RV32-NEXT: li a4, 40
; RV32-NEXT: mul a3, a3, a4
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: vsrl.vi v8, v8, 1, v0.t
+; RV32-NEXT: addi a3, sp, 32
+; RV32-NEXT: vlse64.v v8, (a3), zero
; RV32-NEXT: csrr a3, vlenb
; RV32-NEXT: li a4, 48
; RV32-NEXT: mul a3, a3, a4
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: addi a3, sp, 40
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v8, (a3), zero
+; RV32-NEXT: csrr a3, vlenb
+; RV32-NEXT: slli a3, a3, 5
+; RV32-NEXT: add a3, sp, a3
+; RV32-NEXT: addi a3, a3, 48
+; RV32-NEXT: vl8r.v v16, (a3) # Unknown-size Folded Reload
+; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vi v8, v16, 1, v0.t
; RV32-NEXT: csrr a3, vlenb
; RV32-NEXT: li a4, 24
; RV32-NEXT: mul a3, a3, a4
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: li a4, 48
+; RV32-NEXT: li a4, 40
; RV32-NEXT: mul a3, a3, a4
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
; RV32-NEXT: vl8r.v v16, (a3) # Unknown-size Folded Reload
-; RV32-NEXT: vand.vv v16, v16, v8, v0.t
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: li a4, 40
+; RV32-NEXT: li a4, 24
; RV32-NEXT: mul a3, a3, a4
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
; RV32-NEXT: vl8r.v v8, (a3) # Unknown-size Folded Reload
-; RV32-NEXT: vsub.vv v8, v8, v16, v0.t
+; RV32-NEXT: vand.vv v8, v8, v16, v0.t
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: li a4, 40
-; RV32-NEXT: mul a3, a3, a4
+; RV32-NEXT: slli a3, a3, 5
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
-; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: addi a3, sp, 32
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v8, (a3), zero
+; RV32-NEXT: vl8r.v v16, (a3) # Unknown-size Folded Reload
+; RV32-NEXT: vsub.vv v16, v16, v8, v0.t
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: li a4, 48
-; RV32-NEXT: mul a3, a3, a4
+; RV32-NEXT: slli a3, a3, 5
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
-; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
+; RV32-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: li a4, 40
-; RV32-NEXT: mul a3, a3, a4
+; RV32-NEXT: slli a3, a3, 5
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
; RV32-NEXT: vl8r.v v16, (a3) # Unknown-size Folded Reload
+; RV32-NEXT: csrr a3, vlenb
+; RV32-NEXT: li a4, 48
+; RV32-NEXT: mul a3, a3, a4
+; RV32-NEXT: add a3, sp, a3
+; RV32-NEXT: addi a3, a3, 48
+; RV32-NEXT: vl8r.v v8, (a3) # Unknown-size Folded Reload
; RV32-NEXT: vand.vv v16, v16, v8, v0.t
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: slli a3, a3, 4
+; RV32-NEXT: li a4, 24
+; RV32-NEXT: mul a3, a3, a4
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
; RV32-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: li a4, 40
-; RV32-NEXT: mul a3, a3, a4
+; RV32-NEXT: slli a3, a3, 5
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
; RV32-NEXT: vl8r.v v16, (a3) # Unknown-size Folded Reload
; RV32-NEXT: vsrl.vi v16, v16, 2, v0.t
; RV32-NEXT: vand.vv v16, v16, v8, v0.t
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: slli a3, a3, 4
+; RV32-NEXT: li a4, 24
+; RV32-NEXT: mul a3, a3, a4
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
; RV32-NEXT: vl8r.v v8, (a3) # Unknown-size Folded Reload
; RV32-NEXT: vadd.vv v8, v8, v16, v0.t
-; RV32-NEXT: vsrl.vi v16, v8, 4, v0.t
-; RV32-NEXT: vadd.vv v16, v8, v16, v0.t
+; RV32-NEXT: csrr a3, vlenb
+; RV32-NEXT: slli a3, a3, 3
+; RV32-NEXT: add a3, sp, a3
+; RV32-NEXT: addi a3, a3, 48
+; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
; RV32-NEXT: addi a3, sp, 24
+; RV32-NEXT: addi a4, sp, 16
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV32-NEXT: vlse64.v v8, (a3), zero
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: li a4, 40
-; RV32-NEXT: mul a3, a3, a4
+; RV32-NEXT: li a5, 24
+; RV32-NEXT: mul a3, a3, a5
+; RV32-NEXT: add a3, sp, a3
+; RV32-NEXT: addi a3, a3, 48
+; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
+; RV32-NEXT: vlse64.v v8, (a4), zero
+; RV32-NEXT: csrr a3, vlenb
+; RV32-NEXT: slli a3, a3, 5
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v8, v16, v8, v0.t
-; RV32-NEXT: addi a3, sp, 16
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a3), zero
; RV32-NEXT: csrr a3, vlenb
; RV32-NEXT: slli a3, a3, 3
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
-; RV32-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
+; RV32-NEXT: vl8r.v v8, (a3) # Unknown-size Folded Reload
; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
-; RV32-NEXT: vmul.vv v8, v8, v16, v0.t
+; RV32-NEXT: vsrl.vi v16, v8, 4, v0.t
+; RV32-NEXT: addi a2, sp, 48
+; RV32-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
+; RV32-NEXT: vl8r.v v16, (a2) # Unknown-size Folded Reload
+; RV32-NEXT: vadd.vv v16, v8, v16, v0.t
+; RV32-NEXT: csrr a2, vlenb
+; RV32-NEXT: li a3, 24
+; RV32-NEXT: mul a2, a2, a3
+; RV32-NEXT: add a2, sp, a2
+; RV32-NEXT: addi a2, a2, 48
+; RV32-NEXT: vl8r.v v8, (a2) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v16, v16, v8, v0.t
+; RV32-NEXT: csrr a2, vlenb
+; RV32-NEXT: slli a2, a2, 5
+; RV32-NEXT: add a2, sp, a2
+; RV32-NEXT: addi a2, a2, 48
+; RV32-NEXT: vl8r.v v8, (a2) # Unknown-size Folded Reload
+; RV32-NEXT: vmul.vv v8, v16, v8, v0.t
; RV32-NEXT: li a2, 56
; RV32-NEXT: vsrl.vx v8, v8, a2, v0.t
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: slli a3, a3, 4
+; RV32-NEXT: slli a3, a3, 3
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
@@ -4700,13 +4859,13 @@ define <32 x i64> @vp_ctlz_zero_undef_v32i64(<32 x i64> %va, <32 x i1> %m, i32 z
; RV32-NEXT: sltu a0, a0, a3
; RV32-NEXT: addi a0, a0, -1
; RV32-NEXT: and a0, a0, a3
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vmv1r.v v0, v24
-; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 5
-; RV32-NEXT: add a0, sp, a0
-; RV32-NEXT: addi a0, a0, 48
-; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: csrr a3, vlenb
+; RV32-NEXT: slli a3, a3, 4
+; RV32-NEXT: add a3, sp, a3
+; RV32-NEXT: addi a3, a3, 48
+; RV32-NEXT: vl8r.v v16, (a3) # Unknown-size Folded Reload
+; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vsrl.vi v8, v16, 1, v0.t
; RV32-NEXT: vor.vv v8, v16, v8, v0.t
; RV32-NEXT: vsrl.vi v16, v8, 2, v0.t
@@ -4724,18 +4883,18 @@ define <32 x i64> @vp_ctlz_zero_undef_v32i64(<32 x i64> %va, <32 x i1> %m, i32 z
; RV32-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; RV32-NEXT: vsrl.vi v8, v8, 1, v0.t
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 5
+; RV32-NEXT: slli a0, a0, 4
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 48
; RV32-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: li a1, 24
+; RV32-NEXT: li a1, 40
; RV32-NEXT: mul a0, a0, a1
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 48
; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 5
+; RV32-NEXT: slli a0, a0, 4
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 48
; RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
@@ -4744,17 +4903,35 @@ define <32 x i64> @vp_ctlz_zero_undef_v32i64(<32 x i64> %va, <32 x i1> %m, i32 z
; RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; RV32-NEXT: vsub.vv v8, v8, v16, v0.t
; RV32-NEXT: csrr a0, vlenb
+; RV32-NEXT: li a1, 40
+; RV32-NEXT: mul a0, a0, a1
+; RV32-NEXT: add a0, sp, a0
+; RV32-NEXT: addi a0, a0, 48
+; RV32-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; RV32-NEXT: csrr a0, vlenb
; RV32-NEXT: li a1, 48
; RV32-NEXT: mul a0, a0, a1
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 48
+; RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: csrr a0, vlenb
+; RV32-NEXT: li a1, 40
+; RV32-NEXT: mul a0, a0, a1
+; RV32-NEXT: add a0, sp, a0
+; RV32-NEXT: addi a0, a0, 48
; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; RV32-NEXT: vand.vv v16, v8, v16, v0.t
+; RV32-NEXT: vand.vv v16, v16, v8, v0.t
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 5
+; RV32-NEXT: slli a0, a0, 4
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 48
; RV32-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
+; RV32-NEXT: csrr a0, vlenb
+; RV32-NEXT: li a1, 40
+; RV32-NEXT: mul a0, a0, a1
+; RV32-NEXT: add a0, sp, a0
+; RV32-NEXT: addi a0, a0, 48
+; RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; RV32-NEXT: vsrl.vi v8, v8, 2, v0.t
; RV32-NEXT: csrr a0, vlenb
; RV32-NEXT: li a1, 48
@@ -4764,7 +4941,7 @@ define <32 x i64> @vp_ctlz_zero_undef_v32i64(<32 x i64> %va, <32 x i1> %m, i32 z
; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; RV32-NEXT: vand.vv v8, v8, v16, v0.t
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 5
+; RV32-NEXT: slli a0, a0, 4
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 48
; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
@@ -4772,21 +4949,21 @@ define <32 x i64> @vp_ctlz_zero_undef_v32i64(<32 x i64> %va, <32 x i1> %m, i32 z
; RV32-NEXT: vsrl.vi v16, v8, 4, v0.t
; RV32-NEXT: vadd.vv v8, v8, v16, v0.t
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: li a1, 40
+; RV32-NEXT: li a1, 24
; RV32-NEXT: mul a0, a0, a1
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 48
; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; RV32-NEXT: vand.vv v8, v8, v16, v0.t
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 3
+; RV32-NEXT: slli a0, a0, 5
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 48
; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; RV32-NEXT: vmul.vv v8, v8, v16, v0.t
; RV32-NEXT: vsrl.vx v16, v8, a2, v0.t
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 4
+; RV32-NEXT: slli a0, a0, 3
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 48
; RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
@@ -4810,8 +4987,8 @@ define <32 x i64> @vp_ctlz_zero_undef_v32i64(<32 x i64> %va, <32 x i1> %m, i32 z
; RV64-NEXT: add a1, sp, a1
; RV64-NEXT: addi a1, a1, 16
; RV64-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
-; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64-NEXT: li a2, 16
+; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64-NEXT: vslidedown.vi v24, v0, 2
; RV64-NEXT: mv a1, a0
; RV64-NEXT: bltu a0, a2, .LBB70_2
@@ -4868,13 +5045,13 @@ define <32 x i64> @vp_ctlz_zero_undef_v32i64(<32 x i64> %va, <32 x i1> %m, i32 z
; RV64-NEXT: sltu a0, a0, a7
; RV64-NEXT: addi a0, a0, -1
; RV64-NEXT: and a0, a0, a7
-; RV64-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV64-NEXT: vmv1r.v v0, v24
-; RV64-NEXT: csrr a0, vlenb
-; RV64-NEXT: slli a0, a0, 3
-; RV64-NEXT: add a0, sp, a0
-; RV64-NEXT: addi a0, a0, 16
-; RV64-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; RV64-NEXT: csrr a7, vlenb
+; RV64-NEXT: slli a7, a7, 3
+; RV64-NEXT: add a7, sp, a7
+; RV64-NEXT: addi a7, a7, 16
+; RV64-NEXT: vl8r.v v8, (a7) # Unknown-size Folded Reload
+; RV64-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV64-NEXT: vsrl.vi v16, v8, 1, v0.t
; RV64-NEXT: vor.vv v16, v8, v16, v0.t
; RV64-NEXT: vsrl.vi v8, v16, 2, v0.t
@@ -4917,10 +5094,14 @@ define <32 x i64> @vp_ctlz_zero_undef_v32i64_unmasked(<32 x i64> %va, i32 zeroex
; RV32-NEXT: addi sp, sp, -48
; RV32-NEXT: .cfi_def_cfa_offset 48
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: slli a1, a1, 5
+; RV32-NEXT: slli a1, a1, 4
; RV32-NEXT: sub sp, sp, a1
-; RV32-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x30, 0x22, 0x11, 0x20, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 48 + 32 * vlenb
-; RV32-NEXT: vmv8r.v v24, v16
+; RV32-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x30, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 48 + 16 * vlenb
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 3
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 48
+; RV32-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
; RV32-NEXT: lui a1, 349525
; RV32-NEXT: addi a1, a1, 1365
; RV32-NEXT: sw a1, 44(sp)
@@ -4944,74 +5125,8 @@ define <32 x i64> @vp_ctlz_zero_undef_v32i64_unmasked(<32 x i64> %va, i32 zeroex
; RV32-NEXT: li a1, 16
; RV32-NEXT: .LBB71_2:
; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; RV32-NEXT: vsrl.vi v16, v8, 1
-; RV32-NEXT: vor.vv v8, v8, v16
-; RV32-NEXT: vsrl.vi v16, v8, 2
-; RV32-NEXT: vor.vv v8, v8, v16
-; RV32-NEXT: vsrl.vi v16, v8, 4
-; RV32-NEXT: vor.vv v8, v8, v16
-; RV32-NEXT: vsrl.vi v16, v8, 8
-; RV32-NEXT: vor.vv v8, v8, v16
-; RV32-NEXT: vsrl.vi v16, v8, 16
-; RV32-NEXT: vor.vv v8, v8, v16
-; RV32-NEXT: li a2, 32
-; RV32-NEXT: vsrl.vx v16, v8, a2
-; RV32-NEXT: vor.vv v8, v8, v16
-; RV32-NEXT: vnot.v v8, v8
-; RV32-NEXT: vsrl.vi v16, v8, 1
-; RV32-NEXT: addi a3, sp, 40
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v0, (a3), zero
-; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: li a4, 24
-; RV32-NEXT: mul a3, a3, a4
-; RV32-NEXT: add a3, sp, a3
-; RV32-NEXT: addi a3, a3, 48
-; RV32-NEXT: vs8r.v v0, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v16, v0
-; RV32-NEXT: vsub.vv v8, v8, v16
-; RV32-NEXT: addi a3, sp, 32
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v0, (a3), zero
-; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v8, v0
-; RV32-NEXT: vsrl.vi v8, v8, 2
-; RV32-NEXT: vand.vv v8, v8, v0
-; RV32-NEXT: vadd.vv v8, v16, v8
-; RV32-NEXT: vsrl.vi v16, v8, 4
-; RV32-NEXT: vadd.vv v8, v8, v16
-; RV32-NEXT: addi a3, sp, 24
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a3), zero
-; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: slli a3, a3, 4
-; RV32-NEXT: add a3, sp, a3
-; RV32-NEXT: addi a3, a3, 48
-; RV32-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v8, v8, v16
-; RV32-NEXT: addi a3, sp, 16
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a3), zero
-; RV32-NEXT: addi a3, sp, 48
-; RV32-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; RV32-NEXT: vmul.vv v8, v8, v16
-; RV32-NEXT: li a1, 56
-; RV32-NEXT: vsrl.vx v8, v8, a1
-; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: slli a3, a3, 3
-; RV32-NEXT: add a3, sp, a3
-; RV32-NEXT: addi a3, a3, 48
-; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: addi a3, a0, -16
-; RV32-NEXT: sltu a0, a0, a3
-; RV32-NEXT: addi a0, a0, -1
-; RV32-NEXT: and a0, a0, a3
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vsrl.vi v8, v24, 1
-; RV32-NEXT: vor.vv v8, v24, v8
+; RV32-NEXT: vsrl.vi v24, v8, 1
+; RV32-NEXT: vor.vv v8, v8, v24
; RV32-NEXT: vsrl.vi v24, v8, 2
; RV32-NEXT: vor.vv v8, v8, v24
; RV32-NEXT: vsrl.vi v24, v8, 4
@@ -5020,41 +5135,84 @@ define <32 x i64> @vp_ctlz_zero_undef_v32i64_unmasked(<32 x i64> %va, i32 zeroex
; RV32-NEXT: vor.vv v8, v8, v24
; RV32-NEXT: vsrl.vi v24, v8, 16
; RV32-NEXT: vor.vv v8, v8, v24
+; RV32-NEXT: li a2, 32
; RV32-NEXT: vsrl.vx v24, v8, a2
; RV32-NEXT: vor.vv v8, v8, v24
-; RV32-NEXT: vnot.v v8, v8
-; RV32-NEXT: vsrl.vi v24, v8, 1
-; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: li a2, 24
-; RV32-NEXT: mul a0, a0, a2
-; RV32-NEXT: add a0, sp, a0
-; RV32-NEXT: addi a0, a0, 48
-; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vnot.v v0, v8
+; RV32-NEXT: addi a3, sp, 40
+; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v16, (a3), zero
+; RV32-NEXT: addi a3, sp, 32
+; RV32-NEXT: vlse64.v v8, (a3), zero
+; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vi v24, v0, 1
; RV32-NEXT: vand.vv v24, v24, v16
-; RV32-NEXT: vsub.vv v8, v8, v24
-; RV32-NEXT: vand.vv v24, v8, v0
-; RV32-NEXT: vsrl.vi v8, v8, 2
-; RV32-NEXT: vand.vv v8, v8, v0
+; RV32-NEXT: vsub.vv v24, v0, v24
+; RV32-NEXT: vand.vv v0, v24, v8
+; RV32-NEXT: vsrl.vi v24, v24, 2
+; RV32-NEXT: vand.vv v24, v24, v8
+; RV32-NEXT: vadd.vv v24, v0, v24
+; RV32-NEXT: vsrl.vi v0, v24, 4
+; RV32-NEXT: vadd.vv v24, v24, v0
+; RV32-NEXT: addi a3, sp, 48
+; RV32-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
+; RV32-NEXT: addi a3, a0, -16
+; RV32-NEXT: sltu a0, a0, a3
+; RV32-NEXT: addi a0, a0, -1
+; RV32-NEXT: and a0, a0, a3
+; RV32-NEXT: csrr a3, vlenb
+; RV32-NEXT: slli a3, a3, 3
+; RV32-NEXT: add a3, sp, a3
+; RV32-NEXT: addi a3, a3, 48
+; RV32-NEXT: vl8r.v v0, (a3) # Unknown-size Folded Reload
+; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vi v24, v0, 1
+; RV32-NEXT: vor.vv v24, v0, v24
+; RV32-NEXT: vsrl.vi v0, v24, 2
+; RV32-NEXT: vor.vv v24, v24, v0
+; RV32-NEXT: vsrl.vi v0, v24, 4
+; RV32-NEXT: vor.vv v24, v24, v0
+; RV32-NEXT: vsrl.vi v0, v24, 8
+; RV32-NEXT: vor.vv v24, v24, v0
+; RV32-NEXT: vsrl.vi v0, v24, 16
+; RV32-NEXT: vor.vv v24, v24, v0
+; RV32-NEXT: vsrl.vx v0, v24, a2
+; RV32-NEXT: vor.vv v24, v24, v0
+; RV32-NEXT: vnot.v v24, v24
+; RV32-NEXT: vsrl.vi v0, v24, 1
+; RV32-NEXT: vand.vv v16, v0, v16
+; RV32-NEXT: addi a2, sp, 24
+; RV32-NEXT: vsub.vv v16, v24, v16
+; RV32-NEXT: vand.vv v24, v16, v8
+; RV32-NEXT: vsrl.vi v16, v16, 2
+; RV32-NEXT: vand.vv v8, v16, v8
+; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v16, (a2), zero
+; RV32-NEXT: addi a2, sp, 16
+; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vadd.vv v8, v24, v8
-; RV32-NEXT: vsrl.vi v24, v8, 4
-; RV32-NEXT: vadd.vv v8, v8, v24
-; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 4
-; RV32-NEXT: add a0, sp, a0
-; RV32-NEXT: addi a0, a0, 48
-; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v24, (a2), zero
+; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vi v0, v8, 4
+; RV32-NEXT: vadd.vv v8, v8, v0
+; RV32-NEXT: addi a2, sp, 48
+; RV32-NEXT: vl8r.v v0, (a2) # Unknown-size Folded Reload
+; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
+; RV32-NEXT: vand.vv v0, v0, v16
+; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vand.vv v8, v8, v16
-; RV32-NEXT: addi a0, sp, 48
-; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; RV32-NEXT: vmul.vv v8, v8, v16
-; RV32-NEXT: vsrl.vx v16, v8, a1
-; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 3
-; RV32-NEXT: add a0, sp, a0
-; RV32-NEXT: addi a0, a0, 48
-; RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
+; RV32-NEXT: vmul.vv v16, v0, v24
+; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vmul.vv v24, v8, v24
+; RV32-NEXT: li a2, 56
+; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vx v8, v16, a2
+; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vx v16, v24, a2
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 5
+; RV32-NEXT: slli a0, a0, 4
; RV32-NEXT: add sp, sp, a0
; RV32-NEXT: addi sp, sp, 48
; RV32-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-ctpop-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-ctpop-vp.ll
index 0b6d8b33394d5..5fceab869ab85 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-ctpop-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-ctpop-vp.ll
@@ -1119,55 +1119,93 @@ declare <15 x i64> @llvm.vp.ctpop.v15i64(<15 x i64>, <15 x i1>, i32)
define <15 x i64> @vp_ctpop_v15i64(<15 x i64> %va, <15 x i1> %m, i32 zeroext %evl) {
; RV32-LABEL: vp_ctpop_v15i64:
; RV32: # %bb.0:
-; RV32-NEXT: addi sp, sp, -32
-; RV32-NEXT: .cfi_def_cfa_offset 32
+; RV32-NEXT: addi sp, sp, -48
+; RV32-NEXT: .cfi_def_cfa_offset 48
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: li a2, 24
+; RV32-NEXT: mul a1, a1, a2
+; RV32-NEXT: sub sp, sp, a1
+; RV32-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x30, 0x22, 0x11, 0x18, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 48 + 24 * vlenb
; RV32-NEXT: lui a1, 349525
; RV32-NEXT: addi a1, a1, 1365
-; RV32-NEXT: sw a1, 28(sp)
-; RV32-NEXT: sw a1, 24(sp)
+; RV32-NEXT: sw a1, 44(sp)
+; RV32-NEXT: sw a1, 40(sp)
; RV32-NEXT: lui a1, 209715
; RV32-NEXT: addi a1, a1, 819
-; RV32-NEXT: sw a1, 20(sp)
-; RV32-NEXT: sw a1, 16(sp)
+; RV32-NEXT: sw a1, 36(sp)
+; RV32-NEXT: sw a1, 32(sp)
; RV32-NEXT: lui a1, 61681
; RV32-NEXT: addi a1, a1, -241
-; RV32-NEXT: sw a1, 12(sp)
-; RV32-NEXT: sw a1, 8(sp)
+; RV32-NEXT: sw a1, 28(sp)
+; RV32-NEXT: sw a1, 24(sp)
; RV32-NEXT: lui a1, 4112
; RV32-NEXT: addi a1, a1, 257
-; RV32-NEXT: sw a1, 4(sp)
-; RV32-NEXT: sw a1, 0(sp)
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vsrl.vi v16, v8, 1, v0.t
-; RV32-NEXT: addi a1, sp, 24
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v24, (a1), zero
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v16, v24, v0.t
-; RV32-NEXT: vsub.vv v8, v8, v16, v0.t
-; RV32-NEXT: addi a1, sp, 16
+; RV32-NEXT: sw a1, 20(sp)
+; RV32-NEXT: sw a1, 16(sp)
+; RV32-NEXT: addi a1, sp, 40
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV32-NEXT: vlse64.v v16, (a1), zero
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 4
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 48
+; RV32-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: addi a1, sp, 32
+; RV32-NEXT: vlse64.v v16, (a1), zero
+; RV32-NEXT: addi a1, sp, 48
+; RV32-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vi v16, v8, 1, v0.t
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 3
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 48
+; RV32-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 4
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 48
+; RV32-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 3
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 48
+; RV32-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v24, v16, v24, v0.t
+; RV32-NEXT: vsub.vv v8, v8, v24, v0.t
+; RV32-NEXT: addi a1, sp, 48
+; RV32-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
; RV32-NEXT: vand.vv v24, v8, v16, v0.t
; RV32-NEXT: vsrl.vi v8, v8, 2, v0.t
; RV32-NEXT: vand.vv v8, v8, v16, v0.t
; RV32-NEXT: vadd.vv v8, v24, v8, v0.t
-; RV32-NEXT: vsrl.vi v16, v8, 4, v0.t
-; RV32-NEXT: vadd.vv v8, v8, v16, v0.t
-; RV32-NEXT: addi a1, sp, 8
+; RV32-NEXT: addi a1, sp, 24
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV32-NEXT: vlse64.v v16, (a1), zero
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 4
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 48
+; RV32-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: addi a1, sp, 16
+; RV32-NEXT: vlse64.v v24, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vi v16, v8, 4, v0.t
+; RV32-NEXT: vadd.vv v8, v8, v16, v0.t
+; RV32-NEXT: csrr a0, vlenb
+; RV32-NEXT: slli a0, a0, 4
+; RV32-NEXT: add a0, sp, a0
+; RV32-NEXT: addi a0, a0, 48
+; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; RV32-NEXT: vand.vv v8, v8, v16, v0.t
-; RV32-NEXT: mv a1, sp
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a1), zero
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vmul.vv v8, v8, v16, v0.t
+; RV32-NEXT: vmul.vv v8, v8, v24, v0.t
; RV32-NEXT: li a0, 56
; RV32-NEXT: vsrl.vx v8, v8, a0, v0.t
-; RV32-NEXT: addi sp, sp, 32
+; RV32-NEXT: csrr a0, vlenb
+; RV32-NEXT: li a1, 24
+; RV32-NEXT: mul a0, a0, a1
+; RV32-NEXT: add sp, sp, a0
+; RV32-NEXT: addi sp, sp, 48
; RV32-NEXT: ret
;
; RV64-LABEL: vp_ctpop_v15i64:
@@ -1228,34 +1266,29 @@ define <15 x i64> @vp_ctpop_v15i64_unmasked(<15 x i64> %va, i32 zeroext %evl) {
; RV32-NEXT: addi a1, a1, 257
; RV32-NEXT: sw a1, 4(sp)
; RV32-NEXT: sw a1, 0(sp)
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vsrl.vi v16, v8, 1
; RV32-NEXT: addi a1, sp, 24
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v16, (a1), zero
+; RV32-NEXT: addi a1, sp, 16
; RV32-NEXT: vlse64.v v24, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v16, v24
+; RV32-NEXT: vsrl.vi v0, v8, 1
+; RV32-NEXT: vand.vv v16, v0, v16
; RV32-NEXT: vsub.vv v8, v8, v16
-; RV32-NEXT: addi a1, sp, 16
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a1), zero
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v24, v8, v16
+; RV32-NEXT: vand.vv v16, v8, v24
; RV32-NEXT: vsrl.vi v8, v8, 2
-; RV32-NEXT: vand.vv v8, v8, v16
-; RV32-NEXT: vadd.vv v8, v24, v8
-; RV32-NEXT: vsrl.vi v16, v8, 4
-; RV32-NEXT: vadd.vv v8, v8, v16
+; RV32-NEXT: vand.vv v8, v8, v24
+; RV32-NEXT: vadd.vv v8, v16, v8
; RV32-NEXT: addi a1, sp, 8
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV32-NEXT: vlse64.v v16, (a1), zero
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v8, v8, v16
; RV32-NEXT: mv a1, sp
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a1), zero
+; RV32-NEXT: vlse64.v v24, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vmul.vv v8, v8, v16
+; RV32-NEXT: vsrl.vi v0, v8, 4
+; RV32-NEXT: vadd.vv v8, v8, v0
+; RV32-NEXT: vand.vv v8, v8, v16
+; RV32-NEXT: vmul.vv v8, v8, v24
; RV32-NEXT: li a0, 56
; RV32-NEXT: vsrl.vx v8, v8, a0
; RV32-NEXT: addi sp, sp, 32
@@ -1303,55 +1336,93 @@ declare <16 x i64> @llvm.vp.ctpop.v16i64(<16 x i64>, <16 x i1>, i32)
define <16 x i64> @vp_ctpop_v16i64(<16 x i64> %va, <16 x i1> %m, i32 zeroext %evl) {
; RV32-LABEL: vp_ctpop_v16i64:
; RV32: # %bb.0:
-; RV32-NEXT: addi sp, sp, -32
-; RV32-NEXT: .cfi_def_cfa_offset 32
+; RV32-NEXT: addi sp, sp, -48
+; RV32-NEXT: .cfi_def_cfa_offset 48
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: li a2, 24
+; RV32-NEXT: mul a1, a1, a2
+; RV32-NEXT: sub sp, sp, a1
+; RV32-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x30, 0x22, 0x11, 0x18, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 48 + 24 * vlenb
; RV32-NEXT: lui a1, 349525
; RV32-NEXT: addi a1, a1, 1365
-; RV32-NEXT: sw a1, 28(sp)
-; RV32-NEXT: sw a1, 24(sp)
+; RV32-NEXT: sw a1, 44(sp)
+; RV32-NEXT: sw a1, 40(sp)
; RV32-NEXT: lui a1, 209715
; RV32-NEXT: addi a1, a1, 819
-; RV32-NEXT: sw a1, 20(sp)
-; RV32-NEXT: sw a1, 16(sp)
+; RV32-NEXT: sw a1, 36(sp)
+; RV32-NEXT: sw a1, 32(sp)
; RV32-NEXT: lui a1, 61681
; RV32-NEXT: addi a1, a1, -241
-; RV32-NEXT: sw a1, 12(sp)
-; RV32-NEXT: sw a1, 8(sp)
+; RV32-NEXT: sw a1, 28(sp)
+; RV32-NEXT: sw a1, 24(sp)
; RV32-NEXT: lui a1, 4112
; RV32-NEXT: addi a1, a1, 257
-; RV32-NEXT: sw a1, 4(sp)
-; RV32-NEXT: sw a1, 0(sp)
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vsrl.vi v16, v8, 1, v0.t
-; RV32-NEXT: addi a1, sp, 24
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v24, (a1), zero
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v16, v24, v0.t
-; RV32-NEXT: vsub.vv v8, v8, v16, v0.t
-; RV32-NEXT: addi a1, sp, 16
+; RV32-NEXT: sw a1, 20(sp)
+; RV32-NEXT: sw a1, 16(sp)
+; RV32-NEXT: addi a1, sp, 40
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV32-NEXT: vlse64.v v16, (a1), zero
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 4
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 48
+; RV32-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: addi a1, sp, 32
+; RV32-NEXT: vlse64.v v16, (a1), zero
+; RV32-NEXT: addi a1, sp, 48
+; RV32-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vi v16, v8, 1, v0.t
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 3
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 48
+; RV32-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 4
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 48
+; RV32-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 3
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 48
+; RV32-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v24, v16, v24, v0.t
+; RV32-NEXT: vsub.vv v8, v8, v24, v0.t
+; RV32-NEXT: addi a1, sp, 48
+; RV32-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
; RV32-NEXT: vand.vv v24, v8, v16, v0.t
; RV32-NEXT: vsrl.vi v8, v8, 2, v0.t
; RV32-NEXT: vand.vv v8, v8, v16, v0.t
; RV32-NEXT: vadd.vv v8, v24, v8, v0.t
-; RV32-NEXT: vsrl.vi v16, v8, 4, v0.t
-; RV32-NEXT: vadd.vv v8, v8, v16, v0.t
-; RV32-NEXT: addi a1, sp, 8
+; RV32-NEXT: addi a1, sp, 24
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV32-NEXT: vlse64.v v16, (a1), zero
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 4
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 48
+; RV32-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: addi a1, sp, 16
+; RV32-NEXT: vlse64.v v24, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vi v16, v8, 4, v0.t
+; RV32-NEXT: vadd.vv v8, v8, v16, v0.t
+; RV32-NEXT: csrr a0, vlenb
+; RV32-NEXT: slli a0, a0, 4
+; RV32-NEXT: add a0, sp, a0
+; RV32-NEXT: addi a0, a0, 48
+; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; RV32-NEXT: vand.vv v8, v8, v16, v0.t
-; RV32-NEXT: mv a1, sp
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a1), zero
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vmul.vv v8, v8, v16, v0.t
+; RV32-NEXT: vmul.vv v8, v8, v24, v0.t
; RV32-NEXT: li a0, 56
; RV32-NEXT: vsrl.vx v8, v8, a0, v0.t
-; RV32-NEXT: addi sp, sp, 32
+; RV32-NEXT: csrr a0, vlenb
+; RV32-NEXT: li a1, 24
+; RV32-NEXT: mul a0, a0, a1
+; RV32-NEXT: add sp, sp, a0
+; RV32-NEXT: addi sp, sp, 48
; RV32-NEXT: ret
;
; RV64-LABEL: vp_ctpop_v16i64:
@@ -1412,34 +1483,29 @@ define <16 x i64> @vp_ctpop_v16i64_unmasked(<16 x i64> %va, i32 zeroext %evl) {
; RV32-NEXT: addi a1, a1, 257
; RV32-NEXT: sw a1, 4(sp)
; RV32-NEXT: sw a1, 0(sp)
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vsrl.vi v16, v8, 1
; RV32-NEXT: addi a1, sp, 24
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v16, (a1), zero
+; RV32-NEXT: addi a1, sp, 16
; RV32-NEXT: vlse64.v v24, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v16, v24
+; RV32-NEXT: vsrl.vi v0, v8, 1
+; RV32-NEXT: vand.vv v16, v0, v16
; RV32-NEXT: vsub.vv v8, v8, v16
-; RV32-NEXT: addi a1, sp, 16
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a1), zero
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v24, v8, v16
+; RV32-NEXT: vand.vv v16, v8, v24
; RV32-NEXT: vsrl.vi v8, v8, 2
-; RV32-NEXT: vand.vv v8, v8, v16
-; RV32-NEXT: vadd.vv v8, v24, v8
-; RV32-NEXT: vsrl.vi v16, v8, 4
-; RV32-NEXT: vadd.vv v8, v8, v16
+; RV32-NEXT: vand.vv v8, v8, v24
+; RV32-NEXT: vadd.vv v8, v16, v8
; RV32-NEXT: addi a1, sp, 8
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV32-NEXT: vlse64.v v16, (a1), zero
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v8, v8, v16
; RV32-NEXT: mv a1, sp
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a1), zero
+; RV32-NEXT: vlse64.v v24, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vmul.vv v8, v8, v16
+; RV32-NEXT: vsrl.vi v0, v8, 4
+; RV32-NEXT: vadd.vv v8, v8, v0
+; RV32-NEXT: vand.vv v8, v8, v16
+; RV32-NEXT: vmul.vv v8, v8, v24
; RV32-NEXT: li a0, 56
; RV32-NEXT: vsrl.vx v8, v8, a0
; RV32-NEXT: addi sp, sp, 32
@@ -1495,11 +1561,16 @@ define <32 x i64> @vp_ctpop_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %ev
; RV32-NEXT: sub sp, sp, a1
; RV32-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x30, 0x22, 0x11, 0x30, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 48 + 48 * vlenb
; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 4
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 48
+; RV32-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: csrr a1, vlenb
; RV32-NEXT: li a2, 40
; RV32-NEXT: mul a1, a1, a2
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 48
-; RV32-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
; RV32-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV32-NEXT: vslidedown.vi v7, v0, 2
; RV32-NEXT: lui a1, 349525
@@ -1524,74 +1595,93 @@ define <32 x i64> @vp_ctpop_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %ev
; RV32-NEXT: # %bb.1:
; RV32-NEXT: li a1, 16
; RV32-NEXT: .LBB34_2:
-; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; RV32-NEXT: vsrl.vi v16, v8, 1, v0.t
; RV32-NEXT: addi a2, sp, 40
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: li a4, 24
-; RV32-NEXT: mul a3, a3, a4
-; RV32-NEXT: add a3, sp, a3
-; RV32-NEXT: addi a3, a3, 48
-; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
; RV32-NEXT: vlse64.v v8, (a2), zero
; RV32-NEXT: csrr a2, vlenb
; RV32-NEXT: slli a2, a2, 5
; RV32-NEXT: add a2, sp, a2
; RV32-NEXT: addi a2, a2, 48
; RV32-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
-; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v16, v8, v0.t
+; RV32-NEXT: addi a2, sp, 32
+; RV32-NEXT: vlse64.v v16, (a2), zero
; RV32-NEXT: csrr a2, vlenb
; RV32-NEXT: li a3, 24
; RV32-NEXT: mul a2, a2, a3
; RV32-NEXT: add a2, sp, a2
; RV32-NEXT: addi a2, a2, 48
-; RV32-NEXT: vl8r.v v8, (a2) # Unknown-size Folded Reload
-; RV32-NEXT: vsub.vv v24, v8, v16, v0.t
-; RV32-NEXT: addi a2, sp, 32
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v8, (a2), zero
+; RV32-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
; RV32-NEXT: csrr a2, vlenb
-; RV32-NEXT: li a3, 24
+; RV32-NEXT: li a3, 40
; RV32-NEXT: mul a2, a2, a3
; RV32-NEXT: add a2, sp, a2
; RV32-NEXT: addi a2, a2, 48
-; RV32-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
+; RV32-NEXT: vl8r.v v8, (a2) # Unknown-size Folded Reload
; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v24, v8, v0.t
+; RV32-NEXT: vsrl.vi v24, v8, 1, v0.t
; RV32-NEXT: csrr a2, vlenb
-; RV32-NEXT: slli a2, a2, 4
+; RV32-NEXT: slli a2, a2, 5
; RV32-NEXT: add a2, sp, a2
; RV32-NEXT: addi a2, a2, 48
-; RV32-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
-; RV32-NEXT: vsrl.vi v16, v24, 2, v0.t
-; RV32-NEXT: vand.vv v16, v16, v8, v0.t
+; RV32-NEXT: vl8r.v v8, (a2) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v24, v24, v8, v0.t
; RV32-NEXT: csrr a2, vlenb
-; RV32-NEXT: slli a2, a2, 4
+; RV32-NEXT: li a3, 40
+; RV32-NEXT: mul a2, a2, a3
; RV32-NEXT: add a2, sp, a2
; RV32-NEXT: addi a2, a2, 48
; RV32-NEXT: vl8r.v v8, (a2) # Unknown-size Folded Reload
-; RV32-NEXT: vadd.vv v16, v8, v16, v0.t
-; RV32-NEXT: vsrl.vi v8, v16, 4, v0.t
-; RV32-NEXT: vadd.vv v16, v16, v8, v0.t
-; RV32-NEXT: addi a2, sp, 24
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v8, (a2), zero
+; RV32-NEXT: vsub.vv v24, v8, v24, v0.t
+; RV32-NEXT: vand.vv v8, v24, v16, v0.t
; RV32-NEXT: csrr a2, vlenb
-; RV32-NEXT: slli a2, a2, 4
+; RV32-NEXT: li a3, 40
+; RV32-NEXT: mul a2, a2, a3
; RV32-NEXT: add a2, sp, a2
; RV32-NEXT: addi a2, a2, 48
; RV32-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
-; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v8, v16, v8, v0.t
-; RV32-NEXT: addi a2, sp, 16
+; RV32-NEXT: vsrl.vi v8, v24, 2, v0.t
+; RV32-NEXT: vand.vv v8, v8, v16, v0.t
+; RV32-NEXT: csrr a2, vlenb
+; RV32-NEXT: li a3, 40
+; RV32-NEXT: mul a2, a2, a3
+; RV32-NEXT: add a2, sp, a2
+; RV32-NEXT: addi a2, a2, 48
+; RV32-NEXT: vl8r.v v16, (a2) # Unknown-size Folded Reload
+; RV32-NEXT: vadd.vv v8, v16, v8, v0.t
+; RV32-NEXT: csrr a2, vlenb
+; RV32-NEXT: slli a2, a2, 3
+; RV32-NEXT: add a2, sp, a2
+; RV32-NEXT: addi a2, a2, 48
+; RV32-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
+; RV32-NEXT: addi a2, sp, 24
+; RV32-NEXT: addi a3, sp, 16
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV32-NEXT: vlse64.v v16, (a2), zero
; RV32-NEXT: addi a2, sp, 48
; RV32-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
+; RV32-NEXT: vlse64.v v8, (a3), zero
+; RV32-NEXT: csrr a2, vlenb
+; RV32-NEXT: li a3, 40
+; RV32-NEXT: mul a2, a2, a3
+; RV32-NEXT: add a2, sp, a2
+; RV32-NEXT: addi a2, a2, 48
+; RV32-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
+; RV32-NEXT: csrr a2, vlenb
+; RV32-NEXT: slli a2, a2, 3
+; RV32-NEXT: add a2, sp, a2
+; RV32-NEXT: addi a2, a2, 48
+; RV32-NEXT: vl8r.v v8, (a2) # Unknown-size Folded Reload
; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; RV32-NEXT: vmul.vv v8, v8, v16, v0.t
+; RV32-NEXT: vsrl.vi v24, v8, 4, v0.t
+; RV32-NEXT: vadd.vv v8, v8, v24, v0.t
+; RV32-NEXT: vand.vv v16, v8, v16, v0.t
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: li a2, 40
+; RV32-NEXT: mul a1, a1, a2
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 48
+; RV32-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vmul.vv v8, v16, v8, v0.t
; RV32-NEXT: li a1, 56
; RV32-NEXT: vsrl.vx v8, v8, a1, v0.t
; RV32-NEXT: csrr a2, vlenb
@@ -1603,14 +1693,13 @@ define <32 x i64> @vp_ctpop_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %ev
; RV32-NEXT: sltu a0, a0, a2
; RV32-NEXT: addi a0, a0, -1
; RV32-NEXT: and a0, a0, a2
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vmv1r.v v0, v7
-; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: li a2, 40
-; RV32-NEXT: mul a0, a0, a2
-; RV32-NEXT: add a0, sp, a0
-; RV32-NEXT: addi a0, a0, 48
-; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: csrr a2, vlenb
+; RV32-NEXT: slli a2, a2, 4
+; RV32-NEXT: add a2, sp, a2
+; RV32-NEXT: addi a2, a2, 48
+; RV32-NEXT: vl8r.v v16, (a2) # Unknown-size Folded Reload
+; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vsrl.vi v24, v16, 1, v0.t
; RV32-NEXT: csrr a0, vlenb
; RV32-NEXT: slli a0, a0, 5
@@ -1625,20 +1714,37 @@ define <32 x i64> @vp_ctpop_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %ev
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 48
; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; RV32-NEXT: vand.vv v24, v8, v16, v0.t
+; RV32-NEXT: vand.vv v16, v8, v16, v0.t
+; RV32-NEXT: csrr a0, vlenb
+; RV32-NEXT: slli a0, a0, 5
+; RV32-NEXT: add a0, sp, a0
+; RV32-NEXT: addi a0, a0, 48
+; RV32-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
; RV32-NEXT: vsrl.vi v8, v8, 2, v0.t
-; RV32-NEXT: vand.vv v8, v8, v16, v0.t
-; RV32-NEXT: vadd.vv v8, v24, v8, v0.t
-; RV32-NEXT: vsrl.vi v16, v8, 4, v0.t
-; RV32-NEXT: vadd.vv v8, v8, v16, v0.t
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 4
+; RV32-NEXT: li a2, 24
+; RV32-NEXT: mul a0, a0, a2
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 48
; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; RV32-NEXT: vand.vv v8, v8, v16, v0.t
+; RV32-NEXT: csrr a0, vlenb
+; RV32-NEXT: slli a0, a0, 5
+; RV32-NEXT: add a0, sp, a0
+; RV32-NEXT: addi a0, a0, 48
+; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vadd.vv v8, v16, v8, v0.t
+; RV32-NEXT: vsrl.vi v16, v8, 4, v0.t
+; RV32-NEXT: vadd.vv v8, v8, v16, v0.t
; RV32-NEXT: addi a0, sp, 48
; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v8, v8, v16, v0.t
+; RV32-NEXT: csrr a0, vlenb
+; RV32-NEXT: li a2, 40
+; RV32-NEXT: mul a0, a0, a2
+; RV32-NEXT: add a0, sp, a0
+; RV32-NEXT: addi a0, a0, 48
+; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; RV32-NEXT: vmul.vv v8, v8, v16, v0.t
; RV32-NEXT: vsrl.vx v16, v8, a1, v0.t
; RV32-NEXT: csrr a0, vlenb
@@ -1666,8 +1772,8 @@ define <32 x i64> @vp_ctpop_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %ev
; RV64-NEXT: add a1, sp, a1
; RV64-NEXT: addi a1, a1, 16
; RV64-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
-; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64-NEXT: li a2, 16
+; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64-NEXT: vslidedown.vi v24, v0, 2
; RV64-NEXT: mv a1, a0
; RV64-NEXT: bltu a0, a2, .LBB34_2
@@ -1710,13 +1816,13 @@ define <32 x i64> @vp_ctpop_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %ev
; RV64-NEXT: sltu a0, a0, a6
; RV64-NEXT: addi a0, a0, -1
; RV64-NEXT: and a0, a0, a6
-; RV64-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV64-NEXT: vmv1r.v v0, v24
-; RV64-NEXT: csrr a0, vlenb
-; RV64-NEXT: slli a0, a0, 3
-; RV64-NEXT: add a0, sp, a0
-; RV64-NEXT: addi a0, a0, 16
-; RV64-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; RV64-NEXT: csrr a6, vlenb
+; RV64-NEXT: slli a6, a6, 3
+; RV64-NEXT: add a6, sp, a6
+; RV64-NEXT: addi a6, a6, 16
+; RV64-NEXT: vl8r.v v8, (a6) # Unknown-size Folded Reload
+; RV64-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV64-NEXT: vsrl.vi v16, v8, 1, v0.t
; RV64-NEXT: vand.vx v16, v16, a1, v0.t
; RV64-NEXT: vsub.vv v16, v8, v16, v0.t
@@ -1746,12 +1852,11 @@ define <32 x i64> @vp_ctpop_v32i64_unmasked(<32 x i64> %va, i32 zeroext %evl) {
; RV32-NEXT: addi sp, sp, -48
; RV32-NEXT: .cfi_def_cfa_offset 48
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a2, 40
-; RV32-NEXT: mul a1, a1, a2
+; RV32-NEXT: slli a1, a1, 4
; RV32-NEXT: sub sp, sp, a1
-; RV32-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x30, 0x22, 0x11, 0x28, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 48 + 40 * vlenb
+; RV32-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x30, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 48 + 16 * vlenb
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: slli a1, a1, 5
+; RV32-NEXT: slli a1, a1, 3
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 48
; RV32-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
@@ -1777,97 +1882,67 @@ define <32 x i64> @vp_ctpop_v32i64_unmasked(<32 x i64> %va, i32 zeroext %evl) {
; RV32-NEXT: # %bb.1:
; RV32-NEXT: li a1, 16
; RV32-NEXT: .LBB35_2:
-; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; RV32-NEXT: vsrl.vi v16, v8, 1
; RV32-NEXT: addi a2, sp, 40
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v24, (a2), zero
-; RV32-NEXT: csrr a2, vlenb
-; RV32-NEXT: li a3, 24
-; RV32-NEXT: mul a2, a2, a3
-; RV32-NEXT: add a2, sp, a2
-; RV32-NEXT: addi a2, a2, 48
-; RV32-NEXT: vs8r.v v24, (a2) # Unknown-size Folded Spill
-; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v16, v24
-; RV32-NEXT: vsub.vv v8, v8, v16
+; RV32-NEXT: vlse64.v v16, (a2), zero
; RV32-NEXT: addi a2, sp, 32
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v0, (a2), zero
+; RV32-NEXT: vlse64.v v24, (a2), zero
; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v8, v0
+; RV32-NEXT: vsrl.vi v0, v8, 1
+; RV32-NEXT: vand.vv v0, v0, v16
+; RV32-NEXT: vsub.vv v8, v8, v0
+; RV32-NEXT: vand.vv v0, v8, v24
; RV32-NEXT: vsrl.vi v8, v8, 2
-; RV32-NEXT: vand.vv v8, v8, v0
-; RV32-NEXT: vadd.vv v8, v16, v8
-; RV32-NEXT: vsrl.vi v16, v8, 4
-; RV32-NEXT: vadd.vv v8, v8, v16
-; RV32-NEXT: addi a2, sp, 24
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a2), zero
-; RV32-NEXT: csrr a2, vlenb
-; RV32-NEXT: slli a2, a2, 4
-; RV32-NEXT: add a2, sp, a2
-; RV32-NEXT: addi a2, a2, 48
-; RV32-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
-; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v8, v16
-; RV32-NEXT: addi a2, sp, 16
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v8, (a2), zero
+; RV32-NEXT: vand.vv v8, v8, v24
+; RV32-NEXT: vadd.vv v8, v0, v8
+; RV32-NEXT: vsrl.vi v0, v8, 4
+; RV32-NEXT: vadd.vv v8, v8, v0
; RV32-NEXT: addi a2, sp, 48
; RV32-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
-; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; RV32-NEXT: vmul.vv v16, v16, v8
-; RV32-NEXT: li a1, 56
-; RV32-NEXT: vsrl.vx v8, v16, a1
-; RV32-NEXT: csrr a2, vlenb
-; RV32-NEXT: slli a2, a2, 3
-; RV32-NEXT: add a2, sp, a2
-; RV32-NEXT: addi a2, a2, 48
-; RV32-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
; RV32-NEXT: addi a2, a0, -16
; RV32-NEXT: sltu a0, a0, a2
; RV32-NEXT: addi a0, a0, -1
; RV32-NEXT: and a0, a0, a2
+; RV32-NEXT: csrr a2, vlenb
+; RV32-NEXT: slli a2, a2, 3
+; RV32-NEXT: add a2, sp, a2
+; RV32-NEXT: addi a2, a2, 48
+; RV32-NEXT: vl8r.v v8, (a2) # Unknown-size Folded Reload
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 5
-; RV32-NEXT: add a0, sp, a0
-; RV32-NEXT: addi a0, a0, 48
-; RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; RV32-NEXT: vsrl.vi v16, v8, 1
-; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: li a2, 24
-; RV32-NEXT: mul a0, a0, a2
-; RV32-NEXT: add a0, sp, a0
-; RV32-NEXT: addi a0, a0, 48
-; RV32-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
-; RV32-NEXT: vand.vv v16, v16, v24
+; RV32-NEXT: vsrl.vi v0, v8, 1
+; RV32-NEXT: vand.vv v16, v0, v16
+; RV32-NEXT: addi a2, sp, 24
; RV32-NEXT: vsub.vv v16, v8, v16
-; RV32-NEXT: vand.vv v8, v16, v0
+; RV32-NEXT: vand.vv v0, v16, v24
; RV32-NEXT: vsrl.vi v16, v16, 2
-; RV32-NEXT: vand.vv v16, v16, v0
-; RV32-NEXT: vadd.vv v8, v8, v16
-; RV32-NEXT: vsrl.vi v16, v8, 4
-; RV32-NEXT: vadd.vv v8, v8, v16
+; RV32-NEXT: vand.vv v16, v16, v24
+; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v24, (a2), zero
+; RV32-NEXT: addi a2, sp, 16
+; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vadd.vv v16, v0, v16
+; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v0, (a2), zero
+; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vi v8, v16, 4
+; RV32-NEXT: vadd.vv v8, v16, v8
+; RV32-NEXT: addi a2, sp, 48
+; RV32-NEXT: vl8r.v v16, (a2) # Unknown-size Folded Reload
+; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
+; RV32-NEXT: vand.vv v16, v16, v24
+; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vand.vv v8, v8, v24
+; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
+; RV32-NEXT: vmul.vv v16, v16, v0
+; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vmul.vv v24, v8, v0
+; RV32-NEXT: li a2, 56
+; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vx v8, v16, a2
+; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vx v16, v24, a2
; RV32-NEXT: csrr a0, vlenb
; RV32-NEXT: slli a0, a0, 4
-; RV32-NEXT: add a0, sp, a0
-; RV32-NEXT: addi a0, a0, 48
-; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; RV32-NEXT: vand.vv v8, v8, v16
-; RV32-NEXT: addi a0, sp, 48
-; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; RV32-NEXT: vmul.vv v8, v8, v16
-; RV32-NEXT: vsrl.vx v16, v8, a1
-; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 3
-; RV32-NEXT: add a0, sp, a0
-; RV32-NEXT: addi a0, a0, 48
-; RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: li a1, 40
-; RV32-NEXT: mul a0, a0, a1
; RV32-NEXT: add sp, sp, a0
; RV32-NEXT: addi sp, sp, 48
; RV32-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-cttz-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-cttz-vp.ll
index f2926fa91e5c2..e7736e7f360f3 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-cttz-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-cttz-vp.ll
@@ -1263,59 +1263,86 @@ declare <15 x i64> @llvm.vp.cttz.v15i64(<15 x i64>, i1 immarg, <15 x i1>, i32)
define <15 x i64> @vp_cttz_v15i64(<15 x i64> %va, <15 x i1> %m, i32 zeroext %evl) {
; RV32-LABEL: vp_cttz_v15i64:
; RV32: # %bb.0:
-; RV32-NEXT: addi sp, sp, -32
-; RV32-NEXT: .cfi_def_cfa_offset 32
+; RV32-NEXT: addi sp, sp, -48
+; RV32-NEXT: .cfi_def_cfa_offset 48
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 4
+; RV32-NEXT: sub sp, sp, a1
+; RV32-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x30, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 48 + 16 * vlenb
; RV32-NEXT: lui a1, 349525
; RV32-NEXT: addi a1, a1, 1365
-; RV32-NEXT: sw a1, 28(sp)
-; RV32-NEXT: sw a1, 24(sp)
+; RV32-NEXT: sw a1, 44(sp)
+; RV32-NEXT: sw a1, 40(sp)
; RV32-NEXT: lui a1, 209715
; RV32-NEXT: addi a1, a1, 819
-; RV32-NEXT: sw a1, 20(sp)
-; RV32-NEXT: sw a1, 16(sp)
+; RV32-NEXT: sw a1, 36(sp)
+; RV32-NEXT: sw a1, 32(sp)
; RV32-NEXT: lui a1, 61681
; RV32-NEXT: addi a1, a1, -241
-; RV32-NEXT: sw a1, 12(sp)
-; RV32-NEXT: sw a1, 8(sp)
+; RV32-NEXT: sw a1, 28(sp)
+; RV32-NEXT: sw a1, 24(sp)
; RV32-NEXT: lui a1, 4112
; RV32-NEXT: addi a1, a1, 257
-; RV32-NEXT: sw a1, 4(sp)
-; RV32-NEXT: sw a1, 0(sp)
+; RV32-NEXT: sw a1, 20(sp)
+; RV32-NEXT: sw a1, 16(sp)
; RV32-NEXT: li a1, 1
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vsub.vx v16, v8, a1, v0.t
; RV32-NEXT: vnot.v v8, v8, v0.t
-; RV32-NEXT: vand.vv v8, v8, v16, v0.t
-; RV32-NEXT: vsrl.vi v16, v8, 1, v0.t
-; RV32-NEXT: addi a1, sp, 24
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v24, (a1), zero
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v16, v24, v0.t
-; RV32-NEXT: vsub.vv v8, v8, v16, v0.t
-; RV32-NEXT: addi a1, sp, 16
+; RV32-NEXT: vand.vv v24, v8, v16, v0.t
+; RV32-NEXT: addi a1, sp, 48
+; RV32-NEXT: vs8r.v v24, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: addi a1, sp, 40
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v8, (a1), zero
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 3
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 48
+; RV32-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: addi a1, sp, 32
; RV32-NEXT: vlse64.v v16, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vi v8, v24, 1, v0.t
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 3
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 48
+; RV32-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v8, v8, v24, v0.t
+; RV32-NEXT: addi a1, sp, 48
+; RV32-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vsub.vv v8, v24, v8, v0.t
; RV32-NEXT: vand.vv v24, v8, v16, v0.t
; RV32-NEXT: vsrl.vi v8, v8, 2, v0.t
; RV32-NEXT: vand.vv v8, v8, v16, v0.t
; RV32-NEXT: vadd.vv v8, v24, v8, v0.t
-; RV32-NEXT: vsrl.vi v16, v8, 4, v0.t
-; RV32-NEXT: vadd.vv v8, v8, v16, v0.t
-; RV32-NEXT: addi a1, sp, 8
+; RV32-NEXT: addi a1, sp, 24
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV32-NEXT: vlse64.v v16, (a1), zero
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 3
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 48
+; RV32-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: addi a1, sp, 16
+; RV32-NEXT: vlse64.v v24, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vi v16, v8, 4, v0.t
+; RV32-NEXT: vadd.vv v8, v8, v16, v0.t
+; RV32-NEXT: csrr a0, vlenb
+; RV32-NEXT: slli a0, a0, 3
+; RV32-NEXT: add a0, sp, a0
+; RV32-NEXT: addi a0, a0, 48
+; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; RV32-NEXT: vand.vv v8, v8, v16, v0.t
-; RV32-NEXT: mv a1, sp
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a1), zero
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vmul.vv v8, v8, v16, v0.t
+; RV32-NEXT: vmul.vv v8, v8, v24, v0.t
; RV32-NEXT: li a0, 56
; RV32-NEXT: vsrl.vx v8, v8, a0, v0.t
-; RV32-NEXT: addi sp, sp, 32
+; RV32-NEXT: csrr a0, vlenb
+; RV32-NEXT: slli a0, a0, 4
+; RV32-NEXT: add sp, sp, a0
+; RV32-NEXT: addi sp, sp, 48
; RV32-NEXT: ret
;
; RV64-LABEL: vp_cttz_v15i64:
@@ -1385,33 +1412,29 @@ define <15 x i64> @vp_cttz_v15i64_unmasked(<15 x i64> %va, i32 zeroext %evl) {
; RV32-NEXT: vsub.vx v16, v8, a1
; RV32-NEXT: vnot.v v8, v8
; RV32-NEXT: vand.vv v8, v8, v16
-; RV32-NEXT: vsrl.vi v16, v8, 1
; RV32-NEXT: addi a1, sp, 24
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v16, (a1), zero
+; RV32-NEXT: addi a1, sp, 16
; RV32-NEXT: vlse64.v v24, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v16, v24
+; RV32-NEXT: vsrl.vi v0, v8, 1
+; RV32-NEXT: vand.vv v16, v0, v16
; RV32-NEXT: vsub.vv v8, v8, v16
-; RV32-NEXT: addi a1, sp, 16
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a1), zero
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v24, v8, v16
+; RV32-NEXT: vand.vv v16, v8, v24
; RV32-NEXT: vsrl.vi v8, v8, 2
-; RV32-NEXT: vand.vv v8, v8, v16
-; RV32-NEXT: vadd.vv v8, v24, v8
-; RV32-NEXT: vsrl.vi v16, v8, 4
-; RV32-NEXT: vadd.vv v8, v8, v16
+; RV32-NEXT: vand.vv v8, v8, v24
+; RV32-NEXT: vadd.vv v8, v16, v8
; RV32-NEXT: addi a1, sp, 8
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV32-NEXT: vlse64.v v16, (a1), zero
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v8, v8, v16
; RV32-NEXT: mv a1, sp
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a1), zero
+; RV32-NEXT: vlse64.v v24, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vmul.vv v8, v8, v16
+; RV32-NEXT: vsrl.vi v0, v8, 4
+; RV32-NEXT: vadd.vv v8, v8, v0
+; RV32-NEXT: vand.vv v8, v8, v16
+; RV32-NEXT: vmul.vv v8, v8, v24
; RV32-NEXT: li a0, 56
; RV32-NEXT: vsrl.vx v8, v8, a0
; RV32-NEXT: addi sp, sp, 32
@@ -1463,59 +1486,86 @@ declare <16 x i64> @llvm.vp.cttz.v16i64(<16 x i64>, i1 immarg, <16 x i1>, i32)
define <16 x i64> @vp_cttz_v16i64(<16 x i64> %va, <16 x i1> %m, i32 zeroext %evl) {
; RV32-LABEL: vp_cttz_v16i64:
; RV32: # %bb.0:
-; RV32-NEXT: addi sp, sp, -32
-; RV32-NEXT: .cfi_def_cfa_offset 32
+; RV32-NEXT: addi sp, sp, -48
+; RV32-NEXT: .cfi_def_cfa_offset 48
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 4
+; RV32-NEXT: sub sp, sp, a1
+; RV32-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x30, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 48 + 16 * vlenb
; RV32-NEXT: lui a1, 349525
; RV32-NEXT: addi a1, a1, 1365
-; RV32-NEXT: sw a1, 28(sp)
-; RV32-NEXT: sw a1, 24(sp)
+; RV32-NEXT: sw a1, 44(sp)
+; RV32-NEXT: sw a1, 40(sp)
; RV32-NEXT: lui a1, 209715
; RV32-NEXT: addi a1, a1, 819
-; RV32-NEXT: sw a1, 20(sp)
-; RV32-NEXT: sw a1, 16(sp)
+; RV32-NEXT: sw a1, 36(sp)
+; RV32-NEXT: sw a1, 32(sp)
; RV32-NEXT: lui a1, 61681
; RV32-NEXT: addi a1, a1, -241
-; RV32-NEXT: sw a1, 12(sp)
-; RV32-NEXT: sw a1, 8(sp)
+; RV32-NEXT: sw a1, 28(sp)
+; RV32-NEXT: sw a1, 24(sp)
; RV32-NEXT: lui a1, 4112
; RV32-NEXT: addi a1, a1, 257
-; RV32-NEXT: sw a1, 4(sp)
-; RV32-NEXT: sw a1, 0(sp)
+; RV32-NEXT: sw a1, 20(sp)
+; RV32-NEXT: sw a1, 16(sp)
; RV32-NEXT: li a1, 1
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vsub.vx v16, v8, a1, v0.t
; RV32-NEXT: vnot.v v8, v8, v0.t
-; RV32-NEXT: vand.vv v8, v8, v16, v0.t
-; RV32-NEXT: vsrl.vi v16, v8, 1, v0.t
-; RV32-NEXT: addi a1, sp, 24
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v24, (a1), zero
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v16, v24, v0.t
-; RV32-NEXT: vsub.vv v8, v8, v16, v0.t
-; RV32-NEXT: addi a1, sp, 16
+; RV32-NEXT: vand.vv v24, v8, v16, v0.t
+; RV32-NEXT: addi a1, sp, 48
+; RV32-NEXT: vs8r.v v24, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: addi a1, sp, 40
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v8, (a1), zero
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 3
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 48
+; RV32-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: addi a1, sp, 32
; RV32-NEXT: vlse64.v v16, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vi v8, v24, 1, v0.t
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 3
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 48
+; RV32-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v8, v8, v24, v0.t
+; RV32-NEXT: addi a1, sp, 48
+; RV32-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vsub.vv v8, v24, v8, v0.t
; RV32-NEXT: vand.vv v24, v8, v16, v0.t
; RV32-NEXT: vsrl.vi v8, v8, 2, v0.t
; RV32-NEXT: vand.vv v8, v8, v16, v0.t
; RV32-NEXT: vadd.vv v8, v24, v8, v0.t
-; RV32-NEXT: vsrl.vi v16, v8, 4, v0.t
-; RV32-NEXT: vadd.vv v8, v8, v16, v0.t
-; RV32-NEXT: addi a1, sp, 8
+; RV32-NEXT: addi a1, sp, 24
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV32-NEXT: vlse64.v v16, (a1), zero
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 3
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 48
+; RV32-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: addi a1, sp, 16
+; RV32-NEXT: vlse64.v v24, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vi v16, v8, 4, v0.t
+; RV32-NEXT: vadd.vv v8, v8, v16, v0.t
+; RV32-NEXT: csrr a0, vlenb
+; RV32-NEXT: slli a0, a0, 3
+; RV32-NEXT: add a0, sp, a0
+; RV32-NEXT: addi a0, a0, 48
+; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; RV32-NEXT: vand.vv v8, v8, v16, v0.t
-; RV32-NEXT: mv a1, sp
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a1), zero
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vmul.vv v8, v8, v16, v0.t
+; RV32-NEXT: vmul.vv v8, v8, v24, v0.t
; RV32-NEXT: li a0, 56
; RV32-NEXT: vsrl.vx v8, v8, a0, v0.t
-; RV32-NEXT: addi sp, sp, 32
+; RV32-NEXT: csrr a0, vlenb
+; RV32-NEXT: slli a0, a0, 4
+; RV32-NEXT: add sp, sp, a0
+; RV32-NEXT: addi sp, sp, 48
; RV32-NEXT: ret
;
; RV64-LABEL: vp_cttz_v16i64:
@@ -1585,33 +1635,29 @@ define <16 x i64> @vp_cttz_v16i64_unmasked(<16 x i64> %va, i32 zeroext %evl) {
; RV32-NEXT: vsub.vx v16, v8, a1
; RV32-NEXT: vnot.v v8, v8
; RV32-NEXT: vand.vv v8, v8, v16
-; RV32-NEXT: vsrl.vi v16, v8, 1
; RV32-NEXT: addi a1, sp, 24
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v16, (a1), zero
+; RV32-NEXT: addi a1, sp, 16
; RV32-NEXT: vlse64.v v24, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v16, v24
+; RV32-NEXT: vsrl.vi v0, v8, 1
+; RV32-NEXT: vand.vv v16, v0, v16
; RV32-NEXT: vsub.vv v8, v8, v16
-; RV32-NEXT: addi a1, sp, 16
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a1), zero
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v24, v8, v16
+; RV32-NEXT: vand.vv v16, v8, v24
; RV32-NEXT: vsrl.vi v8, v8, 2
-; RV32-NEXT: vand.vv v8, v8, v16
-; RV32-NEXT: vadd.vv v8, v24, v8
-; RV32-NEXT: vsrl.vi v16, v8, 4
-; RV32-NEXT: vadd.vv v8, v8, v16
+; RV32-NEXT: vand.vv v8, v8, v24
+; RV32-NEXT: vadd.vv v8, v16, v8
; RV32-NEXT: addi a1, sp, 8
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV32-NEXT: vlse64.v v16, (a1), zero
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v8, v8, v16
; RV32-NEXT: mv a1, sp
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a1), zero
+; RV32-NEXT: vlse64.v v24, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vmul.vv v8, v8, v16
+; RV32-NEXT: vsrl.vi v0, v8, 4
+; RV32-NEXT: vadd.vv v8, v8, v0
+; RV32-NEXT: vand.vv v8, v8, v16
+; RV32-NEXT: vmul.vv v8, v8, v24
; RV32-NEXT: li a0, 56
; RV32-NEXT: vsrl.vx v8, v8, a0
; RV32-NEXT: addi sp, sp, 32
@@ -1671,7 +1717,7 @@ define <32 x i64> @vp_cttz_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %evl
; RV32-NEXT: sub sp, sp, a1
; RV32-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x30, 0x22, 0x11, 0x38, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 48 + 56 * vlenb
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: slli a1, a1, 5
+; RV32-NEXT: slli a1, a1, 4
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 48
; RV32-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
@@ -1705,111 +1751,145 @@ define <32 x i64> @vp_cttz_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %evl
; RV32-NEXT: vnot.v v8, v8, v0.t
; RV32-NEXT: vand.vv v8, v8, v16, v0.t
; RV32-NEXT: csrr a3, vlenb
+; RV32-NEXT: slli a3, a3, 5
+; RV32-NEXT: add a3, sp, a3
+; RV32-NEXT: addi a3, a3, 48
+; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
+; RV32-NEXT: addi a3, sp, 40
+; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v8, (a3), zero
+; RV32-NEXT: csrr a3, vlenb
; RV32-NEXT: li a4, 40
; RV32-NEXT: mul a3, a3, a4
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: vsrl.vi v8, v8, 1, v0.t
+; RV32-NEXT: addi a3, sp, 32
+; RV32-NEXT: vlse64.v v8, (a3), zero
; RV32-NEXT: csrr a3, vlenb
; RV32-NEXT: li a4, 48
; RV32-NEXT: mul a3, a3, a4
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: addi a3, sp, 40
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v8, (a3), zero
+; RV32-NEXT: csrr a3, vlenb
+; RV32-NEXT: slli a3, a3, 5
+; RV32-NEXT: add a3, sp, a3
+; RV32-NEXT: addi a3, a3, 48
+; RV32-NEXT: vl8r.v v16, (a3) # Unknown-size Folded Reload
+; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vi v8, v16, 1, v0.t
; RV32-NEXT: csrr a3, vlenb
; RV32-NEXT: li a4, 24
; RV32-NEXT: mul a3, a3, a4
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: li a4, 48
+; RV32-NEXT: li a4, 40
; RV32-NEXT: mul a3, a3, a4
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
; RV32-NEXT: vl8r.v v16, (a3) # Unknown-size Folded Reload
-; RV32-NEXT: vand.vv v16, v16, v8, v0.t
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: li a4, 40
+; RV32-NEXT: li a4, 24
; RV32-NEXT: mul a3, a3, a4
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
; RV32-NEXT: vl8r.v v8, (a3) # Unknown-size Folded Reload
-; RV32-NEXT: vsub.vv v8, v8, v16, v0.t
+; RV32-NEXT: vand.vv v8, v8, v16, v0.t
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: li a4, 40
-; RV32-NEXT: mul a3, a3, a4
+; RV32-NEXT: slli a3, a3, 5
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
-; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: addi a3, sp, 32
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v8, (a3), zero
+; RV32-NEXT: vl8r.v v16, (a3) # Unknown-size Folded Reload
+; RV32-NEXT: vsub.vv v16, v16, v8, v0.t
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: li a4, 48
-; RV32-NEXT: mul a3, a3, a4
+; RV32-NEXT: slli a3, a3, 5
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
-; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
+; RV32-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: li a4, 40
-; RV32-NEXT: mul a3, a3, a4
+; RV32-NEXT: slli a3, a3, 5
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
; RV32-NEXT: vl8r.v v16, (a3) # Unknown-size Folded Reload
+; RV32-NEXT: csrr a3, vlenb
+; RV32-NEXT: li a4, 48
+; RV32-NEXT: mul a3, a3, a4
+; RV32-NEXT: add a3, sp, a3
+; RV32-NEXT: addi a3, a3, 48
+; RV32-NEXT: vl8r.v v8, (a3) # Unknown-size Folded Reload
; RV32-NEXT: vand.vv v16, v16, v8, v0.t
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: slli a3, a3, 4
+; RV32-NEXT: li a4, 24
+; RV32-NEXT: mul a3, a3, a4
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
; RV32-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: li a4, 40
-; RV32-NEXT: mul a3, a3, a4
+; RV32-NEXT: slli a3, a3, 5
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
; RV32-NEXT: vl8r.v v16, (a3) # Unknown-size Folded Reload
; RV32-NEXT: vsrl.vi v16, v16, 2, v0.t
; RV32-NEXT: vand.vv v16, v16, v8, v0.t
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: slli a3, a3, 4
+; RV32-NEXT: li a4, 24
+; RV32-NEXT: mul a3, a3, a4
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
; RV32-NEXT: vl8r.v v8, (a3) # Unknown-size Folded Reload
; RV32-NEXT: vadd.vv v8, v8, v16, v0.t
-; RV32-NEXT: vsrl.vi v16, v8, 4, v0.t
-; RV32-NEXT: vadd.vv v16, v8, v16, v0.t
+; RV32-NEXT: csrr a3, vlenb
+; RV32-NEXT: slli a3, a3, 3
+; RV32-NEXT: add a3, sp, a3
+; RV32-NEXT: addi a3, a3, 48
+; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
; RV32-NEXT: addi a3, sp, 24
+; RV32-NEXT: addi a4, sp, 16
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV32-NEXT: vlse64.v v8, (a3), zero
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: li a4, 40
-; RV32-NEXT: mul a3, a3, a4
+; RV32-NEXT: li a5, 24
+; RV32-NEXT: mul a3, a3, a5
+; RV32-NEXT: add a3, sp, a3
+; RV32-NEXT: addi a3, a3, 48
+; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
+; RV32-NEXT: vlse64.v v8, (a4), zero
+; RV32-NEXT: csrr a3, vlenb
+; RV32-NEXT: slli a3, a3, 5
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v8, v16, v8, v0.t
-; RV32-NEXT: addi a3, sp, 16
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a3), zero
; RV32-NEXT: csrr a3, vlenb
; RV32-NEXT: slli a3, a3, 3
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
-; RV32-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
+; RV32-NEXT: vl8r.v v8, (a3) # Unknown-size Folded Reload
; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
-; RV32-NEXT: vmul.vv v8, v8, v16, v0.t
+; RV32-NEXT: vsrl.vi v16, v8, 4, v0.t
+; RV32-NEXT: addi a2, sp, 48
+; RV32-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
+; RV32-NEXT: vl8r.v v16, (a2) # Unknown-size Folded Reload
+; RV32-NEXT: vadd.vv v16, v8, v16, v0.t
+; RV32-NEXT: csrr a2, vlenb
+; RV32-NEXT: li a3, 24
+; RV32-NEXT: mul a2, a2, a3
+; RV32-NEXT: add a2, sp, a2
+; RV32-NEXT: addi a2, a2, 48
+; RV32-NEXT: vl8r.v v8, (a2) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v16, v16, v8, v0.t
+; RV32-NEXT: csrr a2, vlenb
+; RV32-NEXT: slli a2, a2, 5
+; RV32-NEXT: add a2, sp, a2
+; RV32-NEXT: addi a2, a2, 48
+; RV32-NEXT: vl8r.v v8, (a2) # Unknown-size Folded Reload
+; RV32-NEXT: vmul.vv v8, v16, v8, v0.t
; RV32-NEXT: li a2, 56
; RV32-NEXT: vsrl.vx v8, v8, a2, v0.t
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: slli a3, a3, 4
+; RV32-NEXT: slli a3, a3, 3
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
@@ -1817,13 +1897,13 @@ define <32 x i64> @vp_cttz_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %evl
; RV32-NEXT: sltu a0, a0, a3
; RV32-NEXT: addi a0, a0, -1
; RV32-NEXT: and a0, a0, a3
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vmv1r.v v0, v24
-; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 5
-; RV32-NEXT: add a0, sp, a0
-; RV32-NEXT: addi a0, a0, 48
-; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: csrr a3, vlenb
+; RV32-NEXT: slli a3, a3, 4
+; RV32-NEXT: add a3, sp, a3
+; RV32-NEXT: addi a3, a3, 48
+; RV32-NEXT: vl8r.v v16, (a3) # Unknown-size Folded Reload
+; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vsub.vx v8, v16, a1, v0.t
; RV32-NEXT: vnot.v v16, v16, v0.t
; RV32-NEXT: vand.vv v8, v16, v8, v0.t
@@ -1831,18 +1911,18 @@ define <32 x i64> @vp_cttz_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %evl
; RV32-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; RV32-NEXT: vsrl.vi v8, v8, 1, v0.t
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 5
+; RV32-NEXT: slli a0, a0, 4
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 48
; RV32-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: li a1, 24
+; RV32-NEXT: li a1, 40
; RV32-NEXT: mul a0, a0, a1
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 48
; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 5
+; RV32-NEXT: slli a0, a0, 4
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 48
; RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
@@ -1851,17 +1931,35 @@ define <32 x i64> @vp_cttz_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %evl
; RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; RV32-NEXT: vsub.vv v8, v8, v16, v0.t
; RV32-NEXT: csrr a0, vlenb
+; RV32-NEXT: li a1, 40
+; RV32-NEXT: mul a0, a0, a1
+; RV32-NEXT: add a0, sp, a0
+; RV32-NEXT: addi a0, a0, 48
+; RV32-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; RV32-NEXT: csrr a0, vlenb
; RV32-NEXT: li a1, 48
; RV32-NEXT: mul a0, a0, a1
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 48
+; RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: csrr a0, vlenb
+; RV32-NEXT: li a1, 40
+; RV32-NEXT: mul a0, a0, a1
+; RV32-NEXT: add a0, sp, a0
+; RV32-NEXT: addi a0, a0, 48
; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; RV32-NEXT: vand.vv v16, v8, v16, v0.t
+; RV32-NEXT: vand.vv v16, v16, v8, v0.t
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 5
+; RV32-NEXT: slli a0, a0, 4
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 48
; RV32-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
+; RV32-NEXT: csrr a0, vlenb
+; RV32-NEXT: li a1, 40
+; RV32-NEXT: mul a0, a0, a1
+; RV32-NEXT: add a0, sp, a0
+; RV32-NEXT: addi a0, a0, 48
+; RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; RV32-NEXT: vsrl.vi v8, v8, 2, v0.t
; RV32-NEXT: csrr a0, vlenb
; RV32-NEXT: li a1, 48
@@ -1871,7 +1969,7 @@ define <32 x i64> @vp_cttz_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %evl
; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; RV32-NEXT: vand.vv v8, v8, v16, v0.t
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 5
+; RV32-NEXT: slli a0, a0, 4
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 48
; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
@@ -1879,21 +1977,21 @@ define <32 x i64> @vp_cttz_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %evl
; RV32-NEXT: vsrl.vi v16, v8, 4, v0.t
; RV32-NEXT: vadd.vv v8, v8, v16, v0.t
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: li a1, 40
+; RV32-NEXT: li a1, 24
; RV32-NEXT: mul a0, a0, a1
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 48
; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; RV32-NEXT: vand.vv v8, v8, v16, v0.t
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 3
+; RV32-NEXT: slli a0, a0, 5
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 48
; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; RV32-NEXT: vmul.vv v8, v8, v16, v0.t
; RV32-NEXT: vsrl.vx v16, v8, a2, v0.t
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 4
+; RV32-NEXT: slli a0, a0, 3
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 48
; RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
@@ -1917,8 +2015,8 @@ define <32 x i64> @vp_cttz_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %evl
; RV64-NEXT: add a1, sp, a1
; RV64-NEXT: addi a1, a1, 16
; RV64-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
-; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64-NEXT: li a1, 16
+; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64-NEXT: vslidedown.vi v24, v0, 2
; RV64-NEXT: mv a2, a0
; RV64-NEXT: bltu a0, a1, .LBB34_2
@@ -1965,13 +2063,13 @@ define <32 x i64> @vp_cttz_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %evl
; RV64-NEXT: sltu a0, a0, a7
; RV64-NEXT: addi a0, a0, -1
; RV64-NEXT: and a0, a0, a7
-; RV64-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV64-NEXT: vmv1r.v v0, v24
-; RV64-NEXT: csrr a0, vlenb
-; RV64-NEXT: slli a0, a0, 3
-; RV64-NEXT: add a0, sp, a0
-; RV64-NEXT: addi a0, a0, 16
-; RV64-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; RV64-NEXT: csrr a7, vlenb
+; RV64-NEXT: slli a7, a7, 3
+; RV64-NEXT: add a7, sp, a7
+; RV64-NEXT: addi a7, a7, 16
+; RV64-NEXT: vl8r.v v8, (a7) # Unknown-size Folded Reload
+; RV64-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV64-NEXT: vsub.vx v16, v8, a1, v0.t
; RV64-NEXT: vnot.v v8, v8, v0.t
; RV64-NEXT: vand.vv v8, v8, v16, v0.t
@@ -2004,10 +2102,14 @@ define <32 x i64> @vp_cttz_v32i64_unmasked(<32 x i64> %va, i32 zeroext %evl) {
; RV32-NEXT: addi sp, sp, -48
; RV32-NEXT: .cfi_def_cfa_offset 48
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: slli a1, a1, 5
+; RV32-NEXT: slli a1, a1, 4
; RV32-NEXT: sub sp, sp, a1
-; RV32-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x30, 0x22, 0x11, 0x20, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 48 + 32 * vlenb
-; RV32-NEXT: vmv8r.v v24, v16
+; RV32-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x30, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 48 + 16 * vlenb
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 3
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 48
+; RV32-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
; RV32-NEXT: lui a1, 349525
; RV32-NEXT: addi a1, a1, 1365
; RV32-NEXT: sw a1, 44(sp)
@@ -2032,96 +2134,73 @@ define <32 x i64> @vp_cttz_v32i64_unmasked(<32 x i64> %va, i32 zeroext %evl) {
; RV32-NEXT: .LBB35_2:
; RV32-NEXT: li a2, 1
; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; RV32-NEXT: vsub.vx v16, v8, a2
+; RV32-NEXT: vsub.vx v24, v8, a2
; RV32-NEXT: vnot.v v8, v8
-; RV32-NEXT: vand.vv v8, v8, v16
-; RV32-NEXT: vsrl.vi v16, v8, 1
+; RV32-NEXT: vand.vv v0, v8, v24
; RV32-NEXT: addi a3, sp, 40
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v0, (a3), zero
-; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: li a4, 24
-; RV32-NEXT: mul a3, a3, a4
-; RV32-NEXT: add a3, sp, a3
-; RV32-NEXT: addi a3, a3, 48
-; RV32-NEXT: vs8r.v v0, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v16, v0
-; RV32-NEXT: vsub.vv v8, v8, v16
-; RV32-NEXT: addi a3, sp, 32
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v0, (a3), zero
-; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v8, v0
-; RV32-NEXT: vsrl.vi v8, v8, 2
-; RV32-NEXT: vand.vv v8, v8, v0
-; RV32-NEXT: vadd.vv v8, v16, v8
-; RV32-NEXT: vsrl.vi v16, v8, 4
-; RV32-NEXT: vadd.vv v8, v8, v16
-; RV32-NEXT: addi a3, sp, 24
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV32-NEXT: vlse64.v v16, (a3), zero
-; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: slli a3, a3, 4
-; RV32-NEXT: add a3, sp, a3
-; RV32-NEXT: addi a3, a3, 48
-; RV32-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
+; RV32-NEXT: addi a3, sp, 32
+; RV32-NEXT: vlse64.v v8, (a3), zero
; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v8, v8, v16
-; RV32-NEXT: addi a3, sp, 16
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a3), zero
+; RV32-NEXT: vsrl.vi v24, v0, 1
+; RV32-NEXT: vand.vv v24, v24, v16
+; RV32-NEXT: vsub.vv v24, v0, v24
+; RV32-NEXT: vand.vv v0, v24, v8
+; RV32-NEXT: vsrl.vi v24, v24, 2
+; RV32-NEXT: vand.vv v24, v24, v8
+; RV32-NEXT: vadd.vv v24, v0, v24
+; RV32-NEXT: vsrl.vi v0, v24, 4
+; RV32-NEXT: vadd.vv v24, v24, v0
; RV32-NEXT: addi a3, sp, 48
-; RV32-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; RV32-NEXT: vmul.vv v8, v8, v16
-; RV32-NEXT: li a1, 56
-; RV32-NEXT: vsrl.vx v8, v8, a1
-; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: slli a3, a3, 3
-; RV32-NEXT: add a3, sp, a3
-; RV32-NEXT: addi a3, a3, 48
-; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
+; RV32-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
; RV32-NEXT: addi a3, a0, -16
; RV32-NEXT: sltu a0, a0, a3
; RV32-NEXT: addi a0, a0, -1
; RV32-NEXT: and a0, a0, a3
+; RV32-NEXT: csrr a3, vlenb
+; RV32-NEXT: slli a3, a3, 3
+; RV32-NEXT: add a3, sp, a3
+; RV32-NEXT: addi a3, a3, 48
+; RV32-NEXT: vl8r.v v0, (a3) # Unknown-size Folded Reload
+; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vsub.vx v24, v0, a2
+; RV32-NEXT: vnot.v v0, v0
+; RV32-NEXT: vand.vv v24, v0, v24
+; RV32-NEXT: vsrl.vi v0, v24, 1
+; RV32-NEXT: vand.vv v16, v0, v16
+; RV32-NEXT: addi a2, sp, 24
+; RV32-NEXT: vsub.vv v16, v24, v16
+; RV32-NEXT: vand.vv v24, v16, v8
+; RV32-NEXT: vsrl.vi v16, v16, 2
+; RV32-NEXT: vand.vv v8, v16, v8
+; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v16, (a2), zero
+; RV32-NEXT: addi a2, sp, 16
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vsub.vx v8, v24, a2
-; RV32-NEXT: vnot.v v24, v24
-; RV32-NEXT: vand.vv v8, v24, v8
-; RV32-NEXT: vsrl.vi v24, v8, 1
-; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: li a2, 24
-; RV32-NEXT: mul a0, a0, a2
-; RV32-NEXT: add a0, sp, a0
-; RV32-NEXT: addi a0, a0, 48
-; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; RV32-NEXT: vand.vv v24, v24, v16
-; RV32-NEXT: vsub.vv v8, v8, v24
-; RV32-NEXT: vand.vv v24, v8, v0
-; RV32-NEXT: vsrl.vi v8, v8, 2
-; RV32-NEXT: vand.vv v8, v8, v0
; RV32-NEXT: vadd.vv v8, v24, v8
-; RV32-NEXT: vsrl.vi v24, v8, 4
-; RV32-NEXT: vadd.vv v8, v8, v24
-; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 4
-; RV32-NEXT: add a0, sp, a0
-; RV32-NEXT: addi a0, a0, 48
-; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v24, (a2), zero
+; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vi v0, v8, 4
+; RV32-NEXT: vadd.vv v8, v8, v0
+; RV32-NEXT: addi a2, sp, 48
+; RV32-NEXT: vl8r.v v0, (a2) # Unknown-size Folded Reload
+; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
+; RV32-NEXT: vand.vv v0, v0, v16
+; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vand.vv v8, v8, v16
-; RV32-NEXT: addi a0, sp, 48
-; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; RV32-NEXT: vmul.vv v8, v8, v16
-; RV32-NEXT: vsrl.vx v16, v8, a1
-; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 3
-; RV32-NEXT: add a0, sp, a0
-; RV32-NEXT: addi a0, a0, 48
-; RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
+; RV32-NEXT: vmul.vv v16, v0, v24
+; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vmul.vv v24, v8, v24
+; RV32-NEXT: li a2, 56
+; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vx v8, v16, a2
+; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vx v16, v24, a2
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 5
+; RV32-NEXT: slli a0, a0, 4
; RV32-NEXT: add sp, sp, a0
; RV32-NEXT: addi sp, sp, 48
; RV32-NEXT: ret
@@ -3420,59 +3499,86 @@ define <8 x i64> @vp_cttz_zero_undef_v8i64_unmasked(<8 x i64> %va, i32 zeroext %
define <15 x i64> @vp_cttz_zero_undef_v15i64(<15 x i64> %va, <15 x i1> %m, i32 zeroext %evl) {
; RV32-LABEL: vp_cttz_zero_undef_v15i64:
; RV32: # %bb.0:
-; RV32-NEXT: addi sp, sp, -32
-; RV32-NEXT: .cfi_def_cfa_offset 32
+; RV32-NEXT: addi sp, sp, -48
+; RV32-NEXT: .cfi_def_cfa_offset 48
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 4
+; RV32-NEXT: sub sp, sp, a1
+; RV32-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x30, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 48 + 16 * vlenb
; RV32-NEXT: lui a1, 349525
; RV32-NEXT: addi a1, a1, 1365
-; RV32-NEXT: sw a1, 28(sp)
-; RV32-NEXT: sw a1, 24(sp)
+; RV32-NEXT: sw a1, 44(sp)
+; RV32-NEXT: sw a1, 40(sp)
; RV32-NEXT: lui a1, 209715
; RV32-NEXT: addi a1, a1, 819
-; RV32-NEXT: sw a1, 20(sp)
-; RV32-NEXT: sw a1, 16(sp)
+; RV32-NEXT: sw a1, 36(sp)
+; RV32-NEXT: sw a1, 32(sp)
; RV32-NEXT: lui a1, 61681
; RV32-NEXT: addi a1, a1, -241
-; RV32-NEXT: sw a1, 12(sp)
-; RV32-NEXT: sw a1, 8(sp)
+; RV32-NEXT: sw a1, 28(sp)
+; RV32-NEXT: sw a1, 24(sp)
; RV32-NEXT: lui a1, 4112
; RV32-NEXT: addi a1, a1, 257
-; RV32-NEXT: sw a1, 4(sp)
-; RV32-NEXT: sw a1, 0(sp)
+; RV32-NEXT: sw a1, 20(sp)
+; RV32-NEXT: sw a1, 16(sp)
; RV32-NEXT: li a1, 1
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vsub.vx v16, v8, a1, v0.t
; RV32-NEXT: vnot.v v8, v8, v0.t
-; RV32-NEXT: vand.vv v8, v8, v16, v0.t
-; RV32-NEXT: vsrl.vi v16, v8, 1, v0.t
-; RV32-NEXT: addi a1, sp, 24
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v24, (a1), zero
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v16, v24, v0.t
-; RV32-NEXT: vsub.vv v8, v8, v16, v0.t
-; RV32-NEXT: addi a1, sp, 16
+; RV32-NEXT: vand.vv v24, v8, v16, v0.t
+; RV32-NEXT: addi a1, sp, 48
+; RV32-NEXT: vs8r.v v24, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: addi a1, sp, 40
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v8, (a1), zero
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 3
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 48
+; RV32-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: addi a1, sp, 32
; RV32-NEXT: vlse64.v v16, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vi v8, v24, 1, v0.t
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 3
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 48
+; RV32-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v8, v8, v24, v0.t
+; RV32-NEXT: addi a1, sp, 48
+; RV32-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vsub.vv v8, v24, v8, v0.t
; RV32-NEXT: vand.vv v24, v8, v16, v0.t
; RV32-NEXT: vsrl.vi v8, v8, 2, v0.t
; RV32-NEXT: vand.vv v8, v8, v16, v0.t
; RV32-NEXT: vadd.vv v8, v24, v8, v0.t
-; RV32-NEXT: vsrl.vi v16, v8, 4, v0.t
-; RV32-NEXT: vadd.vv v8, v8, v16, v0.t
-; RV32-NEXT: addi a1, sp, 8
+; RV32-NEXT: addi a1, sp, 24
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV32-NEXT: vlse64.v v16, (a1), zero
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 3
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 48
+; RV32-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: addi a1, sp, 16
+; RV32-NEXT: vlse64.v v24, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vi v16, v8, 4, v0.t
+; RV32-NEXT: vadd.vv v8, v8, v16, v0.t
+; RV32-NEXT: csrr a0, vlenb
+; RV32-NEXT: slli a0, a0, 3
+; RV32-NEXT: add a0, sp, a0
+; RV32-NEXT: addi a0, a0, 48
+; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; RV32-NEXT: vand.vv v8, v8, v16, v0.t
-; RV32-NEXT: mv a1, sp
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a1), zero
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vmul.vv v8, v8, v16, v0.t
+; RV32-NEXT: vmul.vv v8, v8, v24, v0.t
; RV32-NEXT: li a0, 56
; RV32-NEXT: vsrl.vx v8, v8, a0, v0.t
-; RV32-NEXT: addi sp, sp, 32
+; RV32-NEXT: csrr a0, vlenb
+; RV32-NEXT: slli a0, a0, 4
+; RV32-NEXT: add sp, sp, a0
+; RV32-NEXT: addi sp, sp, 48
; RV32-NEXT: ret
;
; RV64-LABEL: vp_cttz_zero_undef_v15i64:
@@ -3542,33 +3648,29 @@ define <15 x i64> @vp_cttz_zero_undef_v15i64_unmasked(<15 x i64> %va, i32 zeroex
; RV32-NEXT: vsub.vx v16, v8, a1
; RV32-NEXT: vnot.v v8, v8
; RV32-NEXT: vand.vv v8, v8, v16
-; RV32-NEXT: vsrl.vi v16, v8, 1
; RV32-NEXT: addi a1, sp, 24
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v16, (a1), zero
+; RV32-NEXT: addi a1, sp, 16
; RV32-NEXT: vlse64.v v24, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v16, v24
+; RV32-NEXT: vsrl.vi v0, v8, 1
+; RV32-NEXT: vand.vv v16, v0, v16
; RV32-NEXT: vsub.vv v8, v8, v16
-; RV32-NEXT: addi a1, sp, 16
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a1), zero
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v24, v8, v16
+; RV32-NEXT: vand.vv v16, v8, v24
; RV32-NEXT: vsrl.vi v8, v8, 2
-; RV32-NEXT: vand.vv v8, v8, v16
-; RV32-NEXT: vadd.vv v8, v24, v8
-; RV32-NEXT: vsrl.vi v16, v8, 4
-; RV32-NEXT: vadd.vv v8, v8, v16
+; RV32-NEXT: vand.vv v8, v8, v24
+; RV32-NEXT: vadd.vv v8, v16, v8
; RV32-NEXT: addi a1, sp, 8
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV32-NEXT: vlse64.v v16, (a1), zero
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v8, v8, v16
; RV32-NEXT: mv a1, sp
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a1), zero
+; RV32-NEXT: vlse64.v v24, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vmul.vv v8, v8, v16
+; RV32-NEXT: vsrl.vi v0, v8, 4
+; RV32-NEXT: vadd.vv v8, v8, v0
+; RV32-NEXT: vand.vv v8, v8, v16
+; RV32-NEXT: vmul.vv v8, v8, v24
; RV32-NEXT: li a0, 56
; RV32-NEXT: vsrl.vx v8, v8, a0
; RV32-NEXT: addi sp, sp, 32
@@ -3618,59 +3720,86 @@ define <15 x i64> @vp_cttz_zero_undef_v15i64_unmasked(<15 x i64> %va, i32 zeroex
define <16 x i64> @vp_cttz_zero_undef_v16i64(<16 x i64> %va, <16 x i1> %m, i32 zeroext %evl) {
; RV32-LABEL: vp_cttz_zero_undef_v16i64:
; RV32: # %bb.0:
-; RV32-NEXT: addi sp, sp, -32
-; RV32-NEXT: .cfi_def_cfa_offset 32
+; RV32-NEXT: addi sp, sp, -48
+; RV32-NEXT: .cfi_def_cfa_offset 48
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 4
+; RV32-NEXT: sub sp, sp, a1
+; RV32-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x30, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 48 + 16 * vlenb
; RV32-NEXT: lui a1, 349525
; RV32-NEXT: addi a1, a1, 1365
-; RV32-NEXT: sw a1, 28(sp)
-; RV32-NEXT: sw a1, 24(sp)
+; RV32-NEXT: sw a1, 44(sp)
+; RV32-NEXT: sw a1, 40(sp)
; RV32-NEXT: lui a1, 209715
; RV32-NEXT: addi a1, a1, 819
-; RV32-NEXT: sw a1, 20(sp)
-; RV32-NEXT: sw a1, 16(sp)
+; RV32-NEXT: sw a1, 36(sp)
+; RV32-NEXT: sw a1, 32(sp)
; RV32-NEXT: lui a1, 61681
; RV32-NEXT: addi a1, a1, -241
-; RV32-NEXT: sw a1, 12(sp)
-; RV32-NEXT: sw a1, 8(sp)
+; RV32-NEXT: sw a1, 28(sp)
+; RV32-NEXT: sw a1, 24(sp)
; RV32-NEXT: lui a1, 4112
; RV32-NEXT: addi a1, a1, 257
-; RV32-NEXT: sw a1, 4(sp)
-; RV32-NEXT: sw a1, 0(sp)
+; RV32-NEXT: sw a1, 20(sp)
+; RV32-NEXT: sw a1, 16(sp)
; RV32-NEXT: li a1, 1
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vsub.vx v16, v8, a1, v0.t
-; RV32-NEXT: vnot.v v8, v8, v0.t
-; RV32-NEXT: vand.vv v8, v8, v16, v0.t
-; RV32-NEXT: vsrl.vi v16, v8, 1, v0.t
-; RV32-NEXT: addi a1, sp, 24
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v24, (a1), zero
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v16, v24, v0.t
-; RV32-NEXT: vsub.vv v8, v8, v16, v0.t
-; RV32-NEXT: addi a1, sp, 16
+; RV32-NEXT: vnot.v v8, v8, v0.t
+; RV32-NEXT: vand.vv v24, v8, v16, v0.t
+; RV32-NEXT: addi a1, sp, 48
+; RV32-NEXT: vs8r.v v24, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: addi a1, sp, 40
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v8, (a1), zero
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 3
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 48
+; RV32-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: addi a1, sp, 32
; RV32-NEXT: vlse64.v v16, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vi v8, v24, 1, v0.t
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 3
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 48
+; RV32-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v8, v8, v24, v0.t
+; RV32-NEXT: addi a1, sp, 48
+; RV32-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vsub.vv v8, v24, v8, v0.t
; RV32-NEXT: vand.vv v24, v8, v16, v0.t
; RV32-NEXT: vsrl.vi v8, v8, 2, v0.t
; RV32-NEXT: vand.vv v8, v8, v16, v0.t
; RV32-NEXT: vadd.vv v8, v24, v8, v0.t
-; RV32-NEXT: vsrl.vi v16, v8, 4, v0.t
-; RV32-NEXT: vadd.vv v8, v8, v16, v0.t
-; RV32-NEXT: addi a1, sp, 8
+; RV32-NEXT: addi a1, sp, 24
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV32-NEXT: vlse64.v v16, (a1), zero
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 3
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 48
+; RV32-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: addi a1, sp, 16
+; RV32-NEXT: vlse64.v v24, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vi v16, v8, 4, v0.t
+; RV32-NEXT: vadd.vv v8, v8, v16, v0.t
+; RV32-NEXT: csrr a0, vlenb
+; RV32-NEXT: slli a0, a0, 3
+; RV32-NEXT: add a0, sp, a0
+; RV32-NEXT: addi a0, a0, 48
+; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; RV32-NEXT: vand.vv v8, v8, v16, v0.t
-; RV32-NEXT: mv a1, sp
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a1), zero
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vmul.vv v8, v8, v16, v0.t
+; RV32-NEXT: vmul.vv v8, v8, v24, v0.t
; RV32-NEXT: li a0, 56
; RV32-NEXT: vsrl.vx v8, v8, a0, v0.t
-; RV32-NEXT: addi sp, sp, 32
+; RV32-NEXT: csrr a0, vlenb
+; RV32-NEXT: slli a0, a0, 4
+; RV32-NEXT: add sp, sp, a0
+; RV32-NEXT: addi sp, sp, 48
; RV32-NEXT: ret
;
; RV64-LABEL: vp_cttz_zero_undef_v16i64:
@@ -3740,33 +3869,29 @@ define <16 x i64> @vp_cttz_zero_undef_v16i64_unmasked(<16 x i64> %va, i32 zeroex
; RV32-NEXT: vsub.vx v16, v8, a1
; RV32-NEXT: vnot.v v8, v8
; RV32-NEXT: vand.vv v8, v8, v16
-; RV32-NEXT: vsrl.vi v16, v8, 1
; RV32-NEXT: addi a1, sp, 24
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v16, (a1), zero
+; RV32-NEXT: addi a1, sp, 16
; RV32-NEXT: vlse64.v v24, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v16, v24
+; RV32-NEXT: vsrl.vi v0, v8, 1
+; RV32-NEXT: vand.vv v16, v0, v16
; RV32-NEXT: vsub.vv v8, v8, v16
-; RV32-NEXT: addi a1, sp, 16
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a1), zero
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v24, v8, v16
+; RV32-NEXT: vand.vv v16, v8, v24
; RV32-NEXT: vsrl.vi v8, v8, 2
-; RV32-NEXT: vand.vv v8, v8, v16
-; RV32-NEXT: vadd.vv v8, v24, v8
-; RV32-NEXT: vsrl.vi v16, v8, 4
-; RV32-NEXT: vadd.vv v8, v8, v16
+; RV32-NEXT: vand.vv v8, v8, v24
+; RV32-NEXT: vadd.vv v8, v16, v8
; RV32-NEXT: addi a1, sp, 8
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV32-NEXT: vlse64.v v16, (a1), zero
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v8, v8, v16
; RV32-NEXT: mv a1, sp
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a1), zero
+; RV32-NEXT: vlse64.v v24, (a1), zero
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vmul.vv v8, v8, v16
+; RV32-NEXT: vsrl.vi v0, v8, 4
+; RV32-NEXT: vadd.vv v8, v8, v0
+; RV32-NEXT: vand.vv v8, v8, v16
+; RV32-NEXT: vmul.vv v8, v8, v24
; RV32-NEXT: li a0, 56
; RV32-NEXT: vsrl.vx v8, v8, a0
; RV32-NEXT: addi sp, sp, 32
@@ -3824,7 +3949,7 @@ define <32 x i64> @vp_cttz_zero_undef_v32i64(<32 x i64> %va, <32 x i1> %m, i32 z
; RV32-NEXT: sub sp, sp, a1
; RV32-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x30, 0x22, 0x11, 0x38, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 48 + 56 * vlenb
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: slli a1, a1, 5
+; RV32-NEXT: slli a1, a1, 4
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 48
; RV32-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
@@ -3858,111 +3983,145 @@ define <32 x i64> @vp_cttz_zero_undef_v32i64(<32 x i64> %va, <32 x i1> %m, i32 z
; RV32-NEXT: vnot.v v8, v8, v0.t
; RV32-NEXT: vand.vv v8, v8, v16, v0.t
; RV32-NEXT: csrr a3, vlenb
+; RV32-NEXT: slli a3, a3, 5
+; RV32-NEXT: add a3, sp, a3
+; RV32-NEXT: addi a3, a3, 48
+; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
+; RV32-NEXT: addi a3, sp, 40
+; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v8, (a3), zero
+; RV32-NEXT: csrr a3, vlenb
; RV32-NEXT: li a4, 40
; RV32-NEXT: mul a3, a3, a4
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: vsrl.vi v8, v8, 1, v0.t
+; RV32-NEXT: addi a3, sp, 32
+; RV32-NEXT: vlse64.v v8, (a3), zero
; RV32-NEXT: csrr a3, vlenb
; RV32-NEXT: li a4, 48
; RV32-NEXT: mul a3, a3, a4
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: addi a3, sp, 40
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v8, (a3), zero
+; RV32-NEXT: csrr a3, vlenb
+; RV32-NEXT: slli a3, a3, 5
+; RV32-NEXT: add a3, sp, a3
+; RV32-NEXT: addi a3, a3, 48
+; RV32-NEXT: vl8r.v v16, (a3) # Unknown-size Folded Reload
+; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vi v8, v16, 1, v0.t
; RV32-NEXT: csrr a3, vlenb
; RV32-NEXT: li a4, 24
; RV32-NEXT: mul a3, a3, a4
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: li a4, 48
+; RV32-NEXT: li a4, 40
; RV32-NEXT: mul a3, a3, a4
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
; RV32-NEXT: vl8r.v v16, (a3) # Unknown-size Folded Reload
-; RV32-NEXT: vand.vv v16, v16, v8, v0.t
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: li a4, 40
+; RV32-NEXT: li a4, 24
; RV32-NEXT: mul a3, a3, a4
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
; RV32-NEXT: vl8r.v v8, (a3) # Unknown-size Folded Reload
-; RV32-NEXT: vsub.vv v8, v8, v16, v0.t
+; RV32-NEXT: vand.vv v8, v8, v16, v0.t
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: li a4, 40
-; RV32-NEXT: mul a3, a3, a4
+; RV32-NEXT: slli a3, a3, 5
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
-; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: addi a3, sp, 32
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v8, (a3), zero
+; RV32-NEXT: vl8r.v v16, (a3) # Unknown-size Folded Reload
+; RV32-NEXT: vsub.vv v16, v16, v8, v0.t
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: li a4, 48
-; RV32-NEXT: mul a3, a3, a4
+; RV32-NEXT: slli a3, a3, 5
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
-; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
+; RV32-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: li a4, 40
-; RV32-NEXT: mul a3, a3, a4
+; RV32-NEXT: slli a3, a3, 5
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
; RV32-NEXT: vl8r.v v16, (a3) # Unknown-size Folded Reload
+; RV32-NEXT: csrr a3, vlenb
+; RV32-NEXT: li a4, 48
+; RV32-NEXT: mul a3, a3, a4
+; RV32-NEXT: add a3, sp, a3
+; RV32-NEXT: addi a3, a3, 48
+; RV32-NEXT: vl8r.v v8, (a3) # Unknown-size Folded Reload
; RV32-NEXT: vand.vv v16, v16, v8, v0.t
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: slli a3, a3, 4
+; RV32-NEXT: li a4, 24
+; RV32-NEXT: mul a3, a3, a4
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
; RV32-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: li a4, 40
-; RV32-NEXT: mul a3, a3, a4
+; RV32-NEXT: slli a3, a3, 5
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
; RV32-NEXT: vl8r.v v16, (a3) # Unknown-size Folded Reload
; RV32-NEXT: vsrl.vi v16, v16, 2, v0.t
; RV32-NEXT: vand.vv v16, v16, v8, v0.t
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: slli a3, a3, 4
+; RV32-NEXT: li a4, 24
+; RV32-NEXT: mul a3, a3, a4
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
; RV32-NEXT: vl8r.v v8, (a3) # Unknown-size Folded Reload
; RV32-NEXT: vadd.vv v8, v8, v16, v0.t
-; RV32-NEXT: vsrl.vi v16, v8, 4, v0.t
-; RV32-NEXT: vadd.vv v16, v8, v16, v0.t
+; RV32-NEXT: csrr a3, vlenb
+; RV32-NEXT: slli a3, a3, 3
+; RV32-NEXT: add a3, sp, a3
+; RV32-NEXT: addi a3, a3, 48
+; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
; RV32-NEXT: addi a3, sp, 24
+; RV32-NEXT: addi a4, sp, 16
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV32-NEXT: vlse64.v v8, (a3), zero
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: li a4, 40
-; RV32-NEXT: mul a3, a3, a4
+; RV32-NEXT: li a5, 24
+; RV32-NEXT: mul a3, a3, a5
+; RV32-NEXT: add a3, sp, a3
+; RV32-NEXT: addi a3, a3, 48
+; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
+; RV32-NEXT: vlse64.v v8, (a4), zero
+; RV32-NEXT: csrr a3, vlenb
+; RV32-NEXT: slli a3, a3, 5
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v8, v16, v8, v0.t
-; RV32-NEXT: addi a3, sp, 16
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a3), zero
; RV32-NEXT: csrr a3, vlenb
; RV32-NEXT: slli a3, a3, 3
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
-; RV32-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
+; RV32-NEXT: vl8r.v v8, (a3) # Unknown-size Folded Reload
; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
-; RV32-NEXT: vmul.vv v8, v8, v16, v0.t
+; RV32-NEXT: vsrl.vi v16, v8, 4, v0.t
+; RV32-NEXT: addi a2, sp, 48
+; RV32-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
+; RV32-NEXT: vl8r.v v16, (a2) # Unknown-size Folded Reload
+; RV32-NEXT: vadd.vv v16, v8, v16, v0.t
+; RV32-NEXT: csrr a2, vlenb
+; RV32-NEXT: li a3, 24
+; RV32-NEXT: mul a2, a2, a3
+; RV32-NEXT: add a2, sp, a2
+; RV32-NEXT: addi a2, a2, 48
+; RV32-NEXT: vl8r.v v8, (a2) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v16, v16, v8, v0.t
+; RV32-NEXT: csrr a2, vlenb
+; RV32-NEXT: slli a2, a2, 5
+; RV32-NEXT: add a2, sp, a2
+; RV32-NEXT: addi a2, a2, 48
+; RV32-NEXT: vl8r.v v8, (a2) # Unknown-size Folded Reload
+; RV32-NEXT: vmul.vv v8, v16, v8, v0.t
; RV32-NEXT: li a2, 56
; RV32-NEXT: vsrl.vx v8, v8, a2, v0.t
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: slli a3, a3, 4
+; RV32-NEXT: slli a3, a3, 3
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
@@ -3970,13 +4129,13 @@ define <32 x i64> @vp_cttz_zero_undef_v32i64(<32 x i64> %va, <32 x i1> %m, i32 z
; RV32-NEXT: sltu a0, a0, a3
; RV32-NEXT: addi a0, a0, -1
; RV32-NEXT: and a0, a0, a3
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vmv1r.v v0, v24
-; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 5
-; RV32-NEXT: add a0, sp, a0
-; RV32-NEXT: addi a0, a0, 48
-; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: csrr a3, vlenb
+; RV32-NEXT: slli a3, a3, 4
+; RV32-NEXT: add a3, sp, a3
+; RV32-NEXT: addi a3, a3, 48
+; RV32-NEXT: vl8r.v v16, (a3) # Unknown-size Folded Reload
+; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vsub.vx v8, v16, a1, v0.t
; RV32-NEXT: vnot.v v16, v16, v0.t
; RV32-NEXT: vand.vv v8, v16, v8, v0.t
@@ -3984,18 +4143,18 @@ define <32 x i64> @vp_cttz_zero_undef_v32i64(<32 x i64> %va, <32 x i1> %m, i32 z
; RV32-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; RV32-NEXT: vsrl.vi v8, v8, 1, v0.t
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 5
+; RV32-NEXT: slli a0, a0, 4
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 48
; RV32-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: li a1, 24
+; RV32-NEXT: li a1, 40
; RV32-NEXT: mul a0, a0, a1
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 48
; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 5
+; RV32-NEXT: slli a0, a0, 4
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 48
; RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
@@ -4004,17 +4163,35 @@ define <32 x i64> @vp_cttz_zero_undef_v32i64(<32 x i64> %va, <32 x i1> %m, i32 z
; RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; RV32-NEXT: vsub.vv v8, v8, v16, v0.t
; RV32-NEXT: csrr a0, vlenb
+; RV32-NEXT: li a1, 40
+; RV32-NEXT: mul a0, a0, a1
+; RV32-NEXT: add a0, sp, a0
+; RV32-NEXT: addi a0, a0, 48
+; RV32-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; RV32-NEXT: csrr a0, vlenb
; RV32-NEXT: li a1, 48
; RV32-NEXT: mul a0, a0, a1
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 48
+; RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: csrr a0, vlenb
+; RV32-NEXT: li a1, 40
+; RV32-NEXT: mul a0, a0, a1
+; RV32-NEXT: add a0, sp, a0
+; RV32-NEXT: addi a0, a0, 48
; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; RV32-NEXT: vand.vv v16, v8, v16, v0.t
+; RV32-NEXT: vand.vv v16, v16, v8, v0.t
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 5
+; RV32-NEXT: slli a0, a0, 4
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 48
; RV32-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
+; RV32-NEXT: csrr a0, vlenb
+; RV32-NEXT: li a1, 40
+; RV32-NEXT: mul a0, a0, a1
+; RV32-NEXT: add a0, sp, a0
+; RV32-NEXT: addi a0, a0, 48
+; RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; RV32-NEXT: vsrl.vi v8, v8, 2, v0.t
; RV32-NEXT: csrr a0, vlenb
; RV32-NEXT: li a1, 48
@@ -4024,7 +4201,7 @@ define <32 x i64> @vp_cttz_zero_undef_v32i64(<32 x i64> %va, <32 x i1> %m, i32 z
; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; RV32-NEXT: vand.vv v8, v8, v16, v0.t
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 5
+; RV32-NEXT: slli a0, a0, 4
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 48
; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
@@ -4032,21 +4209,21 @@ define <32 x i64> @vp_cttz_zero_undef_v32i64(<32 x i64> %va, <32 x i1> %m, i32 z
; RV32-NEXT: vsrl.vi v16, v8, 4, v0.t
; RV32-NEXT: vadd.vv v8, v8, v16, v0.t
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: li a1, 40
+; RV32-NEXT: li a1, 24
; RV32-NEXT: mul a0, a0, a1
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 48
; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; RV32-NEXT: vand.vv v8, v8, v16, v0.t
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 3
+; RV32-NEXT: slli a0, a0, 5
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 48
; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; RV32-NEXT: vmul.vv v8, v8, v16, v0.t
; RV32-NEXT: vsrl.vx v16, v8, a2, v0.t
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 4
+; RV32-NEXT: slli a0, a0, 3
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 48
; RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
@@ -4070,8 +4247,8 @@ define <32 x i64> @vp_cttz_zero_undef_v32i64(<32 x i64> %va, <32 x i1> %m, i32 z
; RV64-NEXT: add a1, sp, a1
; RV64-NEXT: addi a1, a1, 16
; RV64-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
-; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64-NEXT: li a1, 16
+; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64-NEXT: vslidedown.vi v24, v0, 2
; RV64-NEXT: mv a2, a0
; RV64-NEXT: bltu a0, a1, .LBB70_2
@@ -4118,13 +4295,13 @@ define <32 x i64> @vp_cttz_zero_undef_v32i64(<32 x i64> %va, <32 x i1> %m, i32 z
; RV64-NEXT: sltu a0, a0, a7
; RV64-NEXT: addi a0, a0, -1
; RV64-NEXT: and a0, a0, a7
-; RV64-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV64-NEXT: vmv1r.v v0, v24
-; RV64-NEXT: csrr a0, vlenb
-; RV64-NEXT: slli a0, a0, 3
-; RV64-NEXT: add a0, sp, a0
-; RV64-NEXT: addi a0, a0, 16
-; RV64-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; RV64-NEXT: csrr a7, vlenb
+; RV64-NEXT: slli a7, a7, 3
+; RV64-NEXT: add a7, sp, a7
+; RV64-NEXT: addi a7, a7, 16
+; RV64-NEXT: vl8r.v v8, (a7) # Unknown-size Folded Reload
+; RV64-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV64-NEXT: vsub.vx v16, v8, a1, v0.t
; RV64-NEXT: vnot.v v8, v8, v0.t
; RV64-NEXT: vand.vv v8, v8, v16, v0.t
@@ -4157,10 +4334,14 @@ define <32 x i64> @vp_cttz_zero_undef_v32i64_unmasked(<32 x i64> %va, i32 zeroex
; RV32-NEXT: addi sp, sp, -48
; RV32-NEXT: .cfi_def_cfa_offset 48
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: slli a1, a1, 5
+; RV32-NEXT: slli a1, a1, 4
; RV32-NEXT: sub sp, sp, a1
-; RV32-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x30, 0x22, 0x11, 0x20, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 48 + 32 * vlenb
-; RV32-NEXT: vmv8r.v v24, v16
+; RV32-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x30, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 48 + 16 * vlenb
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a1, a1, 3
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 48
+; RV32-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
; RV32-NEXT: lui a1, 349525
; RV32-NEXT: addi a1, a1, 1365
; RV32-NEXT: sw a1, 44(sp)
@@ -4185,96 +4366,73 @@ define <32 x i64> @vp_cttz_zero_undef_v32i64_unmasked(<32 x i64> %va, i32 zeroex
; RV32-NEXT: .LBB71_2:
; RV32-NEXT: li a2, 1
; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; RV32-NEXT: vsub.vx v16, v8, a2
+; RV32-NEXT: vsub.vx v24, v8, a2
; RV32-NEXT: vnot.v v8, v8
-; RV32-NEXT: vand.vv v8, v8, v16
-; RV32-NEXT: vsrl.vi v16, v8, 1
+; RV32-NEXT: vand.vv v0, v8, v24
; RV32-NEXT: addi a3, sp, 40
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v0, (a3), zero
-; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: li a4, 24
-; RV32-NEXT: mul a3, a3, a4
-; RV32-NEXT: add a3, sp, a3
-; RV32-NEXT: addi a3, a3, 48
-; RV32-NEXT: vs8r.v v0, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v16, v0
-; RV32-NEXT: vsub.vv v8, v8, v16
-; RV32-NEXT: addi a3, sp, 32
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v0, (a3), zero
-; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v16, v8, v0
-; RV32-NEXT: vsrl.vi v8, v8, 2
-; RV32-NEXT: vand.vv v8, v8, v0
-; RV32-NEXT: vadd.vv v8, v16, v8
-; RV32-NEXT: vsrl.vi v16, v8, 4
-; RV32-NEXT: vadd.vv v8, v8, v16
-; RV32-NEXT: addi a3, sp, 24
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV32-NEXT: vlse64.v v16, (a3), zero
-; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: slli a3, a3, 4
-; RV32-NEXT: add a3, sp, a3
-; RV32-NEXT: addi a3, a3, 48
-; RV32-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
+; RV32-NEXT: addi a3, sp, 32
+; RV32-NEXT: vlse64.v v8, (a3), zero
; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v8, v8, v16
-; RV32-NEXT: addi a3, sp, 16
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v16, (a3), zero
+; RV32-NEXT: vsrl.vi v24, v0, 1
+; RV32-NEXT: vand.vv v24, v24, v16
+; RV32-NEXT: vsub.vv v24, v0, v24
+; RV32-NEXT: vand.vv v0, v24, v8
+; RV32-NEXT: vsrl.vi v24, v24, 2
+; RV32-NEXT: vand.vv v24, v24, v8
+; RV32-NEXT: vadd.vv v24, v0, v24
+; RV32-NEXT: vsrl.vi v0, v24, 4
+; RV32-NEXT: vadd.vv v24, v24, v0
; RV32-NEXT: addi a3, sp, 48
-; RV32-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; RV32-NEXT: vmul.vv v8, v8, v16
-; RV32-NEXT: li a1, 56
-; RV32-NEXT: vsrl.vx v8, v8, a1
-; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: slli a3, a3, 3
-; RV32-NEXT: add a3, sp, a3
-; RV32-NEXT: addi a3, a3, 48
-; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
+; RV32-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
; RV32-NEXT: addi a3, a0, -16
; RV32-NEXT: sltu a0, a0, a3
; RV32-NEXT: addi a0, a0, -1
; RV32-NEXT: and a0, a0, a3
+; RV32-NEXT: csrr a3, vlenb
+; RV32-NEXT: slli a3, a3, 3
+; RV32-NEXT: add a3, sp, a3
+; RV32-NEXT: addi a3, a3, 48
+; RV32-NEXT: vl8r.v v0, (a3) # Unknown-size Folded Reload
+; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vsub.vx v24, v0, a2
+; RV32-NEXT: vnot.v v0, v0
+; RV32-NEXT: vand.vv v24, v0, v24
+; RV32-NEXT: vsrl.vi v0, v24, 1
+; RV32-NEXT: vand.vv v16, v0, v16
+; RV32-NEXT: addi a2, sp, 24
+; RV32-NEXT: vsub.vv v16, v24, v16
+; RV32-NEXT: vand.vv v24, v16, v8
+; RV32-NEXT: vsrl.vi v16, v16, 2
+; RV32-NEXT: vand.vv v8, v16, v8
+; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v16, (a2), zero
+; RV32-NEXT: addi a2, sp, 16
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vsub.vx v8, v24, a2
-; RV32-NEXT: vnot.v v24, v24
-; RV32-NEXT: vand.vv v8, v24, v8
-; RV32-NEXT: vsrl.vi v24, v8, 1
-; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: li a2, 24
-; RV32-NEXT: mul a0, a0, a2
-; RV32-NEXT: add a0, sp, a0
-; RV32-NEXT: addi a0, a0, 48
-; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; RV32-NEXT: vand.vv v24, v24, v16
-; RV32-NEXT: vsub.vv v8, v8, v24
-; RV32-NEXT: vand.vv v24, v8, v0
-; RV32-NEXT: vsrl.vi v8, v8, 2
-; RV32-NEXT: vand.vv v8, v8, v0
; RV32-NEXT: vadd.vv v8, v24, v8
-; RV32-NEXT: vsrl.vi v24, v8, 4
-; RV32-NEXT: vadd.vv v8, v8, v24
-; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 4
-; RV32-NEXT: add a0, sp, a0
-; RV32-NEXT: addi a0, a0, 48
-; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v24, (a2), zero
+; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vi v0, v8, 4
+; RV32-NEXT: vadd.vv v8, v8, v0
+; RV32-NEXT: addi a2, sp, 48
+; RV32-NEXT: vl8r.v v0, (a2) # Unknown-size Folded Reload
+; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
+; RV32-NEXT: vand.vv v0, v0, v16
+; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vand.vv v8, v8, v16
-; RV32-NEXT: addi a0, sp, 48
-; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; RV32-NEXT: vmul.vv v8, v8, v16
-; RV32-NEXT: vsrl.vx v16, v8, a1
-; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 3
-; RV32-NEXT: add a0, sp, a0
-; RV32-NEXT: addi a0, a0, 48
-; RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
+; RV32-NEXT: vmul.vv v16, v0, v24
+; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vmul.vv v24, v8, v24
+; RV32-NEXT: li a2, 56
+; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vx v8, v16, a2
+; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vx v16, v24, a2
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 5
+; RV32-NEXT: slli a0, a0, 4
; RV32-NEXT: add sp, sp, a0
; RV32-NEXT: addi sp, sp, 48
; RV32-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-deinterleave-load.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-deinterleave-load.ll
index 1587f770f87ca..9f8de22b25c2d 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-deinterleave-load.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-deinterleave-load.ll
@@ -15,16 +15,16 @@ define {<16 x i1>, <16 x i1>} @vector_deinterleave_load_v16i1_v32i1(ptr %p) {
; CHECK-NEXT: vmerge.vim v10, v8, 1, v0
; CHECK-NEXT: vid.v v9
; CHECK-NEXT: vadd.vv v11, v9, v9
-; CHECK-NEXT: vrgather.vv v9, v10, v11
; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; CHECK-NEXT: vslidedown.vi v0, v0, 2
; CHECK-NEXT: vsetivli zero, 16, e8, m1, ta, ma
+; CHECK-NEXT: vrgather.vv v9, v10, v11
; CHECK-NEXT: vmerge.vim v8, v8, 1, v0
-; CHECK-NEXT: vadd.vi v12, v11, -16
; CHECK-NEXT: li a0, -256
; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; CHECK-NEXT: vmv.s.x v0, a0
; CHECK-NEXT: vsetvli zero, zero, e8, m1, ta, mu
+; CHECK-NEXT: vadd.vi v12, v11, -16
; CHECK-NEXT: vrgather.vv v9, v8, v12, v0.t
; CHECK-NEXT: vmsne.vi v9, v9, 0
; CHECK-NEXT: vadd.vi v12, v11, 1
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-extract-i1.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-extract-i1.ll
index dccb62877af3c..386c71cf665ce 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-extract-i1.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-extract-i1.ll
@@ -326,9 +326,9 @@ define i1 @extractelt_v256i1(ptr %x, i64 %idx) nounwind {
; RV32-NEXT: andi sp, sp, -128
; RV32-NEXT: andi a1, a1, 255
; RV32-NEXT: li a2, 128
+; RV32-NEXT: addi a3, a0, 128
; RV32-NEXT: vsetvli zero, a2, e8, m8, ta, ma
-; RV32-NEXT: addi a2, a0, 128
-; RV32-NEXT: vle8.v v16, (a2)
+; RV32-NEXT: vle8.v v16, (a3)
; RV32-NEXT: vle8.v v24, (a0)
; RV32-NEXT: mv a0, sp
; RV32-NEXT: add a1, a0, a1
@@ -357,9 +357,9 @@ define i1 @extractelt_v256i1(ptr %x, i64 %idx) nounwind {
; RV64-NEXT: andi sp, sp, -128
; RV64-NEXT: andi a1, a1, 255
; RV64-NEXT: li a2, 128
+; RV64-NEXT: addi a3, a0, 128
; RV64-NEXT: vsetvli zero, a2, e8, m8, ta, ma
-; RV64-NEXT: addi a2, a0, 128
-; RV64-NEXT: vle8.v v16, (a2)
+; RV64-NEXT: vle8.v v16, (a3)
; RV64-NEXT: vle8.v v24, (a0)
; RV64-NEXT: mv a0, sp
; RV64-NEXT: add a1, a0, a1
@@ -388,9 +388,9 @@ define i1 @extractelt_v256i1(ptr %x, i64 %idx) nounwind {
; RV32ZBS-NEXT: andi sp, sp, -128
; RV32ZBS-NEXT: andi a1, a1, 255
; RV32ZBS-NEXT: li a2, 128
+; RV32ZBS-NEXT: addi a3, a0, 128
; RV32ZBS-NEXT: vsetvli zero, a2, e8, m8, ta, ma
-; RV32ZBS-NEXT: addi a2, a0, 128
-; RV32ZBS-NEXT: vle8.v v16, (a2)
+; RV32ZBS-NEXT: vle8.v v16, (a3)
; RV32ZBS-NEXT: vle8.v v24, (a0)
; RV32ZBS-NEXT: mv a0, sp
; RV32ZBS-NEXT: add a1, a0, a1
@@ -419,9 +419,9 @@ define i1 @extractelt_v256i1(ptr %x, i64 %idx) nounwind {
; RV64ZBS-NEXT: andi sp, sp, -128
; RV64ZBS-NEXT: andi a1, a1, 255
; RV64ZBS-NEXT: li a2, 128
+; RV64ZBS-NEXT: addi a3, a0, 128
; RV64ZBS-NEXT: vsetvli zero, a2, e8, m8, ta, ma
-; RV64ZBS-NEXT: addi a2, a0, 128
-; RV64ZBS-NEXT: vle8.v v16, (a2)
+; RV64ZBS-NEXT: vle8.v v16, (a3)
; RV64ZBS-NEXT: vle8.v v24, (a0)
; RV64ZBS-NEXT: mv a0, sp
; RV64ZBS-NEXT: add a1, a0, a1
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-extract.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-extract.ll
index e969da6fd45b7..1dd14709c839d 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-extract.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-extract.ll
@@ -721,9 +721,9 @@ define i32 @extractelt_v64i32_idx(ptr %x, i32 zeroext %idx) nounwind {
; RV32-NEXT: andi a1, a1, 63
; RV32-NEXT: slli a1, a1, 2
; RV32-NEXT: li a2, 32
+; RV32-NEXT: addi a3, a0, 128
; RV32-NEXT: vsetvli zero, a2, e32, m8, ta, ma
-; RV32-NEXT: addi a2, a0, 128
-; RV32-NEXT: vle32.v v8, (a2)
+; RV32-NEXT: vle32.v v8, (a3)
; RV32-NEXT: vle32.v v16, (a0)
; RV32-NEXT: mv a0, sp
; RV32-NEXT: add a1, a0, a1
@@ -749,9 +749,9 @@ define i32 @extractelt_v64i32_idx(ptr %x, i32 zeroext %idx) nounwind {
; RV64-NEXT: andi a1, a1, 63
; RV64-NEXT: slli a1, a1, 2
; RV64-NEXT: li a2, 32
+; RV64-NEXT: addi a3, a0, 128
; RV64-NEXT: vsetvli zero, a2, e32, m8, ta, ma
-; RV64-NEXT: addi a2, a0, 128
-; RV64-NEXT: vle32.v v8, (a2)
+; RV64-NEXT: vle32.v v8, (a3)
; RV64-NEXT: vle32.v v16, (a0)
; RV64-NEXT: mv a0, sp
; RV64-NEXT: add a1, a0, a1
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-floor-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-floor-vp.ll
index 287dd510674d1..c1b4c5fda6c64 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-floor-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-floor-vp.ll
@@ -204,8 +204,8 @@ define <8 x half> @vp_floor_v8f16(<8 x half> %va, <8 x i1> %m, i32 zeroext %evl)
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m2, ta, mu
; ZVFHMIN-NEXT: vmflt.vf v9, v12, fa5, v0.t
; ZVFHMIN-NEXT: fsrmi a0, 2
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v9
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; ZVFHMIN-NEXT: vfcvt.x.f.v v12, v10, v0.t
; ZVFHMIN-NEXT: fsrm a0
; ZVFHMIN-NEXT: vfcvt.f.x.v v12, v12, v0.t
@@ -261,16 +261,16 @@ declare <16 x half> @llvm.vp.floor.v16f16(<16 x half>, <16 x i1>, i32)
define <16 x half> @vp_floor_v16f16(<16 x half> %va, <16 x i1> %m, i32 zeroext %evl) {
; ZVFH-LABEL: vp_floor_v16f16:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vmv1r.v v10, v0
; ZVFH-NEXT: lui a1, %hi(.LCPI6_0)
; ZVFH-NEXT: flh fa5, %lo(.LCPI6_0)(a1)
+; ZVFH-NEXT: vmv1r.v v10, v0
; ZVFH-NEXT: vsetvli zero, a0, e16, m2, ta, ma
; ZVFH-NEXT: vfabs.v v12, v8, v0.t
; ZVFH-NEXT: vsetvli zero, zero, e16, m2, ta, mu
; ZVFH-NEXT: vmflt.vf v10, v12, fa5, v0.t
; ZVFH-NEXT: fsrmi a0, 2
-; ZVFH-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; ZVFH-NEXT: vmv1r.v v0, v10
+; ZVFH-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; ZVFH-NEXT: vfcvt.x.f.v v12, v8, v0.t
; ZVFH-NEXT: fsrm a0
; ZVFH-NEXT: vfcvt.f.x.v v12, v12, v0.t
@@ -290,8 +290,8 @@ define <16 x half> @vp_floor_v16f16(<16 x half> %va, <16 x i1> %m, i32 zeroext %
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m4, ta, mu
; ZVFHMIN-NEXT: vmflt.vf v10, v16, fa5, v0.t
; ZVFHMIN-NEXT: fsrmi a0, 2
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v10
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; ZVFHMIN-NEXT: vfcvt.x.f.v v16, v12, v0.t
; ZVFHMIN-NEXT: fsrm a0
; ZVFHMIN-NEXT: vfcvt.f.x.v v16, v16, v0.t
@@ -439,8 +439,8 @@ define <8 x float> @vp_floor_v8f32(<8 x float> %va, <8 x i1> %m, i32 zeroext %ev
; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, mu
; CHECK-NEXT: vmflt.vf v10, v12, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 2
-; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
@@ -483,8 +483,8 @@ define <16 x float> @vp_floor_v16f32(<16 x float> %va, <16 x i1> %m, i32 zeroext
; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, mu
; CHECK-NEXT: vmflt.vf v12, v16, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 2
-; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
@@ -561,16 +561,16 @@ declare <4 x double> @llvm.vp.floor.v4f64(<4 x double>, <4 x i1>, i32)
define <4 x double> @vp_floor_v4f64(<4 x double> %va, <4 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_floor_v4f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v10, v0
; CHECK-NEXT: lui a1, %hi(.LCPI18_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI18_0)(a1)
+; CHECK-NEXT: vmv1r.v v10, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, ma
; CHECK-NEXT: vfabs.v v12, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, mu
; CHECK-NEXT: vmflt.vf v10, v12, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 2
-; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
@@ -605,16 +605,16 @@ declare <8 x double> @llvm.vp.floor.v8f64(<8 x double>, <8 x i1>, i32)
define <8 x double> @vp_floor_v8f64(<8 x double> %va, <8 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_floor_v8f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v12, v0
; CHECK-NEXT: lui a1, %hi(.LCPI20_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI20_0)(a1)
+; CHECK-NEXT: vmv1r.v v12, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, ma
; CHECK-NEXT: vfabs.v v16, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, mu
; CHECK-NEXT: vmflt.vf v12, v16, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 2
-; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
@@ -649,16 +649,16 @@ declare <15 x double> @llvm.vp.floor.v15f64(<15 x double>, <15 x i1>, i32)
define <15 x double> @vp_floor_v15f64(<15 x double> %va, <15 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_floor_v15f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: lui a1, %hi(.LCPI22_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI22_0)(a1)
+; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
; CHECK-NEXT: vmflt.vf v16, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 2
-; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
@@ -693,16 +693,16 @@ declare <16 x double> @llvm.vp.floor.v16f64(<16 x double>, <16 x i1>, i32)
define <16 x double> @vp_floor_v16f64(<16 x double> %va, <16 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_floor_v16f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: lui a1, %hi(.LCPI24_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI24_0)(a1)
+; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
; CHECK-NEXT: vmflt.vf v16, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 2
-; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
@@ -737,69 +737,59 @@ declare <32 x double> @llvm.vp.floor.v32f64(<32 x double>, <32 x i1>, i32)
define <32 x double> @vp_floor_v32f64(<32 x double> %va, <32 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_floor_v32f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: addi sp, sp, -16
-; CHECK-NEXT: .cfi_def_cfa_offset 16
-; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a1, a1, 4
-; CHECK-NEXT: sub sp, sp, a1
-; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
-; CHECK-NEXT: vmv1r.v v25, v0
-; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a1, a1, 3
-; CHECK-NEXT: add a1, sp, a1
-; CHECK-NEXT: addi a1, a1, 16
-; CHECK-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
-; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
+; CHECK-NEXT: vmv1r.v v6, v0
; CHECK-NEXT: li a2, 16
-; CHECK-NEXT: vslidedown.vi v24, v0, 2
+; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
+; CHECK-NEXT: vslidedown.vi v7, v0, 2
; CHECK-NEXT: mv a1, a0
; CHECK-NEXT: bltu a0, a2, .LBB26_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: li a1, 16
; CHECK-NEXT: .LBB26_2:
+; CHECK-NEXT: addi sp, sp, -16
+; CHECK-NEXT: .cfi_def_cfa_offset 16
+; CHECK-NEXT: csrr a2, vlenb
+; CHECK-NEXT: slli a2, a2, 3
+; CHECK-NEXT: sub sp, sp, a2
+; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
; CHECK-NEXT: lui a2, %hi(.LCPI26_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI26_0)(a2)
+; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v25
-; CHECK-NEXT: vfabs.v v16, v8, v0.t
+; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vmflt.vf v25, v16, fa5, v0.t
+; CHECK-NEXT: vmflt.vf v6, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a1, 2
+; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v25
-; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
+; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
+; CHECK-NEXT: addi a2, sp, 16
+; CHECK-NEXT: vs8r.v v24, (a2) # Unknown-size Folded Spill
; CHECK-NEXT: fsrm a1
-; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
-; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vfsgnj.vv v8, v16, v8, v0.t
; CHECK-NEXT: addi a1, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
+; CHECK-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v24, v8, v0.t
; CHECK-NEXT: addi a1, a0, -16
; CHECK-NEXT: sltu a0, a0, a1
; CHECK-NEXT: addi a0, a0, -1
; CHECK-NEXT: and a0, a0, a1
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v24
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vfabs.v v8, v16, v0.t
+; CHECK-NEXT: vfabs.v v24, v16, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vmflt.vf v24, v8, fa5, v0.t
+; CHECK-NEXT: vmflt.vf v7, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 2
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v24
-; CHECK-NEXT: vfcvt.x.f.v v8, v16, v0.t
+; CHECK-NEXT: vfcvt.x.f.v v24, v16, v0.t
; CHECK-NEXT: fsrm a0
-; CHECK-NEXT: vfcvt.f.x.v v8, v8, v0.t
+; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vfsgnj.vv v16, v8, v16, v0.t
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vfsgnj.vv v16, v24, v16, v0.t
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 4
+; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add sp, sp, a0
; CHECK-NEXT: addi sp, sp, 16
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fmaximum-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fmaximum-vp.ll
index edb33158e32eb..51eb63f5f9221 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fmaximum-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fmaximum-vp.ll
@@ -177,8 +177,8 @@ define <8 x half> @vfmax_vv_v8f16(<8 x half> %va, <8 x half> %vb, <8 x i1> %m, i
; ZVFHMIN-NEXT: vmfeq.vv v8, v12, v12, v0.t
; ZVFHMIN-NEXT: vsetivli zero, 8, e16, m1, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v14, v9
-; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m2, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v8
+; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m2, ta, ma
; ZVFHMIN-NEXT: vmerge.vvm v16, v12, v14, v0
; ZVFHMIN-NEXT: vmv1r.v v0, v10
; ZVFHMIN-NEXT: vmfeq.vv v8, v14, v14, v0.t
@@ -253,8 +253,8 @@ define <16 x half> @vfmax_vv_v16f16(<16 x half> %va, <16 x half> %vb, <16 x i1>
; ZVFHMIN-NEXT: vmfeq.vv v8, v16, v16, v0.t
; ZVFHMIN-NEXT: vsetivli zero, 16, e16, m2, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v20, v10
-; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m4, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v8
+; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m4, ta, ma
; ZVFHMIN-NEXT: vmerge.vvm v24, v16, v20, v0
; ZVFHMIN-NEXT: vmv1r.v v0, v12
; ZVFHMIN-NEXT: vmfeq.vv v8, v20, v20, v0.t
@@ -608,7 +608,6 @@ define <32 x double> @vfmax_vv_v32f64(<32 x double> %va, <32 x double> %vb, <32
; CHECK-NEXT: slli a1, a1, 5
; CHECK-NEXT: sub sp, sp, a1
; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x20, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 32 * vlenb
-; CHECK-NEXT: vmv1r.v v6, v0
; CHECK-NEXT: addi a1, a0, 128
; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; CHECK-NEXT: vle64.v v24, (a1)
@@ -618,28 +617,28 @@ define <32 x double> @vfmax_vv_v32f64(<32 x double> %va, <32 x double> %vb, <32
; CHECK-NEXT: add a1, sp, a1
; CHECK-NEXT: addi a1, a1, 16
; CHECK-NEXT: vs8r.v v24, (a1) # Unknown-size Folded Spill
-; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
-; CHECK-NEXT: vslidedown.vi v7, v0, 2
-; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; CHECK-NEXT: vle64.v v24, (a0)
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 4
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: li a1, 16
+; CHECK-NEXT: vmv1r.v v6, v0
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: li a1, 16
+; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
+; CHECK-NEXT: vslidedown.vi v7, v0, 2
; CHECK-NEXT: mv a0, a2
; CHECK-NEXT: bltu a2, a1, .LBB24_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: li a0, 16
; CHECK-NEXT: .LBB24_2:
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v6
+; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vmfeq.vv v26, v8, v8, v0.t
; CHECK-NEXT: vmv1r.v v0, v26
; CHECK-NEXT: csrr a0, vlenb
@@ -666,13 +665,13 @@ define <32 x double> @vfmax_vv_v32f64(<32 x double> %va, <32 x double> %vb, <32
; CHECK-NEXT: sltu a1, a2, a0
; CHECK-NEXT: addi a1, a1, -1
; CHECK-NEXT: and a0, a1, a0
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v7
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: csrr a1, vlenb
+; CHECK-NEXT: slli a1, a1, 3
+; CHECK-NEXT: add a1, sp, a1
+; CHECK-NEXT: addi a1, a1, 16
+; CHECK-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vmfeq.vv v25, v16, v16, v0.t
; CHECK-NEXT: vmv1r.v v0, v25
; CHECK-NEXT: csrr a0, vlenb
@@ -759,9 +758,9 @@ define <32 x double> @vfmax_vv_v32f64_unmasked(<32 x double> %va, <32 x double>
; CHECK-NEXT: sltu a1, a2, a0
; CHECK-NEXT: addi a1, a1, -1
; CHECK-NEXT: and a0, a1, a0
+; CHECK-NEXT: addi a1, sp, 16
+; CHECK-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vmfeq.vv v0, v16, v16
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 4
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fminimum-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fminimum-vp.ll
index 48649c43f782a..03e0ac42c442c 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fminimum-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fminimum-vp.ll
@@ -177,8 +177,8 @@ define <8 x half> @vfmin_vv_v8f16(<8 x half> %va, <8 x half> %vb, <8 x i1> %m, i
; ZVFHMIN-NEXT: vmfeq.vv v8, v12, v12, v0.t
; ZVFHMIN-NEXT: vsetivli zero, 8, e16, m1, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v14, v9
-; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m2, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v8
+; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m2, ta, ma
; ZVFHMIN-NEXT: vmerge.vvm v16, v12, v14, v0
; ZVFHMIN-NEXT: vmv1r.v v0, v10
; ZVFHMIN-NEXT: vmfeq.vv v8, v14, v14, v0.t
@@ -253,8 +253,8 @@ define <16 x half> @vfmin_vv_v16f16(<16 x half> %va, <16 x half> %vb, <16 x i1>
; ZVFHMIN-NEXT: vmfeq.vv v8, v16, v16, v0.t
; ZVFHMIN-NEXT: vsetivli zero, 16, e16, m2, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v20, v10
-; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m4, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v8
+; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m4, ta, ma
; ZVFHMIN-NEXT: vmerge.vvm v24, v16, v20, v0
; ZVFHMIN-NEXT: vmv1r.v v0, v12
; ZVFHMIN-NEXT: vmfeq.vv v8, v20, v20, v0.t
@@ -608,7 +608,6 @@ define <32 x double> @vfmin_vv_v32f64(<32 x double> %va, <32 x double> %vb, <32
; CHECK-NEXT: slli a1, a1, 5
; CHECK-NEXT: sub sp, sp, a1
; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x20, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 32 * vlenb
-; CHECK-NEXT: vmv1r.v v6, v0
; CHECK-NEXT: addi a1, a0, 128
; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; CHECK-NEXT: vle64.v v24, (a1)
@@ -618,28 +617,28 @@ define <32 x double> @vfmin_vv_v32f64(<32 x double> %va, <32 x double> %vb, <32
; CHECK-NEXT: add a1, sp, a1
; CHECK-NEXT: addi a1, a1, 16
; CHECK-NEXT: vs8r.v v24, (a1) # Unknown-size Folded Spill
-; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
-; CHECK-NEXT: vslidedown.vi v7, v0, 2
-; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; CHECK-NEXT: vle64.v v24, (a0)
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 4
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: li a1, 16
+; CHECK-NEXT: vmv1r.v v6, v0
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: li a1, 16
+; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
+; CHECK-NEXT: vslidedown.vi v7, v0, 2
; CHECK-NEXT: mv a0, a2
; CHECK-NEXT: bltu a2, a1, .LBB24_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: li a0, 16
; CHECK-NEXT: .LBB24_2:
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v6
+; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vmfeq.vv v26, v8, v8, v0.t
; CHECK-NEXT: vmv1r.v v0, v26
; CHECK-NEXT: csrr a0, vlenb
@@ -666,13 +665,13 @@ define <32 x double> @vfmin_vv_v32f64(<32 x double> %va, <32 x double> %vb, <32
; CHECK-NEXT: sltu a1, a2, a0
; CHECK-NEXT: addi a1, a1, -1
; CHECK-NEXT: and a0, a1, a0
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v7
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: csrr a1, vlenb
+; CHECK-NEXT: slli a1, a1, 3
+; CHECK-NEXT: add a1, sp, a1
+; CHECK-NEXT: addi a1, a1, 16
+; CHECK-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vmfeq.vv v25, v16, v16, v0.t
; CHECK-NEXT: vmv1r.v v0, v25
; CHECK-NEXT: csrr a0, vlenb
@@ -759,9 +758,9 @@ define <32 x double> @vfmin_vv_v32f64_unmasked(<32 x double> %va, <32 x double>
; CHECK-NEXT: sltu a1, a2, a0
; CHECK-NEXT: addi a1, a1, -1
; CHECK-NEXT: and a0, a1, a0
+; CHECK-NEXT: addi a1, sp, 16
+; CHECK-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vmfeq.vv v0, v16, v16
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 4
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fnearbyint-constrained-sdnode.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fnearbyint-constrained-sdnode.ll
index 1b50214bbf164..9e9a8b8a4b644 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fnearbyint-constrained-sdnode.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fnearbyint-constrained-sdnode.ll
@@ -19,9 +19,9 @@ define <2 x half> @nearbyint_v2f16(<2 x half> %v) strictfp {
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v9, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v9, v9, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e16, mf4, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%r = call <2 x half> @llvm.experimental.constrained.nearbyint.v2f16(<2 x half> %v, metadata !"round.dynamic", metadata !"fpexcept.strict")
ret <2 x half> %r
@@ -42,9 +42,9 @@ define <4 x half> @nearbyint_v4f16(<4 x half> %v) strictfp {
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v9, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v9, v9, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e16, mf2, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%r = call <4 x half> @llvm.experimental.constrained.nearbyint.v4f16(<4 x half> %v, metadata !"round.dynamic", metadata !"fpexcept.strict")
ret <4 x half> %r
@@ -65,9 +65,9 @@ define <8 x half> @nearbyint_v8f16(<8 x half> %v) strictfp {
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v9, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v9, v9, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e16, m1, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%r = call <8 x half> @llvm.experimental.constrained.nearbyint.v8f16(<8 x half> %v, metadata !"round.dynamic", metadata !"fpexcept.strict")
ret <8 x half> %r
@@ -88,9 +88,9 @@ define <16 x half> @nearbyint_v16f16(<16 x half> %v) strictfp {
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v10, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v10, v10, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v10, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%r = call <16 x half> @llvm.experimental.constrained.nearbyint.v16f16(<16 x half> %v, metadata !"round.dynamic", metadata !"fpexcept.strict")
ret <16 x half> %r
@@ -112,9 +112,9 @@ define <32 x half> @nearbyint_v32f16(<32 x half> %v) strictfp {
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e16, m4, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v12, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%r = call <32 x half> @llvm.experimental.constrained.nearbyint.v32f16(<32 x half> %v, metadata !"round.dynamic", metadata !"fpexcept.strict")
ret <32 x half> %r
@@ -135,9 +135,9 @@ define <2 x float> @nearbyint_v2f32(<2 x float> %v) strictfp {
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v9, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v9, v9, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e32, mf2, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%r = call <2 x float> @llvm.experimental.constrained.nearbyint.v2f32(<2 x float> %v, metadata !"round.dynamic", metadata !"fpexcept.strict")
ret <2 x float> %r
@@ -158,9 +158,9 @@ define <4 x float> @nearbyint_v4f32(<4 x float> %v) strictfp {
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v9, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v9, v9, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e32, m1, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%r = call <4 x float> @llvm.experimental.constrained.nearbyint.v4f32(<4 x float> %v, metadata !"round.dynamic", metadata !"fpexcept.strict")
ret <4 x float> %r
@@ -181,9 +181,9 @@ define <8 x float> @nearbyint_v8f32(<8 x float> %v) strictfp {
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v10, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v10, v10, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v10, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%r = call <8 x float> @llvm.experimental.constrained.nearbyint.v8f32(<8 x float> %v, metadata !"round.dynamic", metadata !"fpexcept.strict")
ret <8 x float> %r
@@ -204,9 +204,9 @@ define <16 x float> @nearbyint_v16f32(<16 x float> %v) strictfp {
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v12, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%r = call <16 x float> @llvm.experimental.constrained.nearbyint.v16f32(<16 x float> %v, metadata !"round.dynamic", metadata !"fpexcept.strict")
ret <16 x float> %r
@@ -227,9 +227,9 @@ define <2 x double> @nearbyint_v2f64(<2 x double> %v) strictfp {
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v9, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v9, v9, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e64, m1, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%r = call <2 x double> @llvm.experimental.constrained.nearbyint.v2f64(<2 x double> %v, metadata !"round.dynamic", metadata !"fpexcept.strict")
ret <2 x double> %r
@@ -250,9 +250,9 @@ define <4 x double> @nearbyint_v4f64(<4 x double> %v) strictfp {
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v10, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v10, v10, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v10, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%r = call <4 x double> @llvm.experimental.constrained.nearbyint.v4f64(<4 x double> %v, metadata !"round.dynamic", metadata !"fpexcept.strict")
ret <4 x double> %r
@@ -273,9 +273,9 @@ define <8 x double> @nearbyint_v8f64(<8 x double> %v) strictfp {
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v12, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%r = call <8 x double> @llvm.experimental.constrained.nearbyint.v8f64(<8 x double> %v, metadata !"round.dynamic", metadata !"fpexcept.strict")
ret <8 x double> %r
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fp-buildvec.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fp-buildvec.ll
index 9e83efd351953..0047dc4fde643 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fp-buildvec.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fp-buildvec.ll
@@ -39,9 +39,10 @@ define <4 x float> @hang_when_merging_stores_after_legalization(<8 x float> %x,
; CHECK-NEXT: vmul.vx v14, v12, a0
; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; CHECK-NEXT: vrgatherei16.vv v12, v8, v14
-; CHECK-NEXT: vsetvli zero, zero, e16, m1, ta, ma
-; CHECK-NEXT: vadd.vi v8, v14, -14
+; CHECK-NEXT: vsetivli zero, 1, e8, mf8, ta, ma
; CHECK-NEXT: vmv.v.i v0, 12
+; CHECK-NEXT: vsetivli zero, 8, e16, m1, ta, ma
+; CHECK-NEXT: vadd.vi v8, v14, -14
; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, mu
; CHECK-NEXT: vrgatherei16.vv v12, v10, v8, v0.t
; CHECK-NEXT: vmv1r.v v8, v12
@@ -1407,8 +1408,8 @@ define <8 x float> @buildvec_v8f32_zvl256(float %e0, float %e1, float %e2, float
; CHECK-NEXT: vfmv.v.f v8, fa4
; CHECK-NEXT: vfslide1down.vf v8, v8, fa5
; CHECK-NEXT: vfslide1down.vf v8, v8, fa6
-; CHECK-NEXT: vfslide1down.vf v8, v8, fa7
; CHECK-NEXT: vmv.v.i v0, 15
+; CHECK-NEXT: vfslide1down.vf v8, v8, fa7
; CHECK-NEXT: vslidedown.vi v8, v9, 4, v0.t
; CHECK-NEXT: ret
%v0 = insertelement <8 x float> poison, float %e0, i64 0
@@ -1458,8 +1459,8 @@ define <8 x double> @buildvec_v8f64_zvl512(double %e0, double %e1, double %e2, d
; CHECK-NEXT: vfmv.v.f v8, fa4
; CHECK-NEXT: vfslide1down.vf v8, v8, fa5
; CHECK-NEXT: vfslide1down.vf v8, v8, fa6
-; CHECK-NEXT: vfslide1down.vf v8, v8, fa7
; CHECK-NEXT: vmv.v.i v0, 15
+; CHECK-NEXT: vfslide1down.vf v8, v8, fa7
; CHECK-NEXT: vslidedown.vi v8, v9, 4, v0.t
; CHECK-NEXT: ret
%v0 = insertelement <8 x double> poison, double %e0, i64 0
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fp-interleave.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fp-interleave.ll
index ed152e64a91ef..f3b124aa34dcb 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fp-interleave.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fp-interleave.ll
@@ -56,9 +56,9 @@ define <4 x double> @interleave_v2f64(<2 x double> %x, <2 x double> %y) {
; RV32-V512-NEXT: vsetivli zero, 4, e16, mf4, ta, ma
; RV32-V512-NEXT: vid.v v10
; RV32-V512-NEXT: vsrl.vi v11, v10, 1
+; RV32-V512-NEXT: vmv.v.i v0, 10
; RV32-V512-NEXT: vsetvli zero, zero, e64, m1, ta, mu
; RV32-V512-NEXT: vrgatherei16.vv v10, v8, v11
-; RV32-V512-NEXT: vmv.v.i v0, 10
; RV32-V512-NEXT: vrgatherei16.vv v10, v9, v11, v0.t
; RV32-V512-NEXT: vmv.v.v v8, v10
; RV32-V512-NEXT: ret
@@ -68,8 +68,8 @@ define <4 x double> @interleave_v2f64(<2 x double> %x, <2 x double> %y) {
; RV64-V512-NEXT: vsetivli zero, 4, e64, m1, ta, mu
; RV64-V512-NEXT: vid.v v10
; RV64-V512-NEXT: vsrl.vi v11, v10, 1
-; RV64-V512-NEXT: vrgather.vv v10, v8, v11
; RV64-V512-NEXT: vmv.v.i v0, 10
+; RV64-V512-NEXT: vrgather.vv v10, v8, v11
; RV64-V512-NEXT: vrgather.vv v10, v9, v11, v0.t
; RV64-V512-NEXT: vmv.v.v v8, v10
; RV64-V512-NEXT: ret
@@ -261,13 +261,13 @@ define <64 x float> @interleave_v32f32(<32 x float> %x, <32 x float> %y) {
; V128-NEXT: vwmaccu.vx v8, a0, v16
; V128-NEXT: lui a1, 699051
; V128-NEXT: addi a1, a1, -1366
-; V128-NEXT: li a2, 32
; V128-NEXT: vmv.s.x v0, a1
-; V128-NEXT: vsetvli zero, a2, e32, m8, ta, ma
+; V128-NEXT: li a1, 32
+; V128-NEXT: vsetvli zero, a1, e32, m8, ta, ma
; V128-NEXT: vmerge.vvm v24, v8, v24, v0
-; V128-NEXT: vsetivli zero, 16, e32, m4, ta, ma
; V128-NEXT: addi a1, sp, 16
; V128-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
+; V128-NEXT: vsetivli zero, 16, e32, m4, ta, ma
; V128-NEXT: vwaddu.vv v0, v16, v8
; V128-NEXT: vwmaccu.vx v0, a0, v8
; V128-NEXT: vmv8r.v v8, v0
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fp-shuffles.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fp-shuffles.ll
index 8dc32d13e4a34..b2d1c33b76913 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fp-shuffles.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fp-shuffles.ll
@@ -93,12 +93,11 @@ define <4 x double> @vrgather_shuffle_vv_v4f64(<4 x double> %x, <4 x double> %y)
; CHECK: # %bb.0:
; CHECK-NEXT: lui a0, %hi(.LCPI6_0)
; CHECK-NEXT: addi a0, a0, %lo(.LCPI6_0)
-; CHECK-NEXT: vsetivli zero, 4, e64, m2, ta, ma
+; CHECK-NEXT: vsetivli zero, 4, e16, mf2, ta, ma
; CHECK-NEXT: vle16.v v14, (a0)
-; CHECK-NEXT: vrgatherei16.vv v12, v8, v14
-; CHECK-NEXT: vsetivli zero, 1, e8, mf8, ta, ma
; CHECK-NEXT: vmv.v.i v0, 8
-; CHECK-NEXT: vsetivli zero, 4, e64, m2, ta, mu
+; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, mu
+; CHECK-NEXT: vrgatherei16.vv v12, v8, v14
; CHECK-NEXT: vrgather.vi v12, v10, 1, v0.t
; CHECK-NEXT: vmv.v.v v8, v12
; CHECK-NEXT: ret
@@ -109,13 +108,13 @@ define <4 x double> @vrgather_shuffle_vv_v4f64(<4 x double> %x, <4 x double> %y)
define <4 x double> @vrgather_shuffle_xv_v4f64(<4 x double> %x) {
; CHECK-LABEL: vrgather_shuffle_xv_v4f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetivli zero, 4, e16, mf2, ta, ma
-; CHECK-NEXT: vid.v v12
; CHECK-NEXT: lui a0, %hi(.LCPI7_0)
; CHECK-NEXT: addi a0, a0, %lo(.LCPI7_0)
+; CHECK-NEXT: vsetivli zero, 4, e16, mf2, ta, ma
; CHECK-NEXT: vlse64.v v10, (a0), zero
-; CHECK-NEXT: vrsub.vi v12, v12, 4
+; CHECK-NEXT: vid.v v12
; CHECK-NEXT: vmv.v.i v0, 12
+; CHECK-NEXT: vrsub.vi v12, v12, 4
; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, mu
; CHECK-NEXT: vrgatherei16.vv v10, v8, v12, v0.t
; CHECK-NEXT: vmv.v.v v8, v10
@@ -129,12 +128,12 @@ define <4 x double> @vrgather_shuffle_vx_v4f64(<4 x double> %x) {
; CHECK: # %bb.0:
; CHECK-NEXT: vsetivli zero, 4, e16, mf2, ta, ma
; CHECK-NEXT: vid.v v12
+; CHECK-NEXT: lui a0, %hi(.LCPI8_0)
+; CHECK-NEXT: addi a0, a0, %lo(.LCPI8_0)
+; CHECK-NEXT: vlse64.v v10, (a0), zero
; CHECK-NEXT: li a0, 3
-; CHECK-NEXT: lui a1, %hi(.LCPI8_0)
-; CHECK-NEXT: addi a1, a1, %lo(.LCPI8_0)
-; CHECK-NEXT: vlse64.v v10, (a1), zero
-; CHECK-NEXT: vmul.vx v12, v12, a0
; CHECK-NEXT: vmv.v.i v0, 3
+; CHECK-NEXT: vmul.vx v12, v12, a0
; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, mu
; CHECK-NEXT: vrgatherei16.vv v10, v8, v12, v0.t
; CHECK-NEXT: vmv.v.v v8, v10
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fp.ll
index 0f003d7af6100..d25312268ada6 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fp.ll
@@ -1199,7 +1199,7 @@ declare <4 x half> @llvm.copysign.v4f16(<4 x half>, <4 x half>)
define void @copysign_neg_trunc_v3f16_v3f32(ptr %x, ptr %y) {
; ZVFH-LABEL: copysign_neg_trunc_v3f16_v3f32:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vsetivli zero, 3, e16, mf2, ta, ma
+; ZVFH-NEXT: vsetivli zero, 3, e32, m1, ta, ma
; ZVFH-NEXT: vle32.v v8, (a1)
; ZVFH-NEXT: vle16.v v9, (a0)
; ZVFH-NEXT: vsetivli zero, 4, e16, mf2, ta, ma
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fp2i-sat.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fp2i-sat.ll
index a8e4af2d7368e..bc46e7d264bc0 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fp2i-sat.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fp2i-sat.ll
@@ -351,25 +351,23 @@ define void @fp2si_v8f64_v8i8(ptr %x, ptr %y) {
; RV32-NEXT: fmin.d fa3, fa3, fa4
; RV32-NEXT: fcvt.w.d a2, fa3, rtz
; RV32-NEXT: and a0, a0, a2
-; RV32-NEXT: vsetivli zero, 8, e8, mf2, ta, ma
-; RV32-NEXT: vslide1down.vx v10, v10, a0
-; RV32-NEXT: vsetivli zero, 1, e64, m2, ta, ma
; RV32-NEXT: vslidedown.vi v8, v8, 3
; RV32-NEXT: vfmv.f.s fa3, v8
-; RV32-NEXT: feq.d a0, fa3, fa3
+; RV32-NEXT: feq.d a2, fa3, fa3
; RV32-NEXT: fmax.d fa3, fa3, fa5
; RV32-NEXT: fmin.d fa3, fa3, fa4
+; RV32-NEXT: fcvt.w.d a3, fa3, rtz
+; RV32-NEXT: fld fa3, 40(sp)
; RV32-NEXT: vsetivli zero, 8, e8, mf2, ta, mu
-; RV32-NEXT: fld fa2, 40(sp)
-; RV32-NEXT: fcvt.w.d a2, fa3, rtz
-; RV32-NEXT: neg a0, a0
-; RV32-NEXT: and a0, a0, a2
-; RV32-NEXT: feq.d a2, fa2, fa2
-; RV32-NEXT: fmax.d fa3, fa2, fa5
+; RV32-NEXT: vslide1down.vx v8, v10, a0
+; RV32-NEXT: neg a0, a2
+; RV32-NEXT: and a0, a0, a3
+; RV32-NEXT: feq.d a2, fa3, fa3
+; RV32-NEXT: fmax.d fa3, fa3, fa5
; RV32-NEXT: fmin.d fa3, fa3, fa4
; RV32-NEXT: fcvt.w.d a3, fa3, rtz
; RV32-NEXT: fld fa3, 32(sp)
-; RV32-NEXT: vslide1down.vx v8, v10, a0
+; RV32-NEXT: vslide1down.vx v8, v8, a0
; RV32-NEXT: neg a0, a2
; RV32-NEXT: and a0, a0, a3
; RV32-NEXT: feq.d a2, fa3, fa3
@@ -395,8 +393,8 @@ define void @fp2si_v8f64_v8i8(ptr %x, ptr %y) {
; RV32-NEXT: fmin.d fa5, fa5, fa4
; RV32-NEXT: fcvt.w.d a2, fa5, rtz
; RV32-NEXT: and a0, a0, a2
-; RV32-NEXT: vslide1down.vx v9, v9, a0
; RV32-NEXT: vmv.v.i v0, 15
+; RV32-NEXT: vslide1down.vx v9, v9, a0
; RV32-NEXT: vslidedown.vi v9, v8, 4, v0.t
; RV32-NEXT: vse8.v v9, (a1)
; RV32-NEXT: addi sp, s0, -128
@@ -452,25 +450,23 @@ define void @fp2si_v8f64_v8i8(ptr %x, ptr %y) {
; RV64-NEXT: fmin.d fa3, fa3, fa4
; RV64-NEXT: fcvt.l.d a2, fa3, rtz
; RV64-NEXT: and a0, a0, a2
-; RV64-NEXT: vsetivli zero, 8, e8, mf2, ta, ma
-; RV64-NEXT: vslide1down.vx v10, v10, a0
-; RV64-NEXT: vsetivli zero, 1, e64, m2, ta, ma
; RV64-NEXT: vslidedown.vi v8, v8, 3
; RV64-NEXT: vfmv.f.s fa3, v8
-; RV64-NEXT: feq.d a0, fa3, fa3
+; RV64-NEXT: feq.d a2, fa3, fa3
; RV64-NEXT: fmax.d fa3, fa3, fa5
; RV64-NEXT: fmin.d fa3, fa3, fa4
+; RV64-NEXT: fcvt.l.d a3, fa3, rtz
+; RV64-NEXT: fld fa3, 40(sp)
; RV64-NEXT: vsetivli zero, 8, e8, mf2, ta, mu
-; RV64-NEXT: fld fa2, 40(sp)
-; RV64-NEXT: fcvt.l.d a2, fa3, rtz
-; RV64-NEXT: neg a0, a0
-; RV64-NEXT: and a0, a0, a2
-; RV64-NEXT: feq.d a2, fa2, fa2
-; RV64-NEXT: fmax.d fa3, fa2, fa5
+; RV64-NEXT: vslide1down.vx v8, v10, a0
+; RV64-NEXT: neg a0, a2
+; RV64-NEXT: and a0, a0, a3
+; RV64-NEXT: feq.d a2, fa3, fa3
+; RV64-NEXT: fmax.d fa3, fa3, fa5
; RV64-NEXT: fmin.d fa3, fa3, fa4
; RV64-NEXT: fcvt.l.d a3, fa3, rtz
; RV64-NEXT: fld fa3, 32(sp)
-; RV64-NEXT: vslide1down.vx v8, v10, a0
+; RV64-NEXT: vslide1down.vx v8, v8, a0
; RV64-NEXT: neg a0, a2
; RV64-NEXT: and a0, a0, a3
; RV64-NEXT: feq.d a2, fa3, fa3
@@ -496,8 +492,8 @@ define void @fp2si_v8f64_v8i8(ptr %x, ptr %y) {
; RV64-NEXT: fmin.d fa5, fa5, fa4
; RV64-NEXT: fcvt.l.d a2, fa5, rtz
; RV64-NEXT: and a0, a0, a2
-; RV64-NEXT: vslide1down.vx v9, v9, a0
; RV64-NEXT: vmv.v.i v0, 15
+; RV64-NEXT: vslide1down.vx v9, v9, a0
; RV64-NEXT: vslidedown.vi v9, v8, 4, v0.t
; RV64-NEXT: vse8.v v9, (a1)
; RV64-NEXT: addi sp, s0, -128
@@ -542,46 +538,43 @@ define void @fp2ui_v8f64_v8i8(ptr %x, ptr %y) {
; RV32-NEXT: fmax.d fa4, fa4, fa3
; RV32-NEXT: fmin.d fa4, fa4, fa5
; RV32-NEXT: fcvt.wu.d a2, fa4, rtz
-; RV32-NEXT: vsetivli zero, 8, e8, mf2, ta, ma
-; RV32-NEXT: vmv.v.x v10, a2
-; RV32-NEXT: vslide1down.vx v10, v10, a0
; RV32-NEXT: vsetivli zero, 1, e64, m2, ta, ma
-; RV32-NEXT: vslidedown.vi v12, v8, 2
-; RV32-NEXT: vfmv.f.s fa4, v12
+; RV32-NEXT: vslidedown.vi v10, v8, 2
+; RV32-NEXT: vfmv.f.s fa4, v10
; RV32-NEXT: fmax.d fa4, fa4, fa3
; RV32-NEXT: fmin.d fa4, fa4, fa5
-; RV32-NEXT: fcvt.wu.d a0, fa4, rtz
-; RV32-NEXT: vsetivli zero, 8, e8, mf2, ta, ma
-; RV32-NEXT: vslide1down.vx v10, v10, a0
-; RV32-NEXT: vsetivli zero, 1, e64, m2, ta, ma
+; RV32-NEXT: fcvt.wu.d a3, fa4, rtz
; RV32-NEXT: vslidedown.vi v8, v8, 3
; RV32-NEXT: vfmv.f.s fa4, v8
-; RV32-NEXT: fmax.d fa4, fa4, fa3
-; RV32-NEXT: vsetivli zero, 8, e8, mf2, ta, mu
; RV32-NEXT: fld fa2, 40(sp)
-; RV32-NEXT: fmin.d fa4, fa4, fa5
-; RV32-NEXT: fcvt.wu.d a0, fa4, rtz
-; RV32-NEXT: fld fa4, 32(sp)
-; RV32-NEXT: fmax.d fa2, fa2, fa3
-; RV32-NEXT: fmin.d fa2, fa2, fa5
-; RV32-NEXT: fcvt.wu.d a2, fa2, rtz
; RV32-NEXT: fmax.d fa4, fa4, fa3
-; RV32-NEXT: fld fa2, 48(sp)
; RV32-NEXT: fmin.d fa4, fa4, fa5
-; RV32-NEXT: fcvt.wu.d a3, fa4, rtz
-; RV32-NEXT: vslide1down.vx v8, v10, a0
+; RV32-NEXT: fcvt.wu.d a4, fa4, rtz
+; RV32-NEXT: fmax.d fa4, fa2, fa3
+; RV32-NEXT: fld fa2, 32(sp)
+; RV32-NEXT: fmin.d fa4, fa4, fa5
+; RV32-NEXT: fcvt.wu.d a5, fa4, rtz
+; RV32-NEXT: vsetivli zero, 8, e8, mf2, ta, mu
+; RV32-NEXT: vmv.v.x v8, a2
; RV32-NEXT: fmax.d fa4, fa2, fa3
; RV32-NEXT: fmin.d fa4, fa4, fa5
+; RV32-NEXT: fcvt.wu.d a2, fa4, rtz
+; RV32-NEXT: fld fa4, 48(sp)
+; RV32-NEXT: vslide1down.vx v8, v8, a0
+; RV32-NEXT: vslide1down.vx v8, v8, a3
+; RV32-NEXT: vslide1down.vx v8, v8, a4
+; RV32-NEXT: fmax.d fa4, fa4, fa3
+; RV32-NEXT: fmin.d fa4, fa4, fa5
; RV32-NEXT: fcvt.wu.d a0, fa4, rtz
; RV32-NEXT: fld fa4, 56(sp)
-; RV32-NEXT: vmv.v.x v9, a3
-; RV32-NEXT: vslide1down.vx v9, v9, a2
+; RV32-NEXT: vmv.v.x v9, a2
+; RV32-NEXT: vslide1down.vx v9, v9, a5
; RV32-NEXT: vslide1down.vx v9, v9, a0
; RV32-NEXT: fmax.d fa4, fa4, fa3
; RV32-NEXT: fmin.d fa5, fa4, fa5
; RV32-NEXT: fcvt.wu.d a0, fa5, rtz
-; RV32-NEXT: vslide1down.vx v9, v9, a0
; RV32-NEXT: vmv.v.i v0, 15
+; RV32-NEXT: vslide1down.vx v9, v9, a0
; RV32-NEXT: vslidedown.vi v9, v8, 4, v0.t
; RV32-NEXT: vse8.v v9, (a1)
; RV32-NEXT: addi sp, s0, -128
@@ -618,46 +611,43 @@ define void @fp2ui_v8f64_v8i8(ptr %x, ptr %y) {
; RV64-NEXT: fmax.d fa4, fa4, fa3
; RV64-NEXT: fmin.d fa4, fa4, fa5
; RV64-NEXT: fcvt.lu.d a2, fa4, rtz
-; RV64-NEXT: vsetivli zero, 8, e8, mf2, ta, ma
-; RV64-NEXT: vmv.v.x v10, a2
-; RV64-NEXT: vslide1down.vx v10, v10, a0
; RV64-NEXT: vsetivli zero, 1, e64, m2, ta, ma
-; RV64-NEXT: vslidedown.vi v12, v8, 2
-; RV64-NEXT: vfmv.f.s fa4, v12
+; RV64-NEXT: vslidedown.vi v10, v8, 2
+; RV64-NEXT: vfmv.f.s fa4, v10
; RV64-NEXT: fmax.d fa4, fa4, fa3
; RV64-NEXT: fmin.d fa4, fa4, fa5
-; RV64-NEXT: fcvt.lu.d a0, fa4, rtz
-; RV64-NEXT: vsetivli zero, 8, e8, mf2, ta, ma
-; RV64-NEXT: vslide1down.vx v10, v10, a0
-; RV64-NEXT: vsetivli zero, 1, e64, m2, ta, ma
+; RV64-NEXT: fcvt.lu.d a3, fa4, rtz
; RV64-NEXT: vslidedown.vi v8, v8, 3
; RV64-NEXT: vfmv.f.s fa4, v8
-; RV64-NEXT: fmax.d fa4, fa4, fa3
-; RV64-NEXT: vsetivli zero, 8, e8, mf2, ta, mu
; RV64-NEXT: fld fa2, 40(sp)
-; RV64-NEXT: fmin.d fa4, fa4, fa5
-; RV64-NEXT: fcvt.lu.d a0, fa4, rtz
-; RV64-NEXT: fld fa4, 32(sp)
-; RV64-NEXT: fmax.d fa2, fa2, fa3
-; RV64-NEXT: fmin.d fa2, fa2, fa5
-; RV64-NEXT: fcvt.lu.d a2, fa2, rtz
; RV64-NEXT: fmax.d fa4, fa4, fa3
-; RV64-NEXT: fld fa2, 48(sp)
; RV64-NEXT: fmin.d fa4, fa4, fa5
-; RV64-NEXT: fcvt.lu.d a3, fa4, rtz
-; RV64-NEXT: vslide1down.vx v8, v10, a0
+; RV64-NEXT: fcvt.lu.d a4, fa4, rtz
; RV64-NEXT: fmax.d fa4, fa2, fa3
+; RV64-NEXT: fld fa2, 32(sp)
+; RV64-NEXT: fmin.d fa4, fa4, fa5
+; RV64-NEXT: fcvt.lu.d a5, fa4, rtz
+; RV64-NEXT: vsetivli zero, 8, e8, mf2, ta, mu
+; RV64-NEXT: vmv.v.x v8, a2
+; RV64-NEXT: fmax.d fa4, fa2, fa3
+; RV64-NEXT: fmin.d fa4, fa4, fa5
+; RV64-NEXT: fcvt.lu.d a2, fa4, rtz
+; RV64-NEXT: fld fa4, 48(sp)
+; RV64-NEXT: vslide1down.vx v8, v8, a0
+; RV64-NEXT: vslide1down.vx v8, v8, a3
+; RV64-NEXT: vslide1down.vx v8, v8, a4
+; RV64-NEXT: fmax.d fa4, fa4, fa3
; RV64-NEXT: fmin.d fa4, fa4, fa5
; RV64-NEXT: fcvt.lu.d a0, fa4, rtz
; RV64-NEXT: fld fa4, 56(sp)
-; RV64-NEXT: vmv.v.x v9, a3
-; RV64-NEXT: vslide1down.vx v9, v9, a2
+; RV64-NEXT: vmv.v.x v9, a2
+; RV64-NEXT: vslide1down.vx v9, v9, a5
; RV64-NEXT: vslide1down.vx v9, v9, a0
; RV64-NEXT: fmax.d fa4, fa4, fa3
; RV64-NEXT: fmin.d fa5, fa4, fa5
; RV64-NEXT: fcvt.lu.d a0, fa5, rtz
-; RV64-NEXT: vslide1down.vx v9, v9, a0
; RV64-NEXT: vmv.v.i v0, 15
+; RV64-NEXT: vslide1down.vx v9, v9, a0
; RV64-NEXT: vslidedown.vi v9, v8, 4, v0.t
; RV64-NEXT: vse8.v v9, (a1)
; RV64-NEXT: addi sp, s0, -128
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fpext-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fpext-vp.ll
index 48cc3f17a6269..f195eeadf0274 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fpext-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fpext-vp.ll
@@ -96,8 +96,8 @@ declare <32 x double> @llvm.vp.fpext.v32f64.v32f32(<32 x float>, <32 x i1>, i32)
define <32 x double> @vfpext_v32f32_v32f64(<32 x float> %a, <32 x i1> %m, i32 zeroext %vl) {
; CHECK-LABEL: vfpext_v32f32_v32f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; CHECK-NEXT: li a2, 16
+; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; CHECK-NEXT: vslidedown.vi v16, v0, 2
; CHECK-NEXT: mv a1, a0
; CHECK-NEXT: bltu a0, a2, .LBB7_2
@@ -112,8 +112,8 @@ define <32 x double> @vfpext_v32f32_v32f64(<32 x float> %a, <32 x i1> %m, i32 ze
; CHECK-NEXT: and a0, a0, a1
; CHECK-NEXT: vsetivli zero, 16, e32, m8, ta, ma
; CHECK-NEXT: vslidedown.vi v8, v8, 16
-; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, ma
; CHECK-NEXT: vfwcvt.f.f.v v16, v8, v0.t
; CHECK-NEXT: vmv8r.v v8, v24
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fptosi-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fptosi-vp.ll
index 49a1b19b58a27..a4050b716e787 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fptosi-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fptosi-vp.ll
@@ -394,8 +394,8 @@ declare <32 x i64> @llvm.vp.fptosi.v32i64.v32f64(<32 x double>, <32 x i1>, i32)
define <32 x i64> @vfptosi_v32i64_v32f64(<32 x double> %va, <32 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfptosi_v32i64_v32f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; CHECK-NEXT: li a2, 16
+; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; CHECK-NEXT: vslidedown.vi v24, v0, 2
; CHECK-NEXT: mv a1, a0
; CHECK-NEXT: bltu a0, a2, .LBB25_2
@@ -408,8 +408,8 @@ define <32 x i64> @vfptosi_v32i64_v32f64(<32 x double> %va, <32 x i1> %m, i32 ze
; CHECK-NEXT: sltu a0, a0, a1
; CHECK-NEXT: addi a0, a0, -1
; CHECK-NEXT: and a0, a0, a1
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfcvt.rtz.x.f.v v16, v16, v0.t
; CHECK-NEXT: ret
%v = call <32 x i64> @llvm.vp.fptosi.v32i64.v32f64(<32 x double> %va, <32 x i1> %m, i32 %evl)
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fptoui-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fptoui-vp.ll
index d44efa2f6133f..b652cdd88c7c2 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fptoui-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fptoui-vp.ll
@@ -394,8 +394,8 @@ declare <32 x i64> @llvm.vp.fptoui.v32i64.v32f64(<32 x double>, <32 x i1>, i32)
define <32 x i64> @vfptoui_v32i64_v32f64(<32 x double> %va, <32 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfptoui_v32i64_v32f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; CHECK-NEXT: li a2, 16
+; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; CHECK-NEXT: vslidedown.vi v24, v0, 2
; CHECK-NEXT: mv a1, a0
; CHECK-NEXT: bltu a0, a2, .LBB25_2
@@ -408,8 +408,8 @@ define <32 x i64> @vfptoui_v32i64_v32f64(<32 x double> %va, <32 x i1> %m, i32 ze
; CHECK-NEXT: sltu a0, a0, a1
; CHECK-NEXT: addi a0, a0, -1
; CHECK-NEXT: and a0, a0, a1
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfcvt.rtz.xu.f.v v16, v16, v0.t
; CHECK-NEXT: ret
%v = call <32 x i64> @llvm.vp.fptoui.v32i64.v32f64(<32 x double> %va, <32 x i1> %m, i32 %evl)
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fptrunc-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fptrunc-vp.ll
index d890bf5412f9f..920eed322363a 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fptrunc-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fptrunc-vp.ll
@@ -98,8 +98,8 @@ define <32 x float> @vfptrunc_v32f32_v32f64(<32 x double> %a, <32 x i1> %m, i32
; CHECK-LABEL: vfptrunc_v32f32_v32f64:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv8r.v v24, v8
-; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; CHECK-NEXT: li a2, 16
+; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; CHECK-NEXT: vslidedown.vi v12, v0, 2
; CHECK-NEXT: mv a1, a0
; CHECK-NEXT: bltu a0, a2, .LBB7_2
@@ -112,8 +112,8 @@ define <32 x float> @vfptrunc_v32f32_v32f64(<32 x double> %a, <32 x i1> %m, i32
; CHECK-NEXT: sltu a0, a0, a1
; CHECK-NEXT: addi a0, a0, -1
; CHECK-NEXT: and a0, a0, a1
-; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, ma
; CHECK-NEXT: vfncvt.f.f.w v24, v16, v0.t
; CHECK-NEXT: li a0, 32
; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-i2fp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-i2fp.ll
index 6ffa6ac250ed7..9c76b83d0974a 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-i2fp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-i2fp.ll
@@ -132,12 +132,12 @@ define <3 x float> @si2fp_v3i1_v3f32(<3 x i1> %x) {
define <3 x float> @si2fp_v3i7_v3f32(<3 x i7> %x) {
; ZVFH32-LABEL: si2fp_v3i7_v3f32:
; ZVFH32: # %bb.0:
-; ZVFH32-NEXT: lw a1, 4(a0)
-; ZVFH32-NEXT: lw a2, 0(a0)
-; ZVFH32-NEXT: vsetivli zero, 4, e8, mf4, ta, ma
+; ZVFH32-NEXT: lw a1, 0(a0)
+; ZVFH32-NEXT: lw a2, 4(a0)
; ZVFH32-NEXT: lw a0, 8(a0)
-; ZVFH32-NEXT: vmv.v.x v8, a2
-; ZVFH32-NEXT: vslide1down.vx v8, v8, a1
+; ZVFH32-NEXT: vsetivli zero, 4, e8, mf4, ta, ma
+; ZVFH32-NEXT: vmv.v.x v8, a1
+; ZVFH32-NEXT: vslide1down.vx v8, v8, a2
; ZVFH32-NEXT: vslide1down.vx v8, v8, a0
; ZVFH32-NEXT: vslidedown.vi v8, v8, 1
; ZVFH32-NEXT: vadd.vv v8, v8, v8
@@ -149,12 +149,12 @@ define <3 x float> @si2fp_v3i7_v3f32(<3 x i7> %x) {
;
; ZVFH64-LABEL: si2fp_v3i7_v3f32:
; ZVFH64: # %bb.0:
-; ZVFH64-NEXT: ld a1, 8(a0)
-; ZVFH64-NEXT: ld a2, 0(a0)
-; ZVFH64-NEXT: vsetivli zero, 4, e8, mf4, ta, ma
+; ZVFH64-NEXT: ld a1, 0(a0)
+; ZVFH64-NEXT: ld a2, 8(a0)
; ZVFH64-NEXT: ld a0, 16(a0)
-; ZVFH64-NEXT: vmv.v.x v8, a2
-; ZVFH64-NEXT: vslide1down.vx v8, v8, a1
+; ZVFH64-NEXT: vsetivli zero, 4, e8, mf4, ta, ma
+; ZVFH64-NEXT: vmv.v.x v8, a1
+; ZVFH64-NEXT: vslide1down.vx v8, v8, a2
; ZVFH64-NEXT: vslide1down.vx v8, v8, a0
; ZVFH64-NEXT: vslidedown.vi v8, v8, 1
; ZVFH64-NEXT: vadd.vv v8, v8, v8
@@ -166,12 +166,12 @@ define <3 x float> @si2fp_v3i7_v3f32(<3 x i7> %x) {
;
; ZVFHMIN32-LABEL: si2fp_v3i7_v3f32:
; ZVFHMIN32: # %bb.0:
-; ZVFHMIN32-NEXT: lw a1, 4(a0)
-; ZVFHMIN32-NEXT: lw a2, 0(a0)
-; ZVFHMIN32-NEXT: vsetivli zero, 4, e8, mf4, ta, ma
+; ZVFHMIN32-NEXT: lw a1, 0(a0)
+; ZVFHMIN32-NEXT: lw a2, 4(a0)
; ZVFHMIN32-NEXT: lw a0, 8(a0)
-; ZVFHMIN32-NEXT: vmv.v.x v8, a2
-; ZVFHMIN32-NEXT: vslide1down.vx v8, v8, a1
+; ZVFHMIN32-NEXT: vsetivli zero, 4, e8, mf4, ta, ma
+; ZVFHMIN32-NEXT: vmv.v.x v8, a1
+; ZVFHMIN32-NEXT: vslide1down.vx v8, v8, a2
; ZVFHMIN32-NEXT: vslide1down.vx v8, v8, a0
; ZVFHMIN32-NEXT: vslidedown.vi v8, v8, 1
; ZVFHMIN32-NEXT: vadd.vv v8, v8, v8
@@ -183,12 +183,12 @@ define <3 x float> @si2fp_v3i7_v3f32(<3 x i7> %x) {
;
; ZVFHMIN64-LABEL: si2fp_v3i7_v3f32:
; ZVFHMIN64: # %bb.0:
-; ZVFHMIN64-NEXT: ld a1, 8(a0)
-; ZVFHMIN64-NEXT: ld a2, 0(a0)
-; ZVFHMIN64-NEXT: vsetivli zero, 4, e8, mf4, ta, ma
+; ZVFHMIN64-NEXT: ld a1, 0(a0)
+; ZVFHMIN64-NEXT: ld a2, 8(a0)
; ZVFHMIN64-NEXT: ld a0, 16(a0)
-; ZVFHMIN64-NEXT: vmv.v.x v8, a2
-; ZVFHMIN64-NEXT: vslide1down.vx v8, v8, a1
+; ZVFHMIN64-NEXT: vsetivli zero, 4, e8, mf4, ta, ma
+; ZVFHMIN64-NEXT: vmv.v.x v8, a1
+; ZVFHMIN64-NEXT: vslide1down.vx v8, v8, a2
; ZVFHMIN64-NEXT: vslide1down.vx v8, v8, a0
; ZVFHMIN64-NEXT: vslidedown.vi v8, v8, 1
; ZVFHMIN64-NEXT: vadd.vv v8, v8, v8
@@ -205,12 +205,12 @@ define <3 x float> @si2fp_v3i7_v3f32(<3 x i7> %x) {
define <3 x float> @ui2fp_v3i7_v3f32(<3 x i7> %x) {
; ZVFH32-LABEL: ui2fp_v3i7_v3f32:
; ZVFH32: # %bb.0:
-; ZVFH32-NEXT: lw a1, 4(a0)
-; ZVFH32-NEXT: lw a2, 0(a0)
-; ZVFH32-NEXT: vsetivli zero, 4, e8, mf4, ta, ma
+; ZVFH32-NEXT: lw a1, 0(a0)
+; ZVFH32-NEXT: lw a2, 4(a0)
; ZVFH32-NEXT: lw a0, 8(a0)
-; ZVFH32-NEXT: vmv.v.x v8, a2
-; ZVFH32-NEXT: vslide1down.vx v8, v8, a1
+; ZVFH32-NEXT: vsetivli zero, 4, e8, mf4, ta, ma
+; ZVFH32-NEXT: vmv.v.x v8, a1
+; ZVFH32-NEXT: vslide1down.vx v8, v8, a2
; ZVFH32-NEXT: vslide1down.vx v8, v8, a0
; ZVFH32-NEXT: vslidedown.vi v8, v8, 1
; ZVFH32-NEXT: li a0, 127
@@ -222,12 +222,12 @@ define <3 x float> @ui2fp_v3i7_v3f32(<3 x i7> %x) {
;
; ZVFH64-LABEL: ui2fp_v3i7_v3f32:
; ZVFH64: # %bb.0:
-; ZVFH64-NEXT: ld a1, 8(a0)
-; ZVFH64-NEXT: ld a2, 0(a0)
-; ZVFH64-NEXT: vsetivli zero, 4, e8, mf4, ta, ma
+; ZVFH64-NEXT: ld a1, 0(a0)
+; ZVFH64-NEXT: ld a2, 8(a0)
; ZVFH64-NEXT: ld a0, 16(a0)
-; ZVFH64-NEXT: vmv.v.x v8, a2
-; ZVFH64-NEXT: vslide1down.vx v8, v8, a1
+; ZVFH64-NEXT: vsetivli zero, 4, e8, mf4, ta, ma
+; ZVFH64-NEXT: vmv.v.x v8, a1
+; ZVFH64-NEXT: vslide1down.vx v8, v8, a2
; ZVFH64-NEXT: vslide1down.vx v8, v8, a0
; ZVFH64-NEXT: vslidedown.vi v8, v8, 1
; ZVFH64-NEXT: li a0, 127
@@ -239,12 +239,12 @@ define <3 x float> @ui2fp_v3i7_v3f32(<3 x i7> %x) {
;
; ZVFHMIN32-LABEL: ui2fp_v3i7_v3f32:
; ZVFHMIN32: # %bb.0:
-; ZVFHMIN32-NEXT: lw a1, 4(a0)
-; ZVFHMIN32-NEXT: lw a2, 0(a0)
-; ZVFHMIN32-NEXT: vsetivli zero, 4, e8, mf4, ta, ma
+; ZVFHMIN32-NEXT: lw a1, 0(a0)
+; ZVFHMIN32-NEXT: lw a2, 4(a0)
; ZVFHMIN32-NEXT: lw a0, 8(a0)
-; ZVFHMIN32-NEXT: vmv.v.x v8, a2
-; ZVFHMIN32-NEXT: vslide1down.vx v8, v8, a1
+; ZVFHMIN32-NEXT: vsetivli zero, 4, e8, mf4, ta, ma
+; ZVFHMIN32-NEXT: vmv.v.x v8, a1
+; ZVFHMIN32-NEXT: vslide1down.vx v8, v8, a2
; ZVFHMIN32-NEXT: vslide1down.vx v8, v8, a0
; ZVFHMIN32-NEXT: vslidedown.vi v8, v8, 1
; ZVFHMIN32-NEXT: li a0, 127
@@ -256,12 +256,12 @@ define <3 x float> @ui2fp_v3i7_v3f32(<3 x i7> %x) {
;
; ZVFHMIN64-LABEL: ui2fp_v3i7_v3f32:
; ZVFHMIN64: # %bb.0:
-; ZVFHMIN64-NEXT: ld a1, 8(a0)
-; ZVFHMIN64-NEXT: ld a2, 0(a0)
-; ZVFHMIN64-NEXT: vsetivli zero, 4, e8, mf4, ta, ma
+; ZVFHMIN64-NEXT: ld a1, 0(a0)
+; ZVFHMIN64-NEXT: ld a2, 8(a0)
; ZVFHMIN64-NEXT: ld a0, 16(a0)
-; ZVFHMIN64-NEXT: vmv.v.x v8, a2
-; ZVFHMIN64-NEXT: vslide1down.vx v8, v8, a1
+; ZVFHMIN64-NEXT: vsetivli zero, 4, e8, mf4, ta, ma
+; ZVFHMIN64-NEXT: vmv.v.x v8, a1
+; ZVFHMIN64-NEXT: vslide1down.vx v8, v8, a2
; ZVFHMIN64-NEXT: vslide1down.vx v8, v8, a0
; ZVFHMIN64-NEXT: vslidedown.vi v8, v8, 1
; ZVFHMIN64-NEXT: li a0, 127
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-insert-subvector.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-insert-subvector.ll
index 53de1a8755355..e81f686a28303 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-insert-subvector.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-insert-subvector.ll
@@ -277,14 +277,14 @@ define void @insert_v8i32_v2i32_0(ptr %vp, ptr %svp) {
define void @insert_v8i32_v2i32_2(ptr %vp, ptr %svp) {
; VLA-LABEL: insert_v8i32_v2i32_2:
; VLA: # %bb.0:
-; VLA-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
-; VLA-NEXT: vle32.v v8, (a1)
; VLA-NEXT: vsetivli zero, 8, e32, m2, ta, ma
-; VLA-NEXT: vle32.v v10, (a0)
+; VLA-NEXT: vle32.v v8, (a0)
+; VLA-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
+; VLA-NEXT: vle32.v v10, (a1)
; VLA-NEXT: vsetivli zero, 4, e32, m2, tu, ma
-; VLA-NEXT: vslideup.vi v10, v8, 2
+; VLA-NEXT: vslideup.vi v8, v10, 2
; VLA-NEXT: vsetivli zero, 8, e32, m2, ta, ma
-; VLA-NEXT: vse32.v v10, (a0)
+; VLA-NEXT: vse32.v v8, (a0)
; VLA-NEXT: ret
;
; VLS-LABEL: insert_v8i32_v2i32_2:
@@ -306,12 +306,13 @@ define void @insert_v8i32_v2i32_2(ptr %vp, ptr %svp) {
define void @insert_v8i32_v2i32_6(ptr %vp, ptr %svp) {
; VLA-LABEL: insert_v8i32_v2i32_6:
; VLA: # %bb.0:
+; VLA-NEXT: vsetivli zero, 8, e32, m2, ta, ma
+; VLA-NEXT: vle32.v v8, (a0)
; VLA-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
-; VLA-NEXT: vle32.v v8, (a1)
+; VLA-NEXT: vle32.v v10, (a1)
; VLA-NEXT: vsetivli zero, 8, e32, m2, ta, ma
-; VLA-NEXT: vle32.v v10, (a0)
-; VLA-NEXT: vslideup.vi v10, v8, 6
-; VLA-NEXT: vse32.v v10, (a0)
+; VLA-NEXT: vslideup.vi v8, v10, 6
+; VLA-NEXT: vse32.v v8, (a0)
; VLA-NEXT: ret
;
; VLS-LABEL: insert_v8i32_v2i32_6:
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-insert.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-insert.ll
index 4954827876c19..776a1e9bab6b2 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-insert.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-insert.ll
@@ -533,11 +533,11 @@ define void @insertelt_c6_v8i64_0_add(ptr %x, ptr %y) {
; CHECK: # %bb.0:
; CHECK-NEXT: vsetivli zero, 8, e64, m4, ta, ma
; CHECK-NEXT: vle64.v v8, (a0)
-; CHECK-NEXT: li a2, 6
+; CHECK-NEXT: vle64.v v12, (a1)
+; CHECK-NEXT: li a1, 6
; CHECK-NEXT: vsetvli zero, zero, e64, m4, tu, ma
-; CHECK-NEXT: vmv.s.x v8, a2
+; CHECK-NEXT: vmv.s.x v8, a1
; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, ma
-; CHECK-NEXT: vle64.v v12, (a1)
; CHECK-NEXT: vadd.vv v8, v8, v12
; CHECK-NEXT: vse64.v v8, (a0)
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-int-buildvec.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-int-buildvec.ll
index 592ce6fc5be0b..4a5d37b2a85a2 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-int-buildvec.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-int-buildvec.ll
@@ -669,13 +669,14 @@ define void @buildvec_seq_v9i8(ptr %x) {
; CHECK-NEXT: vsetivli zero, 1, e16, m1, ta, ma
; CHECK-NEXT: vmv.s.x v0, a1
; CHECK-NEXT: vsetivli zero, 16, e8, m1, ta, ma
-; CHECK-NEXT: vmv.v.i v8, 3
-; CHECK-NEXT: vmerge.vim v8, v8, 1, v0
+; CHECK-NEXT: vmv.v.i v9, 3
; CHECK-NEXT: li a1, 146
; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, ma
-; CHECK-NEXT: vmv.s.x v0, a1
+; CHECK-NEXT: vmv.s.x v8, a1
; CHECK-NEXT: vsetvli zero, zero, e8, m1, ta, ma
-; CHECK-NEXT: vmerge.vim v8, v8, 2, v0
+; CHECK-NEXT: vmerge.vim v9, v9, 1, v0
+; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vmerge.vim v8, v9, 2, v0
; CHECK-NEXT: vsetivli zero, 9, e8, m1, ta, ma
; CHECK-NEXT: vse8.v v8, (a0)
; CHECK-NEXT: ret
@@ -1183,42 +1184,42 @@ define <8 x i64> @v8xi64_exact_undef_prefix(i64 %a, i64 %b, i64 %c, i64 %d) vsca
define <16 x i8> @buildvec_v16i8_loads_contigous(ptr %p) {
; CHECK-LABEL: buildvec_v16i8_loads_contigous:
; CHECK: # %bb.0:
-; CHECK-NEXT: lbu a1, 1(a0)
-; CHECK-NEXT: lbu a2, 2(a0)
-; CHECK-NEXT: lbu a3, 3(a0)
-; CHECK-NEXT: lbu a4, 4(a0)
-; CHECK-NEXT: lbu a5, 5(a0)
-; CHECK-NEXT: lbu a6, 6(a0)
-; CHECK-NEXT: lbu a7, 7(a0)
-; CHECK-NEXT: lbu t0, 9(a0)
-; CHECK-NEXT: lbu t1, 10(a0)
-; CHECK-NEXT: lbu t2, 11(a0)
-; CHECK-NEXT: lbu t3, 12(a0)
-; CHECK-NEXT: lbu t4, 13(a0)
-; CHECK-NEXT: lbu t5, 14(a0)
-; CHECK-NEXT: lbu t6, 15(a0)
+; CHECK-NEXT: addi a1, a0, 8
+; CHECK-NEXT: lbu a2, 1(a0)
+; CHECK-NEXT: lbu a3, 2(a0)
+; CHECK-NEXT: lbu a4, 3(a0)
+; CHECK-NEXT: lbu a5, 4(a0)
+; CHECK-NEXT: lbu a6, 5(a0)
+; CHECK-NEXT: lbu a7, 6(a0)
+; CHECK-NEXT: lbu t0, 7(a0)
+; CHECK-NEXT: lbu t1, 9(a0)
+; CHECK-NEXT: lbu t2, 10(a0)
+; CHECK-NEXT: lbu t3, 11(a0)
+; CHECK-NEXT: lbu t4, 12(a0)
; CHECK-NEXT: vsetivli zero, 16, e8, m1, ta, ma
; CHECK-NEXT: vlse8.v v8, (a0), zero
-; CHECK-NEXT: addi a0, a0, 8
-; CHECK-NEXT: vslide1down.vx v8, v8, a1
+; CHECK-NEXT: lbu t5, 13(a0)
+; CHECK-NEXT: lbu t6, 14(a0)
+; CHECK-NEXT: lbu a0, 15(a0)
; CHECK-NEXT: vslide1down.vx v8, v8, a2
; CHECK-NEXT: vslide1down.vx v8, v8, a3
; CHECK-NEXT: vslide1down.vx v8, v8, a4
-; CHECK-NEXT: vlse8.v v9, (a0), zero
; CHECK-NEXT: vslide1down.vx v8, v8, a5
+; CHECK-NEXT: vlse8.v v9, (a1), zero
; CHECK-NEXT: vslide1down.vx v8, v8, a6
-; CHECK-NEXT: vslide1down.vx v10, v8, a7
-; CHECK-NEXT: vslide1down.vx v8, v9, t0
-; CHECK-NEXT: vslide1down.vx v8, v8, t1
+; CHECK-NEXT: vslide1down.vx v8, v8, a7
+; CHECK-NEXT: vslide1down.vx v10, v8, t0
+; CHECK-NEXT: vslide1down.vx v8, v9, t1
; CHECK-NEXT: vslide1down.vx v8, v8, t2
; CHECK-NEXT: vslide1down.vx v8, v8, t3
; CHECK-NEXT: vslide1down.vx v8, v8, t4
; CHECK-NEXT: vslide1down.vx v8, v8, t5
; CHECK-NEXT: vslide1down.vx v8, v8, t6
-; CHECK-NEXT: li a0, 255
+; CHECK-NEXT: li a1, 255
; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, ma
-; CHECK-NEXT: vmv.s.x v0, a0
+; CHECK-NEXT: vmv.s.x v0, a1
; CHECK-NEXT: vsetvli zero, zero, e8, m1, ta, mu
+; CHECK-NEXT: vslide1down.vx v8, v8, a0
; CHECK-NEXT: vslidedown.vi v8, v10, 8, v0.t
; CHECK-NEXT: ret
%p2 = getelementptr i8, ptr %p, i32 1
@@ -1277,42 +1278,42 @@ define <16 x i8> @buildvec_v16i8_loads_contigous(ptr %p) {
define <16 x i8> @buildvec_v16i8_loads_gather(ptr %p) {
; CHECK-LABEL: buildvec_v16i8_loads_gather:
; CHECK: # %bb.0:
-; CHECK-NEXT: lbu a1, 1(a0)
-; CHECK-NEXT: lbu a2, 22(a0)
-; CHECK-NEXT: lbu a3, 31(a0)
-; CHECK-NEXT: lbu a4, 44(a0)
-; CHECK-NEXT: lbu a5, 55(a0)
-; CHECK-NEXT: lbu a6, 623(a0)
-; CHECK-NEXT: lbu a7, 75(a0)
-; CHECK-NEXT: lbu t0, 93(a0)
-; CHECK-NEXT: lbu t1, 105(a0)
-; CHECK-NEXT: lbu t2, 161(a0)
-; CHECK-NEXT: lbu t3, 124(a0)
-; CHECK-NEXT: lbu t4, 163(a0)
-; CHECK-NEXT: lbu t5, 144(a0)
-; CHECK-NEXT: lbu t6, 154(a0)
+; CHECK-NEXT: addi a1, a0, 82
+; CHECK-NEXT: lbu a2, 1(a0)
+; CHECK-NEXT: lbu a3, 22(a0)
+; CHECK-NEXT: lbu a4, 31(a0)
+; CHECK-NEXT: lbu a5, 44(a0)
+; CHECK-NEXT: lbu a6, 55(a0)
+; CHECK-NEXT: lbu a7, 623(a0)
+; CHECK-NEXT: lbu t0, 75(a0)
+; CHECK-NEXT: lbu t1, 93(a0)
+; CHECK-NEXT: lbu t2, 105(a0)
+; CHECK-NEXT: lbu t3, 161(a0)
+; CHECK-NEXT: lbu t4, 124(a0)
; CHECK-NEXT: vsetivli zero, 16, e8, m1, ta, ma
; CHECK-NEXT: vlse8.v v8, (a0), zero
-; CHECK-NEXT: addi a0, a0, 82
-; CHECK-NEXT: vslide1down.vx v8, v8, a1
+; CHECK-NEXT: lbu t5, 163(a0)
+; CHECK-NEXT: lbu t6, 144(a0)
+; CHECK-NEXT: lbu a0, 154(a0)
; CHECK-NEXT: vslide1down.vx v8, v8, a2
; CHECK-NEXT: vslide1down.vx v8, v8, a3
; CHECK-NEXT: vslide1down.vx v8, v8, a4
-; CHECK-NEXT: vlse8.v v9, (a0), zero
; CHECK-NEXT: vslide1down.vx v8, v8, a5
+; CHECK-NEXT: vlse8.v v9, (a1), zero
; CHECK-NEXT: vslide1down.vx v8, v8, a6
-; CHECK-NEXT: vslide1down.vx v10, v8, a7
-; CHECK-NEXT: vslide1down.vx v8, v9, t0
-; CHECK-NEXT: vslide1down.vx v8, v8, t1
+; CHECK-NEXT: vslide1down.vx v8, v8, a7
+; CHECK-NEXT: vslide1down.vx v10, v8, t0
+; CHECK-NEXT: vslide1down.vx v8, v9, t1
; CHECK-NEXT: vslide1down.vx v8, v8, t2
; CHECK-NEXT: vslide1down.vx v8, v8, t3
; CHECK-NEXT: vslide1down.vx v8, v8, t4
; CHECK-NEXT: vslide1down.vx v8, v8, t5
; CHECK-NEXT: vslide1down.vx v8, v8, t6
-; CHECK-NEXT: li a0, 255
+; CHECK-NEXT: li a1, 255
; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, ma
-; CHECK-NEXT: vmv.s.x v0, a0
+; CHECK-NEXT: vmv.s.x v0, a1
; CHECK-NEXT: vsetvli zero, zero, e8, m1, ta, mu
+; CHECK-NEXT: vslide1down.vx v8, v8, a0
; CHECK-NEXT: vslidedown.vi v8, v10, 8, v0.t
; CHECK-NEXT: ret
%p2 = getelementptr i8, ptr %p, i32 1
@@ -1375,17 +1376,17 @@ define <16 x i8> @buildvec_v16i8_undef_low_half(ptr %p) {
; CHECK-NEXT: lbu a3, 105(a0)
; CHECK-NEXT: lbu a4, 161(a0)
; CHECK-NEXT: lbu a5, 124(a0)
-; CHECK-NEXT: lbu a6, 163(a0)
-; CHECK-NEXT: lbu a7, 144(a0)
-; CHECK-NEXT: lbu a0, 154(a0)
; CHECK-NEXT: vsetivli zero, 16, e8, m1, ta, ma
; CHECK-NEXT: vlse8.v v8, (a1), zero
+; CHECK-NEXT: lbu a1, 163(a0)
+; CHECK-NEXT: lbu a6, 144(a0)
+; CHECK-NEXT: lbu a0, 154(a0)
; CHECK-NEXT: vslide1down.vx v8, v8, a2
; CHECK-NEXT: vslide1down.vx v8, v8, a3
; CHECK-NEXT: vslide1down.vx v8, v8, a4
; CHECK-NEXT: vslide1down.vx v8, v8, a5
+; CHECK-NEXT: vslide1down.vx v8, v8, a1
; CHECK-NEXT: vslide1down.vx v8, v8, a6
-; CHECK-NEXT: vslide1down.vx v8, v8, a7
; CHECK-NEXT: vslide1down.vx v8, v8, a0
; CHECK-NEXT: ret
%p9 = getelementptr i8, ptr %p, i32 82
@@ -1424,18 +1425,18 @@ define <16 x i8> @buildvec_v16i8_undef_high_half(ptr %p) {
; CHECK-NEXT: lbu a2, 22(a0)
; CHECK-NEXT: lbu a3, 31(a0)
; CHECK-NEXT: lbu a4, 44(a0)
-; CHECK-NEXT: lbu a5, 55(a0)
-; CHECK-NEXT: lbu a6, 623(a0)
-; CHECK-NEXT: lbu a7, 75(a0)
; CHECK-NEXT: vsetivli zero, 16, e8, m1, ta, ma
; CHECK-NEXT: vlse8.v v8, (a0), zero
+; CHECK-NEXT: lbu a5, 55(a0)
+; CHECK-NEXT: lbu a6, 623(a0)
+; CHECK-NEXT: lbu a0, 75(a0)
; CHECK-NEXT: vslide1down.vx v8, v8, a1
; CHECK-NEXT: vslide1down.vx v8, v8, a2
; CHECK-NEXT: vslide1down.vx v8, v8, a3
; CHECK-NEXT: vslide1down.vx v8, v8, a4
; CHECK-NEXT: vslide1down.vx v8, v8, a5
; CHECK-NEXT: vslide1down.vx v8, v8, a6
-; CHECK-NEXT: vslide1down.vx v8, v8, a7
+; CHECK-NEXT: vslide1down.vx v8, v8, a0
; CHECK-NEXT: vslidedown.vi v8, v8, 8
; CHECK-NEXT: ret
%p2 = getelementptr i8, ptr %p, i32 1
@@ -1470,29 +1471,29 @@ define <16 x i8> @buildvec_v16i8_undef_edges(ptr %p) {
; CHECK-LABEL: buildvec_v16i8_undef_edges:
; CHECK: # %bb.0:
; CHECK-NEXT: addi a1, a0, 31
-; CHECK-NEXT: lbu a2, 44(a0)
-; CHECK-NEXT: lbu a3, 55(a0)
-; CHECK-NEXT: lbu a4, 623(a0)
-; CHECK-NEXT: lbu a5, 75(a0)
-; CHECK-NEXT: lbu a6, 93(a0)
-; CHECK-NEXT: lbu a7, 105(a0)
-; CHECK-NEXT: lbu t0, 161(a0)
+; CHECK-NEXT: addi a2, a0, 82
+; CHECK-NEXT: lbu a3, 44(a0)
+; CHECK-NEXT: lbu a4, 55(a0)
+; CHECK-NEXT: lbu a5, 623(a0)
+; CHECK-NEXT: lbu a6, 75(a0)
; CHECK-NEXT: vsetivli zero, 16, e8, m1, ta, ma
; CHECK-NEXT: vlse8.v v8, (a1), zero
-; CHECK-NEXT: addi a0, a0, 82
-; CHECK-NEXT: vslide1down.vx v8, v8, a2
-; CHECK-NEXT: vlse8.v v9, (a0), zero
+; CHECK-NEXT: lbu a1, 93(a0)
+; CHECK-NEXT: lbu a7, 105(a0)
+; CHECK-NEXT: lbu a0, 161(a0)
; CHECK-NEXT: vslide1down.vx v8, v8, a3
+; CHECK-NEXT: vlse8.v v9, (a2), zero
; CHECK-NEXT: vslide1down.vx v8, v8, a4
-; CHECK-NEXT: vslide1down.vx v10, v8, a5
-; CHECK-NEXT: vslide1down.vx v8, v9, a6
+; CHECK-NEXT: vslide1down.vx v8, v8, a5
+; CHECK-NEXT: vslide1down.vx v10, v8, a6
+; CHECK-NEXT: vslide1down.vx v8, v9, a1
; CHECK-NEXT: vslide1down.vx v8, v8, a7
-; CHECK-NEXT: vslide1down.vx v8, v8, t0
-; CHECK-NEXT: vslidedown.vi v8, v8, 4
+; CHECK-NEXT: vslide1down.vx v8, v8, a0
; CHECK-NEXT: li a0, 255
; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; CHECK-NEXT: vmv.s.x v0, a0
; CHECK-NEXT: vsetvli zero, zero, e8, m1, ta, mu
+; CHECK-NEXT: vslidedown.vi v8, v8, 4
; CHECK-NEXT: vslidedown.vi v8, v10, 8, v0.t
; CHECK-NEXT: ret
%p4 = getelementptr i8, ptr %p, i32 31
@@ -1530,34 +1531,34 @@ define <16 x i8> @buildvec_v16i8_undef_edges(ptr %p) {
define <16 x i8> @buildvec_v16i8_loads_undef_scattered(ptr %p) {
; CHECK-LABEL: buildvec_v16i8_loads_undef_scattered:
; CHECK: # %bb.0:
-; CHECK-NEXT: lbu a1, 1(a0)
-; CHECK-NEXT: lbu a2, 44(a0)
-; CHECK-NEXT: lbu a3, 55(a0)
-; CHECK-NEXT: lbu a4, 75(a0)
-; CHECK-NEXT: lbu a5, 93(a0)
-; CHECK-NEXT: lbu a6, 124(a0)
-; CHECK-NEXT: lbu a7, 144(a0)
-; CHECK-NEXT: lbu t0, 154(a0)
+; CHECK-NEXT: addi a1, a0, 82
+; CHECK-NEXT: lbu a2, 1(a0)
+; CHECK-NEXT: lbu a3, 44(a0)
+; CHECK-NEXT: lbu a4, 55(a0)
+; CHECK-NEXT: lbu a5, 75(a0)
+; CHECK-NEXT: lbu a6, 93(a0)
; CHECK-NEXT: vsetivli zero, 16, e8, m1, ta, ma
; CHECK-NEXT: vlse8.v v8, (a0), zero
-; CHECK-NEXT: addi a0, a0, 82
-; CHECK-NEXT: vslide1down.vx v8, v8, a1
-; CHECK-NEXT: vslidedown.vi v8, v8, 2
+; CHECK-NEXT: lbu a7, 124(a0)
+; CHECK-NEXT: lbu t0, 144(a0)
+; CHECK-NEXT: lbu a0, 154(a0)
; CHECK-NEXT: vslide1down.vx v8, v8, a2
-; CHECK-NEXT: vlse8.v v9, (a0), zero
+; CHECK-NEXT: vslidedown.vi v8, v8, 2
; CHECK-NEXT: vslide1down.vx v8, v8, a3
+; CHECK-NEXT: vlse8.v v9, (a1), zero
+; CHECK-NEXT: vslide1down.vx v8, v8, a4
; CHECK-NEXT: vslidedown.vi v8, v8, 1
-; CHECK-NEXT: vslide1down.vx v10, v8, a4
-; CHECK-NEXT: vslide1down.vx v8, v9, a5
+; CHECK-NEXT: vslide1down.vx v10, v8, a5
+; CHECK-NEXT: vslide1down.vx v8, v9, a6
; CHECK-NEXT: vslidedown.vi v8, v8, 2
-; CHECK-NEXT: vslide1down.vx v8, v8, a6
-; CHECK-NEXT: vslidedown.vi v8, v8, 1
; CHECK-NEXT: vslide1down.vx v8, v8, a7
+; CHECK-NEXT: vslidedown.vi v8, v8, 1
; CHECK-NEXT: vslide1down.vx v8, v8, t0
-; CHECK-NEXT: li a0, 255
+; CHECK-NEXT: li a1, 255
; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, ma
-; CHECK-NEXT: vmv.s.x v0, a0
+; CHECK-NEXT: vmv.s.x v0, a1
; CHECK-NEXT: vsetvli zero, zero, e8, m1, ta, mu
+; CHECK-NEXT: vslide1down.vx v8, v8, a0
; CHECK-NEXT: vslidedown.vi v8, v10, 8, v0.t
; CHECK-NEXT: ret
%p2 = getelementptr i8, ptr %p, i32 1
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-int-explodevector.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-int-explodevector.ll
index 4509642fdef17..e0c676788dccc 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-int-explodevector.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-int-explodevector.ll
@@ -828,112 +828,104 @@ define i64 @explode_8xi64(<8 x i64> %v) {
define i64 @explode_16xi64(<16 x i64> %v) {
; RV32-LABEL: explode_16xi64:
; RV32: # %bb.0:
-; RV32-NEXT: addi sp, sp, -64
-; RV32-NEXT: .cfi_def_cfa_offset 64
-; RV32-NEXT: sw ra, 60(sp) # 4-byte Folded Spill
-; RV32-NEXT: sw s0, 56(sp) # 4-byte Folded Spill
-; RV32-NEXT: sw s1, 52(sp) # 4-byte Folded Spill
-; RV32-NEXT: sw s2, 48(sp) # 4-byte Folded Spill
-; RV32-NEXT: sw s3, 44(sp) # 4-byte Folded Spill
-; RV32-NEXT: sw s4, 40(sp) # 4-byte Folded Spill
-; RV32-NEXT: sw s5, 36(sp) # 4-byte Folded Spill
-; RV32-NEXT: sw s6, 32(sp) # 4-byte Folded Spill
-; RV32-NEXT: sw s7, 28(sp) # 4-byte Folded Spill
-; RV32-NEXT: sw s8, 24(sp) # 4-byte Folded Spill
-; RV32-NEXT: sw s9, 20(sp) # 4-byte Folded Spill
-; RV32-NEXT: sw s10, 16(sp) # 4-byte Folded Spill
-; RV32-NEXT: sw s11, 12(sp) # 4-byte Folded Spill
-; RV32-NEXT: .cfi_offset ra, -4
-; RV32-NEXT: .cfi_offset s0, -8
-; RV32-NEXT: .cfi_offset s1, -12
-; RV32-NEXT: .cfi_offset s2, -16
-; RV32-NEXT: .cfi_offset s3, -20
-; RV32-NEXT: .cfi_offset s4, -24
-; RV32-NEXT: .cfi_offset s5, -28
-; RV32-NEXT: .cfi_offset s6, -32
-; RV32-NEXT: .cfi_offset s7, -36
-; RV32-NEXT: .cfi_offset s8, -40
-; RV32-NEXT: .cfi_offset s9, -44
-; RV32-NEXT: .cfi_offset s10, -48
-; RV32-NEXT: .cfi_offset s11, -52
+; RV32-NEXT: addi sp, sp, -48
+; RV32-NEXT: .cfi_def_cfa_offset 48
+; RV32-NEXT: sw s0, 44(sp) # 4-byte Folded Spill
+; RV32-NEXT: sw s1, 40(sp) # 4-byte Folded Spill
+; RV32-NEXT: sw s2, 36(sp) # 4-byte Folded Spill
+; RV32-NEXT: sw s3, 32(sp) # 4-byte Folded Spill
+; RV32-NEXT: sw s4, 28(sp) # 4-byte Folded Spill
+; RV32-NEXT: sw s5, 24(sp) # 4-byte Folded Spill
+; RV32-NEXT: sw s6, 20(sp) # 4-byte Folded Spill
+; RV32-NEXT: sw s7, 16(sp) # 4-byte Folded Spill
+; RV32-NEXT: sw s8, 12(sp) # 4-byte Folded Spill
+; RV32-NEXT: sw s9, 8(sp) # 4-byte Folded Spill
+; RV32-NEXT: sw s10, 4(sp) # 4-byte Folded Spill
+; RV32-NEXT: sw s11, 0(sp) # 4-byte Folded Spill
+; RV32-NEXT: .cfi_offset s0, -4
+; RV32-NEXT: .cfi_offset s1, -8
+; RV32-NEXT: .cfi_offset s2, -12
+; RV32-NEXT: .cfi_offset s3, -16
+; RV32-NEXT: .cfi_offset s4, -20
+; RV32-NEXT: .cfi_offset s5, -24
+; RV32-NEXT: .cfi_offset s6, -28
+; RV32-NEXT: .cfi_offset s7, -32
+; RV32-NEXT: .cfi_offset s8, -36
+; RV32-NEXT: .cfi_offset s9, -40
+; RV32-NEXT: .cfi_offset s10, -44
+; RV32-NEXT: .cfi_offset s11, -48
; RV32-NEXT: vsetivli zero, 1, e64, m8, ta, ma
; RV32-NEXT: vslidedown.vi v16, v8, 2
-; RV32-NEXT: li a3, 32
-; RV32-NEXT: vsrl.vx v24, v16, a3
-; RV32-NEXT: vmv.x.s a0, v24
-; RV32-NEXT: vmv.x.s a1, v16
-; RV32-NEXT: sw a1, 8(sp) # 4-byte Folded Spill
-; RV32-NEXT: vslidedown.vi v16, v8, 3
-; RV32-NEXT: vsrl.vx v24, v16, a3
+; RV32-NEXT: li a0, 32
+; RV32-NEXT: vsrl.vx v24, v16, a0
; RV32-NEXT: vmv.x.s a1, v24
-; RV32-NEXT: sw a1, 4(sp) # 4-byte Folded Spill
+; RV32-NEXT: vmv.x.s a2, v16
+; RV32-NEXT: vslidedown.vi v16, v8, 3
+; RV32-NEXT: vsrl.vx v24, v16, a0
+; RV32-NEXT: vmv.x.s a3, v24
; RV32-NEXT: vmv.x.s a4, v16
; RV32-NEXT: vslidedown.vi v16, v8, 4
-; RV32-NEXT: vsrl.vx v24, v16, a3
+; RV32-NEXT: vsrl.vx v24, v16, a0
; RV32-NEXT: vmv.x.s a5, v24
; RV32-NEXT: vmv.x.s a6, v16
; RV32-NEXT: vslidedown.vi v16, v8, 5
-; RV32-NEXT: vsrl.vx v24, v16, a3
+; RV32-NEXT: vsrl.vx v24, v16, a0
; RV32-NEXT: vmv.x.s a7, v24
; RV32-NEXT: vmv.x.s t0, v16
; RV32-NEXT: vslidedown.vi v16, v8, 6
-; RV32-NEXT: vsrl.vx v24, v16, a3
+; RV32-NEXT: vsrl.vx v24, v16, a0
; RV32-NEXT: vmv.x.s t1, v24
; RV32-NEXT: vmv.x.s t2, v16
; RV32-NEXT: vslidedown.vi v16, v8, 7
-; RV32-NEXT: vsrl.vx v24, v16, a3
+; RV32-NEXT: vsrl.vx v24, v16, a0
; RV32-NEXT: vmv.x.s t3, v24
; RV32-NEXT: vmv.x.s t4, v16
; RV32-NEXT: vslidedown.vi v16, v8, 8
-; RV32-NEXT: vsrl.vx v24, v16, a3
+; RV32-NEXT: vsrl.vx v24, v16, a0
; RV32-NEXT: vmv.x.s t5, v24
; RV32-NEXT: vmv.x.s t6, v16
; RV32-NEXT: vslidedown.vi v16, v8, 9
-; RV32-NEXT: vsrl.vx v24, v16, a3
+; RV32-NEXT: vsrl.vx v24, v16, a0
; RV32-NEXT: vmv.x.s s0, v24
; RV32-NEXT: vmv.x.s s1, v16
; RV32-NEXT: vslidedown.vi v16, v8, 10
-; RV32-NEXT: vsrl.vx v24, v16, a3
+; RV32-NEXT: vsrl.vx v24, v16, a0
; RV32-NEXT: vmv.x.s s2, v24
; RV32-NEXT: vmv.x.s s3, v16
; RV32-NEXT: vslidedown.vi v16, v8, 11
-; RV32-NEXT: vsrl.vx v24, v16, a3
+; RV32-NEXT: vsrl.vx v24, v16, a0
; RV32-NEXT: vmv.x.s s4, v24
; RV32-NEXT: vmv.x.s s5, v16
; RV32-NEXT: vslidedown.vi v16, v8, 12
-; RV32-NEXT: vsrl.vx v24, v16, a3
+; RV32-NEXT: vsrl.vx v24, v16, a0
; RV32-NEXT: vmv.x.s s6, v24
; RV32-NEXT: vmv.x.s s7, v16
; RV32-NEXT: vslidedown.vi v16, v8, 13
-; RV32-NEXT: vsrl.vx v24, v16, a3
-; RV32-NEXT: vmv.x.s s8, v24
-; RV32-NEXT: vmv.x.s s9, v16
+; RV32-NEXT: vsrl.vx v24, v16, a0
+; RV32-NEXT: vmv.x.s s9, v24
+; RV32-NEXT: vmv.x.s s8, v16
; RV32-NEXT: vslidedown.vi v16, v8, 14
-; RV32-NEXT: vsrl.vx v24, v16, a3
-; RV32-NEXT: vmv.x.s s10, v24
-; RV32-NEXT: vmv.x.s s11, v16
-; RV32-NEXT: vslidedown.vi v16, v8, 15
-; RV32-NEXT: vsrl.vx v24, v16, a3
-; RV32-NEXT: vmv.x.s ra, v24
-; RV32-NEXT: vmv.s.x v9, zero
-; RV32-NEXT: vmv.x.s a2, v16
+; RV32-NEXT: vsrl.vx v24, v16, a0
+; RV32-NEXT: vmv.s.x v17, zero
; RV32-NEXT: vsetivli zero, 2, e64, m1, ta, ma
-; RV32-NEXT: vredxor.vs v8, v8, v9
+; RV32-NEXT: vredxor.vs v17, v8, v17
+; RV32-NEXT: vsetivli zero, 1, e64, m8, ta, ma
+; RV32-NEXT: vslidedown.vi v8, v8, 15
; RV32-NEXT: vsetivli zero, 1, e64, m1, ta, ma
-; RV32-NEXT: vsrl.vx v9, v8, a3
-; RV32-NEXT: vmv.x.s a3, v9
-; RV32-NEXT: add a3, a3, a0
-; RV32-NEXT: vmv.x.s a1, v8
-; RV32-NEXT: lw a0, 8(sp) # 4-byte Folded Reload
-; RV32-NEXT: add a0, a1, a0
-; RV32-NEXT: sltu a1, a0, a1
-; RV32-NEXT: add a1, a3, a1
-; RV32-NEXT: lw a3, 4(sp) # 4-byte Folded Reload
-; RV32-NEXT: add a1, a1, a3
-; RV32-NEXT: add a4, a0, a4
-; RV32-NEXT: sltu a0, a4, a0
-; RV32-NEXT: add a0, a0, a5
+; RV32-NEXT: vsrl.vx v18, v17, a0
+; RV32-NEXT: vmv.x.s s10, v18
+; RV32-NEXT: vmv.x.s s11, v17
+; RV32-NEXT: vsetivli zero, 1, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vx v0, v8, a0
+; RV32-NEXT: add a1, s10, a1
+; RV32-NEXT: add a2, s11, a2
+; RV32-NEXT: sltu a0, a2, s11
; RV32-NEXT: add a0, a1, a0
+; RV32-NEXT: add a0, a0, a3
+; RV32-NEXT: add a4, a2, a4
+; RV32-NEXT: sltu a1, a4, a2
+; RV32-NEXT: add a1, a1, a5
+; RV32-NEXT: add a0, a0, a1
; RV32-NEXT: add a6, a4, a6
; RV32-NEXT: sltu a1, a6, a4
; RV32-NEXT: add a1, a1, a7
@@ -968,33 +960,36 @@ define i64 @explode_16xi64(<16 x i64> %v) {
; RV32-NEXT: add a0, a0, a1
; RV32-NEXT: add s7, s5, s7
; RV32-NEXT: sltu a1, s7, s5
-; RV32-NEXT: add a1, a1, s8
+; RV32-NEXT: add a1, a1, s9
; RV32-NEXT: add a0, a0, a1
-; RV32-NEXT: add s9, s7, s9
-; RV32-NEXT: sltu a1, s9, s7
-; RV32-NEXT: add a1, a1, s10
+; RV32-NEXT: vmv.x.s a1, v24
+; RV32-NEXT: add s8, s7, s8
+; RV32-NEXT: sltu a2, s8, s7
+; RV32-NEXT: add a1, a2, a1
+; RV32-NEXT: vmv.x.s a2, v16
; RV32-NEXT: add a0, a0, a1
-; RV32-NEXT: add s11, s9, s11
-; RV32-NEXT: sltu a1, s11, s9
-; RV32-NEXT: add a1, a1, ra
+; RV32-NEXT: vmv.x.s a1, v0
+; RV32-NEXT: add a2, s8, a2
+; RV32-NEXT: sltu a3, a2, s8
+; RV32-NEXT: add a1, a3, a1
; RV32-NEXT: add a1, a0, a1
-; RV32-NEXT: add a0, s11, a2
-; RV32-NEXT: sltu a2, a0, s11
+; RV32-NEXT: vmv.x.s a0, v8
+; RV32-NEXT: add a0, a2, a0
+; RV32-NEXT: sltu a2, a0, a2
; RV32-NEXT: add a1, a1, a2
-; RV32-NEXT: lw ra, 60(sp) # 4-byte Folded Reload
-; RV32-NEXT: lw s0, 56(sp) # 4-byte Folded Reload
-; RV32-NEXT: lw s1, 52(sp) # 4-byte Folded Reload
-; RV32-NEXT: lw s2, 48(sp) # 4-byte Folded Reload
-; RV32-NEXT: lw s3, 44(sp) # 4-byte Folded Reload
-; RV32-NEXT: lw s4, 40(sp) # 4-byte Folded Reload
-; RV32-NEXT: lw s5, 36(sp) # 4-byte Folded Reload
-; RV32-NEXT: lw s6, 32(sp) # 4-byte Folded Reload
-; RV32-NEXT: lw s7, 28(sp) # 4-byte Folded Reload
-; RV32-NEXT: lw s8, 24(sp) # 4-byte Folded Reload
-; RV32-NEXT: lw s9, 20(sp) # 4-byte Folded Reload
-; RV32-NEXT: lw s10, 16(sp) # 4-byte Folded Reload
-; RV32-NEXT: lw s11, 12(sp) # 4-byte Folded Reload
-; RV32-NEXT: addi sp, sp, 64
+; RV32-NEXT: lw s0, 44(sp) # 4-byte Folded Reload
+; RV32-NEXT: lw s1, 40(sp) # 4-byte Folded Reload
+; RV32-NEXT: lw s2, 36(sp) # 4-byte Folded Reload
+; RV32-NEXT: lw s3, 32(sp) # 4-byte Folded Reload
+; RV32-NEXT: lw s4, 28(sp) # 4-byte Folded Reload
+; RV32-NEXT: lw s5, 24(sp) # 4-byte Folded Reload
+; RV32-NEXT: lw s6, 20(sp) # 4-byte Folded Reload
+; RV32-NEXT: lw s7, 16(sp) # 4-byte Folded Reload
+; RV32-NEXT: lw s8, 12(sp) # 4-byte Folded Reload
+; RV32-NEXT: lw s9, 8(sp) # 4-byte Folded Reload
+; RV32-NEXT: lw s10, 4(sp) # 4-byte Folded Reload
+; RV32-NEXT: lw s11, 0(sp) # 4-byte Folded Reload
+; RV32-NEXT: addi sp, sp, 48
; RV32-NEXT: ret
;
; RV64-LABEL: explode_16xi64:
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-int-interleave.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-int-interleave.ll
index 40ff8b50d99d8..2ea90203b2103 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-int-interleave.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-int-interleave.ll
@@ -69,9 +69,9 @@ define <4 x i64> @interleave_v2i64(<2 x i64> %x, <2 x i64> %y) {
; RV32-V512-NEXT: vsetivli zero, 4, e16, mf4, ta, ma
; RV32-V512-NEXT: vid.v v10
; RV32-V512-NEXT: vsrl.vi v11, v10, 1
+; RV32-V512-NEXT: vmv.v.i v0, 10
; RV32-V512-NEXT: vsetvli zero, zero, e64, m1, ta, mu
; RV32-V512-NEXT: vrgatherei16.vv v10, v8, v11
-; RV32-V512-NEXT: vmv.v.i v0, 10
; RV32-V512-NEXT: vrgatherei16.vv v10, v9, v11, v0.t
; RV32-V512-NEXT: vmv.v.v v8, v10
; RV32-V512-NEXT: ret
@@ -81,8 +81,8 @@ define <4 x i64> @interleave_v2i64(<2 x i64> %x, <2 x i64> %y) {
; RV64-V512-NEXT: vsetivli zero, 4, e64, m1, ta, mu
; RV64-V512-NEXT: vid.v v10
; RV64-V512-NEXT: vsrl.vi v11, v10, 1
-; RV64-V512-NEXT: vrgather.vv v10, v8, v11
; RV64-V512-NEXT: vmv.v.i v0, 10
+; RV64-V512-NEXT: vrgather.vv v10, v8, v11
; RV64-V512-NEXT: vrgather.vv v10, v9, v11, v0.t
; RV64-V512-NEXT: vmv.v.v v8, v10
; RV64-V512-NEXT: ret
@@ -195,8 +195,8 @@ define <4 x i32> @interleave_v4i32_offset_1(<4 x i32> %x, <4 x i32> %y) {
; V128-NEXT: vsetivli zero, 4, e32, m1, ta, mu
; V128-NEXT: vid.v v8
; V128-NEXT: vsrl.vi v8, v8, 1
-; V128-NEXT: vadd.vi v8, v8, 1
; V128-NEXT: vmv.v.i v0, 10
+; V128-NEXT: vadd.vi v8, v8, 1
; V128-NEXT: vrgather.vv v10, v9, v8, v0.t
; V128-NEXT: vmv.v.v v8, v10
; V128-NEXT: ret
@@ -210,8 +210,8 @@ define <4 x i32> @interleave_v4i32_offset_1(<4 x i32> %x, <4 x i32> %y) {
; V512-NEXT: vsetivli zero, 4, e32, mf2, ta, mu
; V512-NEXT: vid.v v8
; V512-NEXT: vsrl.vi v8, v8, 1
-; V512-NEXT: vadd.vi v8, v8, 1
; V512-NEXT: vmv.v.i v0, 10
+; V512-NEXT: vadd.vi v8, v8, 1
; V512-NEXT: vrgather.vv v10, v9, v8, v0.t
; V512-NEXT: vmv1r.v v8, v10
; V512-NEXT: ret
@@ -426,13 +426,13 @@ define <64 x i32> @interleave_v32i32(<32 x i32> %x, <32 x i32> %y) {
; V128-NEXT: vwmaccu.vx v8, a0, v16
; V128-NEXT: lui a1, 699051
; V128-NEXT: addi a1, a1, -1366
-; V128-NEXT: li a2, 32
; V128-NEXT: vmv.s.x v0, a1
-; V128-NEXT: vsetvli zero, a2, e32, m8, ta, ma
+; V128-NEXT: li a1, 32
+; V128-NEXT: vsetvli zero, a1, e32, m8, ta, ma
; V128-NEXT: vmerge.vvm v24, v8, v24, v0
-; V128-NEXT: vsetivli zero, 16, e32, m4, ta, ma
; V128-NEXT: addi a1, sp, 16
; V128-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
+; V128-NEXT: vsetivli zero, 16, e32, m4, ta, ma
; V128-NEXT: vwaddu.vv v0, v16, v8
; V128-NEXT: vwmaccu.vx v0, a0, v8
; V128-NEXT: vmv8r.v v8, v0
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-int-shuffles.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-int-shuffles.ll
index aba69dc846201..517a131bd5267 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-int-shuffles.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-int-shuffles.ll
@@ -89,8 +89,8 @@ define <4 x i16> @vrgather_shuffle_vv_v4i16(<4 x i16> %x, <4 x i16> %y) {
; CHECK-NEXT: addi a0, a0, %lo(.LCPI6_0)
; CHECK-NEXT: vsetivli zero, 4, e16, mf2, ta, mu
; CHECK-NEXT: vle16.v v11, (a0)
-; CHECK-NEXT: vrgather.vv v10, v8, v11
; CHECK-NEXT: vmv.v.i v0, 8
+; CHECK-NEXT: vrgather.vv v10, v8, v11
; CHECK-NEXT: vrgather.vi v10, v9, 1, v0.t
; CHECK-NEXT: vmv1r.v v8, v10
; CHECK-NEXT: ret
@@ -162,22 +162,21 @@ define <8 x i64> @vrgather_shuffle_vv_v8i64(<8 x i64> %x, <8 x i64> %y) {
; RV32: # %bb.0:
; RV32-NEXT: vsetivli zero, 8, e16, m1, ta, ma
; RV32-NEXT: vmv.v.i v16, 2
-; RV32-NEXT: lui a0, %hi(.LCPI11_0)
-; RV32-NEXT: addi a0, a0, %lo(.LCPI11_0)
-; RV32-NEXT: vle16.v v20, (a0)
; RV32-NEXT: li a0, 5
+; RV32-NEXT: lui a1, %hi(.LCPI11_0)
+; RV32-NEXT: addi a1, a1, %lo(.LCPI11_0)
+; RV32-NEXT: vle16.v v20, (a1)
; RV32-NEXT: vslide1down.vx v21, v16, a0
-; RV32-NEXT: vsetvli zero, zero, e64, m4, ta, mu
-; RV32-NEXT: vrgatherei16.vv v16, v8, v20
; RV32-NEXT: li a0, 164
; RV32-NEXT: vmv.s.x v0, a0
+; RV32-NEXT: vsetvli zero, zero, e64, m4, ta, mu
+; RV32-NEXT: vrgatherei16.vv v16, v8, v20
; RV32-NEXT: vrgatherei16.vv v16, v12, v21, v0.t
; RV32-NEXT: vmv.v.v v8, v16
; RV32-NEXT: ret
;
; RV64-LABEL: vrgather_shuffle_vv_v8i64:
; RV64: # %bb.0:
-; RV64-NEXT: vmv4r.v v16, v8
; RV64-NEXT: lui a0, 327683
; RV64-NEXT: slli a0, a0, 3
; RV64-NEXT: addi a0, a0, 1
@@ -186,7 +185,7 @@ define <8 x i64> @vrgather_shuffle_vv_v8i64(<8 x i64> %x, <8 x i64> %y) {
; RV64-NEXT: vsetivli zero, 2, e64, m1, ta, ma
; RV64-NEXT: vmv.v.x v20, a0
; RV64-NEXT: vsetivli zero, 8, e64, m4, ta, ma
-; RV64-NEXT: vrgatherei16.vv v8, v16, v20
+; RV64-NEXT: vrgatherei16.vv v16, v8, v20
; RV64-NEXT: li a0, 164
; RV64-NEXT: vmv.s.x v0, a0
; RV64-NEXT: lui a0, 163841
@@ -194,9 +193,10 @@ define <8 x i64> @vrgather_shuffle_vv_v8i64(<8 x i64> %x, <8 x i64> %y) {
; RV64-NEXT: addi a0, a0, 1
; RV64-NEXT: slli a0, a0, 17
; RV64-NEXT: vsetivli zero, 2, e64, m1, ta, ma
-; RV64-NEXT: vmv.v.x v16, a0
+; RV64-NEXT: vmv.v.x v8, a0
; RV64-NEXT: vsetivli zero, 8, e64, m4, ta, mu
-; RV64-NEXT: vrgatherei16.vv v8, v12, v16, v0.t
+; RV64-NEXT: vrgatherei16.vv v16, v12, v8, v0.t
+; RV64-NEXT: vmv.v.v v8, v16
; RV64-NEXT: ret
%s = shufflevector <8 x i64> %x, <8 x i64> %y, <8 x i32> <i32 1, i32 2, i32 10, i32 5, i32 1, i32 10, i32 3, i32 13>
ret <8 x i64> %s
@@ -210,13 +210,13 @@ define <8 x i64> @vrgather_shuffle_xv_v8i64(<8 x i64> %x) {
; RV32-NEXT: vsetivli zero, 8, e64, m4, ta, mu
; RV32-NEXT: vle16.v v16, (a0)
; RV32-NEXT: vmv.v.i v20, -1
-; RV32-NEXT: vrgatherei16.vv v12, v20, v16
; RV32-NEXT: lui a0, %hi(.LCPI12_1)
; RV32-NEXT: addi a0, a0, %lo(.LCPI12_1)
-; RV32-NEXT: vle16.v v16, (a0)
+; RV32-NEXT: vle16.v v17, (a0)
; RV32-NEXT: li a0, 113
; RV32-NEXT: vmv.s.x v0, a0
-; RV32-NEXT: vrgatherei16.vv v12, v8, v16, v0.t
+; RV32-NEXT: vrgatherei16.vv v12, v20, v16
+; RV32-NEXT: vrgatherei16.vv v12, v8, v17, v0.t
; RV32-NEXT: vmv.v.v v8, v12
; RV32-NEXT: ret
;
@@ -367,10 +367,10 @@ define <8 x i8> @splat_ve4_ins_i1ve3(<8 x i8> %v) {
define <8 x i8> @splat_ve2_we0(<8 x i8> %v, <8 x i8> %w) {
; CHECK-LABEL: splat_ve2_we0:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetivli zero, 8, e8, mf2, ta, mu
-; CHECK-NEXT: vrgather.vi v10, v8, 2
; CHECK-NEXT: li a0, 66
+; CHECK-NEXT: vsetivli zero, 8, e8, mf2, ta, mu
; CHECK-NEXT: vmv.s.x v0, a0
+; CHECK-NEXT: vrgather.vi v10, v8, 2
; CHECK-NEXT: vrgather.vi v10, v9, 0, v0.t
; CHECK-NEXT: vmv1r.v v8, v10
; CHECK-NEXT: ret
@@ -386,10 +386,10 @@ define <8 x i8> @splat_ve2_we0_ins_i0ve4(<8 x i8> %v, <8 x i8> %w) {
; CHECK-NEXT: li a0, 4
; CHECK-NEXT: vsetvli zero, zero, e8, mf2, tu, ma
; CHECK-NEXT: vmv.s.x v11, a0
-; CHECK-NEXT: vsetvli zero, zero, e8, mf2, ta, mu
-; CHECK-NEXT: vrgather.vv v10, v8, v11
; CHECK-NEXT: li a0, 66
; CHECK-NEXT: vmv.s.x v0, a0
+; CHECK-NEXT: vsetvli zero, zero, e8, mf2, ta, mu
+; CHECK-NEXT: vrgather.vv v10, v8, v11
; CHECK-NEXT: vrgather.vi v10, v9, 0, v0.t
; CHECK-NEXT: vmv1r.v v8, v10
; CHECK-NEXT: ret
@@ -402,10 +402,10 @@ define <8 x i8> @splat_ve2_we0_ins_i0we4(<8 x i8> %v, <8 x i8> %w) {
; CHECK: # %bb.0:
; CHECK-NEXT: vsetivli zero, 8, e8, mf2, ta, ma
; CHECK-NEXT: vrgather.vi v10, v8, 2
-; CHECK-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
-; CHECK-NEXT: vmv.v.i v8, 4
; CHECK-NEXT: li a0, 67
; CHECK-NEXT: vmv.s.x v0, a0
+; CHECK-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
+; CHECK-NEXT: vmv.v.i v8, 4
; CHECK-NEXT: vsetivli zero, 8, e8, mf2, ta, mu
; CHECK-NEXT: vrgather.vv v10, v9, v8, v0.t
; CHECK-NEXT: vmv1r.v v8, v10
@@ -421,10 +421,10 @@ define <8 x i8> @splat_ve2_we0_ins_i2ve4(<8 x i8> %v, <8 x i8> %w) {
; CHECK-NEXT: addi a0, a0, 514
; CHECK-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
; CHECK-NEXT: vmv.v.x v11, a0
-; CHECK-NEXT: vsetivli zero, 8, e8, mf2, ta, mu
-; CHECK-NEXT: vrgather.vv v10, v8, v11
; CHECK-NEXT: li a0, 66
; CHECK-NEXT: vmv.s.x v0, a0
+; CHECK-NEXT: vsetivli zero, 8, e8, mf2, ta, mu
+; CHECK-NEXT: vrgather.vv v10, v8, v11
; CHECK-NEXT: vrgather.vi v10, v9, 0, v0.t
; CHECK-NEXT: vmv1r.v v8, v10
; CHECK-NEXT: ret
@@ -440,10 +440,10 @@ define <8 x i8> @splat_ve2_we0_ins_i2we4(<8 x i8> %v, <8 x i8> %w) {
; CHECK-NEXT: vmv.v.i v11, 0
; CHECK-NEXT: vsetivli zero, 3, e8, mf2, tu, ma
; CHECK-NEXT: vslideup.vi v11, v10, 2
-; CHECK-NEXT: vsetivli zero, 8, e8, mf2, ta, mu
-; CHECK-NEXT: vrgather.vi v10, v8, 2
; CHECK-NEXT: li a0, 70
; CHECK-NEXT: vmv.s.x v0, a0
+; CHECK-NEXT: vsetivli zero, 8, e8, mf2, ta, mu
+; CHECK-NEXT: vrgather.vi v10, v8, 2
; CHECK-NEXT: vrgather.vv v10, v9, v11, v0.t
; CHECK-NEXT: vmv1r.v v8, v10
; CHECK-NEXT: ret
@@ -463,10 +463,10 @@ define <8 x i8> @splat_ve2_we0_ins_i2ve4_i5we6(<8 x i8> %v, <8 x i8> %w) {
; CHECK-NEXT: addi a0, a0, 2
; CHECK-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
; CHECK-NEXT: vmv.v.x v12, a0
-; CHECK-NEXT: vsetivli zero, 8, e8, mf2, ta, mu
-; CHECK-NEXT: vrgather.vv v10, v8, v12
; CHECK-NEXT: li a0, 98
; CHECK-NEXT: vmv.s.x v0, a0
+; CHECK-NEXT: vsetivli zero, 8, e8, mf2, ta, mu
+; CHECK-NEXT: vrgather.vv v10, v8, v12
; CHECK-NEXT: vrgather.vv v10, v9, v11, v0.t
; CHECK-NEXT: vmv1r.v v8, v10
; CHECK-NEXT: ret
@@ -666,8 +666,8 @@ define <8 x i8> @merge_start_into_start(<8 x i8> %v, <8 x i8> %w) {
define <8 x i8> @merge_slidedown(<8 x i8> %v, <8 x i8> %w) {
; CHECK-LABEL: merge_slidedown:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetivli zero, 8, e8, mf2, ta, ma
; CHECK-NEXT: li a0, 195
+; CHECK-NEXT: vsetivli zero, 8, e8, mf2, ta, ma
; CHECK-NEXT: vmv.s.x v0, a0
; CHECK-NEXT: vslidedown.vi v8, v8, 1
; CHECK-NEXT: vmerge.vvm v8, v8, v9, v0
@@ -680,10 +680,10 @@ define <8 x i8> @merge_slidedown(<8 x i8> %v, <8 x i8> %w) {
define <8 x i8> @merge_non_contiguous_slideup_slidedown(<8 x i8> %v, <8 x i8> %w) {
; CHECK-LABEL: merge_non_contiguous_slideup_slidedown:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetivli zero, 8, e8, mf2, ta, mu
-; CHECK-NEXT: vslidedown.vi v8, v8, 2
; CHECK-NEXT: li a0, -22
+; CHECK-NEXT: vsetivli zero, 8, e8, mf2, ta, mu
; CHECK-NEXT: vmv.s.x v0, a0
+; CHECK-NEXT: vslidedown.vi v8, v8, 2
; CHECK-NEXT: vslideup.vi v8, v9, 1, v0.t
; CHECK-NEXT: ret
%res = shufflevector <8 x i8> %v, <8 x i8> %w, <8 x i32> <i32 2, i32 8, i32 4, i32 10, i32 6, i32 12, i32 13, i32 14>
@@ -694,13 +694,13 @@ define <8 x i8> @merge_non_contiguous_slideup_slidedown(<8 x i8> %v, <8 x i8> %w
define <8 x i8> @unmergable(<8 x i8> %v, <8 x i8> %w) {
; CHECK-LABEL: unmergable:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetivli zero, 8, e8, mf2, ta, mu
-; CHECK-NEXT: vslidedown.vi v8, v8, 2
; CHECK-NEXT: lui a0, %hi(.LCPI46_0)
; CHECK-NEXT: addi a0, a0, %lo(.LCPI46_0)
+; CHECK-NEXT: vsetivli zero, 8, e8, mf2, ta, mu
; CHECK-NEXT: vle8.v v10, (a0)
; CHECK-NEXT: li a0, -22
; CHECK-NEXT: vmv.s.x v0, a0
+; CHECK-NEXT: vslidedown.vi v8, v8, 2
; CHECK-NEXT: vrgather.vv v8, v9, v10, v0.t
; CHECK-NEXT: ret
%res = shufflevector <8 x i8> %v, <8 x i8> %w, <8 x i32> <i32 2, i32 9, i32 4, i32 11, i32 6, i32 13, i32 8, i32 15>
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-int.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-int.ll
index 635869904832c..79c36a629465d 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-int.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-int.ll
@@ -1100,46 +1100,46 @@ define void @mulhu_v16i8(ptr %x) {
; CHECK-LABEL: mulhu_v16i8:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetivli zero, 16, e16, m2, ta, ma
-; CHECK-NEXT: vle8.v v8, (a0)
+; CHECK-NEXT: vle8.v v9, (a0)
; CHECK-NEXT: lui a1, 3
; CHECK-NEXT: addi a1, a1, -2044
; CHECK-NEXT: vmv.s.x v0, a1
; CHECK-NEXT: vsetvli zero, zero, e8, m1, ta, ma
-; CHECK-NEXT: vmv.v.i v9, 0
-; CHECK-NEXT: li a1, -128
-; CHECK-NEXT: vmerge.vxm v10, v9, a1, v0
+; CHECK-NEXT: vmv.v.i v10, 0
; CHECK-NEXT: lui a1, 1
; CHECK-NEXT: addi a2, a1, 32
; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, ma
-; CHECK-NEXT: vmv.s.x v0, a2
-; CHECK-NEXT: vsetvli zero, zero, e8, m1, ta, ma
+; CHECK-NEXT: vmv.s.x v8, a2
; CHECK-NEXT: lui a2, %hi(.LCPI65_0)
; CHECK-NEXT: addi a2, a2, %lo(.LCPI65_0)
; CHECK-NEXT: vle8.v v11, (a2)
-; CHECK-NEXT: vmerge.vim v9, v9, 1, v0
-; CHECK-NEXT: vsrl.vv v9, v8, v9
-; CHECK-NEXT: vmulhu.vv v9, v9, v11
-; CHECK-NEXT: vsub.vv v8, v8, v9
-; CHECK-NEXT: vmulhu.vv v8, v8, v10
-; CHECK-NEXT: vadd.vv v8, v8, v9
+; CHECK-NEXT: li a2, -128
+; CHECK-NEXT: vsetvli zero, zero, e8, m1, ta, ma
+; CHECK-NEXT: vmerge.vxm v12, v10, a2, v0
+; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vmerge.vim v8, v10, 1, v0
+; CHECK-NEXT: vsrl.vv v8, v9, v8
+; CHECK-NEXT: vmulhu.vv v8, v8, v11
+; CHECK-NEXT: vsub.vv v9, v9, v8
+; CHECK-NEXT: vmulhu.vv v9, v9, v12
+; CHECK-NEXT: vadd.vv v9, v9, v8
; CHECK-NEXT: li a2, 513
; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; CHECK-NEXT: vmv.s.x v0, a2
; CHECK-NEXT: vsetvli zero, zero, e8, m1, ta, ma
-; CHECK-NEXT: vmv.v.i v9, 4
-; CHECK-NEXT: vmerge.vim v9, v9, 1, v0
+; CHECK-NEXT: vmv.v.i v8, 4
+; CHECK-NEXT: vmerge.vim v10, v8, 1, v0
; CHECK-NEXT: addi a1, a1, 78
; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; CHECK-NEXT: vmv.s.x v0, a1
-; CHECK-NEXT: vsetvli zero, zero, e8, m1, ta, ma
-; CHECK-NEXT: vmerge.vim v9, v9, 3, v0
; CHECK-NEXT: lui a1, 8
; CHECK-NEXT: addi a1, a1, 304
-; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, ma
-; CHECK-NEXT: vmv.s.x v0, a1
+; CHECK-NEXT: vmv.s.x v8, a1
; CHECK-NEXT: vsetvli zero, zero, e8, m1, ta, ma
-; CHECK-NEXT: vmerge.vim v9, v9, 2, v0
-; CHECK-NEXT: vsrl.vv v8, v8, v9
+; CHECK-NEXT: vmerge.vim v10, v10, 3, v0
+; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vmerge.vim v8, v10, 2, v0
+; CHECK-NEXT: vsrl.vv v8, v9, v8
; CHECK-NEXT: vse8.v v8, (a0)
; CHECK-NEXT: ret
%a = load <16 x i8>, ptr %x
@@ -1158,16 +1158,16 @@ define void @mulhu_v8i16(ptr %x) {
; CHECK-NEXT: vmv.v.i v10, 0
; CHECK-NEXT: vsetvli zero, zero, e16, m1, tu, ma
; CHECK-NEXT: vmv.s.x v10, a1
+; CHECK-NEXT: lui a1, %hi(.LCPI66_0)
+; CHECK-NEXT: addi a1, a1, %lo(.LCPI66_0)
; CHECK-NEXT: vsetvli zero, zero, e16, m1, ta, ma
-; CHECK-NEXT: vmv.v.i v11, 1
+; CHECK-NEXT: vle16.v v11, (a1)
+; CHECK-NEXT: vmv.v.i v12, 1
; CHECK-NEXT: vsetivli zero, 7, e16, m1, tu, ma
-; CHECK-NEXT: vslideup.vi v9, v11, 6
+; CHECK-NEXT: vslideup.vi v9, v12, 6
; CHECK-NEXT: vsetivli zero, 8, e16, m1, ta, ma
-; CHECK-NEXT: lui a1, %hi(.LCPI66_0)
-; CHECK-NEXT: addi a1, a1, %lo(.LCPI66_0)
-; CHECK-NEXT: vle16.v v12, (a1)
; CHECK-NEXT: vsrl.vv v9, v8, v9
-; CHECK-NEXT: vmulhu.vv v9, v9, v12
+; CHECK-NEXT: vmulhu.vv v9, v9, v11
; CHECK-NEXT: vsub.vv v8, v8, v9
; CHECK-NEXT: vmulhu.vv v8, v8, v10
; CHECK-NEXT: vadd.vv v8, v8, v9
@@ -1176,7 +1176,7 @@ define void @mulhu_v8i16(ptr %x) {
; CHECK-NEXT: vmv.v.i v9, 3
; CHECK-NEXT: vmerge.vim v9, v9, 2, v0
; CHECK-NEXT: vsetivli zero, 7, e16, m1, tu, ma
-; CHECK-NEXT: vslideup.vi v9, v11, 6
+; CHECK-NEXT: vslideup.vi v9, v12, 6
; CHECK-NEXT: vsetivli zero, 8, e16, m1, ta, ma
; CHECK-NEXT: vsrl.vv v8, v8, v9
; CHECK-NEXT: vse16.v v8, (a0)
@@ -1222,18 +1222,18 @@ define void @mulhu_v4i32(ptr %x) {
; CHECK: # %bb.0:
; CHECK-NEXT: vsetivli zero, 4, e32, m1, ta, ma
; CHECK-NEXT: vle32.v v8, (a0)
-; CHECK-NEXT: lui a1, 524288
-; CHECK-NEXT: vmv.s.x v9, a1
-; CHECK-NEXT: vmv.v.i v10, 0
-; CHECK-NEXT: vsetivli zero, 3, e32, m1, tu, ma
-; CHECK-NEXT: vslideup.vi v10, v9, 2
; CHECK-NEXT: lui a1, %hi(.LCPI68_0)
; CHECK-NEXT: addi a1, a1, %lo(.LCPI68_0)
-; CHECK-NEXT: vsetivli zero, 4, e32, m1, ta, ma
; CHECK-NEXT: vle32.v v9, (a1)
+; CHECK-NEXT: lui a1, 524288
+; CHECK-NEXT: vmv.s.x v10, a1
+; CHECK-NEXT: vmv.v.i v11, 0
+; CHECK-NEXT: vsetivli zero, 3, e32, m1, tu, ma
+; CHECK-NEXT: vslideup.vi v11, v10, 2
+; CHECK-NEXT: vsetivli zero, 4, e32, m1, ta, ma
; CHECK-NEXT: vmulhu.vv v9, v8, v9
; CHECK-NEXT: vsub.vv v8, v8, v9
-; CHECK-NEXT: vmulhu.vv v8, v8, v10
+; CHECK-NEXT: vmulhu.vv v8, v8, v11
; CHECK-NEXT: vadd.vv v8, v8, v9
; CHECK-NEXT: lui a1, 4128
; CHECK-NEXT: addi a1, a1, 514
@@ -1455,13 +1455,13 @@ define void @mulhs_v2i64(ptr %x) {
; RV64-LABEL: mulhs_v2i64:
; RV64: # %bb.0:
; RV64-NEXT: vsetivli zero, 2, e64, m1, ta, ma
+; RV64-NEXT: vle64.v v8, (a0)
; RV64-NEXT: lui a1, 349525
; RV64-NEXT: addiw a1, a1, 1365
-; RV64-NEXT: slli a2, a1, 32
-; RV64-NEXT: add a1, a1, a2
; RV64-NEXT: lui a2, %hi(.LCPI74_0)
; RV64-NEXT: ld a2, %lo(.LCPI74_0)(a2)
-; RV64-NEXT: vle64.v v8, (a0)
+; RV64-NEXT: slli a3, a1, 32
+; RV64-NEXT: add a1, a1, a3
; RV64-NEXT: vmv.v.x v9, a1
; RV64-NEXT: vsetvli zero, zero, e64, m1, tu, ma
; RV64-NEXT: vmv.s.x v9, a2
@@ -3260,49 +3260,47 @@ define void @mulhu_v32i8(ptr %x) {
; CHECK: # %bb.0:
; CHECK-NEXT: li a1, 32
; CHECK-NEXT: vsetvli zero, a1, e8, m2, ta, ma
-; CHECK-NEXT: vle8.v v8, (a0)
-; CHECK-NEXT: vmv.v.i v10, 0
+; CHECK-NEXT: vle8.v v10, (a0)
+; CHECK-NEXT: vmv.v.i v12, 0
; CHECK-NEXT: lui a1, 163907
; CHECK-NEXT: addi a1, a1, -2044
; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; CHECK-NEXT: vmv.s.x v0, a1
-; CHECK-NEXT: li a1, -128
-; CHECK-NEXT: vsetvli zero, zero, e8, m2, ta, ma
-; CHECK-NEXT: vmerge.vxm v12, v10, a1, v0
; CHECK-NEXT: lui a1, 66049
; CHECK-NEXT: addi a1, a1, 32
-; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; CHECK-NEXT: vmv.s.x v0, a1
-; CHECK-NEXT: vsetvli zero, zero, e8, m2, ta, ma
+; CHECK-NEXT: vmv.s.x v8, a1
; CHECK-NEXT: lui a1, %hi(.LCPI181_0)
; CHECK-NEXT: addi a1, a1, %lo(.LCPI181_0)
; CHECK-NEXT: vle8.v v14, (a1)
-; CHECK-NEXT: vmerge.vim v10, v10, 1, v0
-; CHECK-NEXT: vsrl.vv v10, v8, v10
-; CHECK-NEXT: vmulhu.vv v10, v10, v14
-; CHECK-NEXT: vsub.vv v8, v8, v10
-; CHECK-NEXT: vmulhu.vv v8, v8, v12
-; CHECK-NEXT: vadd.vv v8, v8, v10
-; CHECK-NEXT: vmv.v.i v10, 4
+; CHECK-NEXT: li a1, -128
+; CHECK-NEXT: vsetvli zero, zero, e8, m2, ta, ma
+; CHECK-NEXT: vmerge.vxm v16, v12, a1, v0
+; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vmerge.vim v8, v12, 1, v0
+; CHECK-NEXT: vsrl.vv v8, v10, v8
+; CHECK-NEXT: vmulhu.vv v8, v8, v14
+; CHECK-NEXT: vsub.vv v10, v10, v8
+; CHECK-NEXT: vmulhu.vv v10, v10, v16
+; CHECK-NEXT: vadd.vv v10, v10, v8
; CHECK-NEXT: lui a1, 8208
; CHECK-NEXT: addi a1, a1, 513
; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; CHECK-NEXT: vmv.s.x v0, a1
; CHECK-NEXT: vsetvli zero, zero, e8, m2, ta, ma
-; CHECK-NEXT: vmerge.vim v10, v10, 1, v0
+; CHECK-NEXT: vmv.v.i v8, 4
+; CHECK-NEXT: vmerge.vim v12, v8, 1, v0
; CHECK-NEXT: lui a1, 66785
; CHECK-NEXT: addi a1, a1, 78
; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; CHECK-NEXT: vmv.s.x v0, a1
-; CHECK-NEXT: vsetvli zero, zero, e8, m2, ta, ma
-; CHECK-NEXT: vmerge.vim v10, v10, 3, v0
; CHECK-NEXT: lui a1, 529160
; CHECK-NEXT: addi a1, a1, 304
-; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; CHECK-NEXT: vmv.s.x v0, a1
+; CHECK-NEXT: vmv.s.x v8, a1
; CHECK-NEXT: vsetvli zero, zero, e8, m2, ta, ma
-; CHECK-NEXT: vmerge.vim v10, v10, 2, v0
-; CHECK-NEXT: vsrl.vv v8, v8, v10
+; CHECK-NEXT: vmerge.vim v12, v12, 3, v0
+; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vmerge.vim v8, v12, 2, v0
+; CHECK-NEXT: vsrl.vv v8, v10, v8
; CHECK-NEXT: vse8.v v8, (a0)
; CHECK-NEXT: ret
%a = load <32 x i8>, ptr %x
@@ -3326,12 +3324,12 @@ define void @mulhu_v16i16(ptr %x) {
; RV32-NEXT: vmv.s.x v8, a1
; RV32-NEXT: vsetvli zero, zero, e8, m1, ta, ma
; RV32-NEXT: vmv.v.i v9, 0
-; RV32-NEXT: vmv1r.v v0, v8
-; RV32-NEXT: vmerge.vim v9, v9, 1, v0
-; RV32-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; RV32-NEXT: lui a1, %hi(.LCPI182_0)
; RV32-NEXT: addi a1, a1, %lo(.LCPI182_0)
; RV32-NEXT: vle16.v v14, (a1)
+; RV32-NEXT: vmv1r.v v0, v8
+; RV32-NEXT: vmerge.vim v9, v9, 1, v0
+; RV32-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; RV32-NEXT: vsext.vf2 v16, v9
; RV32-NEXT: vsrl.vv v16, v10, v16
; RV32-NEXT: vmulhu.vv v14, v16, v14
@@ -3361,27 +3359,27 @@ define void @mulhu_v16i16(ptr %x) {
; RV64-NEXT: vmv.v.i v10, 0
; RV64-NEXT: lui a1, 1048568
; RV64-NEXT: vmerge.vxm v10, v10, a1, v0
+; RV64-NEXT: lui a1, %hi(.LCPI182_0)
+; RV64-NEXT: addi a1, a1, %lo(.LCPI182_0)
+; RV64-NEXT: vle16.v v12, (a1)
; RV64-NEXT: li a1, 1
; RV64-NEXT: slli a1, a1, 48
; RV64-NEXT: vsetivli zero, 2, e64, m1, ta, ma
-; RV64-NEXT: vmv.v.x v12, a1
+; RV64-NEXT: vmv.v.x v14, a1
; RV64-NEXT: vsetivli zero, 16, e16, m2, ta, ma
-; RV64-NEXT: lui a1, %hi(.LCPI182_0)
-; RV64-NEXT: addi a1, a1, %lo(.LCPI182_0)
-; RV64-NEXT: vle16.v v14, (a1)
-; RV64-NEXT: vsext.vf2 v16, v12
-; RV64-NEXT: vsrl.vv v12, v8, v16
-; RV64-NEXT: vmulhu.vv v12, v12, v14
-; RV64-NEXT: vsub.vv v8, v8, v12
-; RV64-NEXT: vmulhu.vv v8, v8, v10
-; RV64-NEXT: vadd.vv v8, v8, v12
+; RV64-NEXT: vsext.vf2 v16, v14
+; RV64-NEXT: vsrl.vv v14, v8, v16
+; RV64-NEXT: vmulhu.vv v12, v14, v12
; RV64-NEXT: lui a1, %hi(.LCPI182_1)
; RV64-NEXT: addi a1, a1, %lo(.LCPI182_1)
; RV64-NEXT: vsetivli zero, 2, e64, m1, ta, ma
-; RV64-NEXT: vlse64.v v10, (a1), zero
+; RV64-NEXT: vlse64.v v14, (a1), zero
; RV64-NEXT: vsetivli zero, 16, e16, m2, ta, ma
-; RV64-NEXT: vsext.vf2 v12, v10
-; RV64-NEXT: vsrl.vv v8, v8, v12
+; RV64-NEXT: vsub.vv v8, v8, v12
+; RV64-NEXT: vmulhu.vv v8, v8, v10
+; RV64-NEXT: vadd.vv v8, v8, v12
+; RV64-NEXT: vsext.vf2 v10, v14
+; RV64-NEXT: vsrl.vv v8, v8, v10
; RV64-NEXT: vse16.v v8, (a0)
; RV64-NEXT: ret
%a = load <16 x i16>, ptr %x
@@ -3433,23 +3431,24 @@ define void @mulhu_v4i64(ptr %x) {
; RV32-NEXT: vle32.v v10, (a1)
; RV32-NEXT: vsetivli zero, 4, e64, m2, ta, ma
; RV32-NEXT: vmulhu.vv v10, v8, v10
-; RV32-NEXT: vsub.vv v8, v8, v10
; RV32-NEXT: lui a1, 524288
; RV32-NEXT: vmv.s.x v12, a1
; RV32-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32-NEXT: vmv.v.i v14, 0
; RV32-NEXT: vsetivli zero, 6, e32, m2, tu, ma
; RV32-NEXT: vslideup.vi v14, v12, 5
+; RV32-NEXT: lui a1, %hi(.LCPI184_1)
+; RV32-NEXT: addi a1, a1, %lo(.LCPI184_1)
+; RV32-NEXT: vsetivli zero, 8, e8, mf2, ta, ma
+; RV32-NEXT: vle8.v v12, (a1)
; RV32-NEXT: vsetivli zero, 4, e64, m2, ta, ma
+; RV32-NEXT: vsub.vv v8, v8, v10
; RV32-NEXT: vmulhu.vv v8, v8, v14
; RV32-NEXT: vadd.vv v8, v8, v10
-; RV32-NEXT: lui a1, %hi(.LCPI184_1)
-; RV32-NEXT: addi a1, a1, %lo(.LCPI184_1)
; RV32-NEXT: vsetivli zero, 8, e32, m2, ta, ma
-; RV32-NEXT: vle8.v v10, (a1)
-; RV32-NEXT: vsext.vf4 v12, v10
+; RV32-NEXT: vsext.vf4 v10, v12
; RV32-NEXT: vsetivli zero, 4, e64, m2, ta, ma
-; RV32-NEXT: vsrl.vv v8, v8, v12
+; RV32-NEXT: vsrl.vv v8, v8, v10
; RV32-NEXT: vse64.v v8, (a0)
; RV32-NEXT: ret
;
@@ -3457,19 +3456,19 @@ define void @mulhu_v4i64(ptr %x) {
; RV64: # %bb.0:
; RV64-NEXT: vsetivli zero, 4, e64, m2, ta, ma
; RV64-NEXT: vle64.v v8, (a0)
+; RV64-NEXT: lui a1, %hi(.LCPI184_0)
+; RV64-NEXT: addi a1, a1, %lo(.LCPI184_0)
+; RV64-NEXT: vle64.v v10, (a1)
; RV64-NEXT: li a1, -1
; RV64-NEXT: slli a1, a1, 63
-; RV64-NEXT: vmv.s.x v10, a1
-; RV64-NEXT: vmv.v.i v12, 0
+; RV64-NEXT: vmv.s.x v12, a1
+; RV64-NEXT: vmv.v.i v14, 0
; RV64-NEXT: vsetivli zero, 3, e64, m2, tu, ma
-; RV64-NEXT: vslideup.vi v12, v10, 2
-; RV64-NEXT: lui a1, %hi(.LCPI184_0)
-; RV64-NEXT: addi a1, a1, %lo(.LCPI184_0)
+; RV64-NEXT: vslideup.vi v14, v12, 2
; RV64-NEXT: vsetivli zero, 4, e64, m2, ta, ma
-; RV64-NEXT: vle64.v v10, (a1)
; RV64-NEXT: vmulhu.vv v10, v8, v10
; RV64-NEXT: vsub.vv v8, v8, v10
-; RV64-NEXT: vmulhu.vv v8, v8, v12
+; RV64-NEXT: vmulhu.vv v8, v8, v14
; RV64-NEXT: vadd.vv v8, v8, v10
; RV64-NEXT: lui a1, 12320
; RV64-NEXT: addi a1, a1, 513
@@ -3488,14 +3487,13 @@ define void @mulhs_v32i8(ptr %x) {
; CHECK-LABEL: mulhs_v32i8:
; CHECK: # %bb.0:
; CHECK-NEXT: li a1, 32
-; CHECK-NEXT: vsetvli zero, a1, e8, m2, ta, ma
+; CHECK-NEXT: vsetvli zero, a1, e32, m8, ta, ma
; CHECK-NEXT: vle8.v v8, (a0)
-; CHECK-NEXT: vmv.v.i v10, 7
; CHECK-NEXT: lui a1, 304453
; CHECK-NEXT: addi a1, a1, -1452
-; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; CHECK-NEXT: vmv.s.x v0, a1
; CHECK-NEXT: vsetvli zero, zero, e8, m2, ta, ma
+; CHECK-NEXT: vmv.v.i v10, 7
; CHECK-NEXT: vmerge.vim v10, v10, 1, v0
; CHECK-NEXT: li a1, -123
; CHECK-NEXT: vmv.v.x v12, a1
@@ -3615,19 +3613,19 @@ define void @mulhs_v4i64(ptr %x) {
;
; RV64-LABEL: mulhs_v4i64:
; RV64: # %bb.0:
-; RV64-NEXT: vsetivli zero, 4, e64, m2, ta, ma
-; RV64-NEXT: vle64.v v8, (a0)
; RV64-NEXT: lui a1, 349525
; RV64-NEXT: addiw a1, a1, 1365
; RV64-NEXT: slli a2, a1, 32
; RV64-NEXT: add a1, a1, a2
-; RV64-NEXT: vmv.v.x v10, a1
-; RV64-NEXT: lui a1, %hi(.LCPI188_0)
-; RV64-NEXT: ld a1, %lo(.LCPI188_0)(a1)
+; RV64-NEXT: lui a2, %hi(.LCPI188_0)
+; RV64-NEXT: ld a2, %lo(.LCPI188_0)(a2)
+; RV64-NEXT: vsetivli zero, 4, e64, m2, ta, ma
+; RV64-NEXT: vle64.v v8, (a0)
; RV64-NEXT: vsetivli zero, 1, e8, mf8, ta, ma
; RV64-NEXT: vmv.v.i v0, 5
; RV64-NEXT: vsetivli zero, 4, e64, m2, ta, ma
-; RV64-NEXT: vmerge.vxm v10, v10, a1, v0
+; RV64-NEXT: vmv.v.x v10, a1
+; RV64-NEXT: vmerge.vxm v10, v10, a2, v0
; RV64-NEXT: vmulh.vv v10, v8, v10
; RV64-NEXT: lui a1, 1044496
; RV64-NEXT: addi a1, a1, -256
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-interleaved-access-zve32x.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-interleaved-access-zve32x.ll
index 8acc70faaa1fc..82e0760d593c2 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-interleaved-access-zve32x.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-interleaved-access-zve32x.ll
@@ -21,8 +21,8 @@ define <4 x i1> @load_large_vector(ptr %p) {
; ZVE32X-NEXT: vmv.s.x v8, a4
; ZVE32X-NEXT: vand.vi v8, v8, 1
; ZVE32X-NEXT: vmsne.vi v0, v8, 0
-; ZVE32X-NEXT: vmv.s.x v8, zero
-; ZVE32X-NEXT: vmerge.vim v9, v8, 1, v0
+; ZVE32X-NEXT: vmv.s.x v9, zero
+; ZVE32X-NEXT: vmerge.vim v8, v9, 1, v0
; ZVE32X-NEXT: xor a0, a0, a7
; ZVE32X-NEXT: snez a0, a0
; ZVE32X-NEXT: vmv.s.x v10, a0
@@ -32,32 +32,38 @@ define <4 x i1> @load_large_vector(ptr %p) {
; ZVE32X-NEXT: vmv.v.i v10, 0
; ZVE32X-NEXT: vmerge.vim v11, v10, 1, v0
; ZVE32X-NEXT: vsetivli zero, 2, e8, mf4, tu, ma
-; ZVE32X-NEXT: vslideup.vi v11, v9, 1
+; ZVE32X-NEXT: vslideup.vi v11, v8, 1
; ZVE32X-NEXT: vsetivli zero, 4, e8, mf4, ta, ma
; ZVE32X-NEXT: vmsne.vi v0, v11, 0
-; ZVE32X-NEXT: vmerge.vim v9, v10, 1, v0
; ZVE32X-NEXT: xor a0, a6, a3
; ZVE32X-NEXT: snez a0, a0
-; ZVE32X-NEXT: vmv.s.x v11, a0
+; ZVE32X-NEXT: vmv.s.x v8, a0
; ZVE32X-NEXT: vsetivli zero, 1, e8, mf4, ta, ma
-; ZVE32X-NEXT: vand.vi v11, v11, 1
-; ZVE32X-NEXT: vmsne.vi v0, v11, 0
-; ZVE32X-NEXT: vmerge.vim v11, v8, 1, v0
+; ZVE32X-NEXT: vand.vi v8, v8, 1
+; ZVE32X-NEXT: vmsne.vi v8, v8, 0
+; ZVE32X-NEXT: vsetivli zero, 4, e8, mf4, ta, ma
+; ZVE32X-NEXT: vmerge.vim v11, v10, 1, v0
+; ZVE32X-NEXT: vmv1r.v v0, v8
+; ZVE32X-NEXT: vsetivli zero, 1, e8, mf4, ta, ma
+; ZVE32X-NEXT: vmerge.vim v8, v9, 1, v0
; ZVE32X-NEXT: vsetivli zero, 3, e8, mf4, tu, ma
-; ZVE32X-NEXT: vslideup.vi v9, v11, 2
+; ZVE32X-NEXT: vslideup.vi v11, v8, 2
; ZVE32X-NEXT: vsetivli zero, 4, e8, mf4, ta, ma
-; ZVE32X-NEXT: vmsne.vi v0, v9, 0
-; ZVE32X-NEXT: vmerge.vim v9, v10, 1, v0
+; ZVE32X-NEXT: vmsne.vi v0, v11, 0
; ZVE32X-NEXT: xor a1, a2, a1
; ZVE32X-NEXT: snez a0, a1
-; ZVE32X-NEXT: vmv.s.x v10, a0
+; ZVE32X-NEXT: vmv.s.x v8, a0
; ZVE32X-NEXT: vsetivli zero, 1, e8, mf4, ta, ma
-; ZVE32X-NEXT: vand.vi v10, v10, 1
-; ZVE32X-NEXT: vmsne.vi v0, v10, 0
-; ZVE32X-NEXT: vmerge.vim v8, v8, 1, v0
+; ZVE32X-NEXT: vand.vi v8, v8, 1
+; ZVE32X-NEXT: vmsne.vi v8, v8, 0
; ZVE32X-NEXT: vsetivli zero, 4, e8, mf4, ta, ma
-; ZVE32X-NEXT: vslideup.vi v9, v8, 3
-; ZVE32X-NEXT: vmsne.vi v0, v9, 0
+; ZVE32X-NEXT: vmerge.vim v10, v10, 1, v0
+; ZVE32X-NEXT: vmv1r.v v0, v8
+; ZVE32X-NEXT: vsetivli zero, 1, e8, mf4, ta, ma
+; ZVE32X-NEXT: vmerge.vim v8, v9, 1, v0
+; ZVE32X-NEXT: vsetivli zero, 4, e8, mf4, ta, ma
+; ZVE32X-NEXT: vslideup.vi v10, v8, 3
+; ZVE32X-NEXT: vmsne.vi v0, v10, 0
; ZVE32X-NEXT: ret
;
; ZVE64X-LABEL: load_large_vector:
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-interleaved-access.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-interleaved-access.ll
index 99364264de829..178a920169ad9 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-interleaved-access.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-interleaved-access.ll
@@ -159,16 +159,16 @@ define {<8 x i64>, <8 x i64>, <8 x i64>, <8 x i64>, <8 x i64>, <8 x i64>} @load_
; RV32-NEXT: addi sp, sp, -16
; RV32-NEXT: .cfi_def_cfa_offset 16
; RV32-NEXT: csrr a2, vlenb
-; RV32-NEXT: li a3, 54
+; RV32-NEXT: li a3, 82
; RV32-NEXT: mul a2, a2, a3
; RV32-NEXT: sub sp, sp, a2
-; RV32-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x36, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 54 * vlenb
+; RV32-NEXT: .cfi_escape 0x0f, 0x0e, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0xd2, 0x00, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 82 * vlenb
; RV32-NEXT: addi a3, a1, 256
; RV32-NEXT: li a2, 32
; RV32-NEXT: vsetvli zero, a2, e32, m8, ta, ma
; RV32-NEXT: vle32.v v16, (a3)
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: li a4, 21
+; RV32-NEXT: li a4, 57
; RV32-NEXT: mul a3, a3, a4
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 16
@@ -177,30 +177,27 @@ define {<8 x i64>, <8 x i64>, <8 x i64>, <8 x i64>, <8 x i64>, <8 x i64>} @load_
; RV32-NEXT: vsetivli zero, 16, e32, m4, ta, ma
; RV32-NEXT: vslideup.vi v8, v16, 4
; RV32-NEXT: csrr a4, vlenb
-; RV32-NEXT: slli a5, a4, 3
-; RV32-NEXT: add a4, a5, a4
+; RV32-NEXT: li a5, 41
+; RV32-NEXT: mul a4, a4, a5
; RV32-NEXT: add a4, sp, a4
; RV32-NEXT: addi a4, a4, 16
; RV32-NEXT: vs4r.v v8, (a4) # Unknown-size Folded Spill
; RV32-NEXT: lui a4, 12
-; RV32-NEXT: vmv.s.x v0, a4
-; RV32-NEXT: csrr a4, vlenb
-; RV32-NEXT: add a4, sp, a4
-; RV32-NEXT: addi a4, a4, 16
-; RV32-NEXT: vs1r.v v0, (a4) # Unknown-size Folded Spill
+; RV32-NEXT: vmv.s.x v1, a4
; RV32-NEXT: vsetivli zero, 16, e32, m8, ta, ma
; RV32-NEXT: vslidedown.vi v16, v16, 16
; RV32-NEXT: csrr a4, vlenb
-; RV32-NEXT: li a5, 37
-; RV32-NEXT: mul a4, a4, a5
+; RV32-NEXT: slli a5, a4, 6
+; RV32-NEXT: add a4, a5, a4
; RV32-NEXT: add a4, sp, a4
; RV32-NEXT: addi a4, a4, 16
; RV32-NEXT: vs8r.v v16, (a4) # Unknown-size Folded Spill
+; RV32-NEXT: vmv1r.v v0, v1
; RV32-NEXT: vsetivli zero, 16, e32, m4, ta, mu
; RV32-NEXT: vslideup.vi v8, v16, 10, v0.t
; RV32-NEXT: csrr a4, vlenb
-; RV32-NEXT: slli a5, a4, 4
-; RV32-NEXT: add a4, a5, a4
+; RV32-NEXT: li a5, 45
+; RV32-NEXT: mul a4, a4, a5
; RV32-NEXT: add a4, sp, a4
; RV32-NEXT: addi a4, a4, 16
; RV32-NEXT: vs4r.v v8, (a4) # Unknown-size Folded Spill
@@ -209,391 +206,429 @@ define {<8 x i64>, <8 x i64>, <8 x i64>, <8 x i64>, <8 x i64>, <8 x i64>} @load_
; RV32-NEXT: vsetvli zero, a2, e32, m8, ta, mu
; RV32-NEXT: vle16.v v8, (a4)
; RV32-NEXT: csrr a4, vlenb
-; RV32-NEXT: li a5, 13
-; RV32-NEXT: mul a4, a4, a5
+; RV32-NEXT: slli a5, a4, 5
+; RV32-NEXT: add a4, a5, a4
+; RV32-NEXT: add a4, sp, a4
+; RV32-NEXT: addi a4, a4, 16
+; RV32-NEXT: vs4r.v v8, (a4) # Unknown-size Folded Spill
+; RV32-NEXT: lui a4, %hi(.LCPI6_1)
+; RV32-NEXT: addi a4, a4, %lo(.LCPI6_1)
+; RV32-NEXT: lui a5, 1
+; RV32-NEXT: vle16.v v8, (a4)
+; RV32-NEXT: csrr a4, vlenb
+; RV32-NEXT: li a6, 25
+; RV32-NEXT: mul a4, a4, a6
; RV32-NEXT: add a4, sp, a4
; RV32-NEXT: addi a4, a4, 16
; RV32-NEXT: vs4r.v v8, (a4) # Unknown-size Folded Spill
-; RV32-NEXT: vle32.v v24, (a1)
+; RV32-NEXT: vle32.v v8, (a1)
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a4, 45
+; RV32-NEXT: li a4, 73
; RV32-NEXT: mul a1, a1, a4
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
-; RV32-NEXT: vs8r.v v24, (a1) # Unknown-size Folded Spill
-; RV32-NEXT: lui a1, %hi(.LCPI6_1)
-; RV32-NEXT: addi a1, a1, %lo(.LCPI6_1)
-; RV32-NEXT: lui a4, 1
-; RV32-NEXT: addi a4, a4, -64
-; RV32-NEXT: vle16.v v8, (a1)
+; RV32-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: vle32.v v24, (a3)
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: slli a5, a1, 2
-; RV32-NEXT: add a1, a5, a1
+; RV32-NEXT: li a3, 49
+; RV32-NEXT: mul a1, a1, a3
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
-; RV32-NEXT: vs4r.v v8, (a1) # Unknown-size Folded Spill
-; RV32-NEXT: vle32.v v16, (a3)
+; RV32-NEXT: vs8r.v v24, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: addi a1, a5, -64
+; RV32-NEXT: vmv.s.x v0, a1
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a3, 29
+; RV32-NEXT: li a3, 37
; RV32-NEXT: mul a1, a1, a3
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
-; RV32-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
-; RV32-NEXT: vmv.s.x v2, a4
+; RV32-NEXT: vs1r.v v0, (a1) # Unknown-size Folded Spill
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a3, 13
-; RV32-NEXT: mul a1, a1, a3
+; RV32-NEXT: slli a3, a1, 5
+; RV32-NEXT: add a1, a3, a1
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
; RV32-NEXT: vl4r.v v4, (a1) # Unknown-size Folded Reload
-; RV32-NEXT: vrgatherei16.vv v8, v24, v4
-; RV32-NEXT: vmv1r.v v0, v2
+; RV32-NEXT: vrgatherei16.vv v16, v8, v4
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: slli a3, a1, 2
-; RV32-NEXT: add a1, a3, a1
+; RV32-NEXT: li a3, 25
+; RV32-NEXT: mul a1, a1, a3
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
-; RV32-NEXT: vl4r.v v24, (a1) # Unknown-size Folded Reload
-; RV32-NEXT: vrgatherei16.vv v8, v16, v24, v0.t
-; RV32-NEXT: vsetivli zero, 12, e32, m4, tu, ma
+; RV32-NEXT: vl4r.v v8, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vrgatherei16.vv v16, v24, v8, v0.t
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: slli a3, a1, 4
-; RV32-NEXT: add a1, a3, a1
+; RV32-NEXT: li a3, 45
+; RV32-NEXT: mul a1, a1, a3
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
-; RV32-NEXT: vl4r.v v12, (a1) # Unknown-size Folded Reload
-; RV32-NEXT: vmv.v.v v12, v8
+; RV32-NEXT: vl4r.v v8, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 12, e32, m4, tu, ma
+; RV32-NEXT: vmv.v.v v8, v16
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: slli a3, a1, 4
-; RV32-NEXT: add a1, a3, a1
+; RV32-NEXT: li a3, 45
+; RV32-NEXT: mul a1, a1, a3
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
-; RV32-NEXT: vs4r.v v12, (a1) # Unknown-size Folded Spill
-; RV32-NEXT: vsetivli zero, 16, e32, m4, ta, mu
+; RV32-NEXT: vs4r.v v8, (a1) # Unknown-size Folded Spill
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a3, 21
+; RV32-NEXT: li a3, 57
; RV32-NEXT: mul a1, a1, a3
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
; RV32-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
-; RV32-NEXT: vmv4r.v v16, v8
-; RV32-NEXT: vslideup.vi v8, v16, 2
+; RV32-NEXT: vsetivli zero, 16, e32, m4, ta, mu
+; RV32-NEXT: vslideup.vi v12, v8, 2
; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: li a3, 21
+; RV32-NEXT: mul a1, a1, a3
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
-; RV32-NEXT: vl1r.v v3, (a1) # Unknown-size Folded Reload
-; RV32-NEXT: vmv1r.v v0, v3
+; RV32-NEXT: vs1r.v v1, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: vmv1r.v v0, v1
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a3, 37
-; RV32-NEXT: mul a1, a1, a3
+; RV32-NEXT: slli a3, a1, 6
+; RV32-NEXT: add a1, a3, a1
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
; RV32-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
-; RV32-NEXT: vslideup.vi v8, v16, 8, v0.t
-; RV32-NEXT: vmv.v.v v20, v8
+; RV32-NEXT: vslideup.vi v12, v16, 8, v0.t
+; RV32-NEXT: vmv.v.v v20, v12
; RV32-NEXT: lui a1, %hi(.LCPI6_2)
; RV32-NEXT: addi a1, a1, %lo(.LCPI6_2)
-; RV32-NEXT: vsetvli zero, a2, e32, m8, ta, mu
-; RV32-NEXT: vle16.v v8, (a1)
+; RV32-NEXT: lui a3, %hi(.LCPI6_3)
+; RV32-NEXT: addi a3, a3, %lo(.LCPI6_3)
+; RV32-NEXT: lui a4, %hi(.LCPI6_4)
+; RV32-NEXT: vsetvli zero, a2, e16, m4, ta, ma
+; RV32-NEXT: vle16.v v4, (a1)
+; RV32-NEXT: vle16.v v16, (a3)
+; RV32-NEXT: addi a1, a4, %lo(.LCPI6_4)
+; RV32-NEXT: vsetivli zero, 16, e16, m2, ta, ma
+; RV32-NEXT: vle16.v v2, (a1)
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a3, 13
+; RV32-NEXT: li a3, 73
; RV32-NEXT: mul a1, a1, a3
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
-; RV32-NEXT: vs4r.v v8, (a1) # Unknown-size Folded Spill
-; RV32-NEXT: lui a1, %hi(.LCPI6_3)
-; RV32-NEXT: addi a1, a1, %lo(.LCPI6_3)
-; RV32-NEXT: vle16.v v8, (a1)
+; RV32-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vsetvli zero, a2, e32, m8, ta, mu
+; RV32-NEXT: vrgatherei16.vv v24, v8, v4
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: slli a3, a1, 2
-; RV32-NEXT: add a1, a3, a1
+; RV32-NEXT: li a3, 37
+; RV32-NEXT: mul a1, a1, a3
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
-; RV32-NEXT: vs4r.v v8, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: vl1r.v v0, (a1) # Unknown-size Folded Reload
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a3, 45
+; RV32-NEXT: li a3, 49
; RV32-NEXT: mul a1, a1, a3
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
-; RV32-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vrgatherei16.vv v24, v8, v16, v0.t
+; RV32-NEXT: vsetivli zero, 12, e32, m4, tu, ma
+; RV32-NEXT: vmv.v.v v20, v24
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a3, 13
+; RV32-NEXT: li a3, 37
; RV32-NEXT: mul a1, a1, a3
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
-; RV32-NEXT: vl4r.v v4, (a1) # Unknown-size Folded Reload
-; RV32-NEXT: vrgatherei16.vv v8, v24, v4
-; RV32-NEXT: vmv1r.v v0, v2
+; RV32-NEXT: vs4r.v v20, (a1) # Unknown-size Folded Spill
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a3, 29
+; RV32-NEXT: li a3, 57
; RV32-NEXT: mul a1, a1, a3
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
; RV32-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 16, e32, m4, ta, mu
+; RV32-NEXT: vrgatherei16.vv v16, v24, v2
+; RV32-NEXT: vmv1r.v v0, v1
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: slli a3, a1, 2
+; RV32-NEXT: slli a3, a1, 6
; RV32-NEXT: add a1, a3, a1
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
-; RV32-NEXT: vl4r.v v4, (a1) # Unknown-size Folded Reload
-; RV32-NEXT: vrgatherei16.vv v8, v24, v4, v0.t
-; RV32-NEXT: vsetivli zero, 12, e32, m4, tu, ma
-; RV32-NEXT: vmv.v.v v20, v8
+; RV32-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vslideup.vi v16, v8, 6, v0.t
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: slli a3, a1, 2
+; RV32-NEXT: slli a3, a1, 5
; RV32-NEXT: add a1, a3, a1
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
-; RV32-NEXT: vs4r.v v20, (a1) # Unknown-size Folded Spill
-; RV32-NEXT: lui a1, %hi(.LCPI6_4)
-; RV32-NEXT: addi a1, a1, %lo(.LCPI6_4)
-; RV32-NEXT: vsetivli zero, 16, e32, m4, ta, mu
-; RV32-NEXT: vle16.v v8, (a1)
-; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a3, 21
-; RV32-NEXT: mul a1, a1, a3
-; RV32-NEXT: add a1, sp, a1
-; RV32-NEXT: addi a1, a1, 16
-; RV32-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
-; RV32-NEXT: vrgatherei16.vv v12, v24, v8
-; RV32-NEXT: vmv1r.v v0, v3
-; RV32-NEXT: vslideup.vi v12, v16, 6, v0.t
-; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a3, 13
-; RV32-NEXT: mul a1, a1, a3
-; RV32-NEXT: add a1, sp, a1
-; RV32-NEXT: addi a1, a1, 16
-; RV32-NEXT: vs4r.v v12, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: vs4r.v v16, (a1) # Unknown-size Folded Spill
; RV32-NEXT: lui a1, %hi(.LCPI6_5)
; RV32-NEXT: addi a1, a1, %lo(.LCPI6_5)
+; RV32-NEXT: lui a3, %hi(.LCPI6_6)
+; RV32-NEXT: addi a3, a3, %lo(.LCPI6_6)
; RV32-NEXT: vsetvli zero, a2, e32, m8, ta, mu
-; RV32-NEXT: vle16.v v24, (a1)
-; RV32-NEXT: lui a1, %hi(.LCPI6_6)
-; RV32-NEXT: addi a1, a1, %lo(.LCPI6_6)
-; RV32-NEXT: li a3, 960
-; RV32-NEXT: vle16.v v4, (a1)
-; RV32-NEXT: vmv.s.x v0, a3
-; RV32-NEXT: addi a1, sp, 16
-; RV32-NEXT: vs1r.v v0, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: vle16.v v16, (a1)
+; RV32-NEXT: vle16.v v4, (a3)
+; RV32-NEXT: li a1, 960
+; RV32-NEXT: vmv.s.x v0, a1
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a3, 45
+; RV32-NEXT: li a3, 13
; RV32-NEXT: mul a1, a1, a3
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
-; RV32-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
-; RV32-NEXT: vrgatherei16.vv v8, v16, v24
+; RV32-NEXT: vs1r.v v0, (a1) # Unknown-size Folded Spill
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a3, 29
+; RV32-NEXT: li a3, 73
; RV32-NEXT: mul a1, a1, a3
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
; RV32-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
-; RV32-NEXT: vrgatherei16.vv v8, v24, v4, v0.t
-; RV32-NEXT: vsetivli zero, 10, e32, m4, tu, ma
+; RV32-NEXT: vrgatherei16.vv v8, v24, v16
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a3, 13
+; RV32-NEXT: li a3, 49
; RV32-NEXT: mul a1, a1, a3
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
-; RV32-NEXT: vl4r.v v12, (a1) # Unknown-size Folded Reload
-; RV32-NEXT: vmv.v.v v12, v8
+; RV32-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vrgatherei16.vv v8, v16, v4, v0.t
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a3, 13
+; RV32-NEXT: li a3, 25
; RV32-NEXT: mul a1, a1, a3
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
-; RV32-NEXT: vs4r.v v12, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
; RV32-NEXT: lui a1, %hi(.LCPI6_7)
; RV32-NEXT: addi a1, a1, %lo(.LCPI6_7)
-; RV32-NEXT: vsetivli zero, 16, e32, m4, ta, mu
+; RV32-NEXT: lui a3, %hi(.LCPI6_8)
+; RV32-NEXT: addi a3, a3, %lo(.LCPI6_8)
+; RV32-NEXT: lui a4, %hi(.LCPI6_9)
+; RV32-NEXT: vsetivli zero, 16, e16, m2, ta, ma
; RV32-NEXT: vle16.v v8, (a1)
+; RV32-NEXT: addi a1, a4, %lo(.LCPI6_9)
+; RV32-NEXT: vsetvli zero, a2, e16, m4, ta, ma
+; RV32-NEXT: vle16.v v24, (a3)
+; RV32-NEXT: vle16.v v28, (a1)
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a3, 21
+; RV32-NEXT: li a3, 57
; RV32-NEXT: mul a1, a1, a3
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
-; RV32-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
-; RV32-NEXT: vrgatherei16.vv v4, v16, v8
-; RV32-NEXT: vmv1r.v v0, v3
+; RV32-NEXT: vl8r.v v0, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 16, e32, m4, ta, mu
+; RV32-NEXT: vrgatherei16.vv v4, v0, v8
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a3, 37
+; RV32-NEXT: li a3, 21
; RV32-NEXT: mul a1, a1, a3
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
-; RV32-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
-; RV32-NEXT: vslideup.vi v4, v8, 4, v0.t
-; RV32-NEXT: lui a1, %hi(.LCPI6_8)
-; RV32-NEXT: addi a1, a1, %lo(.LCPI6_8)
-; RV32-NEXT: vsetvli zero, a2, e32, m8, ta, mu
-; RV32-NEXT: vle16.v v0, (a1)
-; RV32-NEXT: lui a1, %hi(.LCPI6_9)
-; RV32-NEXT: addi a1, a1, %lo(.LCPI6_9)
-; RV32-NEXT: vle16.v v8, (a1)
+; RV32-NEXT: vl1r.v v0, (a1) # Unknown-size Folded Reload
; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a3, a1, 6
+; RV32-NEXT: add a1, a3, a1
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
-; RV32-NEXT: vs4r.v v8, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vslideup.vi v4, v8, 4, v0.t
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a3, 45
+; RV32-NEXT: li a3, 21
; RV32-NEXT: mul a1, a1, a3
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
-; RV32-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
-; RV32-NEXT: vrgatherei16.vv v8, v16, v0
-; RV32-NEXT: addi a1, sp, 16
-; RV32-NEXT: vl1r.v v0, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vs4r.v v4, (a1) # Unknown-size Folded Spill
; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: li a3, 73
+; RV32-NEXT: mul a1, a1, a3
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
-; RV32-NEXT: vl4r.v v16, (a1) # Unknown-size Folded Reload
-; RV32-NEXT: vrgatherei16.vv v8, v24, v16, v0.t
-; RV32-NEXT: vsetivli zero, 10, e32, m4, tu, ma
-; RV32-NEXT: vmv.v.v v4, v8
+; RV32-NEXT: vl8r.v v0, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vsetvli zero, a2, e32, m8, ta, mu
+; RV32-NEXT: vrgatherei16.vv v8, v0, v24
; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: li a3, 13
+; RV32-NEXT: mul a1, a1, a3
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
-; RV32-NEXT: vs4r.v v4, (a1) # Unknown-size Folded Spill
-; RV32-NEXT: vsetivli zero, 16, e32, m4, ta, mu
+; RV32-NEXT: vl1r.v v0, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vrgatherei16.vv v8, v16, v28, v0.t
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a3, 21
+; RV32-NEXT: li a3, 13
; RV32-NEXT: mul a1, a1, a3
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
-; RV32-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
-; RV32-NEXT: vslideup.vi v12, v8, 6
+; RV32-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
; RV32-NEXT: lui a1, %hi(.LCPI6_10)
; RV32-NEXT: addi a1, a1, %lo(.LCPI6_10)
+; RV32-NEXT: vsetivli zero, 16, e32, m4, ta, mu
; RV32-NEXT: vle16.v v8, (a1)
; RV32-NEXT: lui a1, 15
-; RV32-NEXT: vmv.s.x v24, a1
-; RV32-NEXT: vmv1r.v v0, v24
+; RV32-NEXT: vmv.s.x v3, a1
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a3, 37
+; RV32-NEXT: li a3, 57
; RV32-NEXT: mul a1, a1, a3
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
; RV32-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vslideup.vi v12, v16, 6
+; RV32-NEXT: vmv1r.v v0, v3
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a3, a1, 6
+; RV32-NEXT: add a1, a3, a1
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 16
+; RV32-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
; RV32-NEXT: vrgatherei16.vv v12, v16, v8, v0.t
-; RV32-NEXT: vmv.v.v v28, v12
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: li a3, 57
+; RV32-NEXT: mul a1, a1, a3
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 16
+; RV32-NEXT: vs4r.v v12, (a1) # Unknown-size Folded Spill
; RV32-NEXT: lui a1, %hi(.LCPI6_11)
; RV32-NEXT: addi a1, a1, %lo(.LCPI6_11)
+; RV32-NEXT: lui a3, %hi(.LCPI6_12)
+; RV32-NEXT: addi a3, a3, %lo(.LCPI6_12)
; RV32-NEXT: vsetvli zero, a2, e32, m8, ta, mu
-; RV32-NEXT: vle16.v v0, (a1)
-; RV32-NEXT: lui a1, %hi(.LCPI6_12)
-; RV32-NEXT: addi a1, a1, %lo(.LCPI6_12)
-; RV32-NEXT: li a3, 1008
-; RV32-NEXT: vle16.v v4, (a1)
-; RV32-NEXT: vmv.s.x v25, a3
-; RV32-NEXT: addi a1, sp, 16
-; RV32-NEXT: vs1r.v v25, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: vle16.v v8, (a1)
+; RV32-NEXT: vle16.v v12, (a3)
+; RV32-NEXT: li a1, 1008
+; RV32-NEXT: vmv.s.x v0, a1
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a3, 45
+; RV32-NEXT: slli a1, a1, 2
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 16
+; RV32-NEXT: vs1r.v v0, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: li a3, 73
; RV32-NEXT: mul a1, a1, a3
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
; RV32-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
-; RV32-NEXT: vrgatherei16.vv v8, v16, v0
-; RV32-NEXT: vmv1r.v v0, v25
+; RV32-NEXT: vrgatherei16.vv v24, v16, v8
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a3, 29
+; RV32-NEXT: li a3, 49
; RV32-NEXT: mul a1, a1, a3
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
; RV32-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
-; RV32-NEXT: vrgatherei16.vv v8, v16, v4, v0.t
-; RV32-NEXT: vsetivli zero, 10, e32, m4, tu, ma
-; RV32-NEXT: vmv.v.v v28, v8
+; RV32-NEXT: vrgatherei16.vv v24, v16, v12, v0.t
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a3, 21
-; RV32-NEXT: mul a1, a1, a3
+; RV32-NEXT: slli a3, a1, 2
+; RV32-NEXT: add a1, a3, a1
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
-; RV32-NEXT: vs4r.v v28, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: vs8r.v v24, (a1) # Unknown-size Folded Spill
; RV32-NEXT: lui a1, %hi(.LCPI6_13)
; RV32-NEXT: addi a1, a1, %lo(.LCPI6_13)
-; RV32-NEXT: vsetivli zero, 16, e32, m4, ta, mu
+; RV32-NEXT: lui a3, %hi(.LCPI6_14)
+; RV32-NEXT: addi a3, a3, %lo(.LCPI6_14)
+; RV32-NEXT: lui a4, %hi(.LCPI6_15)
+; RV32-NEXT: vsetivli zero, 16, e16, m2, ta, ma
+; RV32-NEXT: vle16.v v20, (a1)
+; RV32-NEXT: addi a1, a4, %lo(.LCPI6_15)
+; RV32-NEXT: vsetvli zero, a2, e16, m4, ta, ma
+; RV32-NEXT: vle16.v v24, (a3)
; RV32-NEXT: vle16.v v8, (a1)
-; RV32-NEXT: vmv1r.v v0, v24
+; RV32-NEXT: addi a1, sp, 16
+; RV32-NEXT: vs4r.v v8, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: vmv1r.v v0, v3
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: li a3, 41
+; RV32-NEXT: mul a1, a1, a3
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 16
+; RV32-NEXT: vl4r.v v16, (a1) # Unknown-size Folded Reload
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: slli a3, a1, 3
+; RV32-NEXT: slli a3, a1, 6
; RV32-NEXT: add a1, a3, a1
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
-; RV32-NEXT: vl4r.v v24, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 16, e32, m4, ta, mu
+; RV32-NEXT: vrgatherei16.vv v16, v8, v20, v0.t
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a3, 37
+; RV32-NEXT: slli a3, a1, 5
+; RV32-NEXT: add a1, a3, a1
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 16
+; RV32-NEXT: vl4r.v v20, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: li a3, 25
; RV32-NEXT: mul a1, a1, a3
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
-; RV32-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
-; RV32-NEXT: vrgatherei16.vv v24, v16, v8, v0.t
-; RV32-NEXT: lui a1, %hi(.LCPI6_14)
-; RV32-NEXT: addi a1, a1, %lo(.LCPI6_14)
+; RV32-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 10, e32, m4, tu, ma
+; RV32-NEXT: vmv.v.v v20, v8
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: li a3, 73
+; RV32-NEXT: mul a1, a1, a3
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 16
+; RV32-NEXT: vl8r.v v0, (a1) # Unknown-size Folded Reload
; RV32-NEXT: vsetvli zero, a2, e32, m8, ta, mu
-; RV32-NEXT: vle16.v v16, (a1)
-; RV32-NEXT: lui a1, %hi(.LCPI6_15)
-; RV32-NEXT: addi a1, a1, %lo(.LCPI6_15)
-; RV32-NEXT: vle16.v v28, (a1)
+; RV32-NEXT: vrgatherei16.vv v8, v0, v24
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a2, 45
+; RV32-NEXT: slli a1, a1, 2
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 16
+; RV32-NEXT: vl1r.v v0, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: li a2, 49
; RV32-NEXT: mul a1, a1, a2
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
-; RV32-NEXT: vl8r.v v0, (a1) # Unknown-size Folded Reload
-; RV32-NEXT: vrgatherei16.vv v8, v0, v16
+; RV32-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
; RV32-NEXT: addi a1, sp, 16
-; RV32-NEXT: vl1r.v v0, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vl4r.v v4, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vrgatherei16.vv v8, v24, v4, v0.t
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a2, 29
+; RV32-NEXT: li a2, 21
; RV32-NEXT: mul a1, a1, a2
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
-; RV32-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
-; RV32-NEXT: vrgatherei16.vv v8, v16, v28, v0.t
+; RV32-NEXT: vl4r.v v24, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: li a2, 13
+; RV32-NEXT: mul a1, a1, a2
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 16
+; RV32-NEXT: vl8r.v v0, (a1) # Unknown-size Folded Reload
; RV32-NEXT: vsetivli zero, 10, e32, m4, tu, ma
-; RV32-NEXT: vmv.v.v v24, v8
+; RV32-NEXT: vmv.v.v v24, v0
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: li a2, 57
+; RV32-NEXT: mul a1, a1, a2
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 16
+; RV32-NEXT: vl4r.v v28, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: slli a2, a1, 2
+; RV32-NEXT: add a1, a2, a1
+; RV32-NEXT: add a1, sp, a1
+; RV32-NEXT: addi a1, a1, 16
+; RV32-NEXT: vl8r.v v0, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vmv.v.v v28, v0
+; RV32-NEXT: vmv.v.v v16, v8
; RV32-NEXT: addi a1, a0, 320
; RV32-NEXT: vsetivli zero, 16, e32, m4, ta, ma
-; RV32-NEXT: vse32.v v24, (a1)
+; RV32-NEXT: vse32.v v16, (a1)
; RV32-NEXT: addi a1, a0, 256
-; RV32-NEXT: csrr a2, vlenb
-; RV32-NEXT: li a3, 21
-; RV32-NEXT: mul a2, a2, a3
-; RV32-NEXT: add a2, sp, a2
-; RV32-NEXT: addi a2, a2, 16
-; RV32-NEXT: vl4r.v v8, (a2) # Unknown-size Folded Reload
-; RV32-NEXT: vse32.v v8, (a1)
+; RV32-NEXT: vse32.v v28, (a1)
; RV32-NEXT: addi a1, a0, 192
-; RV32-NEXT: csrr a2, vlenb
-; RV32-NEXT: add a2, sp, a2
-; RV32-NEXT: addi a2, a2, 16
-; RV32-NEXT: vl4r.v v8, (a2) # Unknown-size Folded Reload
-; RV32-NEXT: vse32.v v8, (a1)
+; RV32-NEXT: vse32.v v24, (a1)
; RV32-NEXT: addi a1, a0, 128
-; RV32-NEXT: csrr a2, vlenb
-; RV32-NEXT: li a3, 13
-; RV32-NEXT: mul a2, a2, a3
-; RV32-NEXT: add a2, sp, a2
-; RV32-NEXT: addi a2, a2, 16
-; RV32-NEXT: vl4r.v v8, (a2) # Unknown-size Folded Reload
-; RV32-NEXT: vse32.v v8, (a1)
+; RV32-NEXT: vse32.v v20, (a1)
; RV32-NEXT: addi a1, a0, 64
; RV32-NEXT: csrr a2, vlenb
-; RV32-NEXT: slli a3, a2, 2
-; RV32-NEXT: add a2, a3, a2
+; RV32-NEXT: li a3, 37
+; RV32-NEXT: mul a2, a2, a3
; RV32-NEXT: add a2, sp, a2
; RV32-NEXT: addi a2, a2, 16
; RV32-NEXT: vl4r.v v8, (a2) # Unknown-size Folded Reload
; RV32-NEXT: vse32.v v8, (a1)
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: slli a2, a1, 4
-; RV32-NEXT: add a1, a2, a1
+; RV32-NEXT: li a2, 45
+; RV32-NEXT: mul a1, a1, a2
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
; RV32-NEXT: vl4r.v v8, (a1) # Unknown-size Folded Reload
; RV32-NEXT: vse32.v v8, (a0)
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: li a1, 54
+; RV32-NEXT: li a1, 82
; RV32-NEXT: mul a0, a0, a1
; RV32-NEXT: add sp, sp, a0
; RV32-NEXT: addi sp, sp, 16
@@ -604,372 +639,422 @@ define {<8 x i64>, <8 x i64>, <8 x i64>, <8 x i64>, <8 x i64>, <8 x i64>} @load_
; RV64-NEXT: addi sp, sp, -16
; RV64-NEXT: .cfi_def_cfa_offset 16
; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: li a3, 56
+; RV64-NEXT: li a3, 74
; RV64-NEXT: mul a2, a2, a3
; RV64-NEXT: sub sp, sp, a2
-; RV64-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x38, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 56 * vlenb
-; RV64-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; RV64-NEXT: .cfi_escape 0x0f, 0x0e, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0xca, 0x00, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 74 * vlenb
; RV64-NEXT: addi a2, a1, 256
+; RV64-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV64-NEXT: vle64.v v16, (a2)
; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: slli a2, a2, 5
+; RV64-NEXT: li a3, 25
+; RV64-NEXT: mul a2, a2, a3
; RV64-NEXT: add a2, sp, a2
; RV64-NEXT: addi a2, a2, 16
; RV64-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
; RV64-NEXT: addi a2, a1, 128
-; RV64-NEXT: vle64.v v8, (a2)
-; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: li a3, 40
-; RV64-NEXT: mul a2, a2, a3
-; RV64-NEXT: add a2, sp, a2
-; RV64-NEXT: addi a2, a2, 16
-; RV64-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
-; RV64-NEXT: vle64.v v24, (a1)
+; RV64-NEXT: vle64.v v8, (a1)
+; RV64-NEXT: csrr a1, vlenb
+; RV64-NEXT: slli a3, a1, 6
+; RV64-NEXT: add a1, a3, a1
+; RV64-NEXT: add a1, sp, a1
+; RV64-NEXT: addi a1, a1, 16
+; RV64-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
; RV64-NEXT: vsetivli zero, 8, e64, m4, ta, ma
-; RV64-NEXT: vrgather.vi v8, v16, 4
+; RV64-NEXT: vrgather.vi v12, v16, 4
; RV64-NEXT: li a1, 128
-; RV64-NEXT: vmv.s.x v4, a1
+; RV64-NEXT: vmv.s.x v8, a1
; RV64-NEXT: vsetivli zero, 8, e64, m8, ta, ma
; RV64-NEXT: vslidedown.vi v16, v16, 8
; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: li a2, 24
-; RV64-NEXT: mul a1, a1, a2
+; RV64-NEXT: li a3, 49
+; RV64-NEXT: mul a1, a1, a3
; RV64-NEXT: add a1, sp, a1
; RV64-NEXT: addi a1, a1, 16
; RV64-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
+; RV64-NEXT: vmv1r.v v0, v8
; RV64-NEXT: vsetivli zero, 8, e64, m4, ta, mu
-; RV64-NEXT: vmv1r.v v0, v4
-; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: li a2, 20
-; RV64-NEXT: mul a1, a1, a2
-; RV64-NEXT: add a1, sp, a1
-; RV64-NEXT: addi a1, a1, 16
-; RV64-NEXT: vs1r.v v4, (a1) # Unknown-size Folded Spill
-; RV64-NEXT: vrgather.vi v8, v16, 2, v0.t
-; RV64-NEXT: vmv.v.v v20, v8
+; RV64-NEXT: vrgather.vi v12, v16, 2, v0.t
; RV64-NEXT: vsetivli zero, 16, e16, m2, ta, ma
+; RV64-NEXT: vid.v v10
; RV64-NEXT: li a1, 6
-; RV64-NEXT: vid.v v8
-; RV64-NEXT: vmul.vx v6, v8, a1
-; RV64-NEXT: vsetvli zero, zero, e64, m8, ta, ma
-; RV64-NEXT: vrgatherei16.vv v8, v24, v6
+; RV64-NEXT: vmul.vx v2, v10, a1
+; RV64-NEXT: li a1, 56
+; RV64-NEXT: vle64.v v16, (a2)
+; RV64-NEXT: csrr a2, vlenb
+; RV64-NEXT: li a3, 57
+; RV64-NEXT: mul a2, a2, a3
+; RV64-NEXT: add a2, sp, a2
+; RV64-NEXT: addi a2, a2, 16
+; RV64-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
+; RV64-NEXT: vmv.s.x v7, a1
+; RV64-NEXT: vadd.vi v10, v2, -16
; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: li a2, 48
-; RV64-NEXT: mul a1, a1, a2
+; RV64-NEXT: slli a2, a1, 6
+; RV64-NEXT: add a1, a2, a1
; RV64-NEXT: add a1, sp, a1
; RV64-NEXT: addi a1, a1, 16
-; RV64-NEXT: vs8r.v v24, (a1) # Unknown-size Folded Spill
-; RV64-NEXT: vsetvli zero, zero, e16, m2, ta, ma
-; RV64-NEXT: li a1, 56
-; RV64-NEXT: vmv.s.x v5, a1
-; RV64-NEXT: vadd.vi v16, v6, -16
+; RV64-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
; RV64-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; RV64-NEXT: vmv1r.v v0, v5
+; RV64-NEXT: vrgatherei16.vv v16, v24, v2
+; RV64-NEXT: vmv1r.v v0, v7
; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: li a2, 40
+; RV64-NEXT: li a2, 57
; RV64-NEXT: mul a1, a1, a2
; RV64-NEXT: add a1, sp, a1
; RV64-NEXT: addi a1, a1, 16
; RV64-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
-; RV64-NEXT: vrgatherei16.vv v8, v24, v16, v0.t
+; RV64-NEXT: vrgatherei16.vv v16, v24, v10, v0.t
; RV64-NEXT: vsetivli zero, 6, e64, m4, tu, ma
-; RV64-NEXT: vmv.v.v v20, v8
+; RV64-NEXT: vmv.v.v v12, v16
; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: slli a1, a1, 4
+; RV64-NEXT: li a2, 21
+; RV64-NEXT: mul a1, a1, a2
; RV64-NEXT: add a1, sp, a1
; RV64-NEXT: addi a1, a1, 16
-; RV64-NEXT: vs4r.v v20, (a1) # Unknown-size Folded Spill
-; RV64-NEXT: vsetivli zero, 8, e64, m4, ta, mu
+; RV64-NEXT: vs4r.v v12, (a1) # Unknown-size Folded Spill
; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: slli a1, a1, 5
+; RV64-NEXT: li a2, 25
+; RV64-NEXT: mul a1, a1, a2
; RV64-NEXT: add a1, sp, a1
; RV64-NEXT: addi a1, a1, 16
; RV64-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
-; RV64-NEXT: vrgather.vi v24, v16, 5
-; RV64-NEXT: vmv1r.v v0, v4
+; RV64-NEXT: vsetivli zero, 8, e64, m4, ta, mu
+; RV64-NEXT: vrgather.vi v12, v16, 5
+; RV64-NEXT: vmv1r.v v0, v8
+; RV64-NEXT: vmv1r.v v6, v8
; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: li a2, 24
+; RV64-NEXT: li a2, 49
; RV64-NEXT: mul a1, a1, a2
; RV64-NEXT: add a1, sp, a1
; RV64-NEXT: addi a1, a1, 16
; RV64-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
-; RV64-NEXT: vrgather.vi v24, v16, 3, v0.t
+; RV64-NEXT: vrgather.vi v12, v16, 3, v0.t
+; RV64-NEXT: vmv.v.v v28, v12
; RV64-NEXT: vsetivli zero, 16, e16, m2, ta, ma
-; RV64-NEXT: vadd.vi v28, v6, 1
-; RV64-NEXT: vsetvli zero, zero, e64, m8, ta, ma
+; RV64-NEXT: vadd.vi v24, v2, 1
+; RV64-NEXT: vadd.vi v26, v2, -15
; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: li a2, 48
-; RV64-NEXT: mul a1, a1, a2
+; RV64-NEXT: slli a2, a1, 6
+; RV64-NEXT: add a1, a2, a1
; RV64-NEXT: add a1, sp, a1
; RV64-NEXT: addi a1, a1, 16
-; RV64-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
-; RV64-NEXT: vrgatherei16.vv v8, v16, v28
-; RV64-NEXT: vsetvli zero, zero, e16, m2, ta, ma
-; RV64-NEXT: vadd.vi v28, v6, -15
+; RV64-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
; RV64-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; RV64-NEXT: vmv1r.v v0, v5
+; RV64-NEXT: vrgatherei16.vv v16, v8, v24
+; RV64-NEXT: vmv1r.v v0, v7
; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: li a2, 40
+; RV64-NEXT: li a2, 57
; RV64-NEXT: mul a1, a1, a2
; RV64-NEXT: add a1, sp, a1
; RV64-NEXT: addi a1, a1, 16
-; RV64-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
-; RV64-NEXT: vrgatherei16.vv v8, v16, v28, v0.t
+; RV64-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
+; RV64-NEXT: vrgatherei16.vv v16, v8, v26, v0.t
; RV64-NEXT: vsetivli zero, 6, e64, m4, tu, ma
-; RV64-NEXT: vmv.v.v v24, v8
+; RV64-NEXT: vmv.v.v v28, v16
; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: li a2, 12
+; RV64-NEXT: slli a2, a1, 4
+; RV64-NEXT: add a1, a2, a1
+; RV64-NEXT: add a1, sp, a1
+; RV64-NEXT: addi a1, a1, 16
+; RV64-NEXT: vs4r.v v28, (a1) # Unknown-size Folded Spill
+; RV64-NEXT: lui a1, 16
+; RV64-NEXT: addi a1, a1, 7
+; RV64-NEXT: vsetivli zero, 4, e32, m1, ta, ma
+; RV64-NEXT: vmv.v.i v9, 6
+; RV64-NEXT: vmv.v.x v10, a1
+; RV64-NEXT: csrr a1, vlenb
+; RV64-NEXT: li a2, 25
; RV64-NEXT: mul a1, a1, a2
; RV64-NEXT: add a1, sp, a1
; RV64-NEXT: addi a1, a1, 16
-; RV64-NEXT: vs4r.v v24, (a1) # Unknown-size Folded Spill
-; RV64-NEXT: vsetivli zero, 16, e16, m2, ta, ma
-; RV64-NEXT: vmv2r.v v26, v6
-; RV64-NEXT: vadd.vi v24, v6, 2
-; RV64-NEXT: vsetvli zero, zero, e64, m8, ta, ma
+; RV64-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
+; RV64-NEXT: vsetivli zero, 8, e64, m4, ta, ma
+; RV64-NEXT: vrgatherei16.vv v12, v16, v9
; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: li a2, 48
+; RV64-NEXT: li a2, 45
; RV64-NEXT: mul a1, a1, a2
; RV64-NEXT: add a1, sp, a1
; RV64-NEXT: addi a1, a1, 16
-; RV64-NEXT: vl8r.v v0, (a1) # Unknown-size Folded Reload
-; RV64-NEXT: vrgatherei16.vv v8, v0, v24
-; RV64-NEXT: vsetvli zero, zero, e16, m2, ta, ma
+; RV64-NEXT: vs4r.v v12, (a1) # Unknown-size Folded Spill
+; RV64-NEXT: vrgatherei16.vv v12, v16, v10
+; RV64-NEXT: csrr a1, vlenb
+; RV64-NEXT: li a2, 41
+; RV64-NEXT: mul a1, a1, a2
+; RV64-NEXT: add a1, sp, a1
+; RV64-NEXT: addi a1, a1, 16
+; RV64-NEXT: vs4r.v v12, (a1) # Unknown-size Folded Spill
+; RV64-NEXT: vmv4r.v v8, v16
+; RV64-NEXT: vrgather.vi v12, v16, 2
+; RV64-NEXT: csrr a1, vlenb
+; RV64-NEXT: li a2, 37
+; RV64-NEXT: mul a1, a1, a2
+; RV64-NEXT: add a1, sp, a1
+; RV64-NEXT: addi a1, a1, 16
+; RV64-NEXT: vs4r.v v12, (a1) # Unknown-size Folded Spill
+; RV64-NEXT: vrgather.vi v12, v16, 3
+; RV64-NEXT: csrr a1, vlenb
+; RV64-NEXT: slli a2, a1, 5
+; RV64-NEXT: add a1, a2, a1
+; RV64-NEXT: add a1, sp, a1
+; RV64-NEXT: addi a1, a1, 16
+; RV64-NEXT: vs4r.v v12, (a1) # Unknown-size Folded Spill
; RV64-NEXT: li a1, 24
-; RV64-NEXT: vmv.s.x v0, a1
+; RV64-NEXT: vmv.s.x v1, a1
+; RV64-NEXT: vsetivli zero, 16, e16, m2, ta, ma
+; RV64-NEXT: vadd.vi v24, v2, 2
+; RV64-NEXT: vadd.vi v4, v2, -14
; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: slli a1, a1, 1
+; RV64-NEXT: slli a2, a1, 6
+; RV64-NEXT: add a1, a2, a1
; RV64-NEXT: add a1, sp, a1
; RV64-NEXT: addi a1, a1, 16
-; RV64-NEXT: vs1r.v v0, (a1) # Unknown-size Folded Spill
-; RV64-NEXT: vadd.vi v24, v26, -14
-; RV64-NEXT: vmv2r.v v6, v26
+; RV64-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
; RV64-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; RV64-NEXT: vrgatherei16.vv v8, v16, v24, v0.t
-; RV64-NEXT: vsetivli zero, 4, e32, m1, ta, ma
-; RV64-NEXT: vmv.v.i v12, 6
-; RV64-NEXT: vsetivli zero, 8, e64, m4, ta, mu
+; RV64-NEXT: vrgatherei16.vv v8, v16, v24
+; RV64-NEXT: vmv1r.v v0, v1
; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: slli a1, a1, 5
+; RV64-NEXT: li a2, 57
+; RV64-NEXT: mul a1, a1, a2
; RV64-NEXT: add a1, sp, a1
; RV64-NEXT: addi a1, a1, 16
; RV64-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
-; RV64-NEXT: vrgatherei16.vv v20, v24, v12
+; RV64-NEXT: vrgatherei16.vv v8, v24, v4, v0.t
; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: li a2, 20
+; RV64-NEXT: li a2, 25
; RV64-NEXT: mul a1, a1, a2
; RV64-NEXT: add a1, sp, a1
; RV64-NEXT: addi a1, a1, 16
-; RV64-NEXT: vl1r.v v0, (a1) # Unknown-size Folded Reload
+; RV64-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
+; RV64-NEXT: vmv1r.v v0, v6
; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: li a2, 24
+; RV64-NEXT: li a2, 49
; RV64-NEXT: mul a1, a1, a2
; RV64-NEXT: add a1, sp, a1
; RV64-NEXT: addi a1, a1, 16
-; RV64-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
-; RV64-NEXT: vrgather.vi v20, v24, 4, v0.t
-; RV64-NEXT: vsetivli zero, 5, e64, m4, tu, ma
-; RV64-NEXT: vmv.v.v v20, v8
+; RV64-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: slli a1, a1, 3
+; RV64-NEXT: li a2, 45
+; RV64-NEXT: mul a1, a1, a2
; RV64-NEXT: add a1, sp, a1
; RV64-NEXT: addi a1, a1, 16
-; RV64-NEXT: vs4r.v v20, (a1) # Unknown-size Folded Spill
-; RV64-NEXT: vsetivli zero, 16, e16, m2, ta, ma
-; RV64-NEXT: vmv2r.v v10, v6
+; RV64-NEXT: vl4r.v v20, (a1) # Unknown-size Folded Reload
+; RV64-NEXT: vsetivli zero, 8, e64, m4, ta, mu
+; RV64-NEXT: vrgather.vi v20, v16, 4, v0.t
; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: li a2, 6
+; RV64-NEXT: li a2, 45
; RV64-NEXT: mul a1, a1, a2
; RV64-NEXT: add a1, sp, a1
; RV64-NEXT: addi a1, a1, 16
-; RV64-NEXT: vs2r.v v6, (a1) # Unknown-size Folded Spill
-; RV64-NEXT: vadd.vi v8, v6, 3
-; RV64-NEXT: vsetvli zero, zero, e64, m8, ta, ma
+; RV64-NEXT: vs4r.v v20, (a1) # Unknown-size Folded Spill
+; RV64-NEXT: vsetivli zero, 16, e16, m2, ta, ma
+; RV64-NEXT: vadd.vi v4, v2, 3
+; RV64-NEXT: vadd.vi v8, v2, -13
; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: li a2, 48
-; RV64-NEXT: mul a1, a1, a2
; RV64-NEXT: add a1, sp, a1
; RV64-NEXT: addi a1, a1, 16
-; RV64-NEXT: vl8r.v v0, (a1) # Unknown-size Folded Reload
-; RV64-NEXT: vrgatherei16.vv v16, v0, v8
-; RV64-NEXT: vsetvli zero, zero, e16, m2, ta, ma
-; RV64-NEXT: vadd.vi v28, v10, -13
-; RV64-NEXT: vsetvli zero, zero, e64, m8, ta, mu
+; RV64-NEXT: vs2r.v v8, (a1) # Unknown-size Folded Spill
; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: slli a1, a1, 1
+; RV64-NEXT: slli a2, a1, 6
+; RV64-NEXT: add a1, a2, a1
; RV64-NEXT: add a1, sp, a1
; RV64-NEXT: addi a1, a1, 16
-; RV64-NEXT: vl1r.v v0, (a1) # Unknown-size Folded Reload
+; RV64-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
+; RV64-NEXT: vsetvli zero, zero, e64, m8, ta, mu
+; RV64-NEXT: vrgatherei16.vv v8, v16, v4
+; RV64-NEXT: vmv1r.v v0, v1
; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: li a2, 40
-; RV64-NEXT: mul a1, a1, a2
; RV64-NEXT: add a1, sp, a1
; RV64-NEXT: addi a1, a1, 16
-; RV64-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
-; RV64-NEXT: vrgatherei16.vv v16, v8, v28, v0.t
-; RV64-NEXT: lui a1, 16
-; RV64-NEXT: addi a1, a1, 7
-; RV64-NEXT: vsetivli zero, 4, e32, m1, ta, ma
-; RV64-NEXT: vmv.v.x v12, a1
-; RV64-NEXT: vsetivli zero, 8, e64, m4, ta, mu
+; RV64-NEXT: vl2r.v v16, (a1) # Unknown-size Folded Reload
+; RV64-NEXT: vrgatherei16.vv v8, v24, v16, v0.t
; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: slli a1, a1, 5
+; RV64-NEXT: slli a2, a1, 3
+; RV64-NEXT: add a1, a2, a1
; RV64-NEXT: add a1, sp, a1
; RV64-NEXT: addi a1, a1, 16
-; RV64-NEXT: vl8r.v v0, (a1) # Unknown-size Folded Reload
-; RV64-NEXT: vmv4r.v v8, v0
-; RV64-NEXT: vrgatherei16.vv v20, v0, v12
+; RV64-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
+; RV64-NEXT: vmv1r.v v0, v6
; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: li a2, 20
+; RV64-NEXT: li a2, 49
; RV64-NEXT: mul a1, a1, a2
; RV64-NEXT: add a1, sp, a1
; RV64-NEXT: addi a1, a1, 16
-; RV64-NEXT: vl1r.v v0, (a1) # Unknown-size Folded Reload
-; RV64-NEXT: vrgather.vi v20, v24, 5, v0.t
-; RV64-NEXT: vsetivli zero, 5, e64, m4, tu, ma
-; RV64-NEXT: vmv.v.v v20, v16
+; RV64-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: li a2, 20
+; RV64-NEXT: li a2, 41
; RV64-NEXT: mul a1, a1, a2
; RV64-NEXT: add a1, sp, a1
; RV64-NEXT: addi a1, a1, 16
-; RV64-NEXT: vs4r.v v20, (a1) # Unknown-size Folded Spill
-; RV64-NEXT: lui a1, 96
-; RV64-NEXT: vsetivli zero, 4, e32, m1, ta, ma
-; RV64-NEXT: vmv.v.x v12, a1
+; RV64-NEXT: vl4r.v v8, (a1) # Unknown-size Folded Reload
; RV64-NEXT: vsetivli zero, 8, e64, m4, ta, mu
-; RV64-NEXT: li a1, 192
-; RV64-NEXT: vmv.s.x v0, a1
+; RV64-NEXT: vrgather.vi v8, v24, 5, v0.t
; RV64-NEXT: csrr a1, vlenb
+; RV64-NEXT: li a2, 41
+; RV64-NEXT: mul a1, a1, a2
; RV64-NEXT: add a1, sp, a1
; RV64-NEXT: addi a1, a1, 16
-; RV64-NEXT: vs1r.v v0, (a1) # Unknown-size Folded Spill
-; RV64-NEXT: vrgather.vi v28, v8, 2
-; RV64-NEXT: vrgatherei16.vv v28, v24, v12, v0.t
-; RV64-NEXT: vsetivli zero, 16, e16, m2, ta, ma
+; RV64-NEXT: vs4r.v v8, (a1) # Unknown-size Folded Spill
+; RV64-NEXT: lui a1, 96
+; RV64-NEXT: li a2, 192
+; RV64-NEXT: vmv.s.x v28, a2
+; RV64-NEXT: vsetivli zero, 4, e32, m1, ta, ma
+; RV64-NEXT: vmv.v.x v8, a1
+; RV64-NEXT: vmv1r.v v0, v28
; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: li a2, 6
+; RV64-NEXT: li a2, 37
; RV64-NEXT: mul a1, a1, a2
; RV64-NEXT: add a1, sp, a1
; RV64-NEXT: addi a1, a1, 16
-; RV64-NEXT: vl2r.v v24, (a1) # Unknown-size Folded Reload
-; RV64-NEXT: vadd.vi v16, v24, 4
-; RV64-NEXT: vsetvli zero, zero, e64, m8, ta, ma
+; RV64-NEXT: vl4r.v v12, (a1) # Unknown-size Folded Reload
+; RV64-NEXT: vsetivli zero, 8, e64, m4, ta, mu
+; RV64-NEXT: vrgatherei16.vv v12, v24, v8, v0.t
; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: li a2, 48
+; RV64-NEXT: li a2, 37
; RV64-NEXT: mul a1, a1, a2
; RV64-NEXT: add a1, sp, a1
; RV64-NEXT: addi a1, a1, 16
-; RV64-NEXT: vl8r.v v0, (a1) # Unknown-size Folded Reload
-; RV64-NEXT: vrgatherei16.vv v8, v0, v16
-; RV64-NEXT: vsetvli zero, zero, e16, m2, ta, ma
+; RV64-NEXT: vs4r.v v12, (a1) # Unknown-size Folded Spill
; RV64-NEXT: li a1, 28
; RV64-NEXT: vmv.s.x v0, a1
; RV64-NEXT: addi a1, sp, 16
; RV64-NEXT: vs1r.v v0, (a1) # Unknown-size Folded Spill
-; RV64-NEXT: vadd.vi v26, v24, -12
+; RV64-NEXT: vsetivli zero, 16, e16, m2, ta, ma
+; RV64-NEXT: vadd.vi v30, v2, 4
+; RV64-NEXT: vadd.vi v6, v2, -12
+; RV64-NEXT: csrr a1, vlenb
+; RV64-NEXT: slli a2, a1, 6
+; RV64-NEXT: add a1, a2, a1
+; RV64-NEXT: add a1, sp, a1
+; RV64-NEXT: addi a1, a1, 16
+; RV64-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
; RV64-NEXT: vsetvli zero, zero, e64, m8, ta, mu
+; RV64-NEXT: vrgatherei16.vv v16, v8, v30
; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: li a2, 40
+; RV64-NEXT: li a2, 57
; RV64-NEXT: mul a1, a1, a2
; RV64-NEXT: add a1, sp, a1
; RV64-NEXT: addi a1, a1, 16
-; RV64-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
-; RV64-NEXT: vrgatherei16.vv v8, v16, v26, v0.t
-; RV64-NEXT: vsetivli zero, 5, e64, m4, tu, ma
-; RV64-NEXT: vmv.v.v v28, v8
+; RV64-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
+; RV64-NEXT: vrgatherei16.vv v16, v8, v6, v0.t
; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: slli a1, a1, 1
; RV64-NEXT: add a1, sp, a1
; RV64-NEXT: addi a1, a1, 16
-; RV64-NEXT: vs4r.v v28, (a1) # Unknown-size Folded Spill
+; RV64-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
; RV64-NEXT: lui a1, 112
; RV64-NEXT: addi a1, a1, 1
; RV64-NEXT: vsetivli zero, 4, e32, m1, ta, ma
; RV64-NEXT: vmv.v.x v12, a1
+; RV64-NEXT: vmv1r.v v0, v28
+; RV64-NEXT: csrr a1, vlenb
+; RV64-NEXT: slli a2, a1, 5
+; RV64-NEXT: add a1, a2, a1
+; RV64-NEXT: add a1, sp, a1
+; RV64-NEXT: addi a1, a1, 16
+; RV64-NEXT: vl4r.v v16, (a1) # Unknown-size Folded Reload
; RV64-NEXT: vsetivli zero, 8, e64, m4, ta, mu
+; RV64-NEXT: vrgatherei16.vv v16, v24, v12, v0.t
; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: slli a1, a1, 5
+; RV64-NEXT: slli a2, a1, 5
+; RV64-NEXT: add a1, a2, a1
; RV64-NEXT: add a1, sp, a1
; RV64-NEXT: addi a1, a1, 16
-; RV64-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
-; RV64-NEXT: vrgather.vi v8, v16, 3
+; RV64-NEXT: vs4r.v v16, (a1) # Unknown-size Folded Spill
; RV64-NEXT: csrr a1, vlenb
+; RV64-NEXT: li a2, 45
+; RV64-NEXT: mul a1, a1, a2
; RV64-NEXT: add a1, sp, a1
; RV64-NEXT: addi a1, a1, 16
-; RV64-NEXT: vl1r.v v0, (a1) # Unknown-size Folded Reload
+; RV64-NEXT: vl4r.v v16, (a1) # Unknown-size Folded Reload
; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: li a2, 24
+; RV64-NEXT: li a2, 25
; RV64-NEXT: mul a1, a1, a2
; RV64-NEXT: add a1, sp, a1
; RV64-NEXT: addi a1, a1, 16
-; RV64-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
-; RV64-NEXT: vrgatherei16.vv v8, v16, v12, v0.t
+; RV64-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; RV64-NEXT: vsetivli zero, 5, e64, m4, tu, ma
+; RV64-NEXT: vmv.v.v v16, v24
+; RV64-NEXT: vmv2r.v v8, v2
; RV64-NEXT: vsetivli zero, 16, e16, m2, ta, ma
-; RV64-NEXT: vadd.vi v12, v24, 5
-; RV64-NEXT: vsetvli zero, zero, e64, m8, ta, ma
+; RV64-NEXT: vadd.vi v12, v2, 5
; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: li a2, 48
-; RV64-NEXT: mul a1, a1, a2
+; RV64-NEXT: slli a2, a1, 6
+; RV64-NEXT: add a1, a2, a1
; RV64-NEXT: add a1, sp, a1
; RV64-NEXT: addi a1, a1, 16
; RV64-NEXT: vl8r.v v0, (a1) # Unknown-size Folded Reload
-; RV64-NEXT: vrgatherei16.vv v16, v0, v12
+; RV64-NEXT: vsetvli zero, zero, e64, m8, ta, ma
+; RV64-NEXT: vrgatherei16.vv v24, v0, v12
; RV64-NEXT: vsetvli zero, zero, e16, m2, ta, ma
-; RV64-NEXT: vadd.vi v12, v24, -11
-; RV64-NEXT: vsetvli zero, zero, e64, m8, ta, mu
+; RV64-NEXT: vadd.vi v2, v8, -11
; RV64-NEXT: addi a1, sp, 16
; RV64-NEXT: vl1r.v v0, (a1) # Unknown-size Folded Reload
; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: li a2, 40
+; RV64-NEXT: li a2, 57
; RV64-NEXT: mul a1, a1, a2
; RV64-NEXT: add a1, sp, a1
; RV64-NEXT: addi a1, a1, 16
-; RV64-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
-; RV64-NEXT: vrgatherei16.vv v16, v24, v12, v0.t
+; RV64-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
+; RV64-NEXT: vsetvli zero, zero, e64, m8, ta, mu
+; RV64-NEXT: vrgatherei16.vv v24, v8, v2, v0.t
+; RV64-NEXT: csrr a1, vlenb
+; RV64-NEXT: li a2, 41
+; RV64-NEXT: mul a1, a1, a2
+; RV64-NEXT: add a1, sp, a1
+; RV64-NEXT: addi a1, a1, 16
+; RV64-NEXT: vl4r.v v12, (a1) # Unknown-size Folded Reload
+; RV64-NEXT: csrr a1, vlenb
+; RV64-NEXT: slli a2, a1, 3
+; RV64-NEXT: add a1, a2, a1
+; RV64-NEXT: add a1, sp, a1
+; RV64-NEXT: addi a1, a1, 16
+; RV64-NEXT: vl8r.v v0, (a1) # Unknown-size Folded Reload
; RV64-NEXT: vsetivli zero, 5, e64, m4, tu, ma
-; RV64-NEXT: vmv.v.v v8, v16
+; RV64-NEXT: vmv.v.v v12, v0
+; RV64-NEXT: csrr a1, vlenb
+; RV64-NEXT: li a2, 37
+; RV64-NEXT: mul a1, a1, a2
+; RV64-NEXT: add a1, sp, a1
+; RV64-NEXT: addi a1, a1, 16
+; RV64-NEXT: vl4r.v v20, (a1) # Unknown-size Folded Reload
+; RV64-NEXT: csrr a1, vlenb
+; RV64-NEXT: add a1, sp, a1
+; RV64-NEXT: addi a1, a1, 16
+; RV64-NEXT: vl8r.v v0, (a1) # Unknown-size Folded Reload
+; RV64-NEXT: vmv.v.v v20, v0
+; RV64-NEXT: csrr a1, vlenb
+; RV64-NEXT: slli a2, a1, 5
+; RV64-NEXT: add a1, a2, a1
+; RV64-NEXT: add a1, sp, a1
+; RV64-NEXT: addi a1, a1, 16
+; RV64-NEXT: vl4r.v v8, (a1) # Unknown-size Folded Reload
+; RV64-NEXT: vmv.v.v v8, v24
; RV64-NEXT: addi a1, a0, 320
; RV64-NEXT: vsetivli zero, 8, e64, m4, ta, ma
; RV64-NEXT: vse64.v v8, (a1)
; RV64-NEXT: addi a1, a0, 256
-; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: slli a2, a2, 1
-; RV64-NEXT: add a2, sp, a2
-; RV64-NEXT: addi a2, a2, 16
-; RV64-NEXT: vl4r.v v8, (a2) # Unknown-size Folded Reload
-; RV64-NEXT: vse64.v v8, (a1)
+; RV64-NEXT: vse64.v v20, (a1)
; RV64-NEXT: addi a1, a0, 192
-; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: li a3, 20
-; RV64-NEXT: mul a2, a2, a3
-; RV64-NEXT: add a2, sp, a2
-; RV64-NEXT: addi a2, a2, 16
-; RV64-NEXT: vl4r.v v8, (a2) # Unknown-size Folded Reload
-; RV64-NEXT: vse64.v v8, (a1)
+; RV64-NEXT: vse64.v v12, (a1)
; RV64-NEXT: addi a1, a0, 128
-; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: slli a2, a2, 3
-; RV64-NEXT: add a2, sp, a2
-; RV64-NEXT: addi a2, a2, 16
-; RV64-NEXT: vl4r.v v8, (a2) # Unknown-size Folded Reload
-; RV64-NEXT: vse64.v v8, (a1)
+; RV64-NEXT: vse64.v v16, (a1)
; RV64-NEXT: addi a1, a0, 64
; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: li a3, 12
-; RV64-NEXT: mul a2, a2, a3
+; RV64-NEXT: slli a3, a2, 4
+; RV64-NEXT: add a2, a3, a2
; RV64-NEXT: add a2, sp, a2
; RV64-NEXT: addi a2, a2, 16
; RV64-NEXT: vl4r.v v8, (a2) # Unknown-size Folded Reload
; RV64-NEXT: vse64.v v8, (a1)
; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: slli a1, a1, 4
+; RV64-NEXT: li a2, 21
+; RV64-NEXT: mul a1, a1, a2
; RV64-NEXT: add a1, sp, a1
; RV64-NEXT: addi a1, a1, 16
; RV64-NEXT: vl4r.v v8, (a1) # Unknown-size Folded Reload
; RV64-NEXT: vse64.v v8, (a0)
; RV64-NEXT: csrr a0, vlenb
-; RV64-NEXT: li a1, 56
+; RV64-NEXT: li a1, 74
; RV64-NEXT: mul a0, a0, a1
; RV64-NEXT: add sp, sp, a0
; RV64-NEXT: addi sp, sp, 16
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-llrint.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-llrint.ll
index d55683e653d24..2ba1a751f9ed5 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-llrint.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-llrint.ll
@@ -57,17 +57,17 @@ define <2 x i64> @llrint_v2i64_v2f32(<2 x float> %x) {
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 16
; RV32-NEXT: vs1r.v v8, (a0) # Unknown-size Folded Spill
-; RV32-NEXT: vsetivli zero, 1, e32, mf2, ta, ma
; RV32-NEXT: addi a0, sp, 16
; RV32-NEXT: vl1r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 1, e32, mf2, ta, ma
; RV32-NEXT: vslidedown.vi v8, v8, 1
; RV32-NEXT: vfmv.f.s fa0, v8
; RV32-NEXT: call llrintf
-; RV32-NEXT: vsetivli zero, 4, e32, m1, ta, ma
; RV32-NEXT: csrr a2, vlenb
; RV32-NEXT: add a2, sp, a2
; RV32-NEXT: addi a2, a2, 16
; RV32-NEXT: vl1r.v v8, (a2) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 4, e32, m1, ta, ma
; RV32-NEXT: vslide1down.vx v8, v8, a0
; RV32-NEXT: vslide1down.vx v8, v8, a1
; RV32-NEXT: csrr a0, vlenb
@@ -118,50 +118,50 @@ define <3 x i64> @llrint_v3i64_v3f32(<3 x float> %x) {
; RV32-NEXT: vslide1down.vx v8, v8, a1
; RV32-NEXT: addi a0, sp, 16
; RV32-NEXT: vs2r.v v8, (a0) # Unknown-size Folded Spill
-; RV32-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32-NEXT: csrr a0, vlenb
; RV32-NEXT: slli a0, a0, 1
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 16
; RV32-NEXT: vl1r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32-NEXT: vslidedown.vi v8, v8, 1
; RV32-NEXT: vfmv.f.s fa0, v8
; RV32-NEXT: call llrintf
-; RV32-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32-NEXT: addi a2, sp, 16
; RV32-NEXT: vl2r.v v8, (a2) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32-NEXT: vslide1down.vx v8, v8, a0
; RV32-NEXT: vslide1down.vx v8, v8, a1
; RV32-NEXT: addi a0, sp, 16
; RV32-NEXT: vs2r.v v8, (a0) # Unknown-size Folded Spill
-; RV32-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32-NEXT: csrr a0, vlenb
; RV32-NEXT: slli a0, a0, 1
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 16
; RV32-NEXT: vl1r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32-NEXT: vslidedown.vi v8, v8, 2
; RV32-NEXT: vfmv.f.s fa0, v8
; RV32-NEXT: call llrintf
-; RV32-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32-NEXT: addi a2, sp, 16
; RV32-NEXT: vl2r.v v8, (a2) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32-NEXT: vslide1down.vx v8, v8, a0
; RV32-NEXT: vslide1down.vx v8, v8, a1
; RV32-NEXT: addi a0, sp, 16
; RV32-NEXT: vs2r.v v8, (a0) # Unknown-size Folded Spill
-; RV32-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32-NEXT: csrr a0, vlenb
; RV32-NEXT: slli a0, a0, 1
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 16
; RV32-NEXT: vl1r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32-NEXT: vslidedown.vi v8, v8, 3
; RV32-NEXT: vfmv.f.s fa0, v8
; RV32-NEXT: call llrintf
-; RV32-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32-NEXT: addi a2, sp, 16
; RV32-NEXT: vl2r.v v8, (a2) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32-NEXT: vslide1down.vx v8, v8, a0
; RV32-NEXT: vslide1down.vx v8, v8, a1
; RV32-NEXT: csrr a0, vlenb
@@ -224,50 +224,50 @@ define <4 x i64> @llrint_v4i64_v4f32(<4 x float> %x) {
; RV32-NEXT: vslide1down.vx v8, v8, a1
; RV32-NEXT: addi a0, sp, 16
; RV32-NEXT: vs2r.v v8, (a0) # Unknown-size Folded Spill
-; RV32-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32-NEXT: csrr a0, vlenb
; RV32-NEXT: slli a0, a0, 1
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 16
; RV32-NEXT: vl1r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32-NEXT: vslidedown.vi v8, v8, 1
; RV32-NEXT: vfmv.f.s fa0, v8
; RV32-NEXT: call llrintf
-; RV32-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32-NEXT: addi a2, sp, 16
; RV32-NEXT: vl2r.v v8, (a2) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32-NEXT: vslide1down.vx v8, v8, a0
; RV32-NEXT: vslide1down.vx v8, v8, a1
; RV32-NEXT: addi a0, sp, 16
; RV32-NEXT: vs2r.v v8, (a0) # Unknown-size Folded Spill
-; RV32-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32-NEXT: csrr a0, vlenb
; RV32-NEXT: slli a0, a0, 1
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 16
; RV32-NEXT: vl1r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32-NEXT: vslidedown.vi v8, v8, 2
; RV32-NEXT: vfmv.f.s fa0, v8
; RV32-NEXT: call llrintf
-; RV32-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32-NEXT: addi a2, sp, 16
; RV32-NEXT: vl2r.v v8, (a2) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32-NEXT: vslide1down.vx v8, v8, a0
; RV32-NEXT: vslide1down.vx v8, v8, a1
; RV32-NEXT: addi a0, sp, 16
; RV32-NEXT: vs2r.v v8, (a0) # Unknown-size Folded Spill
-; RV32-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32-NEXT: csrr a0, vlenb
; RV32-NEXT: slli a0, a0, 1
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 16
; RV32-NEXT: vl1r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32-NEXT: vslidedown.vi v8, v8, 3
; RV32-NEXT: vfmv.f.s fa0, v8
; RV32-NEXT: call llrintf
-; RV32-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32-NEXT: addi a2, sp, 16
; RV32-NEXT: vl2r.v v8, (a2) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32-NEXT: vslide1down.vx v8, v8, a0
; RV32-NEXT: vslide1down.vx v8, v8, a1
; RV32-NEXT: csrr a0, vlenb
@@ -328,57 +328,57 @@ define <8 x i64> @llrint_v8i64_v8f32(<8 x float> %x) {
; RV32-NEXT: call llrintf
; RV32-NEXT: sw a1, 68(sp)
; RV32-NEXT: sw a0, 64(sp)
-; RV32-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32-NEXT: addi a0, sp, 192
; RV32-NEXT: vl2r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32-NEXT: vslidedown.vi v8, v8, 7
; RV32-NEXT: vfmv.f.s fa0, v8
; RV32-NEXT: call llrintf
; RV32-NEXT: sw a1, 124(sp)
; RV32-NEXT: sw a0, 120(sp)
-; RV32-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32-NEXT: addi a0, sp, 192
; RV32-NEXT: vl2r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32-NEXT: vslidedown.vi v8, v8, 6
; RV32-NEXT: vfmv.f.s fa0, v8
; RV32-NEXT: call llrintf
; RV32-NEXT: sw a1, 116(sp)
; RV32-NEXT: sw a0, 112(sp)
-; RV32-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32-NEXT: addi a0, sp, 192
; RV32-NEXT: vl2r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32-NEXT: vslidedown.vi v8, v8, 5
; RV32-NEXT: vfmv.f.s fa0, v8
; RV32-NEXT: call llrintf
; RV32-NEXT: sw a1, 108(sp)
; RV32-NEXT: sw a0, 104(sp)
-; RV32-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32-NEXT: addi a0, sp, 192
; RV32-NEXT: vl2r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32-NEXT: vslidedown.vi v8, v8, 4
; RV32-NEXT: vfmv.f.s fa0, v8
; RV32-NEXT: call llrintf
; RV32-NEXT: sw a1, 100(sp)
; RV32-NEXT: sw a0, 96(sp)
-; RV32-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32-NEXT: addi a0, sp, 192
; RV32-NEXT: vl2r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32-NEXT: vslidedown.vi v8, v8, 3
; RV32-NEXT: vfmv.f.s fa0, v8
; RV32-NEXT: call llrintf
; RV32-NEXT: sw a1, 92(sp)
; RV32-NEXT: sw a0, 88(sp)
-; RV32-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32-NEXT: addi a0, sp, 192
; RV32-NEXT: vl2r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32-NEXT: vslidedown.vi v8, v8, 2
; RV32-NEXT: vfmv.f.s fa0, v8
; RV32-NEXT: call llrintf
; RV32-NEXT: sw a1, 84(sp)
; RV32-NEXT: sw a0, 80(sp)
-; RV32-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32-NEXT: addi a0, sp, 192
; RV32-NEXT: vl2r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32-NEXT: vslidedown.vi v8, v8, 1
; RV32-NEXT: vfmv.f.s fa0, v8
; RV32-NEXT: call llrintf
@@ -502,64 +502,64 @@ define <16 x i64> @llrint_v16i64_v16f32(<16 x float> %x) {
; RV32-NEXT: call llrintf
; RV32-NEXT: sw a1, 196(sp)
; RV32-NEXT: sw a0, 192(sp)
-; RV32-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32-NEXT: addi a0, sp, 384
; RV32-NEXT: vl4r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32-NEXT: vfmv.f.s fa0, v8
; RV32-NEXT: call llrintf
; RV32-NEXT: sw a1, 132(sp)
; RV32-NEXT: sw a0, 128(sp)
-; RV32-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32-NEXT: addi a0, sp, 384
; RV32-NEXT: vl4r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32-NEXT: vslidedown.vi v8, v8, 3
; RV32-NEXT: vfmv.f.s fa0, v8
; RV32-NEXT: call llrintf
; RV32-NEXT: sw a1, 156(sp)
; RV32-NEXT: sw a0, 152(sp)
-; RV32-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32-NEXT: addi a0, sp, 384
; RV32-NEXT: vl4r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32-NEXT: vslidedown.vi v8, v8, 2
; RV32-NEXT: vfmv.f.s fa0, v8
; RV32-NEXT: call llrintf
; RV32-NEXT: sw a1, 148(sp)
; RV32-NEXT: sw a0, 144(sp)
-; RV32-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32-NEXT: addi a0, sp, 384
; RV32-NEXT: vl4r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32-NEXT: vslidedown.vi v8, v8, 1
; RV32-NEXT: vfmv.f.s fa0, v8
; RV32-NEXT: call llrintf
; RV32-NEXT: sw a1, 140(sp)
; RV32-NEXT: sw a0, 136(sp)
-; RV32-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32-NEXT: addi a0, sp, 384
; RV32-NEXT: vl4r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32-NEXT: vslidedown.vi v8, v8, 7
; RV32-NEXT: vfmv.f.s fa0, v8
; RV32-NEXT: call llrintf
; RV32-NEXT: sw a1, 188(sp)
; RV32-NEXT: sw a0, 184(sp)
-; RV32-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32-NEXT: addi a0, sp, 384
; RV32-NEXT: vl4r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32-NEXT: vslidedown.vi v8, v8, 6
; RV32-NEXT: vfmv.f.s fa0, v8
; RV32-NEXT: call llrintf
; RV32-NEXT: sw a1, 180(sp)
; RV32-NEXT: sw a0, 176(sp)
-; RV32-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32-NEXT: addi a0, sp, 384
; RV32-NEXT: vl4r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32-NEXT: vslidedown.vi v8, v8, 5
; RV32-NEXT: vfmv.f.s fa0, v8
; RV32-NEXT: call llrintf
; RV32-NEXT: sw a1, 172(sp)
; RV32-NEXT: sw a0, 168(sp)
-; RV32-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32-NEXT: addi a0, sp, 384
; RV32-NEXT: vl4r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32-NEXT: vslidedown.vi v8, v8, 4
; RV32-NEXT: vfmv.f.s fa0, v8
; RV32-NEXT: call llrintf
@@ -711,17 +711,17 @@ define <2 x i64> @llrint_v2i64_v2f64(<2 x double> %x) {
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 16
; RV32-NEXT: vs1r.v v8, (a0) # Unknown-size Folded Spill
-; RV32-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; RV32-NEXT: addi a0, sp, 16
; RV32-NEXT: vl1r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; RV32-NEXT: vslidedown.vi v8, v8, 1
; RV32-NEXT: vfmv.f.s fa0, v8
; RV32-NEXT: call llrint
-; RV32-NEXT: vsetivli zero, 4, e32, m1, ta, ma
; RV32-NEXT: csrr a2, vlenb
; RV32-NEXT: add a2, sp, a2
; RV32-NEXT: addi a2, a2, 16
; RV32-NEXT: vl1r.v v8, (a2) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 4, e32, m1, ta, ma
; RV32-NEXT: vslide1down.vx v8, v8, a0
; RV32-NEXT: vslide1down.vx v8, v8, a1
; RV32-NEXT: csrr a0, vlenb
@@ -772,50 +772,50 @@ define <4 x i64> @llrint_v4i64_v4f64(<4 x double> %x) {
; RV32-NEXT: vslide1down.vx v8, v8, a1
; RV32-NEXT: addi a0, sp, 16
; RV32-NEXT: vs2r.v v8, (a0) # Unknown-size Folded Spill
-; RV32-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; RV32-NEXT: csrr a0, vlenb
; RV32-NEXT: slli a0, a0, 1
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 16
; RV32-NEXT: vl2r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; RV32-NEXT: vslidedown.vi v8, v8, 1
; RV32-NEXT: vfmv.f.s fa0, v8
; RV32-NEXT: call llrint
-; RV32-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32-NEXT: addi a2, sp, 16
; RV32-NEXT: vl2r.v v8, (a2) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32-NEXT: vslide1down.vx v8, v8, a0
; RV32-NEXT: vslide1down.vx v8, v8, a1
; RV32-NEXT: addi a0, sp, 16
; RV32-NEXT: vs2r.v v8, (a0) # Unknown-size Folded Spill
-; RV32-NEXT: vsetivli zero, 1, e64, m2, ta, ma
; RV32-NEXT: csrr a0, vlenb
; RV32-NEXT: slli a0, a0, 1
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 16
; RV32-NEXT: vl2r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 1, e64, m2, ta, ma
; RV32-NEXT: vslidedown.vi v8, v8, 2
; RV32-NEXT: vfmv.f.s fa0, v8
; RV32-NEXT: call llrint
-; RV32-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32-NEXT: addi a2, sp, 16
; RV32-NEXT: vl2r.v v8, (a2) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32-NEXT: vslide1down.vx v8, v8, a0
; RV32-NEXT: vslide1down.vx v8, v8, a1
; RV32-NEXT: addi a0, sp, 16
; RV32-NEXT: vs2r.v v8, (a0) # Unknown-size Folded Spill
-; RV32-NEXT: vsetivli zero, 1, e64, m2, ta, ma
; RV32-NEXT: csrr a0, vlenb
; RV32-NEXT: slli a0, a0, 1
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 16
; RV32-NEXT: vl2r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 1, e64, m2, ta, ma
; RV32-NEXT: vslidedown.vi v8, v8, 3
; RV32-NEXT: vfmv.f.s fa0, v8
; RV32-NEXT: call llrint
-; RV32-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32-NEXT: addi a2, sp, 16
; RV32-NEXT: vl2r.v v8, (a2) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32-NEXT: vslide1down.vx v8, v8, a0
; RV32-NEXT: vslide1down.vx v8, v8, a1
; RV32-NEXT: csrr a0, vlenb
@@ -890,32 +890,32 @@ define <8 x i64> @llrint_v8i64_v8f64(<8 x double> %x) {
; RV32-NEXT: call llrint
; RV32-NEXT: sw a1, 164(sp)
; RV32-NEXT: sw a0, 160(sp)
-; RV32-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; RV32-NEXT: addi a0, sp, 256
; RV32-NEXT: vl4r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; RV32-NEXT: vfmv.f.s fa0, v8
; RV32-NEXT: call llrint
; RV32-NEXT: sw a1, 132(sp)
; RV32-NEXT: sw a0, 128(sp)
-; RV32-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; RV32-NEXT: addi a0, sp, 256
; RV32-NEXT: vl4r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; RV32-NEXT: vslidedown.vi v8, v8, 1
; RV32-NEXT: vfmv.f.s fa0, v8
; RV32-NEXT: call llrint
; RV32-NEXT: sw a1, 140(sp)
; RV32-NEXT: sw a0, 136(sp)
-; RV32-NEXT: vsetivli zero, 1, e64, m2, ta, ma
; RV32-NEXT: addi a0, sp, 256
; RV32-NEXT: vl4r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 1, e64, m2, ta, ma
; RV32-NEXT: vslidedown.vi v8, v8, 3
; RV32-NEXT: vfmv.f.s fa0, v8
; RV32-NEXT: call llrint
; RV32-NEXT: sw a1, 156(sp)
; RV32-NEXT: sw a0, 152(sp)
-; RV32-NEXT: vsetivli zero, 1, e64, m2, ta, ma
; RV32-NEXT: addi a0, sp, 256
; RV32-NEXT: vl4r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vsetivli zero, 1, e64, m2, ta, ma
; RV32-NEXT: vslidedown.vi v8, v8, 2
; RV32-NEXT: vfmv.f.s fa0, v8
; RV32-NEXT: call llrint
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-lrint.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-lrint.ll
index 35baa6808db60..c377851d0c174 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-lrint.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-lrint.ll
@@ -801,30 +801,27 @@ define <8 x iXLen> @lrint_v8f64(<8 x double> %x) {
; RV32-NEXT: fcvt.w.d a0, fa5
; RV32-NEXT: vfmv.f.s fa5, v8
; RV32-NEXT: fcvt.w.d a1, fa5
-; RV32-NEXT: vsetivli zero, 8, e32, m2, ta, ma
-; RV32-NEXT: vmv.v.x v10, a1
-; RV32-NEXT: vslide1down.vx v10, v10, a0
-; RV32-NEXT: vsetivli zero, 1, e64, m2, ta, ma
-; RV32-NEXT: vslidedown.vi v12, v8, 2
-; RV32-NEXT: vfmv.f.s fa5, v12
-; RV32-NEXT: fcvt.w.d a0, fa5
-; RV32-NEXT: vsetivli zero, 8, e32, m2, ta, ma
-; RV32-NEXT: vslide1down.vx v10, v10, a0
; RV32-NEXT: vsetivli zero, 1, e64, m2, ta, ma
+; RV32-NEXT: vslidedown.vi v10, v8, 2
+; RV32-NEXT: vfmv.f.s fa5, v10
+; RV32-NEXT: fcvt.w.d a2, fa5
; RV32-NEXT: vslidedown.vi v8, v8, 3
-; RV32-NEXT: vfmv.f.s fa5, v8
-; RV32-NEXT: vsetivli zero, 8, e32, m2, ta, ma
-; RV32-NEXT: fld fa4, 32(sp)
+; RV32-NEXT: fld fa5, 32(sp)
+; RV32-NEXT: vfmv.f.s fa4, v8
; RV32-NEXT: fld fa3, 40(sp)
-; RV32-NEXT: fcvt.w.d a0, fa5
+; RV32-NEXT: fcvt.w.d a3, fa4
+; RV32-NEXT: fcvt.w.d a4, fa5
+; RV32-NEXT: vsetivli zero, 8, e32, m2, ta, ma
+; RV32-NEXT: vmv.v.x v8, a1
+; RV32-NEXT: fcvt.w.d a1, fa3
; RV32-NEXT: fld fa5, 48(sp)
-; RV32-NEXT: fcvt.w.d a1, fa4
-; RV32-NEXT: fcvt.w.d a2, fa3
-; RV32-NEXT: vslide1down.vx v8, v10, a0
+; RV32-NEXT: vslide1down.vx v8, v8, a0
+; RV32-NEXT: vslide1down.vx v8, v8, a2
+; RV32-NEXT: vslide1down.vx v8, v8, a3
; RV32-NEXT: fcvt.w.d a0, fa5
; RV32-NEXT: fld fa5, 56(sp)
+; RV32-NEXT: vslide1down.vx v8, v8, a4
; RV32-NEXT: vslide1down.vx v8, v8, a1
-; RV32-NEXT: vslide1down.vx v8, v8, a2
; RV32-NEXT: vslide1down.vx v8, v8, a0
; RV32-NEXT: fcvt.w.d a0, fa5
; RV32-NEXT: vslide1down.vx v8, v8, a0
@@ -854,30 +851,27 @@ define <8 x iXLen> @lrint_v8f64(<8 x double> %x) {
; RV64-i32-NEXT: fcvt.l.d a0, fa5
; RV64-i32-NEXT: vfmv.f.s fa5, v8
; RV64-i32-NEXT: fcvt.l.d a1, fa5
-; RV64-i32-NEXT: vsetivli zero, 8, e32, m2, ta, ma
-; RV64-i32-NEXT: vmv.v.x v10, a1
-; RV64-i32-NEXT: vslide1down.vx v10, v10, a0
-; RV64-i32-NEXT: vsetivli zero, 1, e64, m2, ta, ma
-; RV64-i32-NEXT: vslidedown.vi v12, v8, 2
-; RV64-i32-NEXT: vfmv.f.s fa5, v12
-; RV64-i32-NEXT: fcvt.l.d a0, fa5
-; RV64-i32-NEXT: vsetivli zero, 8, e32, m2, ta, ma
-; RV64-i32-NEXT: vslide1down.vx v10, v10, a0
; RV64-i32-NEXT: vsetivli zero, 1, e64, m2, ta, ma
+; RV64-i32-NEXT: vslidedown.vi v10, v8, 2
+; RV64-i32-NEXT: vfmv.f.s fa5, v10
+; RV64-i32-NEXT: fcvt.l.d a2, fa5
; RV64-i32-NEXT: vslidedown.vi v8, v8, 3
-; RV64-i32-NEXT: vfmv.f.s fa5, v8
-; RV64-i32-NEXT: vsetivli zero, 8, e32, m2, ta, ma
-; RV64-i32-NEXT: fld fa4, 32(sp)
+; RV64-i32-NEXT: fld fa5, 32(sp)
+; RV64-i32-NEXT: vfmv.f.s fa4, v8
; RV64-i32-NEXT: fld fa3, 40(sp)
-; RV64-i32-NEXT: fcvt.l.d a0, fa5
+; RV64-i32-NEXT: fcvt.l.d a3, fa4
+; RV64-i32-NEXT: fcvt.l.d a4, fa5
+; RV64-i32-NEXT: vsetivli zero, 8, e32, m2, ta, ma
+; RV64-i32-NEXT: vmv.v.x v8, a1
+; RV64-i32-NEXT: fcvt.l.d a1, fa3
; RV64-i32-NEXT: fld fa5, 48(sp)
-; RV64-i32-NEXT: fcvt.l.d a1, fa4
-; RV64-i32-NEXT: fcvt.l.d a2, fa3
-; RV64-i32-NEXT: vslide1down.vx v8, v10, a0
+; RV64-i32-NEXT: vslide1down.vx v8, v8, a0
+; RV64-i32-NEXT: vslide1down.vx v8, v8, a2
+; RV64-i32-NEXT: vslide1down.vx v8, v8, a3
; RV64-i32-NEXT: fcvt.l.d a0, fa5
; RV64-i32-NEXT: fld fa5, 56(sp)
+; RV64-i32-NEXT: vslide1down.vx v8, v8, a4
; RV64-i32-NEXT: vslide1down.vx v8, v8, a1
-; RV64-i32-NEXT: vslide1down.vx v8, v8, a2
; RV64-i32-NEXT: vslide1down.vx v8, v8, a0
; RV64-i32-NEXT: fcvt.l.d a0, fa5
; RV64-i32-NEXT: vslide1down.vx v8, v8, a0
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-mask-buildvec.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-mask-buildvec.ll
index 023d707f07bff..36dd8c07d11cd 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-mask-buildvec.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-mask-buildvec.ll
@@ -140,8 +140,9 @@ define <4 x i1> @buildvec_mask_v4i1() {
define <4 x i1> @buildvec_mask_nonconst_v4i1(i1 %x, i1 %y) {
; CHECK-LABEL: buildvec_mask_nonconst_v4i1:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetivli zero, 4, e8, mf4, ta, ma
+; CHECK-NEXT: vsetivli zero, 1, e8, mf8, ta, ma
; CHECK-NEXT: vmv.v.i v0, 3
+; CHECK-NEXT: vsetivli zero, 4, e8, mf4, ta, ma
; CHECK-NEXT: vmv.v.x v8, a1
; CHECK-NEXT: vmerge.vxm v8, v8, a0, v0
; CHECK-NEXT: vand.vi v8, v8, 1
@@ -150,8 +151,9 @@ define <4 x i1> @buildvec_mask_nonconst_v4i1(i1 %x, i1 %y) {
;
; ZVE32F-LABEL: buildvec_mask_nonconst_v4i1:
; ZVE32F: # %bb.0:
-; ZVE32F-NEXT: vsetivli zero, 4, e8, mf4, ta, ma
+; ZVE32F-NEXT: vsetivli zero, 1, e8, mf4, ta, ma
; ZVE32F-NEXT: vmv.v.i v0, 3
+; ZVE32F-NEXT: vsetivli zero, 4, e8, mf4, ta, ma
; ZVE32F-NEXT: vmv.v.x v8, a1
; ZVE32F-NEXT: vmerge.vxm v8, v8, a0, v0
; ZVE32F-NEXT: vand.vi v8, v8, 1
@@ -245,8 +247,8 @@ define <8 x i1> @buildvec_mask_v8i1() {
define <8 x i1> @buildvec_mask_nonconst_v8i1(i1 %x, i1 %y) {
; CHECK-LABEL: buildvec_mask_nonconst_v8i1:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetivli zero, 8, e8, mf2, ta, ma
; CHECK-NEXT: li a2, 19
+; CHECK-NEXT: vsetivli zero, 8, e8, mf2, ta, ma
; CHECK-NEXT: vmv.s.x v0, a2
; CHECK-NEXT: vmv.v.x v8, a1
; CHECK-NEXT: vmerge.vxm v8, v8, a0, v0
@@ -256,8 +258,8 @@ define <8 x i1> @buildvec_mask_nonconst_v8i1(i1 %x, i1 %y) {
;
; ZVE32F-LABEL: buildvec_mask_nonconst_v8i1:
; ZVE32F: # %bb.0:
-; ZVE32F-NEXT: vsetivli zero, 8, e8, mf2, ta, ma
; ZVE32F-NEXT: li a2, 19
+; ZVE32F-NEXT: vsetivli zero, 8, e8, mf2, ta, ma
; ZVE32F-NEXT: vmv.s.x v0, a2
; ZVE32F-NEXT: vmv.v.x v8, a1
; ZVE32F-NEXT: vmerge.vxm v8, v8, a0, v0
@@ -286,8 +288,8 @@ define <8 x i1> @buildvec_mask_nonconst_v8i1_2(i1 %x, i1 %y, i1 %z, i1 %w) {
; CHECK-NEXT: vslide1down.vx v9, v9, a1
; CHECK-NEXT: vslide1down.vx v8, v8, a3
; CHECK-NEXT: vslide1down.vx v8, v8, zero
-; CHECK-NEXT: vslide1down.vx v8, v8, a2
; CHECK-NEXT: vmv.v.i v0, 15
+; CHECK-NEXT: vslide1down.vx v8, v8, a2
; CHECK-NEXT: vslidedown.vi v8, v9, 4, v0.t
; CHECK-NEXT: vand.vi v8, v8, 1
; CHECK-NEXT: vmsne.vi v0, v8, 0
@@ -303,8 +305,8 @@ define <8 x i1> @buildvec_mask_nonconst_v8i1_2(i1 %x, i1 %y, i1 %z, i1 %w) {
; ZVE32F-NEXT: vslide1down.vx v9, v9, a1
; ZVE32F-NEXT: vslide1down.vx v8, v8, a3
; ZVE32F-NEXT: vslide1down.vx v8, v8, zero
-; ZVE32F-NEXT: vslide1down.vx v8, v8, a2
; ZVE32F-NEXT: vmv.v.i v0, 15
+; ZVE32F-NEXT: vslide1down.vx v8, v8, a2
; ZVE32F-NEXT: vslidedown.vi v8, v9, 4, v0.t
; ZVE32F-NEXT: vand.vi v8, v8, 1
; ZVE32F-NEXT: vmsne.vi v0, v8, 0
@@ -331,8 +333,8 @@ define <8 x i1> @buildvec_mask_optsize_nonconst_v8i1_2(i1 %x, i1 %y, i1 %z, i1 %
; CHECK-NEXT: vslide1down.vx v9, v9, a1
; CHECK-NEXT: vslide1down.vx v8, v8, a3
; CHECK-NEXT: vslide1down.vx v8, v8, zero
-; CHECK-NEXT: vslide1down.vx v8, v8, a2
; CHECK-NEXT: vmv.v.i v0, 15
+; CHECK-NEXT: vslide1down.vx v8, v8, a2
; CHECK-NEXT: vslidedown.vi v8, v9, 4, v0.t
; CHECK-NEXT: vand.vi v8, v8, 1
; CHECK-NEXT: vmsne.vi v0, v8, 0
@@ -348,8 +350,8 @@ define <8 x i1> @buildvec_mask_optsize_nonconst_v8i1_2(i1 %x, i1 %y, i1 %z, i1 %
; ZVE32F-NEXT: vslide1down.vx v9, v9, a1
; ZVE32F-NEXT: vslide1down.vx v8, v8, a3
; ZVE32F-NEXT: vslide1down.vx v8, v8, zero
-; ZVE32F-NEXT: vslide1down.vx v8, v8, a2
; ZVE32F-NEXT: vmv.v.i v0, 15
+; ZVE32F-NEXT: vslide1down.vx v8, v8, a2
; ZVE32F-NEXT: vslidedown.vi v8, v9, 4, v0.t
; ZVE32F-NEXT: vand.vi v8, v8, 1
; ZVE32F-NEXT: vmsne.vi v0, v8, 0
@@ -375,8 +377,8 @@ define <8 x i1> @buildvec_mask_optsize_nonconst_v8i1(i1 %x, i1 %y) optsize {
; CHECK-NEXT: vslide1down.vx v9, v9, a1
; CHECK-NEXT: vslide1down.vx v8, v8, a1
; CHECK-NEXT: vslide1down.vx v8, v8, a1
-; CHECK-NEXT: vslide1down.vx v8, v8, a1
; CHECK-NEXT: vmv.v.i v0, 15
+; CHECK-NEXT: vslide1down.vx v8, v8, a1
; CHECK-NEXT: vslidedown.vi v8, v9, 4, v0.t
; CHECK-NEXT: vand.vi v8, v8, 1
; CHECK-NEXT: vmsne.vi v0, v8, 0
@@ -391,8 +393,8 @@ define <8 x i1> @buildvec_mask_optsize_nonconst_v8i1(i1 %x, i1 %y) optsize {
; ZVE32F-NEXT: vslide1down.vx v9, v9, a1
; ZVE32F-NEXT: vslide1down.vx v8, v8, a1
; ZVE32F-NEXT: vslide1down.vx v8, v8, a1
-; ZVE32F-NEXT: vslide1down.vx v8, v8, a1
; ZVE32F-NEXT: vmv.v.i v0, 15
+; ZVE32F-NEXT: vslide1down.vx v8, v8, a1
; ZVE32F-NEXT: vslidedown.vi v8, v9, 4, v0.t
; ZVE32F-NEXT: vand.vi v8, v8, 1
; ZVE32F-NEXT: vmsne.vi v0, v8, 0
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-mask-splat.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-mask-splat.ll
index 7fc442c88d101..979785dd2c024 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-mask-splat.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-mask-splat.ll
@@ -24,11 +24,11 @@ define void @splat_zeros_v2i1(ptr %x) {
define void @splat_v1i1(ptr %x, i1 %y) {
; CHECK-LABEL: splat_v1i1:
; CHECK: # %bb.0:
+; CHECK-NEXT: andi a1, a1, 1
; CHECK-NEXT: vsetivli zero, 1, e8, mf8, ta, ma
+; CHECK-NEXT: vmv.s.x v8, a1
+; CHECK-NEXT: vmsne.vi v0, v8, 0
; CHECK-NEXT: vmv.s.x v8, zero
-; CHECK-NEXT: andi a1, a1, 1
-; CHECK-NEXT: vmv.s.x v9, a1
-; CHECK-NEXT: vmsne.vi v0, v9, 0
; CHECK-NEXT: vmerge.vim v8, v8, 1, v0
; CHECK-NEXT: vsetivli zero, 8, e8, mf2, ta, ma
; CHECK-NEXT: vmv.v.i v9, 0
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-masked-gather.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-masked-gather.ll
index f42f32e246585..d791fc2d1709c 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-masked-gather.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-masked-gather.ll
@@ -519,17 +519,17 @@ define <4 x i8> @mgather_truemask_v4i8(<4 x ptr> %ptrs, <4 x i8> %passthru) {
; RV64ZVE32F-LABEL: mgather_truemask_v4i8:
; RV64ZVE32F: # %bb.0:
; RV64ZVE32F-NEXT: ld a1, 8(a0)
-; RV64ZVE32F-NEXT: ld a2, 16(a0)
-; RV64ZVE32F-NEXT: ld a3, 24(a0)
-; RV64ZVE32F-NEXT: ld a0, 0(a0)
+; RV64ZVE32F-NEXT: ld a2, 0(a0)
+; RV64ZVE32F-NEXT: ld a3, 16(a0)
+; RV64ZVE32F-NEXT: ld a0, 24(a0)
; RV64ZVE32F-NEXT: lbu a1, 0(a1)
-; RV64ZVE32F-NEXT: lbu a2, 0(a2)
-; RV64ZVE32F-NEXT: lbu a3, 0(a3)
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf4, ta, ma
-; RV64ZVE32F-NEXT: vlse8.v v8, (a0), zero
+; RV64ZVE32F-NEXT: vlse8.v v8, (a2), zero
+; RV64ZVE32F-NEXT: lbu a2, 0(a3)
+; RV64ZVE32F-NEXT: lbu a0, 0(a0)
; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a1
; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a2
-; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a3
+; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a0
; RV64ZVE32F-NEXT: ret
%v = call <4 x i8> @llvm.masked.gather.v4i8.v4p0(<4 x ptr> %ptrs, i32 1, <4 x i1> splat (i1 1), <4 x i8> %passthru)
ret <4 x i8> %v
@@ -711,8 +711,8 @@ define <8 x i8> @mgather_baseidx_v8i8(ptr %base, <8 x i8> %idxs, <8 x i1> %m, <8
; RV64ZVE32F-NEXT: .LBB12_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB12_12
; RV64ZVE32F-NEXT: # %bb.5: # %else5
@@ -734,8 +734,8 @@ define <8 x i8> @mgather_baseidx_v8i8(ptr %base, <8 x i8> %idxs, <8 x i1> %m, <8
; RV64ZVE32F-NEXT: vsetivli zero, 6, e8, mf2, tu, ma
; RV64ZVE32F-NEXT: vslideup.vi v9, v8, 5
; RV64ZVE32F-NEXT: .LBB12_9: # %else14
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v10, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB12_15
; RV64ZVE32F-NEXT: # %bb.10: # %else17
@@ -1208,17 +1208,17 @@ define <4 x i16> @mgather_truemask_v4i16(<4 x ptr> %ptrs, <4 x i16> %passthru) {
; RV64ZVE32F-LABEL: mgather_truemask_v4i16:
; RV64ZVE32F: # %bb.0:
; RV64ZVE32F-NEXT: ld a1, 8(a0)
-; RV64ZVE32F-NEXT: ld a2, 16(a0)
-; RV64ZVE32F-NEXT: ld a3, 24(a0)
-; RV64ZVE32F-NEXT: ld a0, 0(a0)
+; RV64ZVE32F-NEXT: ld a2, 0(a0)
+; RV64ZVE32F-NEXT: ld a3, 16(a0)
+; RV64ZVE32F-NEXT: ld a0, 24(a0)
; RV64ZVE32F-NEXT: lh a1, 0(a1)
-; RV64ZVE32F-NEXT: lh a2, 0(a2)
-; RV64ZVE32F-NEXT: lh a3, 0(a3)
; RV64ZVE32F-NEXT: vsetivli zero, 4, e16, mf2, ta, ma
-; RV64ZVE32F-NEXT: vlse16.v v8, (a0), zero
+; RV64ZVE32F-NEXT: vlse16.v v8, (a2), zero
+; RV64ZVE32F-NEXT: lh a2, 0(a3)
+; RV64ZVE32F-NEXT: lh a0, 0(a0)
; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a1
; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a2
-; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a3
+; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a0
; RV64ZVE32F-NEXT: ret
%v = call <4 x i16> @llvm.masked.gather.v4i16.v4p0(<4 x ptr> %ptrs, i32 2, <4 x i1> splat (i1 1), <4 x i16> %passthru)
ret <4 x i16> %v
@@ -1405,8 +1405,8 @@ define <8 x i16> @mgather_baseidx_v8i8_v8i16(ptr %base, <8 x i8> %idxs, <8 x i1>
; RV64ZVE32F-NEXT: .LBB23_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB23_12
; RV64ZVE32F-NEXT: # %bb.5: # %else5
@@ -1430,8 +1430,8 @@ define <8 x i16> @mgather_baseidx_v8i8_v8i16(ptr %base, <8 x i8> %idxs, <8 x i1>
; RV64ZVE32F-NEXT: vsetivli zero, 6, e16, m1, tu, ma
; RV64ZVE32F-NEXT: vslideup.vi v9, v8, 5
; RV64ZVE32F-NEXT: .LBB23_9: # %else14
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v10, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB23_15
; RV64ZVE32F-NEXT: # %bb.10: # %else17
@@ -1556,8 +1556,8 @@ define <8 x i16> @mgather_baseidx_sext_v8i8_v8i16(ptr %base, <8 x i8> %idxs, <8
; RV64ZVE32F-NEXT: .LBB24_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB24_12
; RV64ZVE32F-NEXT: # %bb.5: # %else5
@@ -1581,8 +1581,8 @@ define <8 x i16> @mgather_baseidx_sext_v8i8_v8i16(ptr %base, <8 x i8> %idxs, <8
; RV64ZVE32F-NEXT: vsetivli zero, 6, e16, m1, tu, ma
; RV64ZVE32F-NEXT: vslideup.vi v9, v8, 5
; RV64ZVE32F-NEXT: .LBB24_9: # %else14
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v10, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB24_15
; RV64ZVE32F-NEXT: # %bb.10: # %else17
@@ -1708,8 +1708,8 @@ define <8 x i16> @mgather_baseidx_zext_v8i8_v8i16(ptr %base, <8 x i8> %idxs, <8
; RV64ZVE32F-NEXT: .LBB25_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB25_12
; RV64ZVE32F-NEXT: # %bb.5: # %else5
@@ -1734,8 +1734,8 @@ define <8 x i16> @mgather_baseidx_zext_v8i8_v8i16(ptr %base, <8 x i8> %idxs, <8
; RV64ZVE32F-NEXT: vsetivli zero, 6, e16, m1, tu, ma
; RV64ZVE32F-NEXT: vslideup.vi v9, v8, 5
; RV64ZVE32F-NEXT: .LBB25_9: # %else14
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v10, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB25_15
; RV64ZVE32F-NEXT: # %bb.10: # %else17
@@ -1863,8 +1863,8 @@ define <8 x i16> @mgather_baseidx_v8i16(ptr %base, <8 x i16> %idxs, <8 x i1> %m,
; RV64ZVE32F-NEXT: .LBB26_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e16, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB26_12
; RV64ZVE32F-NEXT: # %bb.5: # %else5
@@ -1887,8 +1887,8 @@ define <8 x i16> @mgather_baseidx_v8i16(ptr %base, <8 x i16> %idxs, <8 x i1> %m,
; RV64ZVE32F-NEXT: vsetivli zero, 6, e16, m1, tu, ma
; RV64ZVE32F-NEXT: vslideup.vi v9, v8, 5
; RV64ZVE32F-NEXT: .LBB26_9: # %else14
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v10, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB26_15
; RV64ZVE32F-NEXT: # %bb.10: # %else17
@@ -2257,17 +2257,17 @@ define <4 x i32> @mgather_truemask_v4i32(<4 x ptr> %ptrs, <4 x i32> %passthru) {
; RV64ZVE32F-LABEL: mgather_truemask_v4i32:
; RV64ZVE32F: # %bb.0:
; RV64ZVE32F-NEXT: ld a1, 8(a0)
-; RV64ZVE32F-NEXT: ld a2, 16(a0)
-; RV64ZVE32F-NEXT: ld a3, 24(a0)
-; RV64ZVE32F-NEXT: ld a0, 0(a0)
+; RV64ZVE32F-NEXT: ld a2, 0(a0)
+; RV64ZVE32F-NEXT: ld a3, 16(a0)
+; RV64ZVE32F-NEXT: ld a0, 24(a0)
; RV64ZVE32F-NEXT: lw a1, 0(a1)
-; RV64ZVE32F-NEXT: lw a2, 0(a2)
-; RV64ZVE32F-NEXT: lw a3, 0(a3)
; RV64ZVE32F-NEXT: vsetivli zero, 4, e32, m1, ta, ma
-; RV64ZVE32F-NEXT: vlse32.v v8, (a0), zero
+; RV64ZVE32F-NEXT: vlse32.v v8, (a2), zero
+; RV64ZVE32F-NEXT: lw a2, 0(a3)
+; RV64ZVE32F-NEXT: lw a0, 0(a0)
; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a1
; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a2
-; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a3
+; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a0
; RV64ZVE32F-NEXT: ret
%v = call <4 x i32> @llvm.masked.gather.v4i32.v4p0(<4 x ptr> %ptrs, i32 4, <4 x i1> splat (i1 1), <4 x i32> %passthru)
ret <4 x i32> %v
@@ -2453,8 +2453,8 @@ define <8 x i32> @mgather_baseidx_v8i8_v8i32(ptr %base, <8 x i8> %idxs, <8 x i1>
; RV64ZVE32F-NEXT: .LBB35_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB35_12
; RV64ZVE32F-NEXT: # %bb.5: # %else5
@@ -2478,8 +2478,8 @@ define <8 x i32> @mgather_baseidx_v8i8_v8i32(ptr %base, <8 x i8> %idxs, <8 x i1>
; RV64ZVE32F-NEXT: vsetivli zero, 6, e32, m2, tu, ma
; RV64ZVE32F-NEXT: vslideup.vi v10, v8, 5
; RV64ZVE32F-NEXT: .LBB35_9: # %else14
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v9, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB35_15
; RV64ZVE32F-NEXT: # %bb.10: # %else17
@@ -2603,8 +2603,8 @@ define <8 x i32> @mgather_baseidx_sext_v8i8_v8i32(ptr %base, <8 x i8> %idxs, <8
; RV64ZVE32F-NEXT: .LBB36_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB36_12
; RV64ZVE32F-NEXT: # %bb.5: # %else5
@@ -2628,8 +2628,8 @@ define <8 x i32> @mgather_baseidx_sext_v8i8_v8i32(ptr %base, <8 x i8> %idxs, <8
; RV64ZVE32F-NEXT: vsetivli zero, 6, e32, m2, tu, ma
; RV64ZVE32F-NEXT: vslideup.vi v10, v8, 5
; RV64ZVE32F-NEXT: .LBB36_9: # %else14
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v9, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB36_15
; RV64ZVE32F-NEXT: # %bb.10: # %else17
@@ -2757,8 +2757,8 @@ define <8 x i32> @mgather_baseidx_zext_v8i8_v8i32(ptr %base, <8 x i8> %idxs, <8
; RV64ZVE32F-NEXT: .LBB37_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB37_12
; RV64ZVE32F-NEXT: # %bb.5: # %else5
@@ -2783,8 +2783,8 @@ define <8 x i32> @mgather_baseidx_zext_v8i8_v8i32(ptr %base, <8 x i8> %idxs, <8
; RV64ZVE32F-NEXT: vsetivli zero, 6, e32, m2, tu, ma
; RV64ZVE32F-NEXT: vslideup.vi v10, v8, 5
; RV64ZVE32F-NEXT: .LBB37_9: # %else14
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v9, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB37_15
; RV64ZVE32F-NEXT: # %bb.10: # %else17
@@ -2915,8 +2915,8 @@ define <8 x i32> @mgather_baseidx_v8i16_v8i32(ptr %base, <8 x i16> %idxs, <8 x i
; RV64ZVE32F-NEXT: .LBB38_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e16, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB38_12
; RV64ZVE32F-NEXT: # %bb.5: # %else5
@@ -2940,8 +2940,8 @@ define <8 x i32> @mgather_baseidx_v8i16_v8i32(ptr %base, <8 x i16> %idxs, <8 x i
; RV64ZVE32F-NEXT: vsetivli zero, 6, e32, m2, tu, ma
; RV64ZVE32F-NEXT: vslideup.vi v10, v8, 5
; RV64ZVE32F-NEXT: .LBB38_9: # %else14
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v9, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB38_15
; RV64ZVE32F-NEXT: # %bb.10: # %else17
@@ -3066,8 +3066,8 @@ define <8 x i32> @mgather_baseidx_sext_v8i16_v8i32(ptr %base, <8 x i16> %idxs, <
; RV64ZVE32F-NEXT: .LBB39_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e16, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB39_12
; RV64ZVE32F-NEXT: # %bb.5: # %else5
@@ -3091,8 +3091,8 @@ define <8 x i32> @mgather_baseidx_sext_v8i16_v8i32(ptr %base, <8 x i16> %idxs, <
; RV64ZVE32F-NEXT: vsetivli zero, 6, e32, m2, tu, ma
; RV64ZVE32F-NEXT: vslideup.vi v10, v8, 5
; RV64ZVE32F-NEXT: .LBB39_9: # %else14
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v9, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB39_15
; RV64ZVE32F-NEXT: # %bb.10: # %else17
@@ -3221,8 +3221,8 @@ define <8 x i32> @mgather_baseidx_zext_v8i16_v8i32(ptr %base, <8 x i16> %idxs, <
; RV64ZVE32F-NEXT: .LBB40_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e16, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: andi a3, a2, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a3, .LBB40_12
; RV64ZVE32F-NEXT: # %bb.5: # %else5
@@ -3247,8 +3247,8 @@ define <8 x i32> @mgather_baseidx_zext_v8i16_v8i32(ptr %base, <8 x i16> %idxs, <
; RV64ZVE32F-NEXT: vsetivli zero, 6, e32, m2, tu, ma
; RV64ZVE32F-NEXT: vslideup.vi v10, v8, 5
; RV64ZVE32F-NEXT: .LBB40_9: # %else14
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: andi a3, a2, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v9, 2
; RV64ZVE32F-NEXT: bnez a3, .LBB40_15
; RV64ZVE32F-NEXT: # %bb.10: # %else17
@@ -3376,8 +3376,8 @@ define <8 x i32> @mgather_baseidx_v8i32(ptr %base, <8 x i32> %idxs, <8 x i1> %m,
; RV64ZVE32F-NEXT: .LBB41_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e32, m2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v12, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e32, m1, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e32, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB41_12
; RV64ZVE32F-NEXT: # %bb.5: # %else5
@@ -3400,8 +3400,8 @@ define <8 x i32> @mgather_baseidx_v8i32(ptr %base, <8 x i32> %idxs, <8 x i1> %m,
; RV64ZVE32F-NEXT: vsetivli zero, 6, e32, m2, tu, ma
; RV64ZVE32F-NEXT: vslideup.vi v10, v8, 5
; RV64ZVE32F-NEXT: .LBB41_9: # %else14
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e32, m1, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e32, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v12, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB41_15
; RV64ZVE32F-NEXT: # %bb.10: # %else17
@@ -4090,13 +4090,13 @@ define <8 x i64> @mgather_baseidx_v8i8_v8i64(ptr %base, <8 x i8> %idxs, <8 x i1>
; RV32ZVE32F-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vsext.vf4 v10, v8
; RV32ZVE32F-NEXT: vsll.vi v8, v10, 3
-; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
; RV32ZVE32F-NEXT: vsetvli zero, zero, e8, mf2, ta, ma
; RV32ZVE32F-NEXT: vmv.x.s t0, v0
-; RV32ZVE32F-NEXT: andi a1, t0, 1
-; RV32ZVE32F-NEXT: beqz a1, .LBB48_7
-; RV32ZVE32F-NEXT: # %bb.1: # %cond.load
+; RV32ZVE32F-NEXT: andi a3, t0, 1
; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
+; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
+; RV32ZVE32F-NEXT: beqz a3, .LBB48_7
+; RV32ZVE32F-NEXT: # %bb.1: # %cond.load
; RV32ZVE32F-NEXT: vmv.x.s a3, v8
; RV32ZVE32F-NEXT: lw a1, 4(a3)
; RV32ZVE32F-NEXT: lw a3, 0(a3)
@@ -4253,8 +4253,8 @@ define <8 x i64> @mgather_baseidx_v8i8_v8i64(ptr %base, <8 x i8> %idxs, <8 x i1>
; RV64ZVE32F-NEXT: .LBB48_5: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a6, a5, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: beqz a6, .LBB48_10
; RV64ZVE32F-NEXT: # %bb.6: # %cond.load4
@@ -4364,13 +4364,13 @@ define <8 x i64> @mgather_baseidx_sext_v8i8_v8i64(ptr %base, <8 x i8> %idxs, <8
; RV32ZVE32F-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vsext.vf4 v10, v8
; RV32ZVE32F-NEXT: vsll.vi v8, v10, 3
-; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
; RV32ZVE32F-NEXT: vsetvli zero, zero, e8, mf2, ta, ma
; RV32ZVE32F-NEXT: vmv.x.s t0, v0
-; RV32ZVE32F-NEXT: andi a1, t0, 1
-; RV32ZVE32F-NEXT: beqz a1, .LBB49_7
-; RV32ZVE32F-NEXT: # %bb.1: # %cond.load
+; RV32ZVE32F-NEXT: andi a3, t0, 1
; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
+; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
+; RV32ZVE32F-NEXT: beqz a3, .LBB49_7
+; RV32ZVE32F-NEXT: # %bb.1: # %cond.load
; RV32ZVE32F-NEXT: vmv.x.s a3, v8
; RV32ZVE32F-NEXT: lw a1, 4(a3)
; RV32ZVE32F-NEXT: lw a3, 0(a3)
@@ -4527,8 +4527,8 @@ define <8 x i64> @mgather_baseidx_sext_v8i8_v8i64(ptr %base, <8 x i8> %idxs, <8
; RV64ZVE32F-NEXT: .LBB49_5: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a6, a5, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: beqz a6, .LBB49_10
; RV64ZVE32F-NEXT: # %bb.6: # %cond.load4
@@ -4640,13 +4640,13 @@ define <8 x i64> @mgather_baseidx_zext_v8i8_v8i64(ptr %base, <8 x i8> %idxs, <8
; RV32ZVE32F-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vzext.vf4 v10, v8
; RV32ZVE32F-NEXT: vsll.vi v8, v10, 3
-; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
; RV32ZVE32F-NEXT: vsetvli zero, zero, e8, mf2, ta, ma
; RV32ZVE32F-NEXT: vmv.x.s t0, v0
-; RV32ZVE32F-NEXT: andi a1, t0, 1
-; RV32ZVE32F-NEXT: beqz a1, .LBB50_7
-; RV32ZVE32F-NEXT: # %bb.1: # %cond.load
+; RV32ZVE32F-NEXT: andi a3, t0, 1
; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
+; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
+; RV32ZVE32F-NEXT: beqz a3, .LBB50_7
+; RV32ZVE32F-NEXT: # %bb.1: # %cond.load
; RV32ZVE32F-NEXT: vmv.x.s a3, v8
; RV32ZVE32F-NEXT: lw a1, 4(a3)
; RV32ZVE32F-NEXT: lw a3, 0(a3)
@@ -4805,8 +4805,8 @@ define <8 x i64> @mgather_baseidx_zext_v8i8_v8i64(ptr %base, <8 x i8> %idxs, <8
; RV64ZVE32F-NEXT: .LBB50_5: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a6, a5, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: beqz a6, .LBB50_10
; RV64ZVE32F-NEXT: # %bb.6: # %cond.load4
@@ -4923,13 +4923,13 @@ define <8 x i64> @mgather_baseidx_v8i16_v8i64(ptr %base, <8 x i16> %idxs, <8 x i
; RV32ZVE32F-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vsext.vf2 v10, v8
; RV32ZVE32F-NEXT: vsll.vi v8, v10, 3
-; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
; RV32ZVE32F-NEXT: vsetvli zero, zero, e8, mf2, ta, ma
; RV32ZVE32F-NEXT: vmv.x.s t0, v0
-; RV32ZVE32F-NEXT: andi a1, t0, 1
-; RV32ZVE32F-NEXT: beqz a1, .LBB51_7
-; RV32ZVE32F-NEXT: # %bb.1: # %cond.load
+; RV32ZVE32F-NEXT: andi a3, t0, 1
; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
+; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
+; RV32ZVE32F-NEXT: beqz a3, .LBB51_7
+; RV32ZVE32F-NEXT: # %bb.1: # %cond.load
; RV32ZVE32F-NEXT: vmv.x.s a3, v8
; RV32ZVE32F-NEXT: lw a1, 4(a3)
; RV32ZVE32F-NEXT: lw a3, 0(a3)
@@ -5087,8 +5087,8 @@ define <8 x i64> @mgather_baseidx_v8i16_v8i64(ptr %base, <8 x i16> %idxs, <8 x i
; RV64ZVE32F-NEXT: .LBB51_5: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e16, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: andi a6, a5, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: beqz a6, .LBB51_10
; RV64ZVE32F-NEXT: # %bb.6: # %cond.load4
@@ -5198,13 +5198,13 @@ define <8 x i64> @mgather_baseidx_sext_v8i16_v8i64(ptr %base, <8 x i16> %idxs, <
; RV32ZVE32F-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vsext.vf2 v10, v8
; RV32ZVE32F-NEXT: vsll.vi v8, v10, 3
-; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
; RV32ZVE32F-NEXT: vsetvli zero, zero, e8, mf2, ta, ma
; RV32ZVE32F-NEXT: vmv.x.s t0, v0
-; RV32ZVE32F-NEXT: andi a1, t0, 1
-; RV32ZVE32F-NEXT: beqz a1, .LBB52_7
-; RV32ZVE32F-NEXT: # %bb.1: # %cond.load
+; RV32ZVE32F-NEXT: andi a3, t0, 1
; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
+; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
+; RV32ZVE32F-NEXT: beqz a3, .LBB52_7
+; RV32ZVE32F-NEXT: # %bb.1: # %cond.load
; RV32ZVE32F-NEXT: vmv.x.s a3, v8
; RV32ZVE32F-NEXT: lw a1, 4(a3)
; RV32ZVE32F-NEXT: lw a3, 0(a3)
@@ -5362,8 +5362,8 @@ define <8 x i64> @mgather_baseidx_sext_v8i16_v8i64(ptr %base, <8 x i16> %idxs, <
; RV64ZVE32F-NEXT: .LBB52_5: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e16, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: andi a6, a5, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: beqz a6, .LBB52_10
; RV64ZVE32F-NEXT: # %bb.6: # %cond.load4
@@ -5475,13 +5475,13 @@ define <8 x i64> @mgather_baseidx_zext_v8i16_v8i64(ptr %base, <8 x i16> %idxs, <
; RV32ZVE32F-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vzext.vf2 v10, v8
; RV32ZVE32F-NEXT: vsll.vi v8, v10, 3
-; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
; RV32ZVE32F-NEXT: vsetvli zero, zero, e8, mf2, ta, ma
; RV32ZVE32F-NEXT: vmv.x.s t0, v0
-; RV32ZVE32F-NEXT: andi a1, t0, 1
-; RV32ZVE32F-NEXT: beqz a1, .LBB53_7
-; RV32ZVE32F-NEXT: # %bb.1: # %cond.load
+; RV32ZVE32F-NEXT: andi a3, t0, 1
; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
+; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
+; RV32ZVE32F-NEXT: beqz a3, .LBB53_7
+; RV32ZVE32F-NEXT: # %bb.1: # %cond.load
; RV32ZVE32F-NEXT: vmv.x.s a3, v8
; RV32ZVE32F-NEXT: lw a1, 4(a3)
; RV32ZVE32F-NEXT: lw a3, 0(a3)
@@ -5643,8 +5643,8 @@ define <8 x i64> @mgather_baseidx_zext_v8i16_v8i64(ptr %base, <8 x i16> %idxs, <
; RV64ZVE32F-NEXT: .LBB53_5: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e16, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: andi a7, a6, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: beqz a7, .LBB53_10
; RV64ZVE32F-NEXT: # %bb.6: # %cond.load4
@@ -5759,13 +5759,13 @@ define <8 x i64> @mgather_baseidx_v8i32_v8i64(ptr %base, <8 x i32> %idxs, <8 x i
; RV32ZVE32F: # %bb.0:
; RV32ZVE32F-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vsll.vi v8, v8, 3
-; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
; RV32ZVE32F-NEXT: vsetvli zero, zero, e8, mf2, ta, ma
; RV32ZVE32F-NEXT: vmv.x.s t0, v0
-; RV32ZVE32F-NEXT: andi a1, t0, 1
-; RV32ZVE32F-NEXT: beqz a1, .LBB54_7
-; RV32ZVE32F-NEXT: # %bb.1: # %cond.load
+; RV32ZVE32F-NEXT: andi a3, t0, 1
; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
+; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
+; RV32ZVE32F-NEXT: beqz a3, .LBB54_7
+; RV32ZVE32F-NEXT: # %bb.1: # %cond.load
; RV32ZVE32F-NEXT: vmv.x.s a3, v8
; RV32ZVE32F-NEXT: lw a1, 4(a3)
; RV32ZVE32F-NEXT: lw a3, 0(a3)
@@ -5923,8 +5923,8 @@ define <8 x i64> @mgather_baseidx_v8i32_v8i64(ptr %base, <8 x i32> %idxs, <8 x i
; RV64ZVE32F-NEXT: .LBB54_5: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e32, m2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e32, m1, ta, ma
; RV64ZVE32F-NEXT: andi a6, a5, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e32, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: beqz a6, .LBB54_10
; RV64ZVE32F-NEXT: # %bb.6: # %cond.load4
@@ -6032,13 +6032,13 @@ define <8 x i64> @mgather_baseidx_sext_v8i32_v8i64(ptr %base, <8 x i32> %idxs, <
; RV32ZVE32F: # %bb.0:
; RV32ZVE32F-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vsll.vi v8, v8, 3
-; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
; RV32ZVE32F-NEXT: vsetvli zero, zero, e8, mf2, ta, ma
; RV32ZVE32F-NEXT: vmv.x.s t0, v0
-; RV32ZVE32F-NEXT: andi a1, t0, 1
-; RV32ZVE32F-NEXT: beqz a1, .LBB55_7
-; RV32ZVE32F-NEXT: # %bb.1: # %cond.load
+; RV32ZVE32F-NEXT: andi a3, t0, 1
; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
+; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
+; RV32ZVE32F-NEXT: beqz a3, .LBB55_7
+; RV32ZVE32F-NEXT: # %bb.1: # %cond.load
; RV32ZVE32F-NEXT: vmv.x.s a3, v8
; RV32ZVE32F-NEXT: lw a1, 4(a3)
; RV32ZVE32F-NEXT: lw a3, 0(a3)
@@ -6196,8 +6196,8 @@ define <8 x i64> @mgather_baseidx_sext_v8i32_v8i64(ptr %base, <8 x i32> %idxs, <
; RV64ZVE32F-NEXT: .LBB55_5: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e32, m2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e32, m1, ta, ma
; RV64ZVE32F-NEXT: andi a6, a5, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e32, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: beqz a6, .LBB55_10
; RV64ZVE32F-NEXT: # %bb.6: # %cond.load4
@@ -6306,13 +6306,13 @@ define <8 x i64> @mgather_baseidx_zext_v8i32_v8i64(ptr %base, <8 x i32> %idxs, <
; RV32ZVE32F: # %bb.0:
; RV32ZVE32F-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vsll.vi v8, v8, 3
-; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
; RV32ZVE32F-NEXT: vsetvli zero, zero, e8, mf2, ta, ma
; RV32ZVE32F-NEXT: vmv.x.s t0, v0
-; RV32ZVE32F-NEXT: andi a1, t0, 1
-; RV32ZVE32F-NEXT: beqz a1, .LBB56_7
-; RV32ZVE32F-NEXT: # %bb.1: # %cond.load
+; RV32ZVE32F-NEXT: andi a3, t0, 1
; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
+; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
+; RV32ZVE32F-NEXT: beqz a3, .LBB56_7
+; RV32ZVE32F-NEXT: # %bb.1: # %cond.load
; RV32ZVE32F-NEXT: vmv.x.s a3, v8
; RV32ZVE32F-NEXT: lw a1, 4(a3)
; RV32ZVE32F-NEXT: lw a3, 0(a3)
@@ -6472,8 +6472,8 @@ define <8 x i64> @mgather_baseidx_zext_v8i32_v8i64(ptr %base, <8 x i32> %idxs, <
; RV64ZVE32F-NEXT: .LBB56_5: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e32, m2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e32, m1, ta, ma
; RV64ZVE32F-NEXT: andi a6, a5, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e32, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: beqz a6, .LBB56_10
; RV64ZVE32F-NEXT: # %bb.6: # %cond.load4
@@ -6589,27 +6589,27 @@ define <8 x i64> @mgather_baseidx_v8i64(ptr %base, <8 x i64> %idxs, <8 x i1> %m,
; RV32ZVE32F-NEXT: lw a4, 56(a2)
; RV32ZVE32F-NEXT: lw a5, 48(a2)
; RV32ZVE32F-NEXT: lw a6, 40(a2)
-; RV32ZVE32F-NEXT: lw a7, 32(a2)
-; RV32ZVE32F-NEXT: lw t0, 24(a2)
-; RV32ZVE32F-NEXT: lw t1, 16(a2)
-; RV32ZVE32F-NEXT: lw t2, 8(a2)
+; RV32ZVE32F-NEXT: lw a7, 8(a2)
; RV32ZVE32F-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vlse32.v v8, (a2), zero
-; RV32ZVE32F-NEXT: vslide1down.vx v8, v8, t2
-; RV32ZVE32F-NEXT: vslide1down.vx v8, v8, t1
-; RV32ZVE32F-NEXT: vslide1down.vx v8, v8, t0
+; RV32ZVE32F-NEXT: lw t0, 16(a2)
+; RV32ZVE32F-NEXT: lw t1, 24(a2)
+; RV32ZVE32F-NEXT: lw a2, 32(a2)
; RV32ZVE32F-NEXT: vslide1down.vx v8, v8, a7
+; RV32ZVE32F-NEXT: vslide1down.vx v8, v8, t0
+; RV32ZVE32F-NEXT: vslide1down.vx v8, v8, t1
+; RV32ZVE32F-NEXT: vslide1down.vx v8, v8, a2
; RV32ZVE32F-NEXT: vslide1down.vx v8, v8, a6
; RV32ZVE32F-NEXT: vslide1down.vx v8, v8, a5
; RV32ZVE32F-NEXT: vslide1down.vx v8, v8, a4
; RV32ZVE32F-NEXT: vsll.vi v8, v8, 3
-; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
; RV32ZVE32F-NEXT: vsetvli zero, zero, e8, mf2, ta, ma
; RV32ZVE32F-NEXT: vmv.x.s t0, v0
-; RV32ZVE32F-NEXT: andi a1, t0, 1
-; RV32ZVE32F-NEXT: beqz a1, .LBB57_7
-; RV32ZVE32F-NEXT: # %bb.1: # %cond.load
+; RV32ZVE32F-NEXT: andi a2, t0, 1
; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
+; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
+; RV32ZVE32F-NEXT: beqz a2, .LBB57_7
+; RV32ZVE32F-NEXT: # %bb.1: # %cond.load
; RV32ZVE32F-NEXT: vmv.x.s a2, v8
; RV32ZVE32F-NEXT: lw a1, 4(a2)
; RV32ZVE32F-NEXT: lw a2, 0(a2)
@@ -7017,14 +7017,14 @@ define <4 x half> @mgather_truemask_v4f16(<4 x ptr> %ptrs, <4 x half> %passthru)
; RV64ZVE32F-LABEL: mgather_truemask_v4f16:
; RV64ZVE32F: # %bb.0:
; RV64ZVE32F-NEXT: ld a1, 8(a0)
-; RV64ZVE32F-NEXT: ld a2, 16(a0)
-; RV64ZVE32F-NEXT: ld a3, 24(a0)
-; RV64ZVE32F-NEXT: ld a0, 0(a0)
+; RV64ZVE32F-NEXT: ld a2, 0(a0)
+; RV64ZVE32F-NEXT: ld a3, 16(a0)
+; RV64ZVE32F-NEXT: ld a0, 24(a0)
; RV64ZVE32F-NEXT: flh fa5, 0(a1)
-; RV64ZVE32F-NEXT: flh fa4, 0(a2)
-; RV64ZVE32F-NEXT: flh fa3, 0(a3)
; RV64ZVE32F-NEXT: vsetivli zero, 4, e16, mf2, ta, ma
-; RV64ZVE32F-NEXT: vlse16.v v8, (a0), zero
+; RV64ZVE32F-NEXT: vlse16.v v8, (a2), zero
+; RV64ZVE32F-NEXT: flh fa4, 0(a3)
+; RV64ZVE32F-NEXT: flh fa3, 0(a0)
; RV64ZVE32F-NEXT: vfslide1down.vf v8, v8, fa5
; RV64ZVE32F-NEXT: vfslide1down.vf v8, v8, fa4
; RV64ZVE32F-NEXT: vfslide1down.vf v8, v8, fa3
@@ -7214,8 +7214,8 @@ define <8 x half> @mgather_baseidx_v8i8_v8f16(ptr %base, <8 x i8> %idxs, <8 x i1
; RV64ZVE32F-NEXT: .LBB64_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB64_12
; RV64ZVE32F-NEXT: # %bb.5: # %else5
@@ -7239,8 +7239,8 @@ define <8 x half> @mgather_baseidx_v8i8_v8f16(ptr %base, <8 x i8> %idxs, <8 x i1
; RV64ZVE32F-NEXT: vsetivli zero, 6, e16, m1, tu, ma
; RV64ZVE32F-NEXT: vslideup.vi v9, v8, 5
; RV64ZVE32F-NEXT: .LBB64_9: # %else14
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v10, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB64_15
; RV64ZVE32F-NEXT: # %bb.10: # %else17
@@ -7365,8 +7365,8 @@ define <8 x half> @mgather_baseidx_sext_v8i8_v8f16(ptr %base, <8 x i8> %idxs, <8
; RV64ZVE32F-NEXT: .LBB65_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB65_12
; RV64ZVE32F-NEXT: # %bb.5: # %else5
@@ -7390,8 +7390,8 @@ define <8 x half> @mgather_baseidx_sext_v8i8_v8f16(ptr %base, <8 x i8> %idxs, <8
; RV64ZVE32F-NEXT: vsetivli zero, 6, e16, m1, tu, ma
; RV64ZVE32F-NEXT: vslideup.vi v9, v8, 5
; RV64ZVE32F-NEXT: .LBB65_9: # %else14
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v10, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB65_15
; RV64ZVE32F-NEXT: # %bb.10: # %else17
@@ -7517,8 +7517,8 @@ define <8 x half> @mgather_baseidx_zext_v8i8_v8f16(ptr %base, <8 x i8> %idxs, <8
; RV64ZVE32F-NEXT: .LBB66_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB66_12
; RV64ZVE32F-NEXT: # %bb.5: # %else5
@@ -7543,8 +7543,8 @@ define <8 x half> @mgather_baseidx_zext_v8i8_v8f16(ptr %base, <8 x i8> %idxs, <8
; RV64ZVE32F-NEXT: vsetivli zero, 6, e16, m1, tu, ma
; RV64ZVE32F-NEXT: vslideup.vi v9, v8, 5
; RV64ZVE32F-NEXT: .LBB66_9: # %else14
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v10, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB66_15
; RV64ZVE32F-NEXT: # %bb.10: # %else17
@@ -7672,8 +7672,8 @@ define <8 x half> @mgather_baseidx_v8f16(ptr %base, <8 x i16> %idxs, <8 x i1> %m
; RV64ZVE32F-NEXT: .LBB67_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e16, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB67_12
; RV64ZVE32F-NEXT: # %bb.5: # %else5
@@ -7696,8 +7696,8 @@ define <8 x half> @mgather_baseidx_v8f16(ptr %base, <8 x i16> %idxs, <8 x i1> %m
; RV64ZVE32F-NEXT: vsetivli zero, 6, e16, m1, tu, ma
; RV64ZVE32F-NEXT: vslideup.vi v9, v8, 5
; RV64ZVE32F-NEXT: .LBB67_9: # %else14
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v10, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB67_15
; RV64ZVE32F-NEXT: # %bb.10: # %else17
@@ -7940,14 +7940,14 @@ define <4 x float> @mgather_truemask_v4f32(<4 x ptr> %ptrs, <4 x float> %passthr
; RV64ZVE32F-LABEL: mgather_truemask_v4f32:
; RV64ZVE32F: # %bb.0:
; RV64ZVE32F-NEXT: ld a1, 8(a0)
-; RV64ZVE32F-NEXT: ld a2, 16(a0)
-; RV64ZVE32F-NEXT: ld a3, 24(a0)
-; RV64ZVE32F-NEXT: ld a0, 0(a0)
+; RV64ZVE32F-NEXT: ld a2, 0(a0)
+; RV64ZVE32F-NEXT: ld a3, 16(a0)
+; RV64ZVE32F-NEXT: ld a0, 24(a0)
; RV64ZVE32F-NEXT: flw fa5, 0(a1)
-; RV64ZVE32F-NEXT: flw fa4, 0(a2)
-; RV64ZVE32F-NEXT: flw fa3, 0(a3)
; RV64ZVE32F-NEXT: vsetivli zero, 4, e32, m1, ta, ma
-; RV64ZVE32F-NEXT: vlse32.v v8, (a0), zero
+; RV64ZVE32F-NEXT: vlse32.v v8, (a2), zero
+; RV64ZVE32F-NEXT: flw fa4, 0(a3)
+; RV64ZVE32F-NEXT: flw fa3, 0(a0)
; RV64ZVE32F-NEXT: vfslide1down.vf v8, v8, fa5
; RV64ZVE32F-NEXT: vfslide1down.vf v8, v8, fa4
; RV64ZVE32F-NEXT: vfslide1down.vf v8, v8, fa3
@@ -8136,8 +8136,8 @@ define <8 x float> @mgather_baseidx_v8i8_v8f32(ptr %base, <8 x i8> %idxs, <8 x i
; RV64ZVE32F-NEXT: .LBB74_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB74_12
; RV64ZVE32F-NEXT: # %bb.5: # %else5
@@ -8161,8 +8161,8 @@ define <8 x float> @mgather_baseidx_v8i8_v8f32(ptr %base, <8 x i8> %idxs, <8 x i
; RV64ZVE32F-NEXT: vsetivli zero, 6, e32, m2, tu, ma
; RV64ZVE32F-NEXT: vslideup.vi v10, v8, 5
; RV64ZVE32F-NEXT: .LBB74_9: # %else14
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v9, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB74_15
; RV64ZVE32F-NEXT: # %bb.10: # %else17
@@ -8286,8 +8286,8 @@ define <8 x float> @mgather_baseidx_sext_v8i8_v8f32(ptr %base, <8 x i8> %idxs, <
; RV64ZVE32F-NEXT: .LBB75_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB75_12
; RV64ZVE32F-NEXT: # %bb.5: # %else5
@@ -8311,8 +8311,8 @@ define <8 x float> @mgather_baseidx_sext_v8i8_v8f32(ptr %base, <8 x i8> %idxs, <
; RV64ZVE32F-NEXT: vsetivli zero, 6, e32, m2, tu, ma
; RV64ZVE32F-NEXT: vslideup.vi v10, v8, 5
; RV64ZVE32F-NEXT: .LBB75_9: # %else14
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v9, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB75_15
; RV64ZVE32F-NEXT: # %bb.10: # %else17
@@ -8440,8 +8440,8 @@ define <8 x float> @mgather_baseidx_zext_v8i8_v8f32(ptr %base, <8 x i8> %idxs, <
; RV64ZVE32F-NEXT: .LBB76_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB76_12
; RV64ZVE32F-NEXT: # %bb.5: # %else5
@@ -8466,8 +8466,8 @@ define <8 x float> @mgather_baseidx_zext_v8i8_v8f32(ptr %base, <8 x i8> %idxs, <
; RV64ZVE32F-NEXT: vsetivli zero, 6, e32, m2, tu, ma
; RV64ZVE32F-NEXT: vslideup.vi v10, v8, 5
; RV64ZVE32F-NEXT: .LBB76_9: # %else14
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v9, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB76_15
; RV64ZVE32F-NEXT: # %bb.10: # %else17
@@ -8598,8 +8598,8 @@ define <8 x float> @mgather_baseidx_v8i16_v8f32(ptr %base, <8 x i16> %idxs, <8 x
; RV64ZVE32F-NEXT: .LBB77_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e16, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB77_12
; RV64ZVE32F-NEXT: # %bb.5: # %else5
@@ -8623,8 +8623,8 @@ define <8 x float> @mgather_baseidx_v8i16_v8f32(ptr %base, <8 x i16> %idxs, <8 x
; RV64ZVE32F-NEXT: vsetivli zero, 6, e32, m2, tu, ma
; RV64ZVE32F-NEXT: vslideup.vi v10, v8, 5
; RV64ZVE32F-NEXT: .LBB77_9: # %else14
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v9, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB77_15
; RV64ZVE32F-NEXT: # %bb.10: # %else17
@@ -8749,8 +8749,8 @@ define <8 x float> @mgather_baseidx_sext_v8i16_v8f32(ptr %base, <8 x i16> %idxs,
; RV64ZVE32F-NEXT: .LBB78_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e16, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB78_12
; RV64ZVE32F-NEXT: # %bb.5: # %else5
@@ -8774,8 +8774,8 @@ define <8 x float> @mgather_baseidx_sext_v8i16_v8f32(ptr %base, <8 x i16> %idxs,
; RV64ZVE32F-NEXT: vsetivli zero, 6, e32, m2, tu, ma
; RV64ZVE32F-NEXT: vslideup.vi v10, v8, 5
; RV64ZVE32F-NEXT: .LBB78_9: # %else14
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v9, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB78_15
; RV64ZVE32F-NEXT: # %bb.10: # %else17
@@ -8904,8 +8904,8 @@ define <8 x float> @mgather_baseidx_zext_v8i16_v8f32(ptr %base, <8 x i16> %idxs,
; RV64ZVE32F-NEXT: .LBB79_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e16, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: andi a3, a2, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a3, .LBB79_12
; RV64ZVE32F-NEXT: # %bb.5: # %else5
@@ -8930,8 +8930,8 @@ define <8 x float> @mgather_baseidx_zext_v8i16_v8f32(ptr %base, <8 x i16> %idxs,
; RV64ZVE32F-NEXT: vsetivli zero, 6, e32, m2, tu, ma
; RV64ZVE32F-NEXT: vslideup.vi v10, v8, 5
; RV64ZVE32F-NEXT: .LBB79_9: # %else14
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: andi a3, a2, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v9, 2
; RV64ZVE32F-NEXT: bnez a3, .LBB79_15
; RV64ZVE32F-NEXT: # %bb.10: # %else17
@@ -9059,8 +9059,8 @@ define <8 x float> @mgather_baseidx_v8f32(ptr %base, <8 x i32> %idxs, <8 x i1> %
; RV64ZVE32F-NEXT: .LBB80_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e32, m2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v12, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e32, m1, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e32, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB80_12
; RV64ZVE32F-NEXT: # %bb.5: # %else5
@@ -9083,8 +9083,8 @@ define <8 x float> @mgather_baseidx_v8f32(ptr %base, <8 x i32> %idxs, <8 x i1> %
; RV64ZVE32F-NEXT: vsetivli zero, 6, e32, m2, tu, ma
; RV64ZVE32F-NEXT: vslideup.vi v10, v8, 5
; RV64ZVE32F-NEXT: .LBB80_9: # %else14
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e32, m1, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e32, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v12, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB80_15
; RV64ZVE32F-NEXT: # %bb.10: # %else17
@@ -9664,31 +9664,32 @@ define <8 x double> @mgather_baseidx_v8i8_v8f64(ptr %base, <8 x i8> %idxs, <8 x
; RV32ZVE32F-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vsext.vf4 v10, v8
; RV32ZVE32F-NEXT: vsll.vi v8, v10, 3
-; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
; RV32ZVE32F-NEXT: vsetvli zero, zero, e8, mf2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a1, v0
-; RV32ZVE32F-NEXT: andi a2, a1, 1
-; RV32ZVE32F-NEXT: bnez a2, .LBB87_10
+; RV32ZVE32F-NEXT: vmv.x.s a2, v0
+; RV32ZVE32F-NEXT: andi a3, a2, 1
+; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
+; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
+; RV32ZVE32F-NEXT: bnez a3, .LBB87_10
; RV32ZVE32F-NEXT: # %bb.1: # %else
-; RV32ZVE32F-NEXT: andi a2, a1, 2
-; RV32ZVE32F-NEXT: bnez a2, .LBB87_11
+; RV32ZVE32F-NEXT: andi a1, a2, 2
+; RV32ZVE32F-NEXT: bnez a1, .LBB87_11
; RV32ZVE32F-NEXT: .LBB87_2: # %else2
-; RV32ZVE32F-NEXT: andi a2, a1, 4
-; RV32ZVE32F-NEXT: bnez a2, .LBB87_12
+; RV32ZVE32F-NEXT: andi a1, a2, 4
+; RV32ZVE32F-NEXT: bnez a1, .LBB87_12
; RV32ZVE32F-NEXT: .LBB87_3: # %else5
-; RV32ZVE32F-NEXT: andi a2, a1, 8
-; RV32ZVE32F-NEXT: bnez a2, .LBB87_13
+; RV32ZVE32F-NEXT: andi a1, a2, 8
+; RV32ZVE32F-NEXT: bnez a1, .LBB87_13
; RV32ZVE32F-NEXT: .LBB87_4: # %else8
-; RV32ZVE32F-NEXT: andi a2, a1, 16
-; RV32ZVE32F-NEXT: bnez a2, .LBB87_14
+; RV32ZVE32F-NEXT: andi a1, a2, 16
+; RV32ZVE32F-NEXT: bnez a1, .LBB87_14
; RV32ZVE32F-NEXT: .LBB87_5: # %else11
-; RV32ZVE32F-NEXT: andi a2, a1, 32
-; RV32ZVE32F-NEXT: bnez a2, .LBB87_15
+; RV32ZVE32F-NEXT: andi a1, a2, 32
+; RV32ZVE32F-NEXT: bnez a1, .LBB87_15
; RV32ZVE32F-NEXT: .LBB87_6: # %else14
-; RV32ZVE32F-NEXT: andi a2, a1, 64
-; RV32ZVE32F-NEXT: bnez a2, .LBB87_16
+; RV32ZVE32F-NEXT: andi a1, a2, 64
+; RV32ZVE32F-NEXT: bnez a1, .LBB87_16
; RV32ZVE32F-NEXT: .LBB87_7: # %else17
-; RV32ZVE32F-NEXT: andi a1, a1, -128
+; RV32ZVE32F-NEXT: andi a1, a2, -128
; RV32ZVE32F-NEXT: beqz a1, .LBB87_9
; RV32ZVE32F-NEXT: .LBB87_8: # %cond.load19
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -9706,52 +9707,51 @@ define <8 x double> @mgather_baseidx_v8i8_v8f64(ptr %base, <8 x i8> %idxs, <8 x
; RV32ZVE32F-NEXT: fsd fa7, 56(a0)
; RV32ZVE32F-NEXT: ret
; RV32ZVE32F-NEXT: .LBB87_10: # %cond.load
-; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a2, v8
-; RV32ZVE32F-NEXT: fld fa0, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 2
-; RV32ZVE32F-NEXT: beqz a2, .LBB87_2
+; RV32ZVE32F-NEXT: vmv.x.s a1, v8
+; RV32ZVE32F-NEXT: fld fa0, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 2
+; RV32ZVE32F-NEXT: beqz a1, .LBB87_2
; RV32ZVE32F-NEXT: .LBB87_11: # %cond.load1
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 1
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa1, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 4
-; RV32ZVE32F-NEXT: beqz a2, .LBB87_3
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa1, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 4
+; RV32ZVE32F-NEXT: beqz a1, .LBB87_3
; RV32ZVE32F-NEXT: .LBB87_12: # %cond.load4
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 2
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa2, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 8
-; RV32ZVE32F-NEXT: beqz a2, .LBB87_4
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa2, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 8
+; RV32ZVE32F-NEXT: beqz a1, .LBB87_4
; RV32ZVE32F-NEXT: .LBB87_13: # %cond.load7
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 3
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa3, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 16
-; RV32ZVE32F-NEXT: beqz a2, .LBB87_5
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa3, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 16
+; RV32ZVE32F-NEXT: beqz a1, .LBB87_5
; RV32ZVE32F-NEXT: .LBB87_14: # %cond.load10
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 4
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa4, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 32
-; RV32ZVE32F-NEXT: beqz a2, .LBB87_6
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa4, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 32
+; RV32ZVE32F-NEXT: beqz a1, .LBB87_6
; RV32ZVE32F-NEXT: .LBB87_15: # %cond.load13
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 5
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa5, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 64
-; RV32ZVE32F-NEXT: beqz a2, .LBB87_7
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa5, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 64
+; RV32ZVE32F-NEXT: beqz a1, .LBB87_7
; RV32ZVE32F-NEXT: .LBB87_16: # %cond.load16
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 6
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa6, 0(a2)
-; RV32ZVE32F-NEXT: andi a1, a1, -128
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa6, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, -128
; RV32ZVE32F-NEXT: bnez a1, .LBB87_8
; RV32ZVE32F-NEXT: j .LBB87_9
;
@@ -9779,8 +9779,8 @@ define <8 x double> @mgather_baseidx_v8i8_v8f64(ptr %base, <8 x i8> %idxs, <8 x
; RV64ZVE32F-NEXT: .LBB87_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a3, a2, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a3, .LBB87_14
; RV64ZVE32F-NEXT: # %bb.5: # %else5
@@ -9879,31 +9879,32 @@ define <8 x double> @mgather_baseidx_sext_v8i8_v8f64(ptr %base, <8 x i8> %idxs,
; RV32ZVE32F-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vsext.vf4 v10, v8
; RV32ZVE32F-NEXT: vsll.vi v8, v10, 3
-; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
; RV32ZVE32F-NEXT: vsetvli zero, zero, e8, mf2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a1, v0
-; RV32ZVE32F-NEXT: andi a2, a1, 1
-; RV32ZVE32F-NEXT: bnez a2, .LBB88_10
+; RV32ZVE32F-NEXT: vmv.x.s a2, v0
+; RV32ZVE32F-NEXT: andi a3, a2, 1
+; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
+; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
+; RV32ZVE32F-NEXT: bnez a3, .LBB88_10
; RV32ZVE32F-NEXT: # %bb.1: # %else
-; RV32ZVE32F-NEXT: andi a2, a1, 2
-; RV32ZVE32F-NEXT: bnez a2, .LBB88_11
+; RV32ZVE32F-NEXT: andi a1, a2, 2
+; RV32ZVE32F-NEXT: bnez a1, .LBB88_11
; RV32ZVE32F-NEXT: .LBB88_2: # %else2
-; RV32ZVE32F-NEXT: andi a2, a1, 4
-; RV32ZVE32F-NEXT: bnez a2, .LBB88_12
+; RV32ZVE32F-NEXT: andi a1, a2, 4
+; RV32ZVE32F-NEXT: bnez a1, .LBB88_12
; RV32ZVE32F-NEXT: .LBB88_3: # %else5
-; RV32ZVE32F-NEXT: andi a2, a1, 8
-; RV32ZVE32F-NEXT: bnez a2, .LBB88_13
+; RV32ZVE32F-NEXT: andi a1, a2, 8
+; RV32ZVE32F-NEXT: bnez a1, .LBB88_13
; RV32ZVE32F-NEXT: .LBB88_4: # %else8
-; RV32ZVE32F-NEXT: andi a2, a1, 16
-; RV32ZVE32F-NEXT: bnez a2, .LBB88_14
+; RV32ZVE32F-NEXT: andi a1, a2, 16
+; RV32ZVE32F-NEXT: bnez a1, .LBB88_14
; RV32ZVE32F-NEXT: .LBB88_5: # %else11
-; RV32ZVE32F-NEXT: andi a2, a1, 32
-; RV32ZVE32F-NEXT: bnez a2, .LBB88_15
+; RV32ZVE32F-NEXT: andi a1, a2, 32
+; RV32ZVE32F-NEXT: bnez a1, .LBB88_15
; RV32ZVE32F-NEXT: .LBB88_6: # %else14
-; RV32ZVE32F-NEXT: andi a2, a1, 64
-; RV32ZVE32F-NEXT: bnez a2, .LBB88_16
+; RV32ZVE32F-NEXT: andi a1, a2, 64
+; RV32ZVE32F-NEXT: bnez a1, .LBB88_16
; RV32ZVE32F-NEXT: .LBB88_7: # %else17
-; RV32ZVE32F-NEXT: andi a1, a1, -128
+; RV32ZVE32F-NEXT: andi a1, a2, -128
; RV32ZVE32F-NEXT: beqz a1, .LBB88_9
; RV32ZVE32F-NEXT: .LBB88_8: # %cond.load19
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -9921,52 +9922,51 @@ define <8 x double> @mgather_baseidx_sext_v8i8_v8f64(ptr %base, <8 x i8> %idxs,
; RV32ZVE32F-NEXT: fsd fa7, 56(a0)
; RV32ZVE32F-NEXT: ret
; RV32ZVE32F-NEXT: .LBB88_10: # %cond.load
-; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a2, v8
-; RV32ZVE32F-NEXT: fld fa0, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 2
-; RV32ZVE32F-NEXT: beqz a2, .LBB88_2
+; RV32ZVE32F-NEXT: vmv.x.s a1, v8
+; RV32ZVE32F-NEXT: fld fa0, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 2
+; RV32ZVE32F-NEXT: beqz a1, .LBB88_2
; RV32ZVE32F-NEXT: .LBB88_11: # %cond.load1
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 1
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa1, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 4
-; RV32ZVE32F-NEXT: beqz a2, .LBB88_3
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa1, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 4
+; RV32ZVE32F-NEXT: beqz a1, .LBB88_3
; RV32ZVE32F-NEXT: .LBB88_12: # %cond.load4
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 2
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa2, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 8
-; RV32ZVE32F-NEXT: beqz a2, .LBB88_4
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa2, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 8
+; RV32ZVE32F-NEXT: beqz a1, .LBB88_4
; RV32ZVE32F-NEXT: .LBB88_13: # %cond.load7
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 3
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa3, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 16
-; RV32ZVE32F-NEXT: beqz a2, .LBB88_5
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa3, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 16
+; RV32ZVE32F-NEXT: beqz a1, .LBB88_5
; RV32ZVE32F-NEXT: .LBB88_14: # %cond.load10
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 4
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa4, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 32
-; RV32ZVE32F-NEXT: beqz a2, .LBB88_6
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa4, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 32
+; RV32ZVE32F-NEXT: beqz a1, .LBB88_6
; RV32ZVE32F-NEXT: .LBB88_15: # %cond.load13
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 5
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa5, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 64
-; RV32ZVE32F-NEXT: beqz a2, .LBB88_7
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa5, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 64
+; RV32ZVE32F-NEXT: beqz a1, .LBB88_7
; RV32ZVE32F-NEXT: .LBB88_16: # %cond.load16
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 6
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa6, 0(a2)
-; RV32ZVE32F-NEXT: andi a1, a1, -128
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa6, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, -128
; RV32ZVE32F-NEXT: bnez a1, .LBB88_8
; RV32ZVE32F-NEXT: j .LBB88_9
;
@@ -9994,8 +9994,8 @@ define <8 x double> @mgather_baseidx_sext_v8i8_v8f64(ptr %base, <8 x i8> %idxs,
; RV64ZVE32F-NEXT: .LBB88_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a3, a2, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a3, .LBB88_14
; RV64ZVE32F-NEXT: # %bb.5: # %else5
@@ -10096,31 +10096,32 @@ define <8 x double> @mgather_baseidx_zext_v8i8_v8f64(ptr %base, <8 x i8> %idxs,
; RV32ZVE32F-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vzext.vf4 v10, v8
; RV32ZVE32F-NEXT: vsll.vi v8, v10, 3
-; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
; RV32ZVE32F-NEXT: vsetvli zero, zero, e8, mf2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a1, v0
-; RV32ZVE32F-NEXT: andi a2, a1, 1
-; RV32ZVE32F-NEXT: bnez a2, .LBB89_10
+; RV32ZVE32F-NEXT: vmv.x.s a2, v0
+; RV32ZVE32F-NEXT: andi a3, a2, 1
+; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
+; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
+; RV32ZVE32F-NEXT: bnez a3, .LBB89_10
; RV32ZVE32F-NEXT: # %bb.1: # %else
-; RV32ZVE32F-NEXT: andi a2, a1, 2
-; RV32ZVE32F-NEXT: bnez a2, .LBB89_11
+; RV32ZVE32F-NEXT: andi a1, a2, 2
+; RV32ZVE32F-NEXT: bnez a1, .LBB89_11
; RV32ZVE32F-NEXT: .LBB89_2: # %else2
-; RV32ZVE32F-NEXT: andi a2, a1, 4
-; RV32ZVE32F-NEXT: bnez a2, .LBB89_12
+; RV32ZVE32F-NEXT: andi a1, a2, 4
+; RV32ZVE32F-NEXT: bnez a1, .LBB89_12
; RV32ZVE32F-NEXT: .LBB89_3: # %else5
-; RV32ZVE32F-NEXT: andi a2, a1, 8
-; RV32ZVE32F-NEXT: bnez a2, .LBB89_13
+; RV32ZVE32F-NEXT: andi a1, a2, 8
+; RV32ZVE32F-NEXT: bnez a1, .LBB89_13
; RV32ZVE32F-NEXT: .LBB89_4: # %else8
-; RV32ZVE32F-NEXT: andi a2, a1, 16
-; RV32ZVE32F-NEXT: bnez a2, .LBB89_14
+; RV32ZVE32F-NEXT: andi a1, a2, 16
+; RV32ZVE32F-NEXT: bnez a1, .LBB89_14
; RV32ZVE32F-NEXT: .LBB89_5: # %else11
-; RV32ZVE32F-NEXT: andi a2, a1, 32
-; RV32ZVE32F-NEXT: bnez a2, .LBB89_15
+; RV32ZVE32F-NEXT: andi a1, a2, 32
+; RV32ZVE32F-NEXT: bnez a1, .LBB89_15
; RV32ZVE32F-NEXT: .LBB89_6: # %else14
-; RV32ZVE32F-NEXT: andi a2, a1, 64
-; RV32ZVE32F-NEXT: bnez a2, .LBB89_16
+; RV32ZVE32F-NEXT: andi a1, a2, 64
+; RV32ZVE32F-NEXT: bnez a1, .LBB89_16
; RV32ZVE32F-NEXT: .LBB89_7: # %else17
-; RV32ZVE32F-NEXT: andi a1, a1, -128
+; RV32ZVE32F-NEXT: andi a1, a2, -128
; RV32ZVE32F-NEXT: beqz a1, .LBB89_9
; RV32ZVE32F-NEXT: .LBB89_8: # %cond.load19
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -10138,52 +10139,51 @@ define <8 x double> @mgather_baseidx_zext_v8i8_v8f64(ptr %base, <8 x i8> %idxs,
; RV32ZVE32F-NEXT: fsd fa7, 56(a0)
; RV32ZVE32F-NEXT: ret
; RV32ZVE32F-NEXT: .LBB89_10: # %cond.load
-; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a2, v8
-; RV32ZVE32F-NEXT: fld fa0, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 2
-; RV32ZVE32F-NEXT: beqz a2, .LBB89_2
+; RV32ZVE32F-NEXT: vmv.x.s a1, v8
+; RV32ZVE32F-NEXT: fld fa0, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 2
+; RV32ZVE32F-NEXT: beqz a1, .LBB89_2
; RV32ZVE32F-NEXT: .LBB89_11: # %cond.load1
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 1
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa1, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 4
-; RV32ZVE32F-NEXT: beqz a2, .LBB89_3
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa1, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 4
+; RV32ZVE32F-NEXT: beqz a1, .LBB89_3
; RV32ZVE32F-NEXT: .LBB89_12: # %cond.load4
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 2
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa2, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 8
-; RV32ZVE32F-NEXT: beqz a2, .LBB89_4
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa2, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 8
+; RV32ZVE32F-NEXT: beqz a1, .LBB89_4
; RV32ZVE32F-NEXT: .LBB89_13: # %cond.load7
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 3
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa3, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 16
-; RV32ZVE32F-NEXT: beqz a2, .LBB89_5
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa3, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 16
+; RV32ZVE32F-NEXT: beqz a1, .LBB89_5
; RV32ZVE32F-NEXT: .LBB89_14: # %cond.load10
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 4
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa4, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 32
-; RV32ZVE32F-NEXT: beqz a2, .LBB89_6
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa4, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 32
+; RV32ZVE32F-NEXT: beqz a1, .LBB89_6
; RV32ZVE32F-NEXT: .LBB89_15: # %cond.load13
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 5
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa5, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 64
-; RV32ZVE32F-NEXT: beqz a2, .LBB89_7
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa5, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 64
+; RV32ZVE32F-NEXT: beqz a1, .LBB89_7
; RV32ZVE32F-NEXT: .LBB89_16: # %cond.load16
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 6
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa6, 0(a2)
-; RV32ZVE32F-NEXT: andi a1, a1, -128
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa6, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, -128
; RV32ZVE32F-NEXT: bnez a1, .LBB89_8
; RV32ZVE32F-NEXT: j .LBB89_9
;
@@ -10213,8 +10213,8 @@ define <8 x double> @mgather_baseidx_zext_v8i8_v8f64(ptr %base, <8 x i8> %idxs,
; RV64ZVE32F-NEXT: .LBB89_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a3, a2, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a3, .LBB89_14
; RV64ZVE32F-NEXT: # %bb.5: # %else5
@@ -10320,31 +10320,32 @@ define <8 x double> @mgather_baseidx_v8i16_v8f64(ptr %base, <8 x i16> %idxs, <8
; RV32ZVE32F-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vsext.vf2 v10, v8
; RV32ZVE32F-NEXT: vsll.vi v8, v10, 3
-; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
; RV32ZVE32F-NEXT: vsetvli zero, zero, e8, mf2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a1, v0
-; RV32ZVE32F-NEXT: andi a2, a1, 1
-; RV32ZVE32F-NEXT: bnez a2, .LBB90_10
+; RV32ZVE32F-NEXT: vmv.x.s a2, v0
+; RV32ZVE32F-NEXT: andi a3, a2, 1
+; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
+; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
+; RV32ZVE32F-NEXT: bnez a3, .LBB90_10
; RV32ZVE32F-NEXT: # %bb.1: # %else
-; RV32ZVE32F-NEXT: andi a2, a1, 2
-; RV32ZVE32F-NEXT: bnez a2, .LBB90_11
+; RV32ZVE32F-NEXT: andi a1, a2, 2
+; RV32ZVE32F-NEXT: bnez a1, .LBB90_11
; RV32ZVE32F-NEXT: .LBB90_2: # %else2
-; RV32ZVE32F-NEXT: andi a2, a1, 4
-; RV32ZVE32F-NEXT: bnez a2, .LBB90_12
+; RV32ZVE32F-NEXT: andi a1, a2, 4
+; RV32ZVE32F-NEXT: bnez a1, .LBB90_12
; RV32ZVE32F-NEXT: .LBB90_3: # %else5
-; RV32ZVE32F-NEXT: andi a2, a1, 8
-; RV32ZVE32F-NEXT: bnez a2, .LBB90_13
+; RV32ZVE32F-NEXT: andi a1, a2, 8
+; RV32ZVE32F-NEXT: bnez a1, .LBB90_13
; RV32ZVE32F-NEXT: .LBB90_4: # %else8
-; RV32ZVE32F-NEXT: andi a2, a1, 16
-; RV32ZVE32F-NEXT: bnez a2, .LBB90_14
+; RV32ZVE32F-NEXT: andi a1, a2, 16
+; RV32ZVE32F-NEXT: bnez a1, .LBB90_14
; RV32ZVE32F-NEXT: .LBB90_5: # %else11
-; RV32ZVE32F-NEXT: andi a2, a1, 32
-; RV32ZVE32F-NEXT: bnez a2, .LBB90_15
+; RV32ZVE32F-NEXT: andi a1, a2, 32
+; RV32ZVE32F-NEXT: bnez a1, .LBB90_15
; RV32ZVE32F-NEXT: .LBB90_6: # %else14
-; RV32ZVE32F-NEXT: andi a2, a1, 64
-; RV32ZVE32F-NEXT: bnez a2, .LBB90_16
+; RV32ZVE32F-NEXT: andi a1, a2, 64
+; RV32ZVE32F-NEXT: bnez a1, .LBB90_16
; RV32ZVE32F-NEXT: .LBB90_7: # %else17
-; RV32ZVE32F-NEXT: andi a1, a1, -128
+; RV32ZVE32F-NEXT: andi a1, a2, -128
; RV32ZVE32F-NEXT: beqz a1, .LBB90_9
; RV32ZVE32F-NEXT: .LBB90_8: # %cond.load19
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -10362,52 +10363,51 @@ define <8 x double> @mgather_baseidx_v8i16_v8f64(ptr %base, <8 x i16> %idxs, <8
; RV32ZVE32F-NEXT: fsd fa7, 56(a0)
; RV32ZVE32F-NEXT: ret
; RV32ZVE32F-NEXT: .LBB90_10: # %cond.load
-; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a2, v8
-; RV32ZVE32F-NEXT: fld fa0, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 2
-; RV32ZVE32F-NEXT: beqz a2, .LBB90_2
+; RV32ZVE32F-NEXT: vmv.x.s a1, v8
+; RV32ZVE32F-NEXT: fld fa0, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 2
+; RV32ZVE32F-NEXT: beqz a1, .LBB90_2
; RV32ZVE32F-NEXT: .LBB90_11: # %cond.load1
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 1
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa1, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 4
-; RV32ZVE32F-NEXT: beqz a2, .LBB90_3
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa1, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 4
+; RV32ZVE32F-NEXT: beqz a1, .LBB90_3
; RV32ZVE32F-NEXT: .LBB90_12: # %cond.load4
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 2
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa2, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 8
-; RV32ZVE32F-NEXT: beqz a2, .LBB90_4
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa2, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 8
+; RV32ZVE32F-NEXT: beqz a1, .LBB90_4
; RV32ZVE32F-NEXT: .LBB90_13: # %cond.load7
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 3
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa3, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 16
-; RV32ZVE32F-NEXT: beqz a2, .LBB90_5
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa3, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 16
+; RV32ZVE32F-NEXT: beqz a1, .LBB90_5
; RV32ZVE32F-NEXT: .LBB90_14: # %cond.load10
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 4
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa4, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 32
-; RV32ZVE32F-NEXT: beqz a2, .LBB90_6
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa4, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 32
+; RV32ZVE32F-NEXT: beqz a1, .LBB90_6
; RV32ZVE32F-NEXT: .LBB90_15: # %cond.load13
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 5
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa5, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 64
-; RV32ZVE32F-NEXT: beqz a2, .LBB90_7
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa5, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 64
+; RV32ZVE32F-NEXT: beqz a1, .LBB90_7
; RV32ZVE32F-NEXT: .LBB90_16: # %cond.load16
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 6
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa6, 0(a2)
-; RV32ZVE32F-NEXT: andi a1, a1, -128
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa6, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, -128
; RV32ZVE32F-NEXT: bnez a1, .LBB90_8
; RV32ZVE32F-NEXT: j .LBB90_9
;
@@ -10436,8 +10436,8 @@ define <8 x double> @mgather_baseidx_v8i16_v8f64(ptr %base, <8 x i16> %idxs, <8
; RV64ZVE32F-NEXT: .LBB90_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e16, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: andi a3, a2, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a3, .LBB90_14
; RV64ZVE32F-NEXT: # %bb.5: # %else5
@@ -10536,31 +10536,32 @@ define <8 x double> @mgather_baseidx_sext_v8i16_v8f64(ptr %base, <8 x i16> %idxs
; RV32ZVE32F-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vsext.vf2 v10, v8
; RV32ZVE32F-NEXT: vsll.vi v8, v10, 3
-; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
; RV32ZVE32F-NEXT: vsetvli zero, zero, e8, mf2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a1, v0
-; RV32ZVE32F-NEXT: andi a2, a1, 1
-; RV32ZVE32F-NEXT: bnez a2, .LBB91_10
+; RV32ZVE32F-NEXT: vmv.x.s a2, v0
+; RV32ZVE32F-NEXT: andi a3, a2, 1
+; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
+; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
+; RV32ZVE32F-NEXT: bnez a3, .LBB91_10
; RV32ZVE32F-NEXT: # %bb.1: # %else
-; RV32ZVE32F-NEXT: andi a2, a1, 2
-; RV32ZVE32F-NEXT: bnez a2, .LBB91_11
+; RV32ZVE32F-NEXT: andi a1, a2, 2
+; RV32ZVE32F-NEXT: bnez a1, .LBB91_11
; RV32ZVE32F-NEXT: .LBB91_2: # %else2
-; RV32ZVE32F-NEXT: andi a2, a1, 4
-; RV32ZVE32F-NEXT: bnez a2, .LBB91_12
+; RV32ZVE32F-NEXT: andi a1, a2, 4
+; RV32ZVE32F-NEXT: bnez a1, .LBB91_12
; RV32ZVE32F-NEXT: .LBB91_3: # %else5
-; RV32ZVE32F-NEXT: andi a2, a1, 8
-; RV32ZVE32F-NEXT: bnez a2, .LBB91_13
+; RV32ZVE32F-NEXT: andi a1, a2, 8
+; RV32ZVE32F-NEXT: bnez a1, .LBB91_13
; RV32ZVE32F-NEXT: .LBB91_4: # %else8
-; RV32ZVE32F-NEXT: andi a2, a1, 16
-; RV32ZVE32F-NEXT: bnez a2, .LBB91_14
+; RV32ZVE32F-NEXT: andi a1, a2, 16
+; RV32ZVE32F-NEXT: bnez a1, .LBB91_14
; RV32ZVE32F-NEXT: .LBB91_5: # %else11
-; RV32ZVE32F-NEXT: andi a2, a1, 32
-; RV32ZVE32F-NEXT: bnez a2, .LBB91_15
+; RV32ZVE32F-NEXT: andi a1, a2, 32
+; RV32ZVE32F-NEXT: bnez a1, .LBB91_15
; RV32ZVE32F-NEXT: .LBB91_6: # %else14
-; RV32ZVE32F-NEXT: andi a2, a1, 64
-; RV32ZVE32F-NEXT: bnez a2, .LBB91_16
+; RV32ZVE32F-NEXT: andi a1, a2, 64
+; RV32ZVE32F-NEXT: bnez a1, .LBB91_16
; RV32ZVE32F-NEXT: .LBB91_7: # %else17
-; RV32ZVE32F-NEXT: andi a1, a1, -128
+; RV32ZVE32F-NEXT: andi a1, a2, -128
; RV32ZVE32F-NEXT: beqz a1, .LBB91_9
; RV32ZVE32F-NEXT: .LBB91_8: # %cond.load19
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -10578,52 +10579,51 @@ define <8 x double> @mgather_baseidx_sext_v8i16_v8f64(ptr %base, <8 x i16> %idxs
; RV32ZVE32F-NEXT: fsd fa7, 56(a0)
; RV32ZVE32F-NEXT: ret
; RV32ZVE32F-NEXT: .LBB91_10: # %cond.load
-; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a2, v8
-; RV32ZVE32F-NEXT: fld fa0, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 2
-; RV32ZVE32F-NEXT: beqz a2, .LBB91_2
+; RV32ZVE32F-NEXT: vmv.x.s a1, v8
+; RV32ZVE32F-NEXT: fld fa0, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 2
+; RV32ZVE32F-NEXT: beqz a1, .LBB91_2
; RV32ZVE32F-NEXT: .LBB91_11: # %cond.load1
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 1
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa1, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 4
-; RV32ZVE32F-NEXT: beqz a2, .LBB91_3
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa1, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 4
+; RV32ZVE32F-NEXT: beqz a1, .LBB91_3
; RV32ZVE32F-NEXT: .LBB91_12: # %cond.load4
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 2
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa2, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 8
-; RV32ZVE32F-NEXT: beqz a2, .LBB91_4
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa2, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 8
+; RV32ZVE32F-NEXT: beqz a1, .LBB91_4
; RV32ZVE32F-NEXT: .LBB91_13: # %cond.load7
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 3
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa3, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 16
-; RV32ZVE32F-NEXT: beqz a2, .LBB91_5
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa3, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 16
+; RV32ZVE32F-NEXT: beqz a1, .LBB91_5
; RV32ZVE32F-NEXT: .LBB91_14: # %cond.load10
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 4
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa4, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 32
-; RV32ZVE32F-NEXT: beqz a2, .LBB91_6
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa4, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 32
+; RV32ZVE32F-NEXT: beqz a1, .LBB91_6
; RV32ZVE32F-NEXT: .LBB91_15: # %cond.load13
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 5
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa5, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 64
-; RV32ZVE32F-NEXT: beqz a2, .LBB91_7
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa5, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 64
+; RV32ZVE32F-NEXT: beqz a1, .LBB91_7
; RV32ZVE32F-NEXT: .LBB91_16: # %cond.load16
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 6
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa6, 0(a2)
-; RV32ZVE32F-NEXT: andi a1, a1, -128
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa6, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, -128
; RV32ZVE32F-NEXT: bnez a1, .LBB91_8
; RV32ZVE32F-NEXT: j .LBB91_9
;
@@ -10652,8 +10652,8 @@ define <8 x double> @mgather_baseidx_sext_v8i16_v8f64(ptr %base, <8 x i16> %idxs
; RV64ZVE32F-NEXT: .LBB91_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e16, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: andi a3, a2, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a3, .LBB91_14
; RV64ZVE32F-NEXT: # %bb.5: # %else5
@@ -10754,31 +10754,32 @@ define <8 x double> @mgather_baseidx_zext_v8i16_v8f64(ptr %base, <8 x i16> %idxs
; RV32ZVE32F-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vzext.vf2 v10, v8
; RV32ZVE32F-NEXT: vsll.vi v8, v10, 3
-; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
; RV32ZVE32F-NEXT: vsetvli zero, zero, e8, mf2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a1, v0
-; RV32ZVE32F-NEXT: andi a2, a1, 1
-; RV32ZVE32F-NEXT: bnez a2, .LBB92_10
+; RV32ZVE32F-NEXT: vmv.x.s a2, v0
+; RV32ZVE32F-NEXT: andi a3, a2, 1
+; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
+; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
+; RV32ZVE32F-NEXT: bnez a3, .LBB92_10
; RV32ZVE32F-NEXT: # %bb.1: # %else
-; RV32ZVE32F-NEXT: andi a2, a1, 2
-; RV32ZVE32F-NEXT: bnez a2, .LBB92_11
+; RV32ZVE32F-NEXT: andi a1, a2, 2
+; RV32ZVE32F-NEXT: bnez a1, .LBB92_11
; RV32ZVE32F-NEXT: .LBB92_2: # %else2
-; RV32ZVE32F-NEXT: andi a2, a1, 4
-; RV32ZVE32F-NEXT: bnez a2, .LBB92_12
+; RV32ZVE32F-NEXT: andi a1, a2, 4
+; RV32ZVE32F-NEXT: bnez a1, .LBB92_12
; RV32ZVE32F-NEXT: .LBB92_3: # %else5
-; RV32ZVE32F-NEXT: andi a2, a1, 8
-; RV32ZVE32F-NEXT: bnez a2, .LBB92_13
+; RV32ZVE32F-NEXT: andi a1, a2, 8
+; RV32ZVE32F-NEXT: bnez a1, .LBB92_13
; RV32ZVE32F-NEXT: .LBB92_4: # %else8
-; RV32ZVE32F-NEXT: andi a2, a1, 16
-; RV32ZVE32F-NEXT: bnez a2, .LBB92_14
+; RV32ZVE32F-NEXT: andi a1, a2, 16
+; RV32ZVE32F-NEXT: bnez a1, .LBB92_14
; RV32ZVE32F-NEXT: .LBB92_5: # %else11
-; RV32ZVE32F-NEXT: andi a2, a1, 32
-; RV32ZVE32F-NEXT: bnez a2, .LBB92_15
+; RV32ZVE32F-NEXT: andi a1, a2, 32
+; RV32ZVE32F-NEXT: bnez a1, .LBB92_15
; RV32ZVE32F-NEXT: .LBB92_6: # %else14
-; RV32ZVE32F-NEXT: andi a2, a1, 64
-; RV32ZVE32F-NEXT: bnez a2, .LBB92_16
+; RV32ZVE32F-NEXT: andi a1, a2, 64
+; RV32ZVE32F-NEXT: bnez a1, .LBB92_16
; RV32ZVE32F-NEXT: .LBB92_7: # %else17
-; RV32ZVE32F-NEXT: andi a1, a1, -128
+; RV32ZVE32F-NEXT: andi a1, a2, -128
; RV32ZVE32F-NEXT: beqz a1, .LBB92_9
; RV32ZVE32F-NEXT: .LBB92_8: # %cond.load19
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -10796,52 +10797,51 @@ define <8 x double> @mgather_baseidx_zext_v8i16_v8f64(ptr %base, <8 x i16> %idxs
; RV32ZVE32F-NEXT: fsd fa7, 56(a0)
; RV32ZVE32F-NEXT: ret
; RV32ZVE32F-NEXT: .LBB92_10: # %cond.load
-; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a2, v8
-; RV32ZVE32F-NEXT: fld fa0, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 2
-; RV32ZVE32F-NEXT: beqz a2, .LBB92_2
+; RV32ZVE32F-NEXT: vmv.x.s a1, v8
+; RV32ZVE32F-NEXT: fld fa0, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 2
+; RV32ZVE32F-NEXT: beqz a1, .LBB92_2
; RV32ZVE32F-NEXT: .LBB92_11: # %cond.load1
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 1
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa1, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 4
-; RV32ZVE32F-NEXT: beqz a2, .LBB92_3
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa1, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 4
+; RV32ZVE32F-NEXT: beqz a1, .LBB92_3
; RV32ZVE32F-NEXT: .LBB92_12: # %cond.load4
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 2
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa2, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 8
-; RV32ZVE32F-NEXT: beqz a2, .LBB92_4
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa2, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 8
+; RV32ZVE32F-NEXT: beqz a1, .LBB92_4
; RV32ZVE32F-NEXT: .LBB92_13: # %cond.load7
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 3
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa3, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 16
-; RV32ZVE32F-NEXT: beqz a2, .LBB92_5
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa3, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 16
+; RV32ZVE32F-NEXT: beqz a1, .LBB92_5
; RV32ZVE32F-NEXT: .LBB92_14: # %cond.load10
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 4
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa4, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 32
-; RV32ZVE32F-NEXT: beqz a2, .LBB92_6
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa4, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 32
+; RV32ZVE32F-NEXT: beqz a1, .LBB92_6
; RV32ZVE32F-NEXT: .LBB92_15: # %cond.load13
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 5
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa5, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 64
-; RV32ZVE32F-NEXT: beqz a2, .LBB92_7
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa5, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 64
+; RV32ZVE32F-NEXT: beqz a1, .LBB92_7
; RV32ZVE32F-NEXT: .LBB92_16: # %cond.load16
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 6
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa6, 0(a2)
-; RV32ZVE32F-NEXT: andi a1, a1, -128
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa6, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, -128
; RV32ZVE32F-NEXT: bnez a1, .LBB92_8
; RV32ZVE32F-NEXT: j .LBB92_9
;
@@ -10874,8 +10874,8 @@ define <8 x double> @mgather_baseidx_zext_v8i16_v8f64(ptr %base, <8 x i16> %idxs
; RV64ZVE32F-NEXT: .LBB92_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e16, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: andi a4, a3, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a4, .LBB92_14
; RV64ZVE32F-NEXT: # %bb.5: # %else5
@@ -10979,31 +10979,32 @@ define <8 x double> @mgather_baseidx_v8i32_v8f64(ptr %base, <8 x i32> %idxs, <8
; RV32ZVE32F: # %bb.0:
; RV32ZVE32F-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vsll.vi v8, v8, 3
-; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
; RV32ZVE32F-NEXT: vsetvli zero, zero, e8, mf2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a1, v0
-; RV32ZVE32F-NEXT: andi a2, a1, 1
-; RV32ZVE32F-NEXT: bnez a2, .LBB93_10
+; RV32ZVE32F-NEXT: vmv.x.s a2, v0
+; RV32ZVE32F-NEXT: andi a3, a2, 1
+; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
+; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
+; RV32ZVE32F-NEXT: bnez a3, .LBB93_10
; RV32ZVE32F-NEXT: # %bb.1: # %else
-; RV32ZVE32F-NEXT: andi a2, a1, 2
-; RV32ZVE32F-NEXT: bnez a2, .LBB93_11
+; RV32ZVE32F-NEXT: andi a1, a2, 2
+; RV32ZVE32F-NEXT: bnez a1, .LBB93_11
; RV32ZVE32F-NEXT: .LBB93_2: # %else2
-; RV32ZVE32F-NEXT: andi a2, a1, 4
-; RV32ZVE32F-NEXT: bnez a2, .LBB93_12
+; RV32ZVE32F-NEXT: andi a1, a2, 4
+; RV32ZVE32F-NEXT: bnez a1, .LBB93_12
; RV32ZVE32F-NEXT: .LBB93_3: # %else5
-; RV32ZVE32F-NEXT: andi a2, a1, 8
-; RV32ZVE32F-NEXT: bnez a2, .LBB93_13
+; RV32ZVE32F-NEXT: andi a1, a2, 8
+; RV32ZVE32F-NEXT: bnez a1, .LBB93_13
; RV32ZVE32F-NEXT: .LBB93_4: # %else8
-; RV32ZVE32F-NEXT: andi a2, a1, 16
-; RV32ZVE32F-NEXT: bnez a2, .LBB93_14
+; RV32ZVE32F-NEXT: andi a1, a2, 16
+; RV32ZVE32F-NEXT: bnez a1, .LBB93_14
; RV32ZVE32F-NEXT: .LBB93_5: # %else11
-; RV32ZVE32F-NEXT: andi a2, a1, 32
-; RV32ZVE32F-NEXT: bnez a2, .LBB93_15
+; RV32ZVE32F-NEXT: andi a1, a2, 32
+; RV32ZVE32F-NEXT: bnez a1, .LBB93_15
; RV32ZVE32F-NEXT: .LBB93_6: # %else14
-; RV32ZVE32F-NEXT: andi a2, a1, 64
-; RV32ZVE32F-NEXT: bnez a2, .LBB93_16
+; RV32ZVE32F-NEXT: andi a1, a2, 64
+; RV32ZVE32F-NEXT: bnez a1, .LBB93_16
; RV32ZVE32F-NEXT: .LBB93_7: # %else17
-; RV32ZVE32F-NEXT: andi a1, a1, -128
+; RV32ZVE32F-NEXT: andi a1, a2, -128
; RV32ZVE32F-NEXT: beqz a1, .LBB93_9
; RV32ZVE32F-NEXT: .LBB93_8: # %cond.load19
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -11021,52 +11022,51 @@ define <8 x double> @mgather_baseidx_v8i32_v8f64(ptr %base, <8 x i32> %idxs, <8
; RV32ZVE32F-NEXT: fsd fa7, 56(a0)
; RV32ZVE32F-NEXT: ret
; RV32ZVE32F-NEXT: .LBB93_10: # %cond.load
-; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a2, v8
-; RV32ZVE32F-NEXT: fld fa0, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 2
-; RV32ZVE32F-NEXT: beqz a2, .LBB93_2
+; RV32ZVE32F-NEXT: vmv.x.s a1, v8
+; RV32ZVE32F-NEXT: fld fa0, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 2
+; RV32ZVE32F-NEXT: beqz a1, .LBB93_2
; RV32ZVE32F-NEXT: .LBB93_11: # %cond.load1
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 1
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa1, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 4
-; RV32ZVE32F-NEXT: beqz a2, .LBB93_3
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa1, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 4
+; RV32ZVE32F-NEXT: beqz a1, .LBB93_3
; RV32ZVE32F-NEXT: .LBB93_12: # %cond.load4
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 2
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa2, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 8
-; RV32ZVE32F-NEXT: beqz a2, .LBB93_4
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa2, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 8
+; RV32ZVE32F-NEXT: beqz a1, .LBB93_4
; RV32ZVE32F-NEXT: .LBB93_13: # %cond.load7
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 3
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa3, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 16
-; RV32ZVE32F-NEXT: beqz a2, .LBB93_5
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa3, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 16
+; RV32ZVE32F-NEXT: beqz a1, .LBB93_5
; RV32ZVE32F-NEXT: .LBB93_14: # %cond.load10
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 4
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa4, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 32
-; RV32ZVE32F-NEXT: beqz a2, .LBB93_6
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa4, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 32
+; RV32ZVE32F-NEXT: beqz a1, .LBB93_6
; RV32ZVE32F-NEXT: .LBB93_15: # %cond.load13
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 5
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa5, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 64
-; RV32ZVE32F-NEXT: beqz a2, .LBB93_7
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa5, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 64
+; RV32ZVE32F-NEXT: beqz a1, .LBB93_7
; RV32ZVE32F-NEXT: .LBB93_16: # %cond.load16
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 6
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa6, 0(a2)
-; RV32ZVE32F-NEXT: andi a1, a1, -128
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa6, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, -128
; RV32ZVE32F-NEXT: bnez a1, .LBB93_8
; RV32ZVE32F-NEXT: j .LBB93_9
;
@@ -11095,8 +11095,8 @@ define <8 x double> @mgather_baseidx_v8i32_v8f64(ptr %base, <8 x i32> %idxs, <8
; RV64ZVE32F-NEXT: .LBB93_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e32, m2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e32, m1, ta, ma
; RV64ZVE32F-NEXT: andi a3, a2, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e32, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a3, .LBB93_14
; RV64ZVE32F-NEXT: # %bb.5: # %else5
@@ -11193,31 +11193,32 @@ define <8 x double> @mgather_baseidx_sext_v8i32_v8f64(ptr %base, <8 x i32> %idxs
; RV32ZVE32F: # %bb.0:
; RV32ZVE32F-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vsll.vi v8, v8, 3
-; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
; RV32ZVE32F-NEXT: vsetvli zero, zero, e8, mf2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a1, v0
-; RV32ZVE32F-NEXT: andi a2, a1, 1
-; RV32ZVE32F-NEXT: bnez a2, .LBB94_10
+; RV32ZVE32F-NEXT: vmv.x.s a2, v0
+; RV32ZVE32F-NEXT: andi a3, a2, 1
+; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
+; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
+; RV32ZVE32F-NEXT: bnez a3, .LBB94_10
; RV32ZVE32F-NEXT: # %bb.1: # %else
-; RV32ZVE32F-NEXT: andi a2, a1, 2
-; RV32ZVE32F-NEXT: bnez a2, .LBB94_11
+; RV32ZVE32F-NEXT: andi a1, a2, 2
+; RV32ZVE32F-NEXT: bnez a1, .LBB94_11
; RV32ZVE32F-NEXT: .LBB94_2: # %else2
-; RV32ZVE32F-NEXT: andi a2, a1, 4
-; RV32ZVE32F-NEXT: bnez a2, .LBB94_12
+; RV32ZVE32F-NEXT: andi a1, a2, 4
+; RV32ZVE32F-NEXT: bnez a1, .LBB94_12
; RV32ZVE32F-NEXT: .LBB94_3: # %else5
-; RV32ZVE32F-NEXT: andi a2, a1, 8
-; RV32ZVE32F-NEXT: bnez a2, .LBB94_13
+; RV32ZVE32F-NEXT: andi a1, a2, 8
+; RV32ZVE32F-NEXT: bnez a1, .LBB94_13
; RV32ZVE32F-NEXT: .LBB94_4: # %else8
-; RV32ZVE32F-NEXT: andi a2, a1, 16
-; RV32ZVE32F-NEXT: bnez a2, .LBB94_14
+; RV32ZVE32F-NEXT: andi a1, a2, 16
+; RV32ZVE32F-NEXT: bnez a1, .LBB94_14
; RV32ZVE32F-NEXT: .LBB94_5: # %else11
-; RV32ZVE32F-NEXT: andi a2, a1, 32
-; RV32ZVE32F-NEXT: bnez a2, .LBB94_15
+; RV32ZVE32F-NEXT: andi a1, a2, 32
+; RV32ZVE32F-NEXT: bnez a1, .LBB94_15
; RV32ZVE32F-NEXT: .LBB94_6: # %else14
-; RV32ZVE32F-NEXT: andi a2, a1, 64
-; RV32ZVE32F-NEXT: bnez a2, .LBB94_16
+; RV32ZVE32F-NEXT: andi a1, a2, 64
+; RV32ZVE32F-NEXT: bnez a1, .LBB94_16
; RV32ZVE32F-NEXT: .LBB94_7: # %else17
-; RV32ZVE32F-NEXT: andi a1, a1, -128
+; RV32ZVE32F-NEXT: andi a1, a2, -128
; RV32ZVE32F-NEXT: beqz a1, .LBB94_9
; RV32ZVE32F-NEXT: .LBB94_8: # %cond.load19
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -11235,52 +11236,51 @@ define <8 x double> @mgather_baseidx_sext_v8i32_v8f64(ptr %base, <8 x i32> %idxs
; RV32ZVE32F-NEXT: fsd fa7, 56(a0)
; RV32ZVE32F-NEXT: ret
; RV32ZVE32F-NEXT: .LBB94_10: # %cond.load
-; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a2, v8
-; RV32ZVE32F-NEXT: fld fa0, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 2
-; RV32ZVE32F-NEXT: beqz a2, .LBB94_2
+; RV32ZVE32F-NEXT: vmv.x.s a1, v8
+; RV32ZVE32F-NEXT: fld fa0, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 2
+; RV32ZVE32F-NEXT: beqz a1, .LBB94_2
; RV32ZVE32F-NEXT: .LBB94_11: # %cond.load1
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 1
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa1, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 4
-; RV32ZVE32F-NEXT: beqz a2, .LBB94_3
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa1, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 4
+; RV32ZVE32F-NEXT: beqz a1, .LBB94_3
; RV32ZVE32F-NEXT: .LBB94_12: # %cond.load4
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 2
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa2, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 8
-; RV32ZVE32F-NEXT: beqz a2, .LBB94_4
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa2, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 8
+; RV32ZVE32F-NEXT: beqz a1, .LBB94_4
; RV32ZVE32F-NEXT: .LBB94_13: # %cond.load7
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 3
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa3, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 16
-; RV32ZVE32F-NEXT: beqz a2, .LBB94_5
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa3, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 16
+; RV32ZVE32F-NEXT: beqz a1, .LBB94_5
; RV32ZVE32F-NEXT: .LBB94_14: # %cond.load10
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 4
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa4, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 32
-; RV32ZVE32F-NEXT: beqz a2, .LBB94_6
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa4, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 32
+; RV32ZVE32F-NEXT: beqz a1, .LBB94_6
; RV32ZVE32F-NEXT: .LBB94_15: # %cond.load13
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 5
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa5, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 64
-; RV32ZVE32F-NEXT: beqz a2, .LBB94_7
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa5, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 64
+; RV32ZVE32F-NEXT: beqz a1, .LBB94_7
; RV32ZVE32F-NEXT: .LBB94_16: # %cond.load16
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 6
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa6, 0(a2)
-; RV32ZVE32F-NEXT: andi a1, a1, -128
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa6, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, -128
; RV32ZVE32F-NEXT: bnez a1, .LBB94_8
; RV32ZVE32F-NEXT: j .LBB94_9
;
@@ -11309,8 +11309,8 @@ define <8 x double> @mgather_baseidx_sext_v8i32_v8f64(ptr %base, <8 x i32> %idxs
; RV64ZVE32F-NEXT: .LBB94_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e32, m2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e32, m1, ta, ma
; RV64ZVE32F-NEXT: andi a3, a2, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e32, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a3, .LBB94_14
; RV64ZVE32F-NEXT: # %bb.5: # %else5
@@ -11408,31 +11408,32 @@ define <8 x double> @mgather_baseidx_zext_v8i32_v8f64(ptr %base, <8 x i32> %idxs
; RV32ZVE32F: # %bb.0:
; RV32ZVE32F-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vsll.vi v8, v8, 3
-; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
; RV32ZVE32F-NEXT: vsetvli zero, zero, e8, mf2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a1, v0
-; RV32ZVE32F-NEXT: andi a2, a1, 1
-; RV32ZVE32F-NEXT: bnez a2, .LBB95_10
+; RV32ZVE32F-NEXT: vmv.x.s a2, v0
+; RV32ZVE32F-NEXT: andi a3, a2, 1
+; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
+; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
+; RV32ZVE32F-NEXT: bnez a3, .LBB95_10
; RV32ZVE32F-NEXT: # %bb.1: # %else
-; RV32ZVE32F-NEXT: andi a2, a1, 2
-; RV32ZVE32F-NEXT: bnez a2, .LBB95_11
+; RV32ZVE32F-NEXT: andi a1, a2, 2
+; RV32ZVE32F-NEXT: bnez a1, .LBB95_11
; RV32ZVE32F-NEXT: .LBB95_2: # %else2
-; RV32ZVE32F-NEXT: andi a2, a1, 4
-; RV32ZVE32F-NEXT: bnez a2, .LBB95_12
+; RV32ZVE32F-NEXT: andi a1, a2, 4
+; RV32ZVE32F-NEXT: bnez a1, .LBB95_12
; RV32ZVE32F-NEXT: .LBB95_3: # %else5
-; RV32ZVE32F-NEXT: andi a2, a1, 8
-; RV32ZVE32F-NEXT: bnez a2, .LBB95_13
+; RV32ZVE32F-NEXT: andi a1, a2, 8
+; RV32ZVE32F-NEXT: bnez a1, .LBB95_13
; RV32ZVE32F-NEXT: .LBB95_4: # %else8
-; RV32ZVE32F-NEXT: andi a2, a1, 16
-; RV32ZVE32F-NEXT: bnez a2, .LBB95_14
+; RV32ZVE32F-NEXT: andi a1, a2, 16
+; RV32ZVE32F-NEXT: bnez a1, .LBB95_14
; RV32ZVE32F-NEXT: .LBB95_5: # %else11
-; RV32ZVE32F-NEXT: andi a2, a1, 32
-; RV32ZVE32F-NEXT: bnez a2, .LBB95_15
+; RV32ZVE32F-NEXT: andi a1, a2, 32
+; RV32ZVE32F-NEXT: bnez a1, .LBB95_15
; RV32ZVE32F-NEXT: .LBB95_6: # %else14
-; RV32ZVE32F-NEXT: andi a2, a1, 64
-; RV32ZVE32F-NEXT: bnez a2, .LBB95_16
+; RV32ZVE32F-NEXT: andi a1, a2, 64
+; RV32ZVE32F-NEXT: bnez a1, .LBB95_16
; RV32ZVE32F-NEXT: .LBB95_7: # %else17
-; RV32ZVE32F-NEXT: andi a1, a1, -128
+; RV32ZVE32F-NEXT: andi a1, a2, -128
; RV32ZVE32F-NEXT: beqz a1, .LBB95_9
; RV32ZVE32F-NEXT: .LBB95_8: # %cond.load19
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -11450,52 +11451,51 @@ define <8 x double> @mgather_baseidx_zext_v8i32_v8f64(ptr %base, <8 x i32> %idxs
; RV32ZVE32F-NEXT: fsd fa7, 56(a0)
; RV32ZVE32F-NEXT: ret
; RV32ZVE32F-NEXT: .LBB95_10: # %cond.load
-; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a2, v8
-; RV32ZVE32F-NEXT: fld fa0, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 2
-; RV32ZVE32F-NEXT: beqz a2, .LBB95_2
+; RV32ZVE32F-NEXT: vmv.x.s a1, v8
+; RV32ZVE32F-NEXT: fld fa0, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 2
+; RV32ZVE32F-NEXT: beqz a1, .LBB95_2
; RV32ZVE32F-NEXT: .LBB95_11: # %cond.load1
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 1
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa1, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 4
-; RV32ZVE32F-NEXT: beqz a2, .LBB95_3
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa1, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 4
+; RV32ZVE32F-NEXT: beqz a1, .LBB95_3
; RV32ZVE32F-NEXT: .LBB95_12: # %cond.load4
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 2
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa2, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 8
-; RV32ZVE32F-NEXT: beqz a2, .LBB95_4
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa2, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 8
+; RV32ZVE32F-NEXT: beqz a1, .LBB95_4
; RV32ZVE32F-NEXT: .LBB95_13: # %cond.load7
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
-; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 3
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa3, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 16
-; RV32ZVE32F-NEXT: beqz a2, .LBB95_5
+; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 3
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa3, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 16
+; RV32ZVE32F-NEXT: beqz a1, .LBB95_5
; RV32ZVE32F-NEXT: .LBB95_14: # %cond.load10
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 4
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa4, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 32
-; RV32ZVE32F-NEXT: beqz a2, .LBB95_6
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa4, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 32
+; RV32ZVE32F-NEXT: beqz a1, .LBB95_6
; RV32ZVE32F-NEXT: .LBB95_15: # %cond.load13
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 5
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa5, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 64
-; RV32ZVE32F-NEXT: beqz a2, .LBB95_7
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa5, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 64
+; RV32ZVE32F-NEXT: beqz a1, .LBB95_7
; RV32ZVE32F-NEXT: .LBB95_16: # %cond.load16
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 6
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa6, 0(a2)
-; RV32ZVE32F-NEXT: andi a1, a1, -128
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa6, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, -128
; RV32ZVE32F-NEXT: bnez a1, .LBB95_8
; RV32ZVE32F-NEXT: j .LBB95_9
;
@@ -11526,8 +11526,8 @@ define <8 x double> @mgather_baseidx_zext_v8i32_v8f64(ptr %base, <8 x i32> %idxs
; RV64ZVE32F-NEXT: .LBB95_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e32, m2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e32, m1, ta, ma
; RV64ZVE32F-NEXT: andi a3, a2, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e32, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a3, .LBB95_14
; RV64ZVE32F-NEXT: # %bb.5: # %else5
@@ -11632,45 +11632,46 @@ define <8 x double> @mgather_baseidx_v8f64(ptr %base, <8 x i64> %idxs, <8 x i1>
; RV32ZVE32F-NEXT: lw a3, 56(a2)
; RV32ZVE32F-NEXT: lw a4, 48(a2)
; RV32ZVE32F-NEXT: lw a5, 40(a2)
-; RV32ZVE32F-NEXT: lw a6, 32(a2)
-; RV32ZVE32F-NEXT: lw a7, 24(a2)
-; RV32ZVE32F-NEXT: lw t0, 16(a2)
-; RV32ZVE32F-NEXT: lw t1, 8(a2)
+; RV32ZVE32F-NEXT: lw a6, 8(a2)
; RV32ZVE32F-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vlse32.v v8, (a2), zero
-; RV32ZVE32F-NEXT: vslide1down.vx v8, v8, t1
-; RV32ZVE32F-NEXT: vslide1down.vx v8, v8, t0
-; RV32ZVE32F-NEXT: vslide1down.vx v8, v8, a7
+; RV32ZVE32F-NEXT: lw a7, 16(a2)
+; RV32ZVE32F-NEXT: lw t0, 24(a2)
+; RV32ZVE32F-NEXT: lw a2, 32(a2)
; RV32ZVE32F-NEXT: vslide1down.vx v8, v8, a6
+; RV32ZVE32F-NEXT: vslide1down.vx v8, v8, a7
+; RV32ZVE32F-NEXT: vslide1down.vx v8, v8, t0
+; RV32ZVE32F-NEXT: vslide1down.vx v8, v8, a2
; RV32ZVE32F-NEXT: vslide1down.vx v8, v8, a5
; RV32ZVE32F-NEXT: vslide1down.vx v8, v8, a4
; RV32ZVE32F-NEXT: vslide1down.vx v8, v8, a3
; RV32ZVE32F-NEXT: vsll.vi v8, v8, 3
-; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
; RV32ZVE32F-NEXT: vsetvli zero, zero, e8, mf2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a1, v0
-; RV32ZVE32F-NEXT: andi a2, a1, 1
-; RV32ZVE32F-NEXT: bnez a2, .LBB96_10
+; RV32ZVE32F-NEXT: vmv.x.s a2, v0
+; RV32ZVE32F-NEXT: andi a3, a2, 1
+; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
+; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
+; RV32ZVE32F-NEXT: bnez a3, .LBB96_10
; RV32ZVE32F-NEXT: # %bb.1: # %else
-; RV32ZVE32F-NEXT: andi a2, a1, 2
-; RV32ZVE32F-NEXT: bnez a2, .LBB96_11
+; RV32ZVE32F-NEXT: andi a1, a2, 2
+; RV32ZVE32F-NEXT: bnez a1, .LBB96_11
; RV32ZVE32F-NEXT: .LBB96_2: # %else2
-; RV32ZVE32F-NEXT: andi a2, a1, 4
-; RV32ZVE32F-NEXT: bnez a2, .LBB96_12
+; RV32ZVE32F-NEXT: andi a1, a2, 4
+; RV32ZVE32F-NEXT: bnez a1, .LBB96_12
; RV32ZVE32F-NEXT: .LBB96_3: # %else5
-; RV32ZVE32F-NEXT: andi a2, a1, 8
-; RV32ZVE32F-NEXT: bnez a2, .LBB96_13
+; RV32ZVE32F-NEXT: andi a1, a2, 8
+; RV32ZVE32F-NEXT: bnez a1, .LBB96_13
; RV32ZVE32F-NEXT: .LBB96_4: # %else8
-; RV32ZVE32F-NEXT: andi a2, a1, 16
-; RV32ZVE32F-NEXT: bnez a2, .LBB96_14
+; RV32ZVE32F-NEXT: andi a1, a2, 16
+; RV32ZVE32F-NEXT: bnez a1, .LBB96_14
; RV32ZVE32F-NEXT: .LBB96_5: # %else11
-; RV32ZVE32F-NEXT: andi a2, a1, 32
-; RV32ZVE32F-NEXT: bnez a2, .LBB96_15
+; RV32ZVE32F-NEXT: andi a1, a2, 32
+; RV32ZVE32F-NEXT: bnez a1, .LBB96_15
; RV32ZVE32F-NEXT: .LBB96_6: # %else14
-; RV32ZVE32F-NEXT: andi a2, a1, 64
-; RV32ZVE32F-NEXT: bnez a2, .LBB96_16
+; RV32ZVE32F-NEXT: andi a1, a2, 64
+; RV32ZVE32F-NEXT: bnez a1, .LBB96_16
; RV32ZVE32F-NEXT: .LBB96_7: # %else17
-; RV32ZVE32F-NEXT: andi a1, a1, -128
+; RV32ZVE32F-NEXT: andi a1, a2, -128
; RV32ZVE32F-NEXT: beqz a1, .LBB96_9
; RV32ZVE32F-NEXT: .LBB96_8: # %cond.load19
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -11688,52 +11689,51 @@ define <8 x double> @mgather_baseidx_v8f64(ptr %base, <8 x i64> %idxs, <8 x i1>
; RV32ZVE32F-NEXT: fsd fa7, 56(a0)
; RV32ZVE32F-NEXT: ret
; RV32ZVE32F-NEXT: .LBB96_10: # %cond.load
-; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a2, v8
-; RV32ZVE32F-NEXT: fld fa0, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 2
-; RV32ZVE32F-NEXT: beqz a2, .LBB96_2
+; RV32ZVE32F-NEXT: vmv.x.s a1, v8
+; RV32ZVE32F-NEXT: fld fa0, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 2
+; RV32ZVE32F-NEXT: beqz a1, .LBB96_2
; RV32ZVE32F-NEXT: .LBB96_11: # %cond.load1
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 1
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa1, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 4
-; RV32ZVE32F-NEXT: beqz a2, .LBB96_3
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa1, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 4
+; RV32ZVE32F-NEXT: beqz a1, .LBB96_3
; RV32ZVE32F-NEXT: .LBB96_12: # %cond.load4
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 2
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa2, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 8
-; RV32ZVE32F-NEXT: beqz a2, .LBB96_4
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa2, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 8
+; RV32ZVE32F-NEXT: beqz a1, .LBB96_4
; RV32ZVE32F-NEXT: .LBB96_13: # %cond.load7
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 3
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa3, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 16
-; RV32ZVE32F-NEXT: beqz a2, .LBB96_5
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa3, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 16
+; RV32ZVE32F-NEXT: beqz a1, .LBB96_5
; RV32ZVE32F-NEXT: .LBB96_14: # %cond.load10
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 4
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa4, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 32
-; RV32ZVE32F-NEXT: beqz a2, .LBB96_6
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa4, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 32
+; RV32ZVE32F-NEXT: beqz a1, .LBB96_6
; RV32ZVE32F-NEXT: .LBB96_15: # %cond.load13
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 5
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa5, 0(a2)
-; RV32ZVE32F-NEXT: andi a2, a1, 64
-; RV32ZVE32F-NEXT: beqz a2, .LBB96_7
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa5, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, 64
+; RV32ZVE32F-NEXT: beqz a1, .LBB96_7
; RV32ZVE32F-NEXT: .LBB96_16: # %cond.load16
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 6
-; RV32ZVE32F-NEXT: vmv.x.s a2, v10
-; RV32ZVE32F-NEXT: fld fa6, 0(a2)
-; RV32ZVE32F-NEXT: andi a1, a1, -128
+; RV32ZVE32F-NEXT: vmv.x.s a1, v10
+; RV32ZVE32F-NEXT: fld fa6, 0(a1)
+; RV32ZVE32F-NEXT: andi a1, a2, -128
; RV32ZVE32F-NEXT: bnez a1, .LBB96_8
; RV32ZVE32F-NEXT: j .LBB96_9
;
@@ -11882,8 +11882,8 @@ define <16 x i8> @mgather_baseidx_v16i8(ptr %base, <16 x i8> %idxs, <16 x i1> %m
; RV64ZVE32F-NEXT: .LBB97_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v11, v8, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB97_25
; RV64ZVE32F-NEXT: # %bb.5: # %else5
@@ -11900,8 +11900,8 @@ define <16 x i8> @mgather_baseidx_v16i8(ptr %base, <16 x i8> %idxs, <16 x i1> %m
; RV64ZVE32F-NEXT: vmv.s.x v11, a2
; RV64ZVE32F-NEXT: vslideup.vi v9, v11, 4
; RV64ZVE32F-NEXT: .LBB97_8: # %else11
-; RV64ZVE32F-NEXT: vsetivli zero, 8, e8, m1, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 32
+; RV64ZVE32F-NEXT: vsetivli zero, 8, e8, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 8
; RV64ZVE32F-NEXT: beqz a2, .LBB97_10
; RV64ZVE32F-NEXT: # %bb.9: # %cond.load13
@@ -11914,8 +11914,8 @@ define <16 x i8> @mgather_baseidx_v16i8(ptr %base, <16 x i8> %idxs, <16 x i1> %m
; RV64ZVE32F-NEXT: vsetivli zero, 6, e8, m1, tu, ma
; RV64ZVE32F-NEXT: vslideup.vi v9, v11, 5
; RV64ZVE32F-NEXT: .LBB97_10: # %else14
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v10, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB97_27
; RV64ZVE32F-NEXT: # %bb.11: # %else17
@@ -11939,8 +11939,8 @@ define <16 x i8> @mgather_baseidx_v16i8(ptr %base, <16 x i8> %idxs, <16 x i1> %m
; RV64ZVE32F-NEXT: .LBB97_15: # %else26
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 1024
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB97_30
; RV64ZVE32F-NEXT: # %bb.16: # %else29
@@ -11962,8 +11962,8 @@ define <16 x i8> @mgather_baseidx_v16i8(ptr %base, <16 x i8> %idxs, <16 x i1> %m
; RV64ZVE32F-NEXT: vsetivli zero, 14, e8, m1, tu, ma
; RV64ZVE32F-NEXT: vslideup.vi v9, v8, 13
; RV64ZVE32F-NEXT: .LBB97_20: # %else38
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: slli a2, a1, 49
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v10, 2
; RV64ZVE32F-NEXT: bgez a2, .LBB97_22
; RV64ZVE32F-NEXT: # %bb.21: # %cond.load40
@@ -12092,22 +12092,22 @@ define <32 x i8> @mgather_baseidx_v32i8(ptr %base, <32 x i8> %idxs, <32 x i1> %m
; RV64V: # %bb.0:
; RV64V-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV64V-NEXT: vsext.vf8 v16, v8
-; RV64V-NEXT: vmv1r.v v12, v10
-; RV64V-NEXT: vsetvli zero, zero, e8, m1, ta, mu
-; RV64V-NEXT: vluxei64.v v12, (a0), v16, v0.t
; RV64V-NEXT: vsetivli zero, 16, e8, m2, ta, ma
-; RV64V-NEXT: vslidedown.vi v10, v10, 16
+; RV64V-NEXT: vslidedown.vi v12, v10, 16
+; RV64V-NEXT: vsetivli zero, 16, e8, m1, ta, mu
+; RV64V-NEXT: vluxei64.v v10, (a0), v16, v0.t
+; RV64V-NEXT: vsetivli zero, 16, e8, m2, ta, ma
; RV64V-NEXT: vslidedown.vi v8, v8, 16
-; RV64V-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV64V-NEXT: vsext.vf8 v16, v8
; RV64V-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64V-NEXT: vslidedown.vi v0, v0, 2
-; RV64V-NEXT: vsetivli zero, 16, e8, m1, ta, mu
-; RV64V-NEXT: vluxei64.v v10, (a0), v16, v0.t
+; RV64V-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; RV64V-NEXT: vsext.vf8 v16, v8
+; RV64V-NEXT: vsetvli zero, zero, e8, m1, ta, mu
+; RV64V-NEXT: vluxei64.v v12, (a0), v16, v0.t
; RV64V-NEXT: li a0, 32
; RV64V-NEXT: vsetvli zero, a0, e8, m2, ta, ma
-; RV64V-NEXT: vslideup.vi v12, v10, 16
-; RV64V-NEXT: vmv.v.v v8, v12
+; RV64V-NEXT: vslideup.vi v10, v12, 16
+; RV64V-NEXT: vmv.v.v v8, v10
; RV64V-NEXT: ret
;
; RV64ZVE32F-LABEL: mgather_baseidx_v32i8:
@@ -12139,8 +12139,8 @@ define <32 x i8> @mgather_baseidx_v32i8(ptr %base, <32 x i8> %idxs, <32 x i1> %m
; RV64ZVE32F-NEXT: .LBB98_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v13, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v12, v8, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB98_49
; RV64ZVE32F-NEXT: # %bb.5: # %else5
@@ -12157,8 +12157,8 @@ define <32 x i8> @mgather_baseidx_v32i8(ptr %base, <32 x i8> %idxs, <32 x i1> %m
; RV64ZVE32F-NEXT: vmv.s.x v12, a2
; RV64ZVE32F-NEXT: vslideup.vi v10, v12, 4
; RV64ZVE32F-NEXT: .LBB98_8: # %else11
-; RV64ZVE32F-NEXT: vsetivli zero, 8, e8, m1, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 32
+; RV64ZVE32F-NEXT: vsetivli zero, 8, e8, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v12, v8, 8
; RV64ZVE32F-NEXT: beqz a2, .LBB98_10
; RV64ZVE32F-NEXT: # %bb.9: # %cond.load13
@@ -12171,8 +12171,8 @@ define <32 x i8> @mgather_baseidx_v32i8(ptr %base, <32 x i8> %idxs, <32 x i1> %m
; RV64ZVE32F-NEXT: vsetivli zero, 6, e8, m1, tu, ma
; RV64ZVE32F-NEXT: vslideup.vi v10, v14, 5
; RV64ZVE32F-NEXT: .LBB98_10: # %else14
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v13, v13, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB98_51
; RV64ZVE32F-NEXT: # %bb.11: # %else17
@@ -12196,8 +12196,8 @@ define <32 x i8> @mgather_baseidx_v32i8(ptr %base, <32 x i8> %idxs, <32 x i1> %m
; RV64ZVE32F-NEXT: .LBB98_15: # %else26
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v13, v12, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 1024
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v12, v12, 2
; RV64ZVE32F-NEXT: beqz a2, .LBB98_17
; RV64ZVE32F-NEXT: # %bb.16: # %cond.load28
@@ -12220,8 +12220,8 @@ define <32 x i8> @mgather_baseidx_v32i8(ptr %base, <32 x i8> %idxs, <32 x i1> %m
; RV64ZVE32F-NEXT: vsetivli zero, 12, e8, m1, tu, ma
; RV64ZVE32F-NEXT: vslideup.vi v10, v12, 11
; RV64ZVE32F-NEXT: .LBB98_19: # %else32
-; RV64ZVE32F-NEXT: vsetivli zero, 16, e8, m2, ta, ma
; RV64ZVE32F-NEXT: slli a2, a1, 51
+; RV64ZVE32F-NEXT: vsetivli zero, 16, e8, m2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 16
; RV64ZVE32F-NEXT: bgez a2, .LBB98_21
; RV64ZVE32F-NEXT: # %bb.20: # %cond.load34
@@ -12244,8 +12244,8 @@ define <32 x i8> @mgather_baseidx_v32i8(ptr %base, <32 x i8> %idxs, <32 x i1> %m
; RV64ZVE32F-NEXT: vsetivli zero, 14, e8, m1, tu, ma
; RV64ZVE32F-NEXT: vslideup.vi v10, v9, 13
; RV64ZVE32F-NEXT: .LBB98_23: # %else38
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: slli a2, a1, 49
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v13, 2
; RV64ZVE32F-NEXT: bltz a2, .LBB98_54
; RV64ZVE32F-NEXT: # %bb.24: # %else41
@@ -12269,8 +12269,8 @@ define <32 x i8> @mgather_baseidx_v32i8(ptr %base, <32 x i8> %idxs, <32 x i1> %m
; RV64ZVE32F-NEXT: .LBB98_28: # %else50
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: slli a2, a1, 45
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v12, v8, 2
; RV64ZVE32F-NEXT: bltz a2, .LBB98_57
; RV64ZVE32F-NEXT: # %bb.29: # %else53
@@ -12287,8 +12287,8 @@ define <32 x i8> @mgather_baseidx_v32i8(ptr %base, <32 x i8> %idxs, <32 x i1> %m
; RV64ZVE32F-NEXT: vmv.s.x v12, a2
; RV64ZVE32F-NEXT: vslideup.vi v10, v12, 20
; RV64ZVE32F-NEXT: .LBB98_32: # %else59
-; RV64ZVE32F-NEXT: vsetivli zero, 8, e8, m1, ta, ma
; RV64ZVE32F-NEXT: slli a2, a1, 42
+; RV64ZVE32F-NEXT: vsetivli zero, 8, e8, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 8
; RV64ZVE32F-NEXT: bgez a2, .LBB98_34
; RV64ZVE32F-NEXT: # %bb.33: # %cond.load61
@@ -12301,8 +12301,8 @@ define <32 x i8> @mgather_baseidx_v32i8(ptr %base, <32 x i8> %idxs, <32 x i1> %m
; RV64ZVE32F-NEXT: vsetivli zero, 22, e8, m2, tu, ma
; RV64ZVE32F-NEXT: vslideup.vi v10, v12, 21
; RV64ZVE32F-NEXT: .LBB98_34: # %else62
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: slli a2, a1, 41
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v9, 2
; RV64ZVE32F-NEXT: bltz a2, .LBB98_59
; RV64ZVE32F-NEXT: # %bb.35: # %else65
@@ -12326,8 +12326,8 @@ define <32 x i8> @mgather_baseidx_v32i8(ptr %base, <32 x i8> %idxs, <32 x i1> %m
; RV64ZVE32F-NEXT: .LBB98_39: # %else74
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: slli a2, a1, 37
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bltz a2, .LBB98_62
; RV64ZVE32F-NEXT: # %bb.40: # %else77
@@ -12349,8 +12349,8 @@ define <32 x i8> @mgather_baseidx_v32i8(ptr %base, <32 x i8> %idxs, <32 x i1> %m
; RV64ZVE32F-NEXT: vsetivli zero, 30, e8, m2, tu, ma
; RV64ZVE32F-NEXT: vslideup.vi v10, v12, 29
; RV64ZVE32F-NEXT: .LBB98_44: # %else86
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: slli a2, a1, 33
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v9, 2
; RV64ZVE32F-NEXT: bgez a2, .LBB98_46
; RV64ZVE32F-NEXT: # %bb.45: # %cond.load88
@@ -12627,29 +12627,40 @@ define <4 x i32> @mgather_unit_stride_load_wide_idx(ptr %base) {
; This looks like a strided load (at i8), but isn't at index type.
define <4 x i32> @mgather_narrow_edge_case(ptr %base) {
-; RV32-LABEL: mgather_narrow_edge_case:
-; RV32: # %bb.0:
-; RV32-NEXT: li a1, -512
-; RV32-NEXT: vsetivli zero, 4, e32, m1, ta, ma
-; RV32-NEXT: vmv.v.i v0, 5
-; RV32-NEXT: vmv.v.x v8, a1
-; RV32-NEXT: vmerge.vim v8, v8, 0, v0
-; RV32-NEXT: vluxei32.v v8, (a0), v8
-; RV32-NEXT: ret
+; RV32V-LABEL: mgather_narrow_edge_case:
+; RV32V: # %bb.0:
+; RV32V-NEXT: li a1, -512
+; RV32V-NEXT: vsetivli zero, 1, e8, mf8, ta, ma
+; RV32V-NEXT: vmv.v.i v0, 5
+; RV32V-NEXT: vsetivli zero, 4, e32, m1, ta, ma
+; RV32V-NEXT: vmv.v.x v8, a1
+; RV32V-NEXT: vmerge.vim v8, v8, 0, v0
+; RV32V-NEXT: vluxei32.v v8, (a0), v8
+; RV32V-NEXT: ret
;
; RV64V-LABEL: mgather_narrow_edge_case:
; RV64V: # %bb.0:
; RV64V-NEXT: li a1, -512
-; RV64V-NEXT: vsetivli zero, 4, e64, m2, ta, ma
-; RV64V-NEXT: vmv.v.x v8, a1
; RV64V-NEXT: vsetivli zero, 1, e8, mf8, ta, ma
; RV64V-NEXT: vmv.v.i v0, 5
; RV64V-NEXT: vsetivli zero, 4, e64, m2, ta, ma
+; RV64V-NEXT: vmv.v.x v8, a1
; RV64V-NEXT: vmerge.vim v10, v8, 0, v0
; RV64V-NEXT: vsetvli zero, zero, e32, m1, ta, ma
; RV64V-NEXT: vluxei64.v v8, (a0), v10
; RV64V-NEXT: ret
;
+; RV32ZVE32F-LABEL: mgather_narrow_edge_case:
+; RV32ZVE32F: # %bb.0:
+; RV32ZVE32F-NEXT: li a1, -512
+; RV32ZVE32F-NEXT: vsetivli zero, 1, e8, mf4, ta, ma
+; RV32ZVE32F-NEXT: vmv.v.i v0, 5
+; RV32ZVE32F-NEXT: vsetivli zero, 4, e32, m1, ta, ma
+; RV32ZVE32F-NEXT: vmv.v.x v8, a1
+; RV32ZVE32F-NEXT: vmerge.vim v8, v8, 0, v0
+; RV32ZVE32F-NEXT: vluxei32.v v8, (a0), v8
+; RV32ZVE32F-NEXT: ret
+;
; RV64ZVE32F-LABEL: mgather_narrow_edge_case:
; RV64ZVE32F: # %bb.0:
; RV64ZVE32F-NEXT: addi a1, a0, -512
@@ -12728,8 +12739,8 @@ define <8 x i16> @mgather_strided_unaligned(ptr %base) {
; RV32-NEXT: vmv.v.x v8, a3
; RV32-NEXT: vslide1down.vx v8, v8, a5
; RV32-NEXT: vslide1down.vx v8, v8, a6
-; RV32-NEXT: vslide1down.vx v8, v8, a7
; RV32-NEXT: vmv.v.i v0, 15
+; RV32-NEXT: vslide1down.vx v8, v8, a7
; RV32-NEXT: vslidedown.vi v8, v9, 4, v0.t
; RV32-NEXT: ret
;
@@ -12803,8 +12814,8 @@ define <8 x i16> @mgather_strided_unaligned(ptr %base) {
; RV64V-NEXT: vmv.v.x v8, a3
; RV64V-NEXT: vslide1down.vx v8, v8, a5
; RV64V-NEXT: vslide1down.vx v8, v8, a6
-; RV64V-NEXT: vslide1down.vx v8, v8, a7
; RV64V-NEXT: vmv.v.i v0, 15
+; RV64V-NEXT: vslide1down.vx v8, v8, a7
; RV64V-NEXT: vslidedown.vi v8, v9, 4, v0.t
; RV64V-NEXT: addi sp, s0, -128
; RV64V-NEXT: ld ra, 120(sp) # 8-byte Folded Reload
@@ -12854,8 +12865,8 @@ define <8 x i16> @mgather_strided_unaligned(ptr %base) {
; RV64ZVE32F-NEXT: vmv.v.x v8, a5
; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a6
; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a7
-; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a0
; RV64ZVE32F-NEXT: vmv.v.i v0, 15
+; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a0
; RV64ZVE32F-NEXT: vslidedown.vi v8, v9, 4, v0.t
; RV64ZVE32F-NEXT: ret
%ptrs = getelementptr inbounds i16, ptr %base, <8 x i32> <i32 0, i32 2, i32 4, i32 6, i32 8, i32 10, i32 12, i32 14>
@@ -12881,23 +12892,23 @@ define <8 x i16> @mgather_strided_2xSEW(ptr %base) {
;
; RV64ZVE32F-LABEL: mgather_strided_2xSEW:
; RV64ZVE32F: # %bb.0:
-; RV64ZVE32F-NEXT: lh a1, 2(a0)
-; RV64ZVE32F-NEXT: lh a2, 8(a0)
-; RV64ZVE32F-NEXT: lh a3, 10(a0)
-; RV64ZVE32F-NEXT: lh a4, 18(a0)
-; RV64ZVE32F-NEXT: lh a5, 24(a0)
-; RV64ZVE32F-NEXT: lh a6, 26(a0)
+; RV64ZVE32F-NEXT: addi a1, a0, 16
+; RV64ZVE32F-NEXT: lh a2, 2(a0)
+; RV64ZVE32F-NEXT: lh a3, 8(a0)
+; RV64ZVE32F-NEXT: lh a4, 10(a0)
+; RV64ZVE32F-NEXT: lh a5, 18(a0)
; RV64ZVE32F-NEXT: vsetivli zero, 8, e16, m1, ta, mu
; RV64ZVE32F-NEXT: vlse16.v v8, (a0), zero
-; RV64ZVE32F-NEXT: addi a0, a0, 16
-; RV64ZVE32F-NEXT: vlse16.v v9, (a0), zero
-; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a1
+; RV64ZVE32F-NEXT: lh a6, 24(a0)
+; RV64ZVE32F-NEXT: lh a0, 26(a0)
+; RV64ZVE32F-NEXT: vlse16.v v9, (a1), zero
; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a2
-; RV64ZVE32F-NEXT: vslide1down.vx v10, v8, a3
-; RV64ZVE32F-NEXT: vslide1down.vx v8, v9, a4
-; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a5
+; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a3
+; RV64ZVE32F-NEXT: vslide1down.vx v10, v8, a4
+; RV64ZVE32F-NEXT: vslide1down.vx v8, v9, a5
; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a6
; RV64ZVE32F-NEXT: vmv.v.i v0, 15
+; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a0
; RV64ZVE32F-NEXT: vslidedown.vi v8, v10, 4, v0.t
; RV64ZVE32F-NEXT: ret
%ptrs = getelementptr inbounds i16, ptr %base, <8 x i32> <i32 0, i32 1, i32 4, i32 5, i32 8, i32 9, i32 12, i32 13>
@@ -12925,24 +12936,24 @@ define <8 x i16> @mgather_strided_2xSEW_with_offset(ptr %base) {
;
; RV64ZVE32F-LABEL: mgather_strided_2xSEW_with_offset:
; RV64ZVE32F: # %bb.0:
-; RV64ZVE32F-NEXT: addi a1, a0, 4
-; RV64ZVE32F-NEXT: lh a2, 6(a0)
-; RV64ZVE32F-NEXT: lh a3, 12(a0)
-; RV64ZVE32F-NEXT: lh a4, 14(a0)
-; RV64ZVE32F-NEXT: lh a5, 22(a0)
-; RV64ZVE32F-NEXT: lh a6, 28(a0)
-; RV64ZVE32F-NEXT: lh a7, 30(a0)
+; RV64ZVE32F-NEXT: addi a1, a0, 20
+; RV64ZVE32F-NEXT: addi a2, a0, 4
+; RV64ZVE32F-NEXT: lh a3, 6(a0)
+; RV64ZVE32F-NEXT: lh a4, 12(a0)
+; RV64ZVE32F-NEXT: lh a5, 14(a0)
+; RV64ZVE32F-NEXT: lh a6, 22(a0)
; RV64ZVE32F-NEXT: vsetivli zero, 8, e16, m1, ta, mu
-; RV64ZVE32F-NEXT: vlse16.v v8, (a1), zero
-; RV64ZVE32F-NEXT: addi a0, a0, 20
-; RV64ZVE32F-NEXT: vlse16.v v9, (a0), zero
-; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a2
+; RV64ZVE32F-NEXT: vlse16.v v8, (a2), zero
+; RV64ZVE32F-NEXT: lh a2, 28(a0)
+; RV64ZVE32F-NEXT: lh a0, 30(a0)
+; RV64ZVE32F-NEXT: vlse16.v v9, (a1), zero
; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a3
-; RV64ZVE32F-NEXT: vslide1down.vx v10, v8, a4
-; RV64ZVE32F-NEXT: vslide1down.vx v8, v9, a5
-; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a6
-; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a7
+; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a4
+; RV64ZVE32F-NEXT: vslide1down.vx v10, v8, a5
+; RV64ZVE32F-NEXT: vslide1down.vx v8, v9, a6
+; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a2
; RV64ZVE32F-NEXT: vmv.v.i v0, 15
+; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a0
; RV64ZVE32F-NEXT: vslidedown.vi v8, v10, 4, v0.t
; RV64ZVE32F-NEXT: ret
%ptrs = getelementptr inbounds i16, ptr %base, <8 x i64> <i64 2, i64 3, i64 6, i64 7, i64 10, i64 11, i64 14, i64 15>
@@ -12970,24 +12981,24 @@ define <8 x i16> @mgather_reverse_unit_strided_2xSEW(ptr %base) {
;
; RV64ZVE32F-LABEL: mgather_reverse_unit_strided_2xSEW:
; RV64ZVE32F: # %bb.0:
-; RV64ZVE32F-NEXT: addi a1, a0, 28
-; RV64ZVE32F-NEXT: lh a2, 30(a0)
-; RV64ZVE32F-NEXT: lh a3, 24(a0)
-; RV64ZVE32F-NEXT: lh a4, 26(a0)
-; RV64ZVE32F-NEXT: lh a5, 22(a0)
-; RV64ZVE32F-NEXT: lh a6, 16(a0)
-; RV64ZVE32F-NEXT: lh a7, 18(a0)
+; RV64ZVE32F-NEXT: addi a1, a0, 20
+; RV64ZVE32F-NEXT: addi a2, a0, 28
+; RV64ZVE32F-NEXT: lh a3, 30(a0)
+; RV64ZVE32F-NEXT: lh a4, 24(a0)
+; RV64ZVE32F-NEXT: lh a5, 26(a0)
+; RV64ZVE32F-NEXT: lh a6, 22(a0)
; RV64ZVE32F-NEXT: vsetivli zero, 8, e16, m1, ta, mu
-; RV64ZVE32F-NEXT: vlse16.v v8, (a1), zero
-; RV64ZVE32F-NEXT: addi a0, a0, 20
-; RV64ZVE32F-NEXT: vlse16.v v9, (a0), zero
-; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a2
+; RV64ZVE32F-NEXT: vlse16.v v8, (a2), zero
+; RV64ZVE32F-NEXT: lh a2, 16(a0)
+; RV64ZVE32F-NEXT: lh a0, 18(a0)
+; RV64ZVE32F-NEXT: vlse16.v v9, (a1), zero
; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a3
-; RV64ZVE32F-NEXT: vslide1down.vx v10, v8, a4
-; RV64ZVE32F-NEXT: vslide1down.vx v8, v9, a5
-; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a6
-; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a7
+; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a4
+; RV64ZVE32F-NEXT: vslide1down.vx v10, v8, a5
+; RV64ZVE32F-NEXT: vslide1down.vx v8, v9, a6
+; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a2
; RV64ZVE32F-NEXT: vmv.v.i v0, 15
+; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a0
; RV64ZVE32F-NEXT: vslidedown.vi v8, v10, 4, v0.t
; RV64ZVE32F-NEXT: ret
%ptrs = getelementptr inbounds i16, ptr %base, <8 x i64> <i64 14, i64 15, i64 12, i64 13, i64 10, i64 11, i64 8, i64 9>
@@ -13015,24 +13026,24 @@ define <8 x i16> @mgather_reverse_strided_2xSEW(ptr %base) {
;
; RV64ZVE32F-LABEL: mgather_reverse_strided_2xSEW:
; RV64ZVE32F: # %bb.0:
-; RV64ZVE32F-NEXT: addi a1, a0, 28
-; RV64ZVE32F-NEXT: lh a2, 30(a0)
-; RV64ZVE32F-NEXT: lh a3, 20(a0)
-; RV64ZVE32F-NEXT: lh a4, 22(a0)
-; RV64ZVE32F-NEXT: lh a5, 14(a0)
-; RV64ZVE32F-NEXT: lh a6, 4(a0)
-; RV64ZVE32F-NEXT: lh a7, 6(a0)
+; RV64ZVE32F-NEXT: addi a1, a0, 12
+; RV64ZVE32F-NEXT: addi a2, a0, 28
+; RV64ZVE32F-NEXT: lh a3, 30(a0)
+; RV64ZVE32F-NEXT: lh a4, 20(a0)
+; RV64ZVE32F-NEXT: lh a5, 22(a0)
+; RV64ZVE32F-NEXT: lh a6, 14(a0)
; RV64ZVE32F-NEXT: vsetivli zero, 8, e16, m1, ta, mu
-; RV64ZVE32F-NEXT: vlse16.v v8, (a1), zero
-; RV64ZVE32F-NEXT: addi a0, a0, 12
-; RV64ZVE32F-NEXT: vlse16.v v9, (a0), zero
-; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a2
+; RV64ZVE32F-NEXT: vlse16.v v8, (a2), zero
+; RV64ZVE32F-NEXT: lh a2, 4(a0)
+; RV64ZVE32F-NEXT: lh a0, 6(a0)
+; RV64ZVE32F-NEXT: vlse16.v v9, (a1), zero
; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a3
-; RV64ZVE32F-NEXT: vslide1down.vx v10, v8, a4
-; RV64ZVE32F-NEXT: vslide1down.vx v8, v9, a5
-; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a6
-; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a7
+; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a4
+; RV64ZVE32F-NEXT: vslide1down.vx v10, v8, a5
+; RV64ZVE32F-NEXT: vslide1down.vx v8, v9, a6
+; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a2
; RV64ZVE32F-NEXT: vmv.v.i v0, 15
+; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a0
; RV64ZVE32F-NEXT: vslidedown.vi v8, v10, 4, v0.t
; RV64ZVE32F-NEXT: ret
%ptrs = getelementptr inbounds i16, ptr %base, <8 x i64> <i64 14, i64 15, i64 10, i64 11, i64 6, i64 7, i64 2, i64 3>
@@ -13059,23 +13070,23 @@ define <8 x i16> @mgather_gather_2xSEW(ptr %base) {
;
; RV64ZVE32F-LABEL: mgather_gather_2xSEW:
; RV64ZVE32F: # %bb.0:
-; RV64ZVE32F-NEXT: lh a1, 2(a0)
-; RV64ZVE32F-NEXT: lh a2, 16(a0)
-; RV64ZVE32F-NEXT: lh a3, 18(a0)
-; RV64ZVE32F-NEXT: lh a4, 10(a0)
-; RV64ZVE32F-NEXT: lh a5, 4(a0)
-; RV64ZVE32F-NEXT: lh a6, 6(a0)
+; RV64ZVE32F-NEXT: addi a1, a0, 8
+; RV64ZVE32F-NEXT: lh a2, 2(a0)
+; RV64ZVE32F-NEXT: lh a3, 16(a0)
+; RV64ZVE32F-NEXT: lh a4, 18(a0)
+; RV64ZVE32F-NEXT: lh a5, 10(a0)
; RV64ZVE32F-NEXT: vsetivli zero, 8, e16, m1, ta, mu
; RV64ZVE32F-NEXT: vlse16.v v8, (a0), zero
-; RV64ZVE32F-NEXT: addi a0, a0, 8
-; RV64ZVE32F-NEXT: vlse16.v v9, (a0), zero
-; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a1
+; RV64ZVE32F-NEXT: lh a6, 4(a0)
+; RV64ZVE32F-NEXT: lh a0, 6(a0)
+; RV64ZVE32F-NEXT: vlse16.v v9, (a1), zero
; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a2
-; RV64ZVE32F-NEXT: vslide1down.vx v10, v8, a3
-; RV64ZVE32F-NEXT: vslide1down.vx v8, v9, a4
-; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a5
+; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a3
+; RV64ZVE32F-NEXT: vslide1down.vx v10, v8, a4
+; RV64ZVE32F-NEXT: vslide1down.vx v8, v9, a5
; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a6
; RV64ZVE32F-NEXT: vmv.v.i v0, 15
+; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a0
; RV64ZVE32F-NEXT: vslidedown.vi v8, v10, 4, v0.t
; RV64ZVE32F-NEXT: ret
%ptrs = getelementptr inbounds i16, ptr %base, <8 x i32> <i32 0, i32 1, i32 8, i32 9, i32 4, i32 5, i32 2, i32 3>
@@ -13105,23 +13116,23 @@ define <8 x i16> @mgather_gather_2xSEW_unaligned(ptr %base) {
;
; RV64ZVE32F-LABEL: mgather_gather_2xSEW_unaligned:
; RV64ZVE32F: # %bb.0:
-; RV64ZVE32F-NEXT: lh a1, 2(a0)
-; RV64ZVE32F-NEXT: lh a2, 18(a0)
-; RV64ZVE32F-NEXT: lh a3, 20(a0)
-; RV64ZVE32F-NEXT: lh a4, 10(a0)
-; RV64ZVE32F-NEXT: lh a5, 4(a0)
-; RV64ZVE32F-NEXT: lh a6, 6(a0)
+; RV64ZVE32F-NEXT: addi a1, a0, 8
+; RV64ZVE32F-NEXT: lh a2, 2(a0)
+; RV64ZVE32F-NEXT: lh a3, 18(a0)
+; RV64ZVE32F-NEXT: lh a4, 20(a0)
+; RV64ZVE32F-NEXT: lh a5, 10(a0)
; RV64ZVE32F-NEXT: vsetivli zero, 8, e16, m1, ta, mu
; RV64ZVE32F-NEXT: vlse16.v v8, (a0), zero
-; RV64ZVE32F-NEXT: addi a0, a0, 8
-; RV64ZVE32F-NEXT: vlse16.v v9, (a0), zero
-; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a1
+; RV64ZVE32F-NEXT: lh a6, 4(a0)
+; RV64ZVE32F-NEXT: lh a0, 6(a0)
+; RV64ZVE32F-NEXT: vlse16.v v9, (a1), zero
; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a2
-; RV64ZVE32F-NEXT: vslide1down.vx v10, v8, a3
-; RV64ZVE32F-NEXT: vslide1down.vx v8, v9, a4
-; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a5
+; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a3
+; RV64ZVE32F-NEXT: vslide1down.vx v10, v8, a4
+; RV64ZVE32F-NEXT: vslide1down.vx v8, v9, a5
; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a6
; RV64ZVE32F-NEXT: vmv.v.i v0, 15
+; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a0
; RV64ZVE32F-NEXT: vslidedown.vi v8, v10, 4, v0.t
; RV64ZVE32F-NEXT: ret
%ptrs = getelementptr inbounds i16, ptr %base, <8 x i32> <i32 0, i32 1, i32 9, i32 10, i32 4, i32 5, i32 2, i32 3>
@@ -13152,23 +13163,23 @@ define <8 x i16> @mgather_gather_2xSEW_unaligned2(ptr %base) {
;
; RV64ZVE32F-LABEL: mgather_gather_2xSEW_unaligned2:
; RV64ZVE32F: # %bb.0:
-; RV64ZVE32F-NEXT: addi a1, a0, 2
-; RV64ZVE32F-NEXT: lh a2, 4(a0)
-; RV64ZVE32F-NEXT: lh a3, 18(a0)
-; RV64ZVE32F-NEXT: lh a4, 20(a0)
-; RV64ZVE32F-NEXT: lh a5, 10(a0)
-; RV64ZVE32F-NEXT: lh a6, 6(a0)
+; RV64ZVE32F-NEXT: addi a1, a0, 8
+; RV64ZVE32F-NEXT: addi a2, a0, 2
+; RV64ZVE32F-NEXT: lh a3, 4(a0)
+; RV64ZVE32F-NEXT: lh a4, 18(a0)
+; RV64ZVE32F-NEXT: lh a5, 20(a0)
; RV64ZVE32F-NEXT: vsetivli zero, 8, e16, m1, ta, mu
-; RV64ZVE32F-NEXT: vlse16.v v8, (a1), zero
-; RV64ZVE32F-NEXT: addi a0, a0, 8
-; RV64ZVE32F-NEXT: vlse16.v v9, (a0), zero
-; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a2
+; RV64ZVE32F-NEXT: vlse16.v v8, (a2), zero
+; RV64ZVE32F-NEXT: lh a2, 10(a0)
+; RV64ZVE32F-NEXT: lh a0, 6(a0)
+; RV64ZVE32F-NEXT: vlse16.v v9, (a1), zero
+; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a3
+; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a4
+; RV64ZVE32F-NEXT: vslide1down.vx v10, v8, a5
+; RV64ZVE32F-NEXT: vslide1down.vx v8, v9, a2
; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a3
-; RV64ZVE32F-NEXT: vslide1down.vx v10, v8, a4
-; RV64ZVE32F-NEXT: vslide1down.vx v8, v9, a5
-; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a2
-; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a6
; RV64ZVE32F-NEXT: vmv.v.i v0, 15
+; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a0
; RV64ZVE32F-NEXT: vslidedown.vi v8, v10, 4, v0.t
; RV64ZVE32F-NEXT: ret
%ptrs = getelementptr inbounds i16, ptr %base, <8 x i32> <i32 1, i32 2, i32 9, i32 10, i32 4, i32 5, i32 2, i32 3>
@@ -13202,23 +13213,23 @@ define <8 x i16> @mgather_gather_4xSEW(ptr %base) {
;
; RV64ZVE32F-LABEL: mgather_gather_4xSEW:
; RV64ZVE32F: # %bb.0:
-; RV64ZVE32F-NEXT: lh a1, 2(a0)
-; RV64ZVE32F-NEXT: lh a2, 4(a0)
-; RV64ZVE32F-NEXT: lh a3, 6(a0)
-; RV64ZVE32F-NEXT: lh a4, 18(a0)
-; RV64ZVE32F-NEXT: lh a5, 20(a0)
-; RV64ZVE32F-NEXT: lh a6, 22(a0)
+; RV64ZVE32F-NEXT: addi a1, a0, 16
+; RV64ZVE32F-NEXT: lh a2, 2(a0)
+; RV64ZVE32F-NEXT: lh a3, 4(a0)
+; RV64ZVE32F-NEXT: lh a4, 6(a0)
+; RV64ZVE32F-NEXT: lh a5, 18(a0)
; RV64ZVE32F-NEXT: vsetivli zero, 8, e16, m1, ta, mu
; RV64ZVE32F-NEXT: vlse16.v v8, (a0), zero
-; RV64ZVE32F-NEXT: addi a0, a0, 16
-; RV64ZVE32F-NEXT: vlse16.v v9, (a0), zero
-; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a1
+; RV64ZVE32F-NEXT: lh a6, 20(a0)
+; RV64ZVE32F-NEXT: lh a0, 22(a0)
+; RV64ZVE32F-NEXT: vlse16.v v9, (a1), zero
; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a2
-; RV64ZVE32F-NEXT: vslide1down.vx v10, v8, a3
-; RV64ZVE32F-NEXT: vslide1down.vx v8, v9, a4
-; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a5
+; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a3
+; RV64ZVE32F-NEXT: vslide1down.vx v10, v8, a4
+; RV64ZVE32F-NEXT: vslide1down.vx v8, v9, a5
; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a6
; RV64ZVE32F-NEXT: vmv.v.i v0, 15
+; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a0
; RV64ZVE32F-NEXT: vslidedown.vi v8, v10, 4, v0.t
; RV64ZVE32F-NEXT: ret
%ptrs = getelementptr inbounds i16, ptr %base, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 8, i32 9, i32 10, i32 11>
@@ -13249,23 +13260,23 @@ define <8 x i16> @mgather_gather_4xSEW_partial_align(ptr %base) {
;
; RV64ZVE32F-LABEL: mgather_gather_4xSEW_partial_align:
; RV64ZVE32F: # %bb.0:
-; RV64ZVE32F-NEXT: lh a1, 2(a0)
-; RV64ZVE32F-NEXT: lh a2, 4(a0)
-; RV64ZVE32F-NEXT: lh a3, 6(a0)
-; RV64ZVE32F-NEXT: lh a4, 18(a0)
-; RV64ZVE32F-NEXT: lh a5, 20(a0)
-; RV64ZVE32F-NEXT: lh a6, 22(a0)
+; RV64ZVE32F-NEXT: addi a1, a0, 16
+; RV64ZVE32F-NEXT: lh a2, 2(a0)
+; RV64ZVE32F-NEXT: lh a3, 4(a0)
+; RV64ZVE32F-NEXT: lh a4, 6(a0)
+; RV64ZVE32F-NEXT: lh a5, 18(a0)
; RV64ZVE32F-NEXT: vsetivli zero, 8, e16, m1, ta, mu
; RV64ZVE32F-NEXT: vlse16.v v8, (a0), zero
-; RV64ZVE32F-NEXT: addi a0, a0, 16
-; RV64ZVE32F-NEXT: vlse16.v v9, (a0), zero
-; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a1
+; RV64ZVE32F-NEXT: lh a6, 20(a0)
+; RV64ZVE32F-NEXT: lh a0, 22(a0)
+; RV64ZVE32F-NEXT: vlse16.v v9, (a1), zero
; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a2
-; RV64ZVE32F-NEXT: vslide1down.vx v10, v8, a3
-; RV64ZVE32F-NEXT: vslide1down.vx v8, v9, a4
-; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a5
+; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a3
+; RV64ZVE32F-NEXT: vslide1down.vx v10, v8, a4
+; RV64ZVE32F-NEXT: vslide1down.vx v8, v9, a5
; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a6
; RV64ZVE32F-NEXT: vmv.v.i v0, 15
+; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a0
; RV64ZVE32F-NEXT: vslidedown.vi v8, v10, 4, v0.t
; RV64ZVE32F-NEXT: ret
%ptrs = getelementptr inbounds i16, ptr %base, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 8, i32 9, i32 10, i32 11>
@@ -13305,23 +13316,23 @@ define <8 x i16> @mgather_shuffle_rotate(ptr %base) {
;
; RV64ZVE32F-LABEL: mgather_shuffle_rotate:
; RV64ZVE32F: # %bb.0:
-; RV64ZVE32F-NEXT: lh a1, 10(a0)
-; RV64ZVE32F-NEXT: lh a2, 12(a0)
-; RV64ZVE32F-NEXT: lh a3, 14(a0)
-; RV64ZVE32F-NEXT: lh a4, 2(a0)
-; RV64ZVE32F-NEXT: lh a5, 4(a0)
-; RV64ZVE32F-NEXT: lh a6, 6(a0)
+; RV64ZVE32F-NEXT: addi a1, a0, 8
+; RV64ZVE32F-NEXT: lh a2, 10(a0)
+; RV64ZVE32F-NEXT: lh a3, 12(a0)
+; RV64ZVE32F-NEXT: lh a4, 14(a0)
+; RV64ZVE32F-NEXT: lh a5, 2(a0)
; RV64ZVE32F-NEXT: vsetivli zero, 8, e16, m1, ta, mu
; RV64ZVE32F-NEXT: vlse16.v v8, (a0), zero
-; RV64ZVE32F-NEXT: addi a0, a0, 8
-; RV64ZVE32F-NEXT: vlse16.v v9, (a0), zero
-; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a4
+; RV64ZVE32F-NEXT: lh a6, 4(a0)
+; RV64ZVE32F-NEXT: lh a0, 6(a0)
+; RV64ZVE32F-NEXT: vlse16.v v9, (a1), zero
; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a5
; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a6
-; RV64ZVE32F-NEXT: vslide1down.vx v9, v9, a1
+; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a0
; RV64ZVE32F-NEXT: vslide1down.vx v9, v9, a2
; RV64ZVE32F-NEXT: vslide1down.vx v9, v9, a3
; RV64ZVE32F-NEXT: vmv.v.i v0, 15
+; RV64ZVE32F-NEXT: vslide1down.vx v9, v9, a4
; RV64ZVE32F-NEXT: vslidedown.vi v8, v9, 4, v0.t
; RV64ZVE32F-NEXT: ret
%ptrs = getelementptr inbounds i16, ptr %base, <8 x i64> <i64 4, i64 5, i64 6, i64 7, i64 0, i64 1, i64 2, i64 3>
@@ -13352,23 +13363,23 @@ define <8 x i16> @mgather_shuffle_vrgather(ptr %base) {
;
; RV64ZVE32F-LABEL: mgather_shuffle_vrgather:
; RV64ZVE32F: # %bb.0:
-; RV64ZVE32F-NEXT: lh a1, 4(a0)
-; RV64ZVE32F-NEXT: lh a2, 6(a0)
-; RV64ZVE32F-NEXT: lh a3, 2(a0)
-; RV64ZVE32F-NEXT: lh a4, 10(a0)
-; RV64ZVE32F-NEXT: lh a5, 12(a0)
-; RV64ZVE32F-NEXT: lh a6, 14(a0)
+; RV64ZVE32F-NEXT: addi a1, a0, 8
+; RV64ZVE32F-NEXT: lh a2, 4(a0)
+; RV64ZVE32F-NEXT: lh a3, 6(a0)
+; RV64ZVE32F-NEXT: lh a4, 2(a0)
+; RV64ZVE32F-NEXT: lh a5, 10(a0)
; RV64ZVE32F-NEXT: vsetivli zero, 8, e16, m1, ta, mu
; RV64ZVE32F-NEXT: vlse16.v v8, (a0), zero
-; RV64ZVE32F-NEXT: addi a0, a0, 8
-; RV64ZVE32F-NEXT: vlse16.v v9, (a0), zero
-; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a1
+; RV64ZVE32F-NEXT: lh a6, 12(a0)
+; RV64ZVE32F-NEXT: lh a0, 14(a0)
+; RV64ZVE32F-NEXT: vlse16.v v9, (a1), zero
; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a2
-; RV64ZVE32F-NEXT: vslide1down.vx v10, v8, a3
-; RV64ZVE32F-NEXT: vslide1down.vx v8, v9, a4
-; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a5
+; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a3
+; RV64ZVE32F-NEXT: vslide1down.vx v10, v8, a4
+; RV64ZVE32F-NEXT: vslide1down.vx v8, v9, a5
; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a6
; RV64ZVE32F-NEXT: vmv.v.i v0, 15
+; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a0
; RV64ZVE32F-NEXT: vslidedown.vi v8, v10, 4, v0.t
; RV64ZVE32F-NEXT: ret
%ptrs = getelementptr inbounds i16, ptr %base, <8 x i64> <i64 0, i64 2, i64 3, i64 1, i64 4, i64 5, i64 6, i64 7>
@@ -13853,12 +13864,12 @@ define <4 x i32> @masked_gather_widen_sew_negative_stride(ptr %base) {
; RV64ZVE32F: # %bb.0:
; RV64ZVE32F-NEXT: addi a1, a0, 136
; RV64ZVE32F-NEXT: lw a2, 140(a0)
-; RV64ZVE32F-NEXT: lw a3, 0(a0)
-; RV64ZVE32F-NEXT: lw a0, 4(a0)
; RV64ZVE32F-NEXT: vsetivli zero, 4, e32, m1, ta, ma
; RV64ZVE32F-NEXT: vlse32.v v8, (a1), zero
+; RV64ZVE32F-NEXT: lw a1, 0(a0)
+; RV64ZVE32F-NEXT: lw a0, 4(a0)
; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a2
-; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a3
+; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a1
; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a0
; RV64ZVE32F-NEXT: ret
%ptrs = getelementptr i32, ptr %base, <4 x i64> <i64 34, i64 35, i64 0, i64 1>
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-masked-load-int.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-masked-load-int.ll
index 4bbda2152a6f9..ad075e4b4e198 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-masked-load-int.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-masked-load-int.ll
@@ -401,14 +401,14 @@ define void @masked_load_v32i64(ptr %a, ptr %m_ptr, ptr %res_ptr) nounwind {
; RV32: # %bb.0:
; RV32-NEXT: addi a3, a1, 128
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vle64.v v16, (a3)
; RV32-NEXT: vle64.v v0, (a1)
+; RV32-NEXT: vle64.v v24, (a3)
; RV32-NEXT: li a1, 32
; RV32-NEXT: vsetvli zero, a1, e32, m8, ta, ma
-; RV32-NEXT: vmv.v.i v24, 0
+; RV32-NEXT: vmv.v.i v16, 0
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vmseq.vv v8, v0, v24
-; RV32-NEXT: vmseq.vv v0, v16, v24
+; RV32-NEXT: vmseq.vv v8, v0, v16
+; RV32-NEXT: vmseq.vv v0, v24, v16
; RV32-NEXT: addi a1, a0, 128
; RV32-NEXT: vle64.v v16, (a1), v0.t
; RV32-NEXT: vmv1r.v v0, v8
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-masked-scatter.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-masked-scatter.ll
index aa815e18ac101..52b2ffbc21c01 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-masked-scatter.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-masked-scatter.ll
@@ -335,18 +335,18 @@ define void @mscatter_truemask_v4i8(<4 x i8> %val, <4 x ptr> %ptrs) {
;
; RV64ZVE32F-LABEL: mscatter_truemask_v4i8:
; RV64ZVE32F: # %bb.0:
-; RV64ZVE32F-NEXT: ld a1, 24(a0)
-; RV64ZVE32F-NEXT: ld a2, 0(a0)
-; RV64ZVE32F-NEXT: ld a3, 16(a0)
-; RV64ZVE32F-NEXT: ld a0, 8(a0)
+; RV64ZVE32F-NEXT: ld a1, 0(a0)
+; RV64ZVE32F-NEXT: ld a2, 24(a0)
+; RV64ZVE32F-NEXT: ld a3, 8(a0)
+; RV64ZVE32F-NEXT: ld a0, 16(a0)
; RV64ZVE32F-NEXT: vsetivli zero, 1, e8, mf4, ta, ma
-; RV64ZVE32F-NEXT: vse8.v v8, (a2)
+; RV64ZVE32F-NEXT: vse8.v v8, (a1)
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 1
-; RV64ZVE32F-NEXT: vse8.v v9, (a0)
-; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 2
; RV64ZVE32F-NEXT: vse8.v v9, (a3)
+; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 2
+; RV64ZVE32F-NEXT: vse8.v v9, (a0)
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 3
-; RV64ZVE32F-NEXT: vse8.v v8, (a1)
+; RV64ZVE32F-NEXT: vse8.v v8, (a2)
; RV64ZVE32F-NEXT: ret
call void @llvm.masked.scatter.v4i8.v4p0(<4 x i8> %val, <4 x ptr> %ptrs, i32 1, <4 x i1> splat (i1 1))
ret void
@@ -504,8 +504,8 @@ define void @mscatter_baseidx_v8i8(<8 x i8> %val, ptr %base, <8 x i8> %idxs, <8
; RV64ZVE32F-NEXT: .LBB9_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v9, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v9, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB9_12
; RV64ZVE32F-NEXT: # %bb.5: # %else4
@@ -526,8 +526,8 @@ define void @mscatter_baseidx_v8i8(<8 x i8> %val, ptr %base, <8 x i8> %idxs, <8
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 5
; RV64ZVE32F-NEXT: vse8.v v9, (a2)
; RV64ZVE32F-NEXT: .LBB9_9: # %else10
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v10, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB9_15
; RV64ZVE32F-NEXT: # %bb.10: # %else12
@@ -689,11 +689,11 @@ define void @mscatter_v2i32_truncstore_v2i16(<2 x i32> %val, <2 x ptr> %ptrs, <2
;
; RV64ZVE32F-LABEL: mscatter_v2i32_truncstore_v2i16:
; RV64ZVE32F: # %bb.0:
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
-; RV64ZVE32F-NEXT: vnsrl.wi v8, v8, 0
-; RV64ZVE32F-NEXT: vsetvli zero, zero, e8, mf4, ta, ma
+; RV64ZVE32F-NEXT: vsetivli zero, 1, e8, m1, ta, ma
; RV64ZVE32F-NEXT: vmv.x.s a2, v0
; RV64ZVE32F-NEXT: andi a3, a2, 1
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
+; RV64ZVE32F-NEXT: vnsrl.wi v8, v8, 0
; RV64ZVE32F-NEXT: bnez a3, .LBB12_3
; RV64ZVE32F-NEXT: # %bb.1: # %else
; RV64ZVE32F-NEXT: andi a2, a2, 2
@@ -747,13 +747,14 @@ define void @mscatter_v2i64_truncstore_v2i16(<2 x i64> %val, <2 x ptr> %ptrs, <2
;
; RV64ZVE32F-LABEL: mscatter_v2i64_truncstore_v2i16:
; RV64ZVE32F: # %bb.0:
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
+; RV64ZVE32F-NEXT: vsetivli zero, 1, e16, m1, ta, ma
; RV64ZVE32F-NEXT: vmv.s.x v9, a1
; RV64ZVE32F-NEXT: vmv.s.x v8, a0
-; RV64ZVE32F-NEXT: vslideup.vi v8, v9, 1
-; RV64ZVE32F-NEXT: vsetvli zero, zero, e8, mf4, ta, ma
+; RV64ZVE32F-NEXT: vsetvli zero, zero, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vmv.x.s a0, v0
; RV64ZVE32F-NEXT: andi a1, a0, 1
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
+; RV64ZVE32F-NEXT: vslideup.vi v8, v9, 1
; RV64ZVE32F-NEXT: bnez a1, .LBB13_3
; RV64ZVE32F-NEXT: # %bb.1: # %else
; RV64ZVE32F-NEXT: andi a0, a0, 2
@@ -852,18 +853,18 @@ define void @mscatter_truemask_v4i16(<4 x i16> %val, <4 x ptr> %ptrs) {
;
; RV64ZVE32F-LABEL: mscatter_truemask_v4i16:
; RV64ZVE32F: # %bb.0:
-; RV64ZVE32F-NEXT: ld a1, 24(a0)
-; RV64ZVE32F-NEXT: ld a2, 0(a0)
-; RV64ZVE32F-NEXT: ld a3, 16(a0)
-; RV64ZVE32F-NEXT: ld a0, 8(a0)
+; RV64ZVE32F-NEXT: ld a1, 0(a0)
+; RV64ZVE32F-NEXT: ld a2, 24(a0)
+; RV64ZVE32F-NEXT: ld a3, 8(a0)
+; RV64ZVE32F-NEXT: ld a0, 16(a0)
; RV64ZVE32F-NEXT: vsetivli zero, 1, e16, mf2, ta, ma
-; RV64ZVE32F-NEXT: vse16.v v8, (a2)
+; RV64ZVE32F-NEXT: vse16.v v8, (a1)
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 1
-; RV64ZVE32F-NEXT: vse16.v v9, (a0)
-; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 2
; RV64ZVE32F-NEXT: vse16.v v9, (a3)
+; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 2
+; RV64ZVE32F-NEXT: vse16.v v9, (a0)
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 3
-; RV64ZVE32F-NEXT: vse16.v v8, (a1)
+; RV64ZVE32F-NEXT: vse16.v v8, (a2)
; RV64ZVE32F-NEXT: ret
call void @llvm.masked.scatter.v4i16.v4p0(<4 x i16> %val, <4 x ptr> %ptrs, i32 2, <4 x i1> splat (i1 1))
ret void
@@ -1025,8 +1026,8 @@ define void @mscatter_baseidx_v8i8_v8i16(<8 x i16> %val, ptr %base, <8 x i8> %id
; RV64ZVE32F-NEXT: .LBB18_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v9, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v9, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB18_12
; RV64ZVE32F-NEXT: # %bb.5: # %else4
@@ -1048,8 +1049,8 @@ define void @mscatter_baseidx_v8i8_v8i16(<8 x i16> %val, ptr %base, <8 x i8> %id
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 5
; RV64ZVE32F-NEXT: vse16.v v9, (a2)
; RV64ZVE32F-NEXT: .LBB18_9: # %else10
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v10, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB18_15
; RV64ZVE32F-NEXT: # %bb.10: # %else12
@@ -1158,8 +1159,8 @@ define void @mscatter_baseidx_sext_v8i8_v8i16(<8 x i16> %val, ptr %base, <8 x i8
; RV64ZVE32F-NEXT: .LBB19_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v9, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v9, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB19_12
; RV64ZVE32F-NEXT: # %bb.5: # %else4
@@ -1181,8 +1182,8 @@ define void @mscatter_baseidx_sext_v8i8_v8i16(<8 x i16> %val, ptr %base, <8 x i8
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 5
; RV64ZVE32F-NEXT: vse16.v v9, (a2)
; RV64ZVE32F-NEXT: .LBB19_9: # %else10
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v10, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB19_15
; RV64ZVE32F-NEXT: # %bb.10: # %else12
@@ -1292,8 +1293,8 @@ define void @mscatter_baseidx_zext_v8i8_v8i16(<8 x i16> %val, ptr %base, <8 x i8
; RV64ZVE32F-NEXT: .LBB20_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v9, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v9, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB20_12
; RV64ZVE32F-NEXT: # %bb.5: # %else4
@@ -1316,8 +1317,8 @@ define void @mscatter_baseidx_zext_v8i8_v8i16(<8 x i16> %val, ptr %base, <8 x i8
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 5
; RV64ZVE32F-NEXT: vse16.v v9, (a2)
; RV64ZVE32F-NEXT: .LBB20_9: # %else10
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v10, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB20_15
; RV64ZVE32F-NEXT: # %bb.10: # %else12
@@ -1430,8 +1431,8 @@ define void @mscatter_baseidx_v8i16(<8 x i16> %val, ptr %base, <8 x i16> %idxs,
; RV64ZVE32F-NEXT: .LBB21_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e16, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v9, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v9, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB21_12
; RV64ZVE32F-NEXT: # %bb.5: # %else4
@@ -1453,8 +1454,8 @@ define void @mscatter_baseidx_v8i16(<8 x i16> %val, ptr %base, <8 x i16> %idxs,
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 5
; RV64ZVE32F-NEXT: vse16.v v9, (a2)
; RV64ZVE32F-NEXT: .LBB21_9: # %else10
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v10, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB21_15
; RV64ZVE32F-NEXT: # %bb.10: # %else12
@@ -1625,11 +1626,12 @@ define void @mscatter_v2i64_truncstore_v2i32(<2 x i64> %val, <2 x ptr> %ptrs, <2
; RV64ZVE32F: # %bb.0:
; RV64ZVE32F-NEXT: vsetivli zero, 2, e32, m1, ta, ma
; RV64ZVE32F-NEXT: vmv.v.x v8, a0
-; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a1
; RV64ZVE32F-NEXT: vsetvli zero, zero, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vmv.x.s a0, v0
-; RV64ZVE32F-NEXT: andi a1, a0, 1
-; RV64ZVE32F-NEXT: bnez a1, .LBB24_3
+; RV64ZVE32F-NEXT: andi a4, a0, 1
+; RV64ZVE32F-NEXT: vsetvli zero, zero, e32, m1, ta, ma
+; RV64ZVE32F-NEXT: vslide1down.vx v8, v8, a1
+; RV64ZVE32F-NEXT: bnez a4, .LBB24_3
; RV64ZVE32F-NEXT: # %bb.1: # %else
; RV64ZVE32F-NEXT: andi a0, a0, 2
; RV64ZVE32F-NEXT: bnez a0, .LBB24_4
@@ -1727,18 +1729,18 @@ define void @mscatter_truemask_v4i32(<4 x i32> %val, <4 x ptr> %ptrs) {
;
; RV64ZVE32F-LABEL: mscatter_truemask_v4i32:
; RV64ZVE32F: # %bb.0:
-; RV64ZVE32F-NEXT: ld a1, 24(a0)
-; RV64ZVE32F-NEXT: ld a2, 0(a0)
-; RV64ZVE32F-NEXT: ld a3, 16(a0)
-; RV64ZVE32F-NEXT: ld a0, 8(a0)
+; RV64ZVE32F-NEXT: ld a1, 0(a0)
+; RV64ZVE32F-NEXT: ld a2, 24(a0)
+; RV64ZVE32F-NEXT: ld a3, 8(a0)
+; RV64ZVE32F-NEXT: ld a0, 16(a0)
; RV64ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
-; RV64ZVE32F-NEXT: vse32.v v8, (a2)
+; RV64ZVE32F-NEXT: vse32.v v8, (a1)
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 1
-; RV64ZVE32F-NEXT: vse32.v v9, (a0)
-; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 2
; RV64ZVE32F-NEXT: vse32.v v9, (a3)
+; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 2
+; RV64ZVE32F-NEXT: vse32.v v9, (a0)
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 3
-; RV64ZVE32F-NEXT: vse32.v v8, (a1)
+; RV64ZVE32F-NEXT: vse32.v v8, (a2)
; RV64ZVE32F-NEXT: ret
call void @llvm.masked.scatter.v4i32.v4p0(<4 x i32> %val, <4 x ptr> %ptrs, i32 4, <4 x i1> splat (i1 1))
ret void
@@ -1903,8 +1905,8 @@ define void @mscatter_baseidx_v8i8_v8i32(<8 x i32> %val, ptr %base, <8 x i8> %id
; RV64ZVE32F-NEXT: .LBB29_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v11, v10, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v10, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB29_12
; RV64ZVE32F-NEXT: # %bb.5: # %else4
@@ -1927,8 +1929,8 @@ define void @mscatter_baseidx_v8i8_v8i32(<8 x i32> %val, ptr %base, <8 x i8> %id
; RV64ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV64ZVE32F-NEXT: vse32.v v12, (a2)
; RV64ZVE32F-NEXT: .LBB29_9: # %else10
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v11, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB29_15
; RV64ZVE32F-NEXT: # %bb.10: # %else12
@@ -2039,8 +2041,8 @@ define void @mscatter_baseidx_sext_v8i8_v8i32(<8 x i32> %val, ptr %base, <8 x i8
; RV64ZVE32F-NEXT: .LBB30_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v11, v10, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v10, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB30_12
; RV64ZVE32F-NEXT: # %bb.5: # %else4
@@ -2063,8 +2065,8 @@ define void @mscatter_baseidx_sext_v8i8_v8i32(<8 x i32> %val, ptr %base, <8 x i8
; RV64ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV64ZVE32F-NEXT: vse32.v v12, (a2)
; RV64ZVE32F-NEXT: .LBB30_9: # %else10
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v11, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB30_15
; RV64ZVE32F-NEXT: # %bb.10: # %else12
@@ -2179,8 +2181,8 @@ define void @mscatter_baseidx_zext_v8i8_v8i32(<8 x i32> %val, ptr %base, <8 x i8
; RV64ZVE32F-NEXT: .LBB31_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v11, v10, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v10, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB31_12
; RV64ZVE32F-NEXT: # %bb.5: # %else4
@@ -2204,8 +2206,8 @@ define void @mscatter_baseidx_zext_v8i8_v8i32(<8 x i32> %val, ptr %base, <8 x i8
; RV64ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV64ZVE32F-NEXT: vse32.v v12, (a2)
; RV64ZVE32F-NEXT: .LBB31_9: # %else10
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v11, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB31_15
; RV64ZVE32F-NEXT: # %bb.10: # %else12
@@ -2323,8 +2325,8 @@ define void @mscatter_baseidx_v8i16_v8i32(<8 x i32> %val, ptr %base, <8 x i16> %
; RV64ZVE32F-NEXT: .LBB32_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e16, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v11, v10, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v10, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB32_12
; RV64ZVE32F-NEXT: # %bb.5: # %else4
@@ -2347,8 +2349,8 @@ define void @mscatter_baseidx_v8i16_v8i32(<8 x i32> %val, ptr %base, <8 x i16> %
; RV64ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV64ZVE32F-NEXT: vse32.v v12, (a2)
; RV64ZVE32F-NEXT: .LBB32_9: # %else10
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v11, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB32_15
; RV64ZVE32F-NEXT: # %bb.10: # %else12
@@ -2460,8 +2462,8 @@ define void @mscatter_baseidx_sext_v8i16_v8i32(<8 x i32> %val, ptr %base, <8 x i
; RV64ZVE32F-NEXT: .LBB33_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e16, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v11, v10, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v10, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB33_12
; RV64ZVE32F-NEXT: # %bb.5: # %else4
@@ -2484,8 +2486,8 @@ define void @mscatter_baseidx_sext_v8i16_v8i32(<8 x i32> %val, ptr %base, <8 x i
; RV64ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV64ZVE32F-NEXT: vse32.v v12, (a2)
; RV64ZVE32F-NEXT: .LBB33_9: # %else10
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v11, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB33_15
; RV64ZVE32F-NEXT: # %bb.10: # %else12
@@ -2601,8 +2603,8 @@ define void @mscatter_baseidx_zext_v8i16_v8i32(<8 x i32> %val, ptr %base, <8 x i
; RV64ZVE32F-NEXT: .LBB34_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e16, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v11, v10, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: andi a3, a2, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v10, 2
; RV64ZVE32F-NEXT: bnez a3, .LBB34_12
; RV64ZVE32F-NEXT: # %bb.5: # %else4
@@ -2626,8 +2628,8 @@ define void @mscatter_baseidx_zext_v8i16_v8i32(<8 x i32> %val, ptr %base, <8 x i
; RV64ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV64ZVE32F-NEXT: vse32.v v12, (a3)
; RV64ZVE32F-NEXT: .LBB34_9: # %else10
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: andi a3, a2, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v11, 2
; RV64ZVE32F-NEXT: bnez a3, .LBB34_15
; RV64ZVE32F-NEXT: # %bb.10: # %else12
@@ -2742,8 +2744,8 @@ define void @mscatter_baseidx_v8i32(<8 x i32> %val, ptr %base, <8 x i32> %idxs,
; RV64ZVE32F-NEXT: .LBB35_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e32, m2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v12, v10, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e32, m1, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e32, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v10, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB35_12
; RV64ZVE32F-NEXT: # %bb.5: # %else4
@@ -2766,8 +2768,8 @@ define void @mscatter_baseidx_v8i32(<8 x i32> %val, ptr %base, <8 x i32> %idxs,
; RV64ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV64ZVE32F-NEXT: vse32.v v10, (a2)
; RV64ZVE32F-NEXT: .LBB35_9: # %else10
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e32, m1, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e32, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v12, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB35_15
; RV64ZVE32F-NEXT: # %bb.10: # %else12
@@ -3074,17 +3076,17 @@ define void @mscatter_truemask_v4i64(<4 x i64> %val, <4 x ptr> %ptrs) {
; RV32ZVE32F-NEXT: lw a3, 20(a0)
; RV32ZVE32F-NEXT: lw a4, 16(a0)
; RV32ZVE32F-NEXT: lw a5, 12(a0)
-; RV32ZVE32F-NEXT: lw a6, 8(a0)
-; RV32ZVE32F-NEXT: lw a7, 0(a0)
-; RV32ZVE32F-NEXT: lw a0, 4(a0)
+; RV32ZVE32F-NEXT: lw a6, 0(a0)
+; RV32ZVE32F-NEXT: lw a7, 4(a0)
+; RV32ZVE32F-NEXT: lw a0, 8(a0)
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vmv.x.s t0, v8
-; RV32ZVE32F-NEXT: sw a7, 0(t0)
-; RV32ZVE32F-NEXT: sw a0, 4(t0)
+; RV32ZVE32F-NEXT: sw a6, 0(t0)
+; RV32ZVE32F-NEXT: sw a7, 4(t0)
; RV32ZVE32F-NEXT: vslidedown.vi v9, v8, 1
-; RV32ZVE32F-NEXT: vmv.x.s a0, v9
-; RV32ZVE32F-NEXT: sw a6, 0(a0)
-; RV32ZVE32F-NEXT: sw a5, 4(a0)
+; RV32ZVE32F-NEXT: vmv.x.s a6, v9
+; RV32ZVE32F-NEXT: sw a0, 0(a6)
+; RV32ZVE32F-NEXT: sw a5, 4(a6)
; RV32ZVE32F-NEXT: vslidedown.vi v9, v8, 2
; RV32ZVE32F-NEXT: vmv.x.s a0, v9
; RV32ZVE32F-NEXT: sw a4, 0(a0)
@@ -3383,42 +3385,43 @@ define void @mscatter_baseidx_v8i8_v8i64(<8 x i64> %val, ptr %base, <8 x i8> %id
; RV32ZVE32F-NEXT: lw a5, 48(a0)
; RV32ZVE32F-NEXT: lw a6, 44(a0)
; RV32ZVE32F-NEXT: lw a7, 40(a0)
-; RV32ZVE32F-NEXT: lw t0, 36(a0)
-; RV32ZVE32F-NEXT: lw t1, 32(a0)
-; RV32ZVE32F-NEXT: lw t2, 28(a0)
-; RV32ZVE32F-NEXT: lw t3, 24(a0)
-; RV32ZVE32F-NEXT: lw t4, 20(a0)
-; RV32ZVE32F-NEXT: lw t5, 16(a0)
-; RV32ZVE32F-NEXT: lw s0, 12(a0)
-; RV32ZVE32F-NEXT: lw t6, 8(a0)
+; RV32ZVE32F-NEXT: lw t1, 36(a0)
+; RV32ZVE32F-NEXT: lw t2, 32(a0)
+; RV32ZVE32F-NEXT: lw t3, 28(a0)
+; RV32ZVE32F-NEXT: lw t4, 24(a0)
+; RV32ZVE32F-NEXT: lw t5, 20(a0)
+; RV32ZVE32F-NEXT: lw t6, 16(a0)
+; RV32ZVE32F-NEXT: lw s1, 12(a0)
+; RV32ZVE32F-NEXT: lw s0, 8(a0)
; RV32ZVE32F-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vsext.vf4 v10, v8
; RV32ZVE32F-NEXT: vsll.vi v8, v10, 3
-; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
; RV32ZVE32F-NEXT: vsetvli zero, zero, e8, mf2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a1, v0
-; RV32ZVE32F-NEXT: andi s1, a1, 1
-; RV32ZVE32F-NEXT: bnez s1, .LBB42_10
+; RV32ZVE32F-NEXT: vmv.x.s t0, v0
+; RV32ZVE32F-NEXT: andi s2, t0, 1
+; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
+; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
+; RV32ZVE32F-NEXT: bnez s2, .LBB42_10
; RV32ZVE32F-NEXT: # %bb.1: # %else
-; RV32ZVE32F-NEXT: andi a0, a1, 2
+; RV32ZVE32F-NEXT: andi a0, t0, 2
; RV32ZVE32F-NEXT: bnez a0, .LBB42_11
; RV32ZVE32F-NEXT: .LBB42_2: # %else2
-; RV32ZVE32F-NEXT: andi a0, a1, 4
+; RV32ZVE32F-NEXT: andi a0, t0, 4
; RV32ZVE32F-NEXT: bnez a0, .LBB42_12
; RV32ZVE32F-NEXT: .LBB42_3: # %else4
-; RV32ZVE32F-NEXT: andi a0, a1, 8
+; RV32ZVE32F-NEXT: andi a0, t0, 8
; RV32ZVE32F-NEXT: bnez a0, .LBB42_13
; RV32ZVE32F-NEXT: .LBB42_4: # %else6
-; RV32ZVE32F-NEXT: andi a0, a1, 16
+; RV32ZVE32F-NEXT: andi a0, t0, 16
; RV32ZVE32F-NEXT: bnez a0, .LBB42_14
; RV32ZVE32F-NEXT: .LBB42_5: # %else8
-; RV32ZVE32F-NEXT: andi a0, a1, 32
+; RV32ZVE32F-NEXT: andi a0, t0, 32
; RV32ZVE32F-NEXT: bnez a0, .LBB42_15
; RV32ZVE32F-NEXT: .LBB42_6: # %else10
-; RV32ZVE32F-NEXT: andi a0, a1, 64
+; RV32ZVE32F-NEXT: andi a0, t0, 64
; RV32ZVE32F-NEXT: bnez a0, .LBB42_16
; RV32ZVE32F-NEXT: .LBB42_7: # %else12
-; RV32ZVE32F-NEXT: andi a0, a1, -128
+; RV32ZVE32F-NEXT: andi a0, t0, -128
; RV32ZVE32F-NEXT: beqz a0, .LBB42_9
; RV32ZVE32F-NEXT: .LBB42_8: # %cond.store13
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -3433,45 +3436,44 @@ define void @mscatter_baseidx_v8i8_v8i64(<8 x i64> %val, ptr %base, <8 x i8> %id
; RV32ZVE32F-NEXT: addi sp, sp, 16
; RV32ZVE32F-NEXT: ret
; RV32ZVE32F-NEXT: .LBB42_10: # %cond.store
-; RV32ZVE32F-NEXT: lw s1, 4(a0)
+; RV32ZVE32F-NEXT: lw a1, 4(a0)
; RV32ZVE32F-NEXT: lw a0, 0(a0)
-; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vmv.x.s s2, v8
-; RV32ZVE32F-NEXT: sw s1, 4(s2)
+; RV32ZVE32F-NEXT: sw a1, 4(s2)
; RV32ZVE32F-NEXT: sw a0, 0(s2)
-; RV32ZVE32F-NEXT: andi a0, a1, 2
+; RV32ZVE32F-NEXT: andi a0, t0, 2
; RV32ZVE32F-NEXT: beqz a0, .LBB42_2
; RV32ZVE32F-NEXT: .LBB42_11: # %cond.store1
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 1
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
-; RV32ZVE32F-NEXT: sw s0, 4(a0)
-; RV32ZVE32F-NEXT: sw t6, 0(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 4
+; RV32ZVE32F-NEXT: sw s1, 4(a0)
+; RV32ZVE32F-NEXT: sw s0, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, t0, 4
; RV32ZVE32F-NEXT: beqz a0, .LBB42_3
; RV32ZVE32F-NEXT: .LBB42_12: # %cond.store3
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 2
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
-; RV32ZVE32F-NEXT: sw t5, 0(a0)
-; RV32ZVE32F-NEXT: sw t4, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 8
+; RV32ZVE32F-NEXT: sw t6, 0(a0)
+; RV32ZVE32F-NEXT: sw t5, 4(a0)
+; RV32ZVE32F-NEXT: andi a0, t0, 8
; RV32ZVE32F-NEXT: beqz a0, .LBB42_4
; RV32ZVE32F-NEXT: .LBB42_13: # %cond.store5
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 3
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
-; RV32ZVE32F-NEXT: sw t3, 0(a0)
-; RV32ZVE32F-NEXT: sw t2, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 16
+; RV32ZVE32F-NEXT: sw t4, 0(a0)
+; RV32ZVE32F-NEXT: sw t3, 4(a0)
+; RV32ZVE32F-NEXT: andi a0, t0, 16
; RV32ZVE32F-NEXT: beqz a0, .LBB42_5
; RV32ZVE32F-NEXT: .LBB42_14: # %cond.store7
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 4
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
-; RV32ZVE32F-NEXT: sw t1, 0(a0)
-; RV32ZVE32F-NEXT: sw t0, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 32
+; RV32ZVE32F-NEXT: sw t2, 0(a0)
+; RV32ZVE32F-NEXT: sw t1, 4(a0)
+; RV32ZVE32F-NEXT: andi a0, t0, 32
; RV32ZVE32F-NEXT: beqz a0, .LBB42_6
; RV32ZVE32F-NEXT: .LBB42_15: # %cond.store9
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -3479,7 +3481,7 @@ define void @mscatter_baseidx_v8i8_v8i64(<8 x i64> %val, ptr %base, <8 x i8> %id
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
; RV32ZVE32F-NEXT: sw a7, 0(a0)
; RV32ZVE32F-NEXT: sw a6, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 64
+; RV32ZVE32F-NEXT: andi a0, t0, 64
; RV32ZVE32F-NEXT: beqz a0, .LBB42_7
; RV32ZVE32F-NEXT: .LBB42_16: # %cond.store11
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -3487,7 +3489,7 @@ define void @mscatter_baseidx_v8i8_v8i64(<8 x i64> %val, ptr %base, <8 x i8> %id
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
; RV32ZVE32F-NEXT: sw a5, 0(a0)
; RV32ZVE32F-NEXT: sw a4, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, -128
+; RV32ZVE32F-NEXT: andi a0, t0, -128
; RV32ZVE32F-NEXT: bnez a0, .LBB42_8
; RV32ZVE32F-NEXT: j .LBB42_9
;
@@ -3523,8 +3525,8 @@ define void @mscatter_baseidx_v8i8_v8i64(<8 x i64> %val, ptr %base, <8 x i8> %id
; RV64ZVE32F-NEXT: .LBB42_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a0, a4, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a0, .LBB42_12
; RV64ZVE32F-NEXT: # %bb.5: # %else4
@@ -3627,42 +3629,43 @@ define void @mscatter_baseidx_sext_v8i8_v8i64(<8 x i64> %val, ptr %base, <8 x i8
; RV32ZVE32F-NEXT: lw a5, 48(a0)
; RV32ZVE32F-NEXT: lw a6, 44(a0)
; RV32ZVE32F-NEXT: lw a7, 40(a0)
-; RV32ZVE32F-NEXT: lw t0, 36(a0)
-; RV32ZVE32F-NEXT: lw t1, 32(a0)
-; RV32ZVE32F-NEXT: lw t2, 28(a0)
-; RV32ZVE32F-NEXT: lw t3, 24(a0)
-; RV32ZVE32F-NEXT: lw t4, 20(a0)
-; RV32ZVE32F-NEXT: lw t5, 16(a0)
-; RV32ZVE32F-NEXT: lw s0, 12(a0)
-; RV32ZVE32F-NEXT: lw t6, 8(a0)
+; RV32ZVE32F-NEXT: lw t1, 36(a0)
+; RV32ZVE32F-NEXT: lw t2, 32(a0)
+; RV32ZVE32F-NEXT: lw t3, 28(a0)
+; RV32ZVE32F-NEXT: lw t4, 24(a0)
+; RV32ZVE32F-NEXT: lw t5, 20(a0)
+; RV32ZVE32F-NEXT: lw t6, 16(a0)
+; RV32ZVE32F-NEXT: lw s1, 12(a0)
+; RV32ZVE32F-NEXT: lw s0, 8(a0)
; RV32ZVE32F-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vsext.vf4 v10, v8
; RV32ZVE32F-NEXT: vsll.vi v8, v10, 3
-; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
; RV32ZVE32F-NEXT: vsetvli zero, zero, e8, mf2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a1, v0
-; RV32ZVE32F-NEXT: andi s1, a1, 1
-; RV32ZVE32F-NEXT: bnez s1, .LBB43_10
+; RV32ZVE32F-NEXT: vmv.x.s t0, v0
+; RV32ZVE32F-NEXT: andi s2, t0, 1
+; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
+; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
+; RV32ZVE32F-NEXT: bnez s2, .LBB43_10
; RV32ZVE32F-NEXT: # %bb.1: # %else
-; RV32ZVE32F-NEXT: andi a0, a1, 2
+; RV32ZVE32F-NEXT: andi a0, t0, 2
; RV32ZVE32F-NEXT: bnez a0, .LBB43_11
; RV32ZVE32F-NEXT: .LBB43_2: # %else2
-; RV32ZVE32F-NEXT: andi a0, a1, 4
+; RV32ZVE32F-NEXT: andi a0, t0, 4
; RV32ZVE32F-NEXT: bnez a0, .LBB43_12
; RV32ZVE32F-NEXT: .LBB43_3: # %else4
-; RV32ZVE32F-NEXT: andi a0, a1, 8
+; RV32ZVE32F-NEXT: andi a0, t0, 8
; RV32ZVE32F-NEXT: bnez a0, .LBB43_13
; RV32ZVE32F-NEXT: .LBB43_4: # %else6
-; RV32ZVE32F-NEXT: andi a0, a1, 16
+; RV32ZVE32F-NEXT: andi a0, t0, 16
; RV32ZVE32F-NEXT: bnez a0, .LBB43_14
; RV32ZVE32F-NEXT: .LBB43_5: # %else8
-; RV32ZVE32F-NEXT: andi a0, a1, 32
+; RV32ZVE32F-NEXT: andi a0, t0, 32
; RV32ZVE32F-NEXT: bnez a0, .LBB43_15
; RV32ZVE32F-NEXT: .LBB43_6: # %else10
-; RV32ZVE32F-NEXT: andi a0, a1, 64
+; RV32ZVE32F-NEXT: andi a0, t0, 64
; RV32ZVE32F-NEXT: bnez a0, .LBB43_16
; RV32ZVE32F-NEXT: .LBB43_7: # %else12
-; RV32ZVE32F-NEXT: andi a0, a1, -128
+; RV32ZVE32F-NEXT: andi a0, t0, -128
; RV32ZVE32F-NEXT: beqz a0, .LBB43_9
; RV32ZVE32F-NEXT: .LBB43_8: # %cond.store13
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -3677,45 +3680,44 @@ define void @mscatter_baseidx_sext_v8i8_v8i64(<8 x i64> %val, ptr %base, <8 x i8
; RV32ZVE32F-NEXT: addi sp, sp, 16
; RV32ZVE32F-NEXT: ret
; RV32ZVE32F-NEXT: .LBB43_10: # %cond.store
-; RV32ZVE32F-NEXT: lw s1, 4(a0)
+; RV32ZVE32F-NEXT: lw a1, 4(a0)
; RV32ZVE32F-NEXT: lw a0, 0(a0)
-; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vmv.x.s s2, v8
-; RV32ZVE32F-NEXT: sw s1, 4(s2)
+; RV32ZVE32F-NEXT: sw a1, 4(s2)
; RV32ZVE32F-NEXT: sw a0, 0(s2)
-; RV32ZVE32F-NEXT: andi a0, a1, 2
+; RV32ZVE32F-NEXT: andi a0, t0, 2
; RV32ZVE32F-NEXT: beqz a0, .LBB43_2
; RV32ZVE32F-NEXT: .LBB43_11: # %cond.store1
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 1
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
-; RV32ZVE32F-NEXT: sw s0, 4(a0)
-; RV32ZVE32F-NEXT: sw t6, 0(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 4
+; RV32ZVE32F-NEXT: sw s1, 4(a0)
+; RV32ZVE32F-NEXT: sw s0, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, t0, 4
; RV32ZVE32F-NEXT: beqz a0, .LBB43_3
; RV32ZVE32F-NEXT: .LBB43_12: # %cond.store3
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 2
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
-; RV32ZVE32F-NEXT: sw t5, 0(a0)
-; RV32ZVE32F-NEXT: sw t4, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 8
+; RV32ZVE32F-NEXT: sw t6, 0(a0)
+; RV32ZVE32F-NEXT: sw t5, 4(a0)
+; RV32ZVE32F-NEXT: andi a0, t0, 8
; RV32ZVE32F-NEXT: beqz a0, .LBB43_4
; RV32ZVE32F-NEXT: .LBB43_13: # %cond.store5
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 3
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
-; RV32ZVE32F-NEXT: sw t3, 0(a0)
-; RV32ZVE32F-NEXT: sw t2, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 16
+; RV32ZVE32F-NEXT: sw t4, 0(a0)
+; RV32ZVE32F-NEXT: sw t3, 4(a0)
+; RV32ZVE32F-NEXT: andi a0, t0, 16
; RV32ZVE32F-NEXT: beqz a0, .LBB43_5
; RV32ZVE32F-NEXT: .LBB43_14: # %cond.store7
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 4
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
-; RV32ZVE32F-NEXT: sw t1, 0(a0)
-; RV32ZVE32F-NEXT: sw t0, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 32
+; RV32ZVE32F-NEXT: sw t2, 0(a0)
+; RV32ZVE32F-NEXT: sw t1, 4(a0)
+; RV32ZVE32F-NEXT: andi a0, t0, 32
; RV32ZVE32F-NEXT: beqz a0, .LBB43_6
; RV32ZVE32F-NEXT: .LBB43_15: # %cond.store9
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -3723,7 +3725,7 @@ define void @mscatter_baseidx_sext_v8i8_v8i64(<8 x i64> %val, ptr %base, <8 x i8
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
; RV32ZVE32F-NEXT: sw a7, 0(a0)
; RV32ZVE32F-NEXT: sw a6, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 64
+; RV32ZVE32F-NEXT: andi a0, t0, 64
; RV32ZVE32F-NEXT: beqz a0, .LBB43_7
; RV32ZVE32F-NEXT: .LBB43_16: # %cond.store11
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -3731,7 +3733,7 @@ define void @mscatter_baseidx_sext_v8i8_v8i64(<8 x i64> %val, ptr %base, <8 x i8
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
; RV32ZVE32F-NEXT: sw a5, 0(a0)
; RV32ZVE32F-NEXT: sw a4, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, -128
+; RV32ZVE32F-NEXT: andi a0, t0, -128
; RV32ZVE32F-NEXT: bnez a0, .LBB43_8
; RV32ZVE32F-NEXT: j .LBB43_9
;
@@ -3767,8 +3769,8 @@ define void @mscatter_baseidx_sext_v8i8_v8i64(<8 x i64> %val, ptr %base, <8 x i8
; RV64ZVE32F-NEXT: .LBB43_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a0, a4, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a0, .LBB43_12
; RV64ZVE32F-NEXT: # %bb.5: # %else4
@@ -3873,42 +3875,43 @@ define void @mscatter_baseidx_zext_v8i8_v8i64(<8 x i64> %val, ptr %base, <8 x i8
; RV32ZVE32F-NEXT: lw a5, 48(a0)
; RV32ZVE32F-NEXT: lw a6, 44(a0)
; RV32ZVE32F-NEXT: lw a7, 40(a0)
-; RV32ZVE32F-NEXT: lw t0, 36(a0)
-; RV32ZVE32F-NEXT: lw t1, 32(a0)
-; RV32ZVE32F-NEXT: lw t2, 28(a0)
-; RV32ZVE32F-NEXT: lw t3, 24(a0)
-; RV32ZVE32F-NEXT: lw t4, 20(a0)
-; RV32ZVE32F-NEXT: lw t5, 16(a0)
-; RV32ZVE32F-NEXT: lw s0, 12(a0)
-; RV32ZVE32F-NEXT: lw t6, 8(a0)
+; RV32ZVE32F-NEXT: lw t1, 36(a0)
+; RV32ZVE32F-NEXT: lw t2, 32(a0)
+; RV32ZVE32F-NEXT: lw t3, 28(a0)
+; RV32ZVE32F-NEXT: lw t4, 24(a0)
+; RV32ZVE32F-NEXT: lw t5, 20(a0)
+; RV32ZVE32F-NEXT: lw t6, 16(a0)
+; RV32ZVE32F-NEXT: lw s1, 12(a0)
+; RV32ZVE32F-NEXT: lw s0, 8(a0)
; RV32ZVE32F-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vzext.vf4 v10, v8
; RV32ZVE32F-NEXT: vsll.vi v8, v10, 3
-; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
; RV32ZVE32F-NEXT: vsetvli zero, zero, e8, mf2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a1, v0
-; RV32ZVE32F-NEXT: andi s1, a1, 1
-; RV32ZVE32F-NEXT: bnez s1, .LBB44_10
+; RV32ZVE32F-NEXT: vmv.x.s t0, v0
+; RV32ZVE32F-NEXT: andi s2, t0, 1
+; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
+; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
+; RV32ZVE32F-NEXT: bnez s2, .LBB44_10
; RV32ZVE32F-NEXT: # %bb.1: # %else
-; RV32ZVE32F-NEXT: andi a0, a1, 2
+; RV32ZVE32F-NEXT: andi a0, t0, 2
; RV32ZVE32F-NEXT: bnez a0, .LBB44_11
; RV32ZVE32F-NEXT: .LBB44_2: # %else2
-; RV32ZVE32F-NEXT: andi a0, a1, 4
+; RV32ZVE32F-NEXT: andi a0, t0, 4
; RV32ZVE32F-NEXT: bnez a0, .LBB44_12
; RV32ZVE32F-NEXT: .LBB44_3: # %else4
-; RV32ZVE32F-NEXT: andi a0, a1, 8
+; RV32ZVE32F-NEXT: andi a0, t0, 8
; RV32ZVE32F-NEXT: bnez a0, .LBB44_13
; RV32ZVE32F-NEXT: .LBB44_4: # %else6
-; RV32ZVE32F-NEXT: andi a0, a1, 16
+; RV32ZVE32F-NEXT: andi a0, t0, 16
; RV32ZVE32F-NEXT: bnez a0, .LBB44_14
; RV32ZVE32F-NEXT: .LBB44_5: # %else8
-; RV32ZVE32F-NEXT: andi a0, a1, 32
+; RV32ZVE32F-NEXT: andi a0, t0, 32
; RV32ZVE32F-NEXT: bnez a0, .LBB44_15
; RV32ZVE32F-NEXT: .LBB44_6: # %else10
-; RV32ZVE32F-NEXT: andi a0, a1, 64
+; RV32ZVE32F-NEXT: andi a0, t0, 64
; RV32ZVE32F-NEXT: bnez a0, .LBB44_16
; RV32ZVE32F-NEXT: .LBB44_7: # %else12
-; RV32ZVE32F-NEXT: andi a0, a1, -128
+; RV32ZVE32F-NEXT: andi a0, t0, -128
; RV32ZVE32F-NEXT: beqz a0, .LBB44_9
; RV32ZVE32F-NEXT: .LBB44_8: # %cond.store13
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -3923,45 +3926,44 @@ define void @mscatter_baseidx_zext_v8i8_v8i64(<8 x i64> %val, ptr %base, <8 x i8
; RV32ZVE32F-NEXT: addi sp, sp, 16
; RV32ZVE32F-NEXT: ret
; RV32ZVE32F-NEXT: .LBB44_10: # %cond.store
-; RV32ZVE32F-NEXT: lw s1, 4(a0)
+; RV32ZVE32F-NEXT: lw a1, 4(a0)
; RV32ZVE32F-NEXT: lw a0, 0(a0)
-; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vmv.x.s s2, v8
-; RV32ZVE32F-NEXT: sw s1, 4(s2)
+; RV32ZVE32F-NEXT: sw a1, 4(s2)
; RV32ZVE32F-NEXT: sw a0, 0(s2)
-; RV32ZVE32F-NEXT: andi a0, a1, 2
+; RV32ZVE32F-NEXT: andi a0, t0, 2
; RV32ZVE32F-NEXT: beqz a0, .LBB44_2
; RV32ZVE32F-NEXT: .LBB44_11: # %cond.store1
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 1
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
-; RV32ZVE32F-NEXT: sw s0, 4(a0)
-; RV32ZVE32F-NEXT: sw t6, 0(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 4
+; RV32ZVE32F-NEXT: sw s1, 4(a0)
+; RV32ZVE32F-NEXT: sw s0, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, t0, 4
; RV32ZVE32F-NEXT: beqz a0, .LBB44_3
; RV32ZVE32F-NEXT: .LBB44_12: # %cond.store3
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 2
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
-; RV32ZVE32F-NEXT: sw t5, 0(a0)
-; RV32ZVE32F-NEXT: sw t4, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 8
+; RV32ZVE32F-NEXT: sw t6, 0(a0)
+; RV32ZVE32F-NEXT: sw t5, 4(a0)
+; RV32ZVE32F-NEXT: andi a0, t0, 8
; RV32ZVE32F-NEXT: beqz a0, .LBB44_4
; RV32ZVE32F-NEXT: .LBB44_13: # %cond.store5
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 3
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
-; RV32ZVE32F-NEXT: sw t3, 0(a0)
-; RV32ZVE32F-NEXT: sw t2, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 16
-; RV32ZVE32F-NEXT: beqz a0, .LBB44_5
+; RV32ZVE32F-NEXT: sw t4, 0(a0)
+; RV32ZVE32F-NEXT: sw t3, 4(a0)
+; RV32ZVE32F-NEXT: andi a0, t0, 16
+; RV32ZVE32F-NEXT: beqz a0, .LBB44_5
; RV32ZVE32F-NEXT: .LBB44_14: # %cond.store7
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 4
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
-; RV32ZVE32F-NEXT: sw t1, 0(a0)
-; RV32ZVE32F-NEXT: sw t0, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 32
+; RV32ZVE32F-NEXT: sw t2, 0(a0)
+; RV32ZVE32F-NEXT: sw t1, 4(a0)
+; RV32ZVE32F-NEXT: andi a0, t0, 32
; RV32ZVE32F-NEXT: beqz a0, .LBB44_6
; RV32ZVE32F-NEXT: .LBB44_15: # %cond.store9
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -3969,7 +3971,7 @@ define void @mscatter_baseidx_zext_v8i8_v8i64(<8 x i64> %val, ptr %base, <8 x i8
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
; RV32ZVE32F-NEXT: sw a7, 0(a0)
; RV32ZVE32F-NEXT: sw a6, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 64
+; RV32ZVE32F-NEXT: andi a0, t0, 64
; RV32ZVE32F-NEXT: beqz a0, .LBB44_7
; RV32ZVE32F-NEXT: .LBB44_16: # %cond.store11
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -3977,7 +3979,7 @@ define void @mscatter_baseidx_zext_v8i8_v8i64(<8 x i64> %val, ptr %base, <8 x i8
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
; RV32ZVE32F-NEXT: sw a5, 0(a0)
; RV32ZVE32F-NEXT: sw a4, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, -128
+; RV32ZVE32F-NEXT: andi a0, t0, -128
; RV32ZVE32F-NEXT: bnez a0, .LBB44_8
; RV32ZVE32F-NEXT: j .LBB44_9
;
@@ -4015,8 +4017,8 @@ define void @mscatter_baseidx_zext_v8i8_v8i64(<8 x i64> %val, ptr %base, <8 x i8
; RV64ZVE32F-NEXT: .LBB44_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a0, a4, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a0, .LBB44_12
; RV64ZVE32F-NEXT: # %bb.5: # %else4
@@ -4126,42 +4128,43 @@ define void @mscatter_baseidx_v8i16_v8i64(<8 x i64> %val, ptr %base, <8 x i16> %
; RV32ZVE32F-NEXT: lw a5, 48(a0)
; RV32ZVE32F-NEXT: lw a6, 44(a0)
; RV32ZVE32F-NEXT: lw a7, 40(a0)
-; RV32ZVE32F-NEXT: lw t0, 36(a0)
-; RV32ZVE32F-NEXT: lw t1, 32(a0)
-; RV32ZVE32F-NEXT: lw t2, 28(a0)
-; RV32ZVE32F-NEXT: lw t3, 24(a0)
-; RV32ZVE32F-NEXT: lw t4, 20(a0)
-; RV32ZVE32F-NEXT: lw t5, 16(a0)
-; RV32ZVE32F-NEXT: lw s0, 12(a0)
-; RV32ZVE32F-NEXT: lw t6, 8(a0)
+; RV32ZVE32F-NEXT: lw t1, 36(a0)
+; RV32ZVE32F-NEXT: lw t2, 32(a0)
+; RV32ZVE32F-NEXT: lw t3, 28(a0)
+; RV32ZVE32F-NEXT: lw t4, 24(a0)
+; RV32ZVE32F-NEXT: lw t5, 20(a0)
+; RV32ZVE32F-NEXT: lw t6, 16(a0)
+; RV32ZVE32F-NEXT: lw s1, 12(a0)
+; RV32ZVE32F-NEXT: lw s0, 8(a0)
; RV32ZVE32F-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vsext.vf2 v10, v8
; RV32ZVE32F-NEXT: vsll.vi v8, v10, 3
-; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
; RV32ZVE32F-NEXT: vsetvli zero, zero, e8, mf2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a1, v0
-; RV32ZVE32F-NEXT: andi s1, a1, 1
-; RV32ZVE32F-NEXT: bnez s1, .LBB45_10
+; RV32ZVE32F-NEXT: vmv.x.s t0, v0
+; RV32ZVE32F-NEXT: andi s2, t0, 1
+; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
+; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
+; RV32ZVE32F-NEXT: bnez s2, .LBB45_10
; RV32ZVE32F-NEXT: # %bb.1: # %else
-; RV32ZVE32F-NEXT: andi a0, a1, 2
+; RV32ZVE32F-NEXT: andi a0, t0, 2
; RV32ZVE32F-NEXT: bnez a0, .LBB45_11
; RV32ZVE32F-NEXT: .LBB45_2: # %else2
-; RV32ZVE32F-NEXT: andi a0, a1, 4
+; RV32ZVE32F-NEXT: andi a0, t0, 4
; RV32ZVE32F-NEXT: bnez a0, .LBB45_12
; RV32ZVE32F-NEXT: .LBB45_3: # %else4
-; RV32ZVE32F-NEXT: andi a0, a1, 8
+; RV32ZVE32F-NEXT: andi a0, t0, 8
; RV32ZVE32F-NEXT: bnez a0, .LBB45_13
; RV32ZVE32F-NEXT: .LBB45_4: # %else6
-; RV32ZVE32F-NEXT: andi a0, a1, 16
+; RV32ZVE32F-NEXT: andi a0, t0, 16
; RV32ZVE32F-NEXT: bnez a0, .LBB45_14
; RV32ZVE32F-NEXT: .LBB45_5: # %else8
-; RV32ZVE32F-NEXT: andi a0, a1, 32
+; RV32ZVE32F-NEXT: andi a0, t0, 32
; RV32ZVE32F-NEXT: bnez a0, .LBB45_15
; RV32ZVE32F-NEXT: .LBB45_6: # %else10
-; RV32ZVE32F-NEXT: andi a0, a1, 64
+; RV32ZVE32F-NEXT: andi a0, t0, 64
; RV32ZVE32F-NEXT: bnez a0, .LBB45_16
; RV32ZVE32F-NEXT: .LBB45_7: # %else12
-; RV32ZVE32F-NEXT: andi a0, a1, -128
+; RV32ZVE32F-NEXT: andi a0, t0, -128
; RV32ZVE32F-NEXT: beqz a0, .LBB45_9
; RV32ZVE32F-NEXT: .LBB45_8: # %cond.store13
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -4176,45 +4179,44 @@ define void @mscatter_baseidx_v8i16_v8i64(<8 x i64> %val, ptr %base, <8 x i16> %
; RV32ZVE32F-NEXT: addi sp, sp, 16
; RV32ZVE32F-NEXT: ret
; RV32ZVE32F-NEXT: .LBB45_10: # %cond.store
-; RV32ZVE32F-NEXT: lw s1, 4(a0)
+; RV32ZVE32F-NEXT: lw a1, 4(a0)
; RV32ZVE32F-NEXT: lw a0, 0(a0)
-; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vmv.x.s s2, v8
-; RV32ZVE32F-NEXT: sw s1, 4(s2)
+; RV32ZVE32F-NEXT: sw a1, 4(s2)
; RV32ZVE32F-NEXT: sw a0, 0(s2)
-; RV32ZVE32F-NEXT: andi a0, a1, 2
+; RV32ZVE32F-NEXT: andi a0, t0, 2
; RV32ZVE32F-NEXT: beqz a0, .LBB45_2
; RV32ZVE32F-NEXT: .LBB45_11: # %cond.store1
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 1
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
-; RV32ZVE32F-NEXT: sw s0, 4(a0)
-; RV32ZVE32F-NEXT: sw t6, 0(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 4
+; RV32ZVE32F-NEXT: sw s1, 4(a0)
+; RV32ZVE32F-NEXT: sw s0, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, t0, 4
; RV32ZVE32F-NEXT: beqz a0, .LBB45_3
; RV32ZVE32F-NEXT: .LBB45_12: # %cond.store3
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 2
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
-; RV32ZVE32F-NEXT: sw t5, 0(a0)
-; RV32ZVE32F-NEXT: sw t4, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 8
+; RV32ZVE32F-NEXT: sw t6, 0(a0)
+; RV32ZVE32F-NEXT: sw t5, 4(a0)
+; RV32ZVE32F-NEXT: andi a0, t0, 8
; RV32ZVE32F-NEXT: beqz a0, .LBB45_4
; RV32ZVE32F-NEXT: .LBB45_13: # %cond.store5
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 3
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
-; RV32ZVE32F-NEXT: sw t3, 0(a0)
-; RV32ZVE32F-NEXT: sw t2, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 16
+; RV32ZVE32F-NEXT: sw t4, 0(a0)
+; RV32ZVE32F-NEXT: sw t3, 4(a0)
+; RV32ZVE32F-NEXT: andi a0, t0, 16
; RV32ZVE32F-NEXT: beqz a0, .LBB45_5
; RV32ZVE32F-NEXT: .LBB45_14: # %cond.store7
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 4
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
-; RV32ZVE32F-NEXT: sw t1, 0(a0)
-; RV32ZVE32F-NEXT: sw t0, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 32
+; RV32ZVE32F-NEXT: sw t2, 0(a0)
+; RV32ZVE32F-NEXT: sw t1, 4(a0)
+; RV32ZVE32F-NEXT: andi a0, t0, 32
; RV32ZVE32F-NEXT: beqz a0, .LBB45_6
; RV32ZVE32F-NEXT: .LBB45_15: # %cond.store9
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -4222,7 +4224,7 @@ define void @mscatter_baseidx_v8i16_v8i64(<8 x i64> %val, ptr %base, <8 x i16> %
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
; RV32ZVE32F-NEXT: sw a7, 0(a0)
; RV32ZVE32F-NEXT: sw a6, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 64
+; RV32ZVE32F-NEXT: andi a0, t0, 64
; RV32ZVE32F-NEXT: beqz a0, .LBB45_7
; RV32ZVE32F-NEXT: .LBB45_16: # %cond.store11
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -4230,7 +4232,7 @@ define void @mscatter_baseidx_v8i16_v8i64(<8 x i64> %val, ptr %base, <8 x i16> %
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
; RV32ZVE32F-NEXT: sw a5, 0(a0)
; RV32ZVE32F-NEXT: sw a4, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, -128
+; RV32ZVE32F-NEXT: andi a0, t0, -128
; RV32ZVE32F-NEXT: bnez a0, .LBB45_8
; RV32ZVE32F-NEXT: j .LBB45_9
;
@@ -4267,8 +4269,8 @@ define void @mscatter_baseidx_v8i16_v8i64(<8 x i64> %val, ptr %base, <8 x i16> %
; RV64ZVE32F-NEXT: .LBB45_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e16, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: andi a0, a4, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a0, .LBB45_12
; RV64ZVE32F-NEXT: # %bb.5: # %else4
@@ -4371,42 +4373,43 @@ define void @mscatter_baseidx_sext_v8i16_v8i64(<8 x i64> %val, ptr %base, <8 x i
; RV32ZVE32F-NEXT: lw a5, 48(a0)
; RV32ZVE32F-NEXT: lw a6, 44(a0)
; RV32ZVE32F-NEXT: lw a7, 40(a0)
-; RV32ZVE32F-NEXT: lw t0, 36(a0)
-; RV32ZVE32F-NEXT: lw t1, 32(a0)
-; RV32ZVE32F-NEXT: lw t2, 28(a0)
-; RV32ZVE32F-NEXT: lw t3, 24(a0)
-; RV32ZVE32F-NEXT: lw t4, 20(a0)
-; RV32ZVE32F-NEXT: lw t5, 16(a0)
-; RV32ZVE32F-NEXT: lw s0, 12(a0)
-; RV32ZVE32F-NEXT: lw t6, 8(a0)
+; RV32ZVE32F-NEXT: lw t1, 36(a0)
+; RV32ZVE32F-NEXT: lw t2, 32(a0)
+; RV32ZVE32F-NEXT: lw t3, 28(a0)
+; RV32ZVE32F-NEXT: lw t4, 24(a0)
+; RV32ZVE32F-NEXT: lw t5, 20(a0)
+; RV32ZVE32F-NEXT: lw t6, 16(a0)
+; RV32ZVE32F-NEXT: lw s1, 12(a0)
+; RV32ZVE32F-NEXT: lw s0, 8(a0)
; RV32ZVE32F-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vsext.vf2 v10, v8
; RV32ZVE32F-NEXT: vsll.vi v8, v10, 3
-; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
; RV32ZVE32F-NEXT: vsetvli zero, zero, e8, mf2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a1, v0
-; RV32ZVE32F-NEXT: andi s1, a1, 1
-; RV32ZVE32F-NEXT: bnez s1, .LBB46_10
+; RV32ZVE32F-NEXT: vmv.x.s t0, v0
+; RV32ZVE32F-NEXT: andi s2, t0, 1
+; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
+; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
+; RV32ZVE32F-NEXT: bnez s2, .LBB46_10
; RV32ZVE32F-NEXT: # %bb.1: # %else
-; RV32ZVE32F-NEXT: andi a0, a1, 2
+; RV32ZVE32F-NEXT: andi a0, t0, 2
; RV32ZVE32F-NEXT: bnez a0, .LBB46_11
; RV32ZVE32F-NEXT: .LBB46_2: # %else2
-; RV32ZVE32F-NEXT: andi a0, a1, 4
+; RV32ZVE32F-NEXT: andi a0, t0, 4
; RV32ZVE32F-NEXT: bnez a0, .LBB46_12
; RV32ZVE32F-NEXT: .LBB46_3: # %else4
-; RV32ZVE32F-NEXT: andi a0, a1, 8
+; RV32ZVE32F-NEXT: andi a0, t0, 8
; RV32ZVE32F-NEXT: bnez a0, .LBB46_13
; RV32ZVE32F-NEXT: .LBB46_4: # %else6
-; RV32ZVE32F-NEXT: andi a0, a1, 16
+; RV32ZVE32F-NEXT: andi a0, t0, 16
; RV32ZVE32F-NEXT: bnez a0, .LBB46_14
; RV32ZVE32F-NEXT: .LBB46_5: # %else8
-; RV32ZVE32F-NEXT: andi a0, a1, 32
+; RV32ZVE32F-NEXT: andi a0, t0, 32
; RV32ZVE32F-NEXT: bnez a0, .LBB46_15
; RV32ZVE32F-NEXT: .LBB46_6: # %else10
-; RV32ZVE32F-NEXT: andi a0, a1, 64
+; RV32ZVE32F-NEXT: andi a0, t0, 64
; RV32ZVE32F-NEXT: bnez a0, .LBB46_16
; RV32ZVE32F-NEXT: .LBB46_7: # %else12
-; RV32ZVE32F-NEXT: andi a0, a1, -128
+; RV32ZVE32F-NEXT: andi a0, t0, -128
; RV32ZVE32F-NEXT: beqz a0, .LBB46_9
; RV32ZVE32F-NEXT: .LBB46_8: # %cond.store13
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -4421,45 +4424,44 @@ define void @mscatter_baseidx_sext_v8i16_v8i64(<8 x i64> %val, ptr %base, <8 x i
; RV32ZVE32F-NEXT: addi sp, sp, 16
; RV32ZVE32F-NEXT: ret
; RV32ZVE32F-NEXT: .LBB46_10: # %cond.store
-; RV32ZVE32F-NEXT: lw s1, 4(a0)
+; RV32ZVE32F-NEXT: lw a1, 4(a0)
; RV32ZVE32F-NEXT: lw a0, 0(a0)
-; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vmv.x.s s2, v8
-; RV32ZVE32F-NEXT: sw s1, 4(s2)
+; RV32ZVE32F-NEXT: sw a1, 4(s2)
; RV32ZVE32F-NEXT: sw a0, 0(s2)
-; RV32ZVE32F-NEXT: andi a0, a1, 2
+; RV32ZVE32F-NEXT: andi a0, t0, 2
; RV32ZVE32F-NEXT: beqz a0, .LBB46_2
; RV32ZVE32F-NEXT: .LBB46_11: # %cond.store1
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 1
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
-; RV32ZVE32F-NEXT: sw s0, 4(a0)
-; RV32ZVE32F-NEXT: sw t6, 0(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 4
+; RV32ZVE32F-NEXT: sw s1, 4(a0)
+; RV32ZVE32F-NEXT: sw s0, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, t0, 4
; RV32ZVE32F-NEXT: beqz a0, .LBB46_3
; RV32ZVE32F-NEXT: .LBB46_12: # %cond.store3
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 2
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
-; RV32ZVE32F-NEXT: sw t5, 0(a0)
-; RV32ZVE32F-NEXT: sw t4, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 8
+; RV32ZVE32F-NEXT: sw t6, 0(a0)
+; RV32ZVE32F-NEXT: sw t5, 4(a0)
+; RV32ZVE32F-NEXT: andi a0, t0, 8
; RV32ZVE32F-NEXT: beqz a0, .LBB46_4
; RV32ZVE32F-NEXT: .LBB46_13: # %cond.store5
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 3
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
-; RV32ZVE32F-NEXT: sw t3, 0(a0)
-; RV32ZVE32F-NEXT: sw t2, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 16
+; RV32ZVE32F-NEXT: sw t4, 0(a0)
+; RV32ZVE32F-NEXT: sw t3, 4(a0)
+; RV32ZVE32F-NEXT: andi a0, t0, 16
; RV32ZVE32F-NEXT: beqz a0, .LBB46_5
; RV32ZVE32F-NEXT: .LBB46_14: # %cond.store7
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 4
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
-; RV32ZVE32F-NEXT: sw t1, 0(a0)
-; RV32ZVE32F-NEXT: sw t0, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 32
+; RV32ZVE32F-NEXT: sw t2, 0(a0)
+; RV32ZVE32F-NEXT: sw t1, 4(a0)
+; RV32ZVE32F-NEXT: andi a0, t0, 32
; RV32ZVE32F-NEXT: beqz a0, .LBB46_6
; RV32ZVE32F-NEXT: .LBB46_15: # %cond.store9
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -4467,7 +4469,7 @@ define void @mscatter_baseidx_sext_v8i16_v8i64(<8 x i64> %val, ptr %base, <8 x i
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
; RV32ZVE32F-NEXT: sw a7, 0(a0)
; RV32ZVE32F-NEXT: sw a6, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 64
+; RV32ZVE32F-NEXT: andi a0, t0, 64
; RV32ZVE32F-NEXT: beqz a0, .LBB46_7
; RV32ZVE32F-NEXT: .LBB46_16: # %cond.store11
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -4475,7 +4477,7 @@ define void @mscatter_baseidx_sext_v8i16_v8i64(<8 x i64> %val, ptr %base, <8 x i
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
; RV32ZVE32F-NEXT: sw a5, 0(a0)
; RV32ZVE32F-NEXT: sw a4, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, -128
+; RV32ZVE32F-NEXT: andi a0, t0, -128
; RV32ZVE32F-NEXT: bnez a0, .LBB46_8
; RV32ZVE32F-NEXT: j .LBB46_9
;
@@ -4512,8 +4514,8 @@ define void @mscatter_baseidx_sext_v8i16_v8i64(<8 x i64> %val, ptr %base, <8 x i
; RV64ZVE32F-NEXT: .LBB46_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e16, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: andi a0, a4, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a0, .LBB46_12
; RV64ZVE32F-NEXT: # %bb.5: # %else4
@@ -4618,42 +4620,43 @@ define void @mscatter_baseidx_zext_v8i16_v8i64(<8 x i64> %val, ptr %base, <8 x i
; RV32ZVE32F-NEXT: lw a5, 48(a0)
; RV32ZVE32F-NEXT: lw a6, 44(a0)
; RV32ZVE32F-NEXT: lw a7, 40(a0)
-; RV32ZVE32F-NEXT: lw t0, 36(a0)
-; RV32ZVE32F-NEXT: lw t1, 32(a0)
-; RV32ZVE32F-NEXT: lw t2, 28(a0)
-; RV32ZVE32F-NEXT: lw t3, 24(a0)
-; RV32ZVE32F-NEXT: lw t4, 20(a0)
-; RV32ZVE32F-NEXT: lw t5, 16(a0)
-; RV32ZVE32F-NEXT: lw s0, 12(a0)
-; RV32ZVE32F-NEXT: lw t6, 8(a0)
+; RV32ZVE32F-NEXT: lw t1, 36(a0)
+; RV32ZVE32F-NEXT: lw t2, 32(a0)
+; RV32ZVE32F-NEXT: lw t3, 28(a0)
+; RV32ZVE32F-NEXT: lw t4, 24(a0)
+; RV32ZVE32F-NEXT: lw t5, 20(a0)
+; RV32ZVE32F-NEXT: lw t6, 16(a0)
+; RV32ZVE32F-NEXT: lw s1, 12(a0)
+; RV32ZVE32F-NEXT: lw s0, 8(a0)
; RV32ZVE32F-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vzext.vf2 v10, v8
; RV32ZVE32F-NEXT: vsll.vi v8, v10, 3
-; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
; RV32ZVE32F-NEXT: vsetvli zero, zero, e8, mf2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a1, v0
-; RV32ZVE32F-NEXT: andi s1, a1, 1
-; RV32ZVE32F-NEXT: bnez s1, .LBB47_10
+; RV32ZVE32F-NEXT: vmv.x.s t0, v0
+; RV32ZVE32F-NEXT: andi s2, t0, 1
+; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
+; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
+; RV32ZVE32F-NEXT: bnez s2, .LBB47_10
; RV32ZVE32F-NEXT: # %bb.1: # %else
-; RV32ZVE32F-NEXT: andi a0, a1, 2
+; RV32ZVE32F-NEXT: andi a0, t0, 2
; RV32ZVE32F-NEXT: bnez a0, .LBB47_11
; RV32ZVE32F-NEXT: .LBB47_2: # %else2
-; RV32ZVE32F-NEXT: andi a0, a1, 4
+; RV32ZVE32F-NEXT: andi a0, t0, 4
; RV32ZVE32F-NEXT: bnez a0, .LBB47_12
; RV32ZVE32F-NEXT: .LBB47_3: # %else4
-; RV32ZVE32F-NEXT: andi a0, a1, 8
+; RV32ZVE32F-NEXT: andi a0, t0, 8
; RV32ZVE32F-NEXT: bnez a0, .LBB47_13
; RV32ZVE32F-NEXT: .LBB47_4: # %else6
-; RV32ZVE32F-NEXT: andi a0, a1, 16
+; RV32ZVE32F-NEXT: andi a0, t0, 16
; RV32ZVE32F-NEXT: bnez a0, .LBB47_14
; RV32ZVE32F-NEXT: .LBB47_5: # %else8
-; RV32ZVE32F-NEXT: andi a0, a1, 32
+; RV32ZVE32F-NEXT: andi a0, t0, 32
; RV32ZVE32F-NEXT: bnez a0, .LBB47_15
; RV32ZVE32F-NEXT: .LBB47_6: # %else10
-; RV32ZVE32F-NEXT: andi a0, a1, 64
+; RV32ZVE32F-NEXT: andi a0, t0, 64
; RV32ZVE32F-NEXT: bnez a0, .LBB47_16
; RV32ZVE32F-NEXT: .LBB47_7: # %else12
-; RV32ZVE32F-NEXT: andi a0, a1, -128
+; RV32ZVE32F-NEXT: andi a0, t0, -128
; RV32ZVE32F-NEXT: beqz a0, .LBB47_9
; RV32ZVE32F-NEXT: .LBB47_8: # %cond.store13
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -4668,45 +4671,44 @@ define void @mscatter_baseidx_zext_v8i16_v8i64(<8 x i64> %val, ptr %base, <8 x i
; RV32ZVE32F-NEXT: addi sp, sp, 16
; RV32ZVE32F-NEXT: ret
; RV32ZVE32F-NEXT: .LBB47_10: # %cond.store
-; RV32ZVE32F-NEXT: lw s1, 4(a0)
+; RV32ZVE32F-NEXT: lw a1, 4(a0)
; RV32ZVE32F-NEXT: lw a0, 0(a0)
-; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vmv.x.s s2, v8
-; RV32ZVE32F-NEXT: sw s1, 4(s2)
+; RV32ZVE32F-NEXT: sw a1, 4(s2)
; RV32ZVE32F-NEXT: sw a0, 0(s2)
-; RV32ZVE32F-NEXT: andi a0, a1, 2
+; RV32ZVE32F-NEXT: andi a0, t0, 2
; RV32ZVE32F-NEXT: beqz a0, .LBB47_2
; RV32ZVE32F-NEXT: .LBB47_11: # %cond.store1
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 1
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
-; RV32ZVE32F-NEXT: sw s0, 4(a0)
-; RV32ZVE32F-NEXT: sw t6, 0(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 4
+; RV32ZVE32F-NEXT: sw s1, 4(a0)
+; RV32ZVE32F-NEXT: sw s0, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, t0, 4
; RV32ZVE32F-NEXT: beqz a0, .LBB47_3
; RV32ZVE32F-NEXT: .LBB47_12: # %cond.store3
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 2
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
-; RV32ZVE32F-NEXT: sw t5, 0(a0)
-; RV32ZVE32F-NEXT: sw t4, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 8
+; RV32ZVE32F-NEXT: sw t6, 0(a0)
+; RV32ZVE32F-NEXT: sw t5, 4(a0)
+; RV32ZVE32F-NEXT: andi a0, t0, 8
; RV32ZVE32F-NEXT: beqz a0, .LBB47_4
; RV32ZVE32F-NEXT: .LBB47_13: # %cond.store5
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 3
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
-; RV32ZVE32F-NEXT: sw t3, 0(a0)
-; RV32ZVE32F-NEXT: sw t2, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 16
+; RV32ZVE32F-NEXT: sw t4, 0(a0)
+; RV32ZVE32F-NEXT: sw t3, 4(a0)
+; RV32ZVE32F-NEXT: andi a0, t0, 16
; RV32ZVE32F-NEXT: beqz a0, .LBB47_5
; RV32ZVE32F-NEXT: .LBB47_14: # %cond.store7
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 4
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
-; RV32ZVE32F-NEXT: sw t1, 0(a0)
-; RV32ZVE32F-NEXT: sw t0, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 32
+; RV32ZVE32F-NEXT: sw t2, 0(a0)
+; RV32ZVE32F-NEXT: sw t1, 4(a0)
+; RV32ZVE32F-NEXT: andi a0, t0, 32
; RV32ZVE32F-NEXT: beqz a0, .LBB47_6
; RV32ZVE32F-NEXT: .LBB47_15: # %cond.store9
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -4714,7 +4716,7 @@ define void @mscatter_baseidx_zext_v8i16_v8i64(<8 x i64> %val, ptr %base, <8 x i
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
; RV32ZVE32F-NEXT: sw a7, 0(a0)
; RV32ZVE32F-NEXT: sw a6, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 64
+; RV32ZVE32F-NEXT: andi a0, t0, 64
; RV32ZVE32F-NEXT: beqz a0, .LBB47_7
; RV32ZVE32F-NEXT: .LBB47_16: # %cond.store11
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -4722,7 +4724,7 @@ define void @mscatter_baseidx_zext_v8i16_v8i64(<8 x i64> %val, ptr %base, <8 x i
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
; RV32ZVE32F-NEXT: sw a5, 0(a0)
; RV32ZVE32F-NEXT: sw a4, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, -128
+; RV32ZVE32F-NEXT: andi a0, t0, -128
; RV32ZVE32F-NEXT: bnez a0, .LBB47_8
; RV32ZVE32F-NEXT: j .LBB47_9
;
@@ -4763,8 +4765,8 @@ define void @mscatter_baseidx_zext_v8i16_v8i64(<8 x i64> %val, ptr %base, <8 x i
; RV64ZVE32F-NEXT: .LBB47_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e16, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: andi a0, a5, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a0, .LBB47_12
; RV64ZVE32F-NEXT: # %bb.5: # %else4
@@ -4872,42 +4874,43 @@ define void @mscatter_baseidx_v8i32_v8i64(<8 x i64> %val, ptr %base, <8 x i32> %
; RV32ZVE32F-NEXT: lw a4, 52(a0)
; RV32ZVE32F-NEXT: lw a5, 48(a0)
; RV32ZVE32F-NEXT: lw a6, 44(a0)
-; RV32ZVE32F-NEXT: lw a7, 40(a0)
-; RV32ZVE32F-NEXT: lw t0, 36(a0)
-; RV32ZVE32F-NEXT: lw t1, 32(a0)
-; RV32ZVE32F-NEXT: lw t2, 28(a0)
-; RV32ZVE32F-NEXT: lw t3, 24(a0)
-; RV32ZVE32F-NEXT: lw t4, 20(a0)
-; RV32ZVE32F-NEXT: lw t5, 16(a0)
-; RV32ZVE32F-NEXT: lw s0, 12(a0)
-; RV32ZVE32F-NEXT: lw t6, 8(a0)
+; RV32ZVE32F-NEXT: lw t0, 40(a0)
+; RV32ZVE32F-NEXT: lw t1, 36(a0)
+; RV32ZVE32F-NEXT: lw t2, 32(a0)
+; RV32ZVE32F-NEXT: lw t3, 28(a0)
+; RV32ZVE32F-NEXT: lw t4, 24(a0)
+; RV32ZVE32F-NEXT: lw t5, 20(a0)
+; RV32ZVE32F-NEXT: lw t6, 16(a0)
+; RV32ZVE32F-NEXT: lw s1, 12(a0)
+; RV32ZVE32F-NEXT: lw s0, 8(a0)
; RV32ZVE32F-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vsll.vi v8, v8, 3
-; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
; RV32ZVE32F-NEXT: vsetvli zero, zero, e8, mf2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a1, v0
-; RV32ZVE32F-NEXT: andi s1, a1, 1
-; RV32ZVE32F-NEXT: bnez s1, .LBB48_10
+; RV32ZVE32F-NEXT: vmv.x.s a7, v0
+; RV32ZVE32F-NEXT: andi s2, a7, 1
+; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
+; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
+; RV32ZVE32F-NEXT: bnez s2, .LBB48_10
; RV32ZVE32F-NEXT: # %bb.1: # %else
-; RV32ZVE32F-NEXT: andi a0, a1, 2
+; RV32ZVE32F-NEXT: andi a0, a7, 2
; RV32ZVE32F-NEXT: bnez a0, .LBB48_11
; RV32ZVE32F-NEXT: .LBB48_2: # %else2
-; RV32ZVE32F-NEXT: andi a0, a1, 4
+; RV32ZVE32F-NEXT: andi a0, a7, 4
; RV32ZVE32F-NEXT: bnez a0, .LBB48_12
; RV32ZVE32F-NEXT: .LBB48_3: # %else4
-; RV32ZVE32F-NEXT: andi a0, a1, 8
+; RV32ZVE32F-NEXT: andi a0, a7, 8
; RV32ZVE32F-NEXT: bnez a0, .LBB48_13
; RV32ZVE32F-NEXT: .LBB48_4: # %else6
-; RV32ZVE32F-NEXT: andi a0, a1, 16
+; RV32ZVE32F-NEXT: andi a0, a7, 16
; RV32ZVE32F-NEXT: bnez a0, .LBB48_14
; RV32ZVE32F-NEXT: .LBB48_5: # %else8
-; RV32ZVE32F-NEXT: andi a0, a1, 32
+; RV32ZVE32F-NEXT: andi a0, a7, 32
; RV32ZVE32F-NEXT: bnez a0, .LBB48_15
; RV32ZVE32F-NEXT: .LBB48_6: # %else10
-; RV32ZVE32F-NEXT: andi a0, a1, 64
+; RV32ZVE32F-NEXT: andi a0, a7, 64
; RV32ZVE32F-NEXT: bnez a0, .LBB48_16
; RV32ZVE32F-NEXT: .LBB48_7: # %else12
-; RV32ZVE32F-NEXT: andi a0, a1, -128
+; RV32ZVE32F-NEXT: andi a0, a7, -128
; RV32ZVE32F-NEXT: beqz a0, .LBB48_9
; RV32ZVE32F-NEXT: .LBB48_8: # %cond.store13
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -4922,53 +4925,52 @@ define void @mscatter_baseidx_v8i32_v8i64(<8 x i64> %val, ptr %base, <8 x i32> %
; RV32ZVE32F-NEXT: addi sp, sp, 16
; RV32ZVE32F-NEXT: ret
; RV32ZVE32F-NEXT: .LBB48_10: # %cond.store
-; RV32ZVE32F-NEXT: lw s1, 4(a0)
+; RV32ZVE32F-NEXT: lw a1, 4(a0)
; RV32ZVE32F-NEXT: lw a0, 0(a0)
-; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vmv.x.s s2, v8
-; RV32ZVE32F-NEXT: sw s1, 4(s2)
+; RV32ZVE32F-NEXT: sw a1, 4(s2)
; RV32ZVE32F-NEXT: sw a0, 0(s2)
-; RV32ZVE32F-NEXT: andi a0, a1, 2
+; RV32ZVE32F-NEXT: andi a0, a7, 2
; RV32ZVE32F-NEXT: beqz a0, .LBB48_2
; RV32ZVE32F-NEXT: .LBB48_11: # %cond.store1
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 1
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
-; RV32ZVE32F-NEXT: sw s0, 4(a0)
-; RV32ZVE32F-NEXT: sw t6, 0(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 4
+; RV32ZVE32F-NEXT: sw s1, 4(a0)
+; RV32ZVE32F-NEXT: sw s0, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a7, 4
; RV32ZVE32F-NEXT: beqz a0, .LBB48_3
; RV32ZVE32F-NEXT: .LBB48_12: # %cond.store3
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 2
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
-; RV32ZVE32F-NEXT: sw t5, 0(a0)
-; RV32ZVE32F-NEXT: sw t4, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 8
+; RV32ZVE32F-NEXT: sw t6, 0(a0)
+; RV32ZVE32F-NEXT: sw t5, 4(a0)
+; RV32ZVE32F-NEXT: andi a0, a7, 8
; RV32ZVE32F-NEXT: beqz a0, .LBB48_4
; RV32ZVE32F-NEXT: .LBB48_13: # %cond.store5
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 3
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
-; RV32ZVE32F-NEXT: sw t3, 0(a0)
-; RV32ZVE32F-NEXT: sw t2, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 16
+; RV32ZVE32F-NEXT: sw t4, 0(a0)
+; RV32ZVE32F-NEXT: sw t3, 4(a0)
+; RV32ZVE32F-NEXT: andi a0, a7, 16
; RV32ZVE32F-NEXT: beqz a0, .LBB48_5
; RV32ZVE32F-NEXT: .LBB48_14: # %cond.store7
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 4
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
-; RV32ZVE32F-NEXT: sw t1, 0(a0)
-; RV32ZVE32F-NEXT: sw t0, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 32
+; RV32ZVE32F-NEXT: sw t2, 0(a0)
+; RV32ZVE32F-NEXT: sw t1, 4(a0)
+; RV32ZVE32F-NEXT: andi a0, a7, 32
; RV32ZVE32F-NEXT: beqz a0, .LBB48_6
; RV32ZVE32F-NEXT: .LBB48_15: # %cond.store9
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 5
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
-; RV32ZVE32F-NEXT: sw a7, 0(a0)
+; RV32ZVE32F-NEXT: sw t0, 0(a0)
; RV32ZVE32F-NEXT: sw a6, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 64
+; RV32ZVE32F-NEXT: andi a0, a7, 64
; RV32ZVE32F-NEXT: beqz a0, .LBB48_7
; RV32ZVE32F-NEXT: .LBB48_16: # %cond.store11
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -4976,7 +4978,7 @@ define void @mscatter_baseidx_v8i32_v8i64(<8 x i64> %val, ptr %base, <8 x i32> %
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
; RV32ZVE32F-NEXT: sw a5, 0(a0)
; RV32ZVE32F-NEXT: sw a4, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, -128
+; RV32ZVE32F-NEXT: andi a0, a7, -128
; RV32ZVE32F-NEXT: bnez a0, .LBB48_8
; RV32ZVE32F-NEXT: j .LBB48_9
;
@@ -5013,8 +5015,8 @@ define void @mscatter_baseidx_v8i32_v8i64(<8 x i64> %val, ptr %base, <8 x i32> %
; RV64ZVE32F-NEXT: .LBB48_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e32, m2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e32, m1, ta, ma
; RV64ZVE32F-NEXT: andi a0, a4, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e32, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a0, .LBB48_12
; RV64ZVE32F-NEXT: # %bb.5: # %else4
@@ -5115,42 +5117,43 @@ define void @mscatter_baseidx_sext_v8i32_v8i64(<8 x i64> %val, ptr %base, <8 x i
; RV32ZVE32F-NEXT: lw a4, 52(a0)
; RV32ZVE32F-NEXT: lw a5, 48(a0)
; RV32ZVE32F-NEXT: lw a6, 44(a0)
-; RV32ZVE32F-NEXT: lw a7, 40(a0)
-; RV32ZVE32F-NEXT: lw t0, 36(a0)
-; RV32ZVE32F-NEXT: lw t1, 32(a0)
-; RV32ZVE32F-NEXT: lw t2, 28(a0)
-; RV32ZVE32F-NEXT: lw t3, 24(a0)
-; RV32ZVE32F-NEXT: lw t4, 20(a0)
-; RV32ZVE32F-NEXT: lw t5, 16(a0)
-; RV32ZVE32F-NEXT: lw s0, 12(a0)
-; RV32ZVE32F-NEXT: lw t6, 8(a0)
+; RV32ZVE32F-NEXT: lw t0, 40(a0)
+; RV32ZVE32F-NEXT: lw t1, 36(a0)
+; RV32ZVE32F-NEXT: lw t2, 32(a0)
+; RV32ZVE32F-NEXT: lw t3, 28(a0)
+; RV32ZVE32F-NEXT: lw t4, 24(a0)
+; RV32ZVE32F-NEXT: lw t5, 20(a0)
+; RV32ZVE32F-NEXT: lw t6, 16(a0)
+; RV32ZVE32F-NEXT: lw s1, 12(a0)
+; RV32ZVE32F-NEXT: lw s0, 8(a0)
; RV32ZVE32F-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vsll.vi v8, v8, 3
-; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
; RV32ZVE32F-NEXT: vsetvli zero, zero, e8, mf2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a1, v0
-; RV32ZVE32F-NEXT: andi s1, a1, 1
-; RV32ZVE32F-NEXT: bnez s1, .LBB49_10
+; RV32ZVE32F-NEXT: vmv.x.s a7, v0
+; RV32ZVE32F-NEXT: andi s2, a7, 1
+; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
+; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
+; RV32ZVE32F-NEXT: bnez s2, .LBB49_10
; RV32ZVE32F-NEXT: # %bb.1: # %else
-; RV32ZVE32F-NEXT: andi a0, a1, 2
+; RV32ZVE32F-NEXT: andi a0, a7, 2
; RV32ZVE32F-NEXT: bnez a0, .LBB49_11
; RV32ZVE32F-NEXT: .LBB49_2: # %else2
-; RV32ZVE32F-NEXT: andi a0, a1, 4
+; RV32ZVE32F-NEXT: andi a0, a7, 4
; RV32ZVE32F-NEXT: bnez a0, .LBB49_12
; RV32ZVE32F-NEXT: .LBB49_3: # %else4
-; RV32ZVE32F-NEXT: andi a0, a1, 8
+; RV32ZVE32F-NEXT: andi a0, a7, 8
; RV32ZVE32F-NEXT: bnez a0, .LBB49_13
; RV32ZVE32F-NEXT: .LBB49_4: # %else6
-; RV32ZVE32F-NEXT: andi a0, a1, 16
+; RV32ZVE32F-NEXT: andi a0, a7, 16
; RV32ZVE32F-NEXT: bnez a0, .LBB49_14
; RV32ZVE32F-NEXT: .LBB49_5: # %else8
-; RV32ZVE32F-NEXT: andi a0, a1, 32
+; RV32ZVE32F-NEXT: andi a0, a7, 32
; RV32ZVE32F-NEXT: bnez a0, .LBB49_15
; RV32ZVE32F-NEXT: .LBB49_6: # %else10
-; RV32ZVE32F-NEXT: andi a0, a1, 64
+; RV32ZVE32F-NEXT: andi a0, a7, 64
; RV32ZVE32F-NEXT: bnez a0, .LBB49_16
; RV32ZVE32F-NEXT: .LBB49_7: # %else12
-; RV32ZVE32F-NEXT: andi a0, a1, -128
+; RV32ZVE32F-NEXT: andi a0, a7, -128
; RV32ZVE32F-NEXT: beqz a0, .LBB49_9
; RV32ZVE32F-NEXT: .LBB49_8: # %cond.store13
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -5165,53 +5168,52 @@ define void @mscatter_baseidx_sext_v8i32_v8i64(<8 x i64> %val, ptr %base, <8 x i
; RV32ZVE32F-NEXT: addi sp, sp, 16
; RV32ZVE32F-NEXT: ret
; RV32ZVE32F-NEXT: .LBB49_10: # %cond.store
-; RV32ZVE32F-NEXT: lw s1, 4(a0)
+; RV32ZVE32F-NEXT: lw a1, 4(a0)
; RV32ZVE32F-NEXT: lw a0, 0(a0)
-; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vmv.x.s s2, v8
-; RV32ZVE32F-NEXT: sw s1, 4(s2)
+; RV32ZVE32F-NEXT: sw a1, 4(s2)
; RV32ZVE32F-NEXT: sw a0, 0(s2)
-; RV32ZVE32F-NEXT: andi a0, a1, 2
+; RV32ZVE32F-NEXT: andi a0, a7, 2
; RV32ZVE32F-NEXT: beqz a0, .LBB49_2
; RV32ZVE32F-NEXT: .LBB49_11: # %cond.store1
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 1
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
-; RV32ZVE32F-NEXT: sw s0, 4(a0)
-; RV32ZVE32F-NEXT: sw t6, 0(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 4
+; RV32ZVE32F-NEXT: sw s1, 4(a0)
+; RV32ZVE32F-NEXT: sw s0, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a7, 4
; RV32ZVE32F-NEXT: beqz a0, .LBB49_3
; RV32ZVE32F-NEXT: .LBB49_12: # %cond.store3
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 2
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
-; RV32ZVE32F-NEXT: sw t5, 0(a0)
-; RV32ZVE32F-NEXT: sw t4, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 8
+; RV32ZVE32F-NEXT: sw t6, 0(a0)
+; RV32ZVE32F-NEXT: sw t5, 4(a0)
+; RV32ZVE32F-NEXT: andi a0, a7, 8
; RV32ZVE32F-NEXT: beqz a0, .LBB49_4
; RV32ZVE32F-NEXT: .LBB49_13: # %cond.store5
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 3
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
-; RV32ZVE32F-NEXT: sw t3, 0(a0)
-; RV32ZVE32F-NEXT: sw t2, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 16
+; RV32ZVE32F-NEXT: sw t4, 0(a0)
+; RV32ZVE32F-NEXT: sw t3, 4(a0)
+; RV32ZVE32F-NEXT: andi a0, a7, 16
; RV32ZVE32F-NEXT: beqz a0, .LBB49_5
; RV32ZVE32F-NEXT: .LBB49_14: # %cond.store7
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 4
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
-; RV32ZVE32F-NEXT: sw t1, 0(a0)
-; RV32ZVE32F-NEXT: sw t0, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 32
+; RV32ZVE32F-NEXT: sw t2, 0(a0)
+; RV32ZVE32F-NEXT: sw t1, 4(a0)
+; RV32ZVE32F-NEXT: andi a0, a7, 32
; RV32ZVE32F-NEXT: beqz a0, .LBB49_6
; RV32ZVE32F-NEXT: .LBB49_15: # %cond.store9
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 5
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
-; RV32ZVE32F-NEXT: sw a7, 0(a0)
+; RV32ZVE32F-NEXT: sw t0, 0(a0)
; RV32ZVE32F-NEXT: sw a6, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 64
+; RV32ZVE32F-NEXT: andi a0, a7, 64
; RV32ZVE32F-NEXT: beqz a0, .LBB49_7
; RV32ZVE32F-NEXT: .LBB49_16: # %cond.store11
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -5219,7 +5221,7 @@ define void @mscatter_baseidx_sext_v8i32_v8i64(<8 x i64> %val, ptr %base, <8 x i
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
; RV32ZVE32F-NEXT: sw a5, 0(a0)
; RV32ZVE32F-NEXT: sw a4, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, -128
+; RV32ZVE32F-NEXT: andi a0, a7, -128
; RV32ZVE32F-NEXT: bnez a0, .LBB49_8
; RV32ZVE32F-NEXT: j .LBB49_9
;
@@ -5256,8 +5258,8 @@ define void @mscatter_baseidx_sext_v8i32_v8i64(<8 x i64> %val, ptr %base, <8 x i
; RV64ZVE32F-NEXT: .LBB49_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e32, m2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e32, m1, ta, ma
; RV64ZVE32F-NEXT: andi a0, a4, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e32, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a0, .LBB49_12
; RV64ZVE32F-NEXT: # %bb.5: # %else4
@@ -5359,42 +5361,43 @@ define void @mscatter_baseidx_zext_v8i32_v8i64(<8 x i64> %val, ptr %base, <8 x i
; RV32ZVE32F-NEXT: lw a4, 52(a0)
; RV32ZVE32F-NEXT: lw a5, 48(a0)
; RV32ZVE32F-NEXT: lw a6, 44(a0)
-; RV32ZVE32F-NEXT: lw a7, 40(a0)
-; RV32ZVE32F-NEXT: lw t0, 36(a0)
-; RV32ZVE32F-NEXT: lw t1, 32(a0)
-; RV32ZVE32F-NEXT: lw t2, 28(a0)
-; RV32ZVE32F-NEXT: lw t3, 24(a0)
-; RV32ZVE32F-NEXT: lw t4, 20(a0)
-; RV32ZVE32F-NEXT: lw t5, 16(a0)
-; RV32ZVE32F-NEXT: lw s0, 12(a0)
-; RV32ZVE32F-NEXT: lw t6, 8(a0)
+; RV32ZVE32F-NEXT: lw t0, 40(a0)
+; RV32ZVE32F-NEXT: lw t1, 36(a0)
+; RV32ZVE32F-NEXT: lw t2, 32(a0)
+; RV32ZVE32F-NEXT: lw t3, 28(a0)
+; RV32ZVE32F-NEXT: lw t4, 24(a0)
+; RV32ZVE32F-NEXT: lw t5, 20(a0)
+; RV32ZVE32F-NEXT: lw t6, 16(a0)
+; RV32ZVE32F-NEXT: lw s1, 12(a0)
+; RV32ZVE32F-NEXT: lw s0, 8(a0)
; RV32ZVE32F-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vsll.vi v8, v8, 3
-; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
; RV32ZVE32F-NEXT: vsetvli zero, zero, e8, mf2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a1, v0
-; RV32ZVE32F-NEXT: andi s1, a1, 1
-; RV32ZVE32F-NEXT: bnez s1, .LBB50_10
+; RV32ZVE32F-NEXT: vmv.x.s a7, v0
+; RV32ZVE32F-NEXT: andi s2, a7, 1
+; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
+; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
+; RV32ZVE32F-NEXT: bnez s2, .LBB50_10
; RV32ZVE32F-NEXT: # %bb.1: # %else
-; RV32ZVE32F-NEXT: andi a0, a1, 2
+; RV32ZVE32F-NEXT: andi a0, a7, 2
; RV32ZVE32F-NEXT: bnez a0, .LBB50_11
; RV32ZVE32F-NEXT: .LBB50_2: # %else2
-; RV32ZVE32F-NEXT: andi a0, a1, 4
+; RV32ZVE32F-NEXT: andi a0, a7, 4
; RV32ZVE32F-NEXT: bnez a0, .LBB50_12
; RV32ZVE32F-NEXT: .LBB50_3: # %else4
-; RV32ZVE32F-NEXT: andi a0, a1, 8
+; RV32ZVE32F-NEXT: andi a0, a7, 8
; RV32ZVE32F-NEXT: bnez a0, .LBB50_13
; RV32ZVE32F-NEXT: .LBB50_4: # %else6
-; RV32ZVE32F-NEXT: andi a0, a1, 16
+; RV32ZVE32F-NEXT: andi a0, a7, 16
; RV32ZVE32F-NEXT: bnez a0, .LBB50_14
; RV32ZVE32F-NEXT: .LBB50_5: # %else8
-; RV32ZVE32F-NEXT: andi a0, a1, 32
+; RV32ZVE32F-NEXT: andi a0, a7, 32
; RV32ZVE32F-NEXT: bnez a0, .LBB50_15
; RV32ZVE32F-NEXT: .LBB50_6: # %else10
-; RV32ZVE32F-NEXT: andi a0, a1, 64
+; RV32ZVE32F-NEXT: andi a0, a7, 64
; RV32ZVE32F-NEXT: bnez a0, .LBB50_16
; RV32ZVE32F-NEXT: .LBB50_7: # %else12
-; RV32ZVE32F-NEXT: andi a0, a1, -128
+; RV32ZVE32F-NEXT: andi a0, a7, -128
; RV32ZVE32F-NEXT: beqz a0, .LBB50_9
; RV32ZVE32F-NEXT: .LBB50_8: # %cond.store13
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -5409,53 +5412,52 @@ define void @mscatter_baseidx_zext_v8i32_v8i64(<8 x i64> %val, ptr %base, <8 x i
; RV32ZVE32F-NEXT: addi sp, sp, 16
; RV32ZVE32F-NEXT: ret
; RV32ZVE32F-NEXT: .LBB50_10: # %cond.store
-; RV32ZVE32F-NEXT: lw s1, 4(a0)
+; RV32ZVE32F-NEXT: lw a1, 4(a0)
; RV32ZVE32F-NEXT: lw a0, 0(a0)
-; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vmv.x.s s2, v8
-; RV32ZVE32F-NEXT: sw s1, 4(s2)
+; RV32ZVE32F-NEXT: sw a1, 4(s2)
; RV32ZVE32F-NEXT: sw a0, 0(s2)
-; RV32ZVE32F-NEXT: andi a0, a1, 2
+; RV32ZVE32F-NEXT: andi a0, a7, 2
; RV32ZVE32F-NEXT: beqz a0, .LBB50_2
; RV32ZVE32F-NEXT: .LBB50_11: # %cond.store1
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 1
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
-; RV32ZVE32F-NEXT: sw s0, 4(a0)
-; RV32ZVE32F-NEXT: sw t6, 0(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 4
+; RV32ZVE32F-NEXT: sw s1, 4(a0)
+; RV32ZVE32F-NEXT: sw s0, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a7, 4
; RV32ZVE32F-NEXT: beqz a0, .LBB50_3
; RV32ZVE32F-NEXT: .LBB50_12: # %cond.store3
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 2
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
-; RV32ZVE32F-NEXT: sw t5, 0(a0)
-; RV32ZVE32F-NEXT: sw t4, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 8
+; RV32ZVE32F-NEXT: sw t6, 0(a0)
+; RV32ZVE32F-NEXT: sw t5, 4(a0)
+; RV32ZVE32F-NEXT: andi a0, a7, 8
; RV32ZVE32F-NEXT: beqz a0, .LBB50_4
; RV32ZVE32F-NEXT: .LBB50_13: # %cond.store5
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 3
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
-; RV32ZVE32F-NEXT: sw t3, 0(a0)
-; RV32ZVE32F-NEXT: sw t2, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 16
+; RV32ZVE32F-NEXT: sw t4, 0(a0)
+; RV32ZVE32F-NEXT: sw t3, 4(a0)
+; RV32ZVE32F-NEXT: andi a0, a7, 16
; RV32ZVE32F-NEXT: beqz a0, .LBB50_5
; RV32ZVE32F-NEXT: .LBB50_14: # %cond.store7
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 4
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
-; RV32ZVE32F-NEXT: sw t1, 0(a0)
-; RV32ZVE32F-NEXT: sw t0, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 32
+; RV32ZVE32F-NEXT: sw t2, 0(a0)
+; RV32ZVE32F-NEXT: sw t1, 4(a0)
+; RV32ZVE32F-NEXT: andi a0, a7, 32
; RV32ZVE32F-NEXT: beqz a0, .LBB50_6
; RV32ZVE32F-NEXT: .LBB50_15: # %cond.store9
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 5
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
-; RV32ZVE32F-NEXT: sw a7, 0(a0)
+; RV32ZVE32F-NEXT: sw t0, 0(a0)
; RV32ZVE32F-NEXT: sw a6, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 64
+; RV32ZVE32F-NEXT: andi a0, a7, 64
; RV32ZVE32F-NEXT: beqz a0, .LBB50_7
; RV32ZVE32F-NEXT: .LBB50_16: # %cond.store11
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -5463,7 +5465,7 @@ define void @mscatter_baseidx_zext_v8i32_v8i64(<8 x i64> %val, ptr %base, <8 x i
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
; RV32ZVE32F-NEXT: sw a5, 0(a0)
; RV32ZVE32F-NEXT: sw a4, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, -128
+; RV32ZVE32F-NEXT: andi a0, a7, -128
; RV32ZVE32F-NEXT: bnez a0, .LBB50_8
; RV32ZVE32F-NEXT: j .LBB50_9
;
@@ -5502,8 +5504,8 @@ define void @mscatter_baseidx_zext_v8i32_v8i64(<8 x i64> %val, ptr %base, <8 x i
; RV64ZVE32F-NEXT: .LBB50_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e32, m2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e32, m1, ta, ma
; RV64ZVE32F-NEXT: andi a0, a4, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e32, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a0, .LBB50_12
; RV64ZVE32F-NEXT: # %bb.5: # %else4
@@ -5598,17 +5600,16 @@ define void @mscatter_baseidx_v8i64(<8 x i64> %val, ptr %base, <8 x i64> %idxs,
;
; RV32ZVE32F-LABEL: mscatter_baseidx_v8i64:
; RV32ZVE32F: # %bb.0:
-; RV32ZVE32F-NEXT: addi sp, sp, -48
-; RV32ZVE32F-NEXT: .cfi_def_cfa_offset 48
-; RV32ZVE32F-NEXT: sw s0, 44(sp) # 4-byte Folded Spill
-; RV32ZVE32F-NEXT: sw s1, 40(sp) # 4-byte Folded Spill
-; RV32ZVE32F-NEXT: sw s2, 36(sp) # 4-byte Folded Spill
-; RV32ZVE32F-NEXT: sw s3, 32(sp) # 4-byte Folded Spill
-; RV32ZVE32F-NEXT: sw s4, 28(sp) # 4-byte Folded Spill
-; RV32ZVE32F-NEXT: sw s5, 24(sp) # 4-byte Folded Spill
-; RV32ZVE32F-NEXT: sw s6, 20(sp) # 4-byte Folded Spill
-; RV32ZVE32F-NEXT: sw s7, 16(sp) # 4-byte Folded Spill
-; RV32ZVE32F-NEXT: sw s8, 12(sp) # 4-byte Folded Spill
+; RV32ZVE32F-NEXT: addi sp, sp, -32
+; RV32ZVE32F-NEXT: .cfi_def_cfa_offset 32
+; RV32ZVE32F-NEXT: sw s0, 28(sp) # 4-byte Folded Spill
+; RV32ZVE32F-NEXT: sw s1, 24(sp) # 4-byte Folded Spill
+; RV32ZVE32F-NEXT: sw s2, 20(sp) # 4-byte Folded Spill
+; RV32ZVE32F-NEXT: sw s3, 16(sp) # 4-byte Folded Spill
+; RV32ZVE32F-NEXT: sw s4, 12(sp) # 4-byte Folded Spill
+; RV32ZVE32F-NEXT: sw s5, 8(sp) # 4-byte Folded Spill
+; RV32ZVE32F-NEXT: sw s6, 4(sp) # 4-byte Folded Spill
+; RV32ZVE32F-NEXT: sw s7, 0(sp) # 4-byte Folded Spill
; RV32ZVE32F-NEXT: .cfi_offset s0, -4
; RV32ZVE32F-NEXT: .cfi_offset s1, -8
; RV32ZVE32F-NEXT: .cfi_offset s2, -12
@@ -5617,7 +5618,6 @@ define void @mscatter_baseidx_v8i64(<8 x i64> %val, ptr %base, <8 x i64> %idxs,
; RV32ZVE32F-NEXT: .cfi_offset s5, -24
; RV32ZVE32F-NEXT: .cfi_offset s6, -28
; RV32ZVE32F-NEXT: .cfi_offset s7, -32
-; RV32ZVE32F-NEXT: .cfi_offset s8, -36
; RV32ZVE32F-NEXT: lw a3, 60(a0)
; RV32ZVE32F-NEXT: lw a4, 56(a0)
; RV32ZVE32F-NEXT: lw a5, 52(a0)
@@ -5635,45 +5635,46 @@ define void @mscatter_baseidx_v8i64(<8 x i64> %val, ptr %base, <8 x i64> %idxs,
; RV32ZVE32F-NEXT: lw s2, 56(a2)
; RV32ZVE32F-NEXT: lw s3, 48(a2)
; RV32ZVE32F-NEXT: lw s4, 40(a2)
-; RV32ZVE32F-NEXT: lw s5, 32(a2)
-; RV32ZVE32F-NEXT: lw s6, 24(a2)
-; RV32ZVE32F-NEXT: lw s7, 16(a2)
-; RV32ZVE32F-NEXT: lw s8, 8(a2)
+; RV32ZVE32F-NEXT: lw s5, 8(a2)
; RV32ZVE32F-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vlse32.v v8, (a2), zero
-; RV32ZVE32F-NEXT: vslide1down.vx v8, v8, s8
-; RV32ZVE32F-NEXT: vslide1down.vx v8, v8, s7
-; RV32ZVE32F-NEXT: vslide1down.vx v8, v8, s6
+; RV32ZVE32F-NEXT: lw s6, 16(a2)
+; RV32ZVE32F-NEXT: lw s7, 24(a2)
+; RV32ZVE32F-NEXT: lw a2, 32(a2)
; RV32ZVE32F-NEXT: vslide1down.vx v8, v8, s5
+; RV32ZVE32F-NEXT: vslide1down.vx v8, v8, s6
+; RV32ZVE32F-NEXT: vslide1down.vx v8, v8, s7
+; RV32ZVE32F-NEXT: vslide1down.vx v8, v8, a2
; RV32ZVE32F-NEXT: vslide1down.vx v8, v8, s4
; RV32ZVE32F-NEXT: vslide1down.vx v8, v8, s3
; RV32ZVE32F-NEXT: vslide1down.vx v8, v8, s2
; RV32ZVE32F-NEXT: vsll.vi v8, v8, 3
-; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
; RV32ZVE32F-NEXT: vsetvli zero, zero, e8, mf2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a1, v0
-; RV32ZVE32F-NEXT: andi a2, a1, 1
-; RV32ZVE32F-NEXT: bnez a2, .LBB51_10
+; RV32ZVE32F-NEXT: vmv.x.s a2, v0
+; RV32ZVE32F-NEXT: andi s2, a2, 1
+; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
+; RV32ZVE32F-NEXT: vadd.vx v8, v8, a1
+; RV32ZVE32F-NEXT: bnez s2, .LBB51_10
; RV32ZVE32F-NEXT: # %bb.1: # %else
-; RV32ZVE32F-NEXT: andi a0, a1, 2
+; RV32ZVE32F-NEXT: andi a0, a2, 2
; RV32ZVE32F-NEXT: bnez a0, .LBB51_11
; RV32ZVE32F-NEXT: .LBB51_2: # %else2
-; RV32ZVE32F-NEXT: andi a0, a1, 4
+; RV32ZVE32F-NEXT: andi a0, a2, 4
; RV32ZVE32F-NEXT: bnez a0, .LBB51_12
; RV32ZVE32F-NEXT: .LBB51_3: # %else4
-; RV32ZVE32F-NEXT: andi a0, a1, 8
+; RV32ZVE32F-NEXT: andi a0, a2, 8
; RV32ZVE32F-NEXT: bnez a0, .LBB51_13
; RV32ZVE32F-NEXT: .LBB51_4: # %else6
-; RV32ZVE32F-NEXT: andi a0, a1, 16
+; RV32ZVE32F-NEXT: andi a0, a2, 16
; RV32ZVE32F-NEXT: bnez a0, .LBB51_14
; RV32ZVE32F-NEXT: .LBB51_5: # %else8
-; RV32ZVE32F-NEXT: andi a0, a1, 32
+; RV32ZVE32F-NEXT: andi a0, a2, 32
; RV32ZVE32F-NEXT: bnez a0, .LBB51_15
; RV32ZVE32F-NEXT: .LBB51_6: # %else10
-; RV32ZVE32F-NEXT: andi a0, a1, 64
+; RV32ZVE32F-NEXT: andi a0, a2, 64
; RV32ZVE32F-NEXT: bnez a0, .LBB51_16
; RV32ZVE32F-NEXT: .LBB51_7: # %else12
-; RV32ZVE32F-NEXT: andi a0, a1, -128
+; RV32ZVE32F-NEXT: andi a0, a2, -128
; RV32ZVE32F-NEXT: beqz a0, .LBB51_9
; RV32ZVE32F-NEXT: .LBB51_8: # %cond.store13
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -5682,25 +5683,23 @@ define void @mscatter_baseidx_v8i64(<8 x i64> %val, ptr %base, <8 x i64> %idxs,
; RV32ZVE32F-NEXT: sw a4, 0(a0)
; RV32ZVE32F-NEXT: sw a3, 4(a0)
; RV32ZVE32F-NEXT: .LBB51_9: # %else14
-; RV32ZVE32F-NEXT: lw s0, 44(sp) # 4-byte Folded Reload
-; RV32ZVE32F-NEXT: lw s1, 40(sp) # 4-byte Folded Reload
-; RV32ZVE32F-NEXT: lw s2, 36(sp) # 4-byte Folded Reload
-; RV32ZVE32F-NEXT: lw s3, 32(sp) # 4-byte Folded Reload
-; RV32ZVE32F-NEXT: lw s4, 28(sp) # 4-byte Folded Reload
-; RV32ZVE32F-NEXT: lw s5, 24(sp) # 4-byte Folded Reload
-; RV32ZVE32F-NEXT: lw s6, 20(sp) # 4-byte Folded Reload
-; RV32ZVE32F-NEXT: lw s7, 16(sp) # 4-byte Folded Reload
-; RV32ZVE32F-NEXT: lw s8, 12(sp) # 4-byte Folded Reload
-; RV32ZVE32F-NEXT: addi sp, sp, 48
+; RV32ZVE32F-NEXT: lw s0, 28(sp) # 4-byte Folded Reload
+; RV32ZVE32F-NEXT: lw s1, 24(sp) # 4-byte Folded Reload
+; RV32ZVE32F-NEXT: lw s2, 20(sp) # 4-byte Folded Reload
+; RV32ZVE32F-NEXT: lw s3, 16(sp) # 4-byte Folded Reload
+; RV32ZVE32F-NEXT: lw s4, 12(sp) # 4-byte Folded Reload
+; RV32ZVE32F-NEXT: lw s5, 8(sp) # 4-byte Folded Reload
+; RV32ZVE32F-NEXT: lw s6, 4(sp) # 4-byte Folded Reload
+; RV32ZVE32F-NEXT: lw s7, 0(sp) # 4-byte Folded Reload
+; RV32ZVE32F-NEXT: addi sp, sp, 32
; RV32ZVE32F-NEXT: ret
; RV32ZVE32F-NEXT: .LBB51_10: # %cond.store
-; RV32ZVE32F-NEXT: lw a2, 4(a0)
+; RV32ZVE32F-NEXT: lw a1, 4(a0)
; RV32ZVE32F-NEXT: lw a0, 0(a0)
-; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vmv.x.s s2, v8
-; RV32ZVE32F-NEXT: sw a2, 4(s2)
+; RV32ZVE32F-NEXT: sw a1, 4(s2)
; RV32ZVE32F-NEXT: sw a0, 0(s2)
-; RV32ZVE32F-NEXT: andi a0, a1, 2
+; RV32ZVE32F-NEXT: andi a0, a2, 2
; RV32ZVE32F-NEXT: beqz a0, .LBB51_2
; RV32ZVE32F-NEXT: .LBB51_11: # %cond.store1
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
@@ -5708,7 +5707,7 @@ define void @mscatter_baseidx_v8i64(<8 x i64> %val, ptr %base, <8 x i64> %idxs,
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
; RV32ZVE32F-NEXT: sw s1, 4(a0)
; RV32ZVE32F-NEXT: sw s0, 0(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 4
+; RV32ZVE32F-NEXT: andi a0, a2, 4
; RV32ZVE32F-NEXT: beqz a0, .LBB51_3
; RV32ZVE32F-NEXT: .LBB51_12: # %cond.store3
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
@@ -5716,7 +5715,7 @@ define void @mscatter_baseidx_v8i64(<8 x i64> %val, ptr %base, <8 x i64> %idxs,
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
; RV32ZVE32F-NEXT: sw t6, 0(a0)
; RV32ZVE32F-NEXT: sw t5, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 8
+; RV32ZVE32F-NEXT: andi a0, a2, 8
; RV32ZVE32F-NEXT: beqz a0, .LBB51_4
; RV32ZVE32F-NEXT: .LBB51_13: # %cond.store5
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
@@ -5724,7 +5723,7 @@ define void @mscatter_baseidx_v8i64(<8 x i64> %val, ptr %base, <8 x i64> %idxs,
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
; RV32ZVE32F-NEXT: sw t4, 0(a0)
; RV32ZVE32F-NEXT: sw t3, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 16
+; RV32ZVE32F-NEXT: andi a0, a2, 16
; RV32ZVE32F-NEXT: beqz a0, .LBB51_5
; RV32ZVE32F-NEXT: .LBB51_14: # %cond.store7
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -5732,7 +5731,7 @@ define void @mscatter_baseidx_v8i64(<8 x i64> %val, ptr %base, <8 x i64> %idxs,
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
; RV32ZVE32F-NEXT: sw t2, 0(a0)
; RV32ZVE32F-NEXT: sw t1, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 32
+; RV32ZVE32F-NEXT: andi a0, a2, 32
; RV32ZVE32F-NEXT: beqz a0, .LBB51_6
; RV32ZVE32F-NEXT: .LBB51_15: # %cond.store9
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -5740,7 +5739,7 @@ define void @mscatter_baseidx_v8i64(<8 x i64> %val, ptr %base, <8 x i64> %idxs,
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
; RV32ZVE32F-NEXT: sw t0, 0(a0)
; RV32ZVE32F-NEXT: sw a7, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, 64
+; RV32ZVE32F-NEXT: andi a0, a2, 64
; RV32ZVE32F-NEXT: beqz a0, .LBB51_7
; RV32ZVE32F-NEXT: .LBB51_16: # %cond.store11
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -5748,7 +5747,7 @@ define void @mscatter_baseidx_v8i64(<8 x i64> %val, ptr %base, <8 x i64> %idxs,
; RV32ZVE32F-NEXT: vmv.x.s a0, v10
; RV32ZVE32F-NEXT: sw a6, 0(a0)
; RV32ZVE32F-NEXT: sw a5, 4(a0)
-; RV32ZVE32F-NEXT: andi a0, a1, -128
+; RV32ZVE32F-NEXT: andi a0, a2, -128
; RV32ZVE32F-NEXT: bnez a0, .LBB51_8
; RV32ZVE32F-NEXT: j .LBB51_9
;
@@ -6022,18 +6021,18 @@ define void @mscatter_truemask_v4f16(<4 x half> %val, <4 x ptr> %ptrs) {
;
; RV64ZVE32F-LABEL: mscatter_truemask_v4f16:
; RV64ZVE32F: # %bb.0:
-; RV64ZVE32F-NEXT: ld a1, 24(a0)
-; RV64ZVE32F-NEXT: ld a2, 0(a0)
-; RV64ZVE32F-NEXT: ld a3, 16(a0)
-; RV64ZVE32F-NEXT: ld a0, 8(a0)
+; RV64ZVE32F-NEXT: ld a1, 0(a0)
+; RV64ZVE32F-NEXT: ld a2, 24(a0)
+; RV64ZVE32F-NEXT: ld a3, 8(a0)
+; RV64ZVE32F-NEXT: ld a0, 16(a0)
; RV64ZVE32F-NEXT: vsetivli zero, 1, e16, mf2, ta, ma
-; RV64ZVE32F-NEXT: vse16.v v8, (a2)
+; RV64ZVE32F-NEXT: vse16.v v8, (a1)
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 1
-; RV64ZVE32F-NEXT: vse16.v v9, (a0)
-; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 2
; RV64ZVE32F-NEXT: vse16.v v9, (a3)
+; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 2
+; RV64ZVE32F-NEXT: vse16.v v9, (a0)
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 3
-; RV64ZVE32F-NEXT: vse16.v v8, (a1)
+; RV64ZVE32F-NEXT: vse16.v v8, (a2)
; RV64ZVE32F-NEXT: ret
call void @llvm.masked.scatter.v4f16.v4p0(<4 x half> %val, <4 x ptr> %ptrs, i32 2, <4 x i1> splat (i1 1))
ret void
@@ -6195,8 +6194,8 @@ define void @mscatter_baseidx_v8i8_v8f16(<8 x half> %val, ptr %base, <8 x i8> %i
; RV64ZVE32F-NEXT: .LBB58_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v9, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v9, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB58_12
; RV64ZVE32F-NEXT: # %bb.5: # %else4
@@ -6218,8 +6217,8 @@ define void @mscatter_baseidx_v8i8_v8f16(<8 x half> %val, ptr %base, <8 x i8> %i
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 5
; RV64ZVE32F-NEXT: vse16.v v9, (a2)
; RV64ZVE32F-NEXT: .LBB58_9: # %else10
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v10, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB58_15
; RV64ZVE32F-NEXT: # %bb.10: # %else12
@@ -6328,8 +6327,8 @@ define void @mscatter_baseidx_sext_v8i8_v8f16(<8 x half> %val, ptr %base, <8 x i
; RV64ZVE32F-NEXT: .LBB59_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v9, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v9, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB59_12
; RV64ZVE32F-NEXT: # %bb.5: # %else4
@@ -6351,8 +6350,8 @@ define void @mscatter_baseidx_sext_v8i8_v8f16(<8 x half> %val, ptr %base, <8 x i
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 5
; RV64ZVE32F-NEXT: vse16.v v9, (a2)
; RV64ZVE32F-NEXT: .LBB59_9: # %else10
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v10, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB59_15
; RV64ZVE32F-NEXT: # %bb.10: # %else12
@@ -6462,8 +6461,8 @@ define void @mscatter_baseidx_zext_v8i8_v8f16(<8 x half> %val, ptr %base, <8 x i
; RV64ZVE32F-NEXT: .LBB60_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v9, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v9, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB60_12
; RV64ZVE32F-NEXT: # %bb.5: # %else4
@@ -6486,8 +6485,8 @@ define void @mscatter_baseidx_zext_v8i8_v8f16(<8 x half> %val, ptr %base, <8 x i
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 5
; RV64ZVE32F-NEXT: vse16.v v9, (a2)
; RV64ZVE32F-NEXT: .LBB60_9: # %else10
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v10, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB60_15
; RV64ZVE32F-NEXT: # %bb.10: # %else12
@@ -6600,8 +6599,8 @@ define void @mscatter_baseidx_v8f16(<8 x half> %val, ptr %base, <8 x i16> %idxs,
; RV64ZVE32F-NEXT: .LBB61_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e16, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v9, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v9, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB61_12
; RV64ZVE32F-NEXT: # %bb.5: # %else4
@@ -6623,8 +6622,8 @@ define void @mscatter_baseidx_v8f16(<8 x half> %val, ptr %base, <8 x i16> %idxs,
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 5
; RV64ZVE32F-NEXT: vse16.v v9, (a2)
; RV64ZVE32F-NEXT: .LBB61_9: # %else10
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v10, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB61_15
; RV64ZVE32F-NEXT: # %bb.10: # %else12
@@ -6844,18 +6843,18 @@ define void @mscatter_truemask_v4f32(<4 x float> %val, <4 x ptr> %ptrs) {
;
; RV64ZVE32F-LABEL: mscatter_truemask_v4f32:
; RV64ZVE32F: # %bb.0:
-; RV64ZVE32F-NEXT: ld a1, 24(a0)
-; RV64ZVE32F-NEXT: ld a2, 0(a0)
-; RV64ZVE32F-NEXT: ld a3, 16(a0)
-; RV64ZVE32F-NEXT: ld a0, 8(a0)
+; RV64ZVE32F-NEXT: ld a1, 0(a0)
+; RV64ZVE32F-NEXT: ld a2, 24(a0)
+; RV64ZVE32F-NEXT: ld a3, 8(a0)
+; RV64ZVE32F-NEXT: ld a0, 16(a0)
; RV64ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
-; RV64ZVE32F-NEXT: vse32.v v8, (a2)
+; RV64ZVE32F-NEXT: vse32.v v8, (a1)
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 1
-; RV64ZVE32F-NEXT: vse32.v v9, (a0)
-; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 2
; RV64ZVE32F-NEXT: vse32.v v9, (a3)
+; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 2
+; RV64ZVE32F-NEXT: vse32.v v9, (a0)
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 3
-; RV64ZVE32F-NEXT: vse32.v v8, (a1)
+; RV64ZVE32F-NEXT: vse32.v v8, (a2)
; RV64ZVE32F-NEXT: ret
call void @llvm.masked.scatter.v4f32.v4p0(<4 x float> %val, <4 x ptr> %ptrs, i32 4, <4 x i1> splat (i1 1))
ret void
@@ -7020,8 +7019,8 @@ define void @mscatter_baseidx_v8i8_v8f32(<8 x float> %val, ptr %base, <8 x i8> %
; RV64ZVE32F-NEXT: .LBB68_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v11, v10, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v10, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB68_12
; RV64ZVE32F-NEXT: # %bb.5: # %else4
@@ -7044,8 +7043,8 @@ define void @mscatter_baseidx_v8i8_v8f32(<8 x float> %val, ptr %base, <8 x i8> %
; RV64ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV64ZVE32F-NEXT: vse32.v v12, (a2)
; RV64ZVE32F-NEXT: .LBB68_9: # %else10
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v11, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB68_15
; RV64ZVE32F-NEXT: # %bb.10: # %else12
@@ -7156,8 +7155,8 @@ define void @mscatter_baseidx_sext_v8i8_v8f32(<8 x float> %val, ptr %base, <8 x
; RV64ZVE32F-NEXT: .LBB69_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v11, v10, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v10, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB69_12
; RV64ZVE32F-NEXT: # %bb.5: # %else4
@@ -7180,8 +7179,8 @@ define void @mscatter_baseidx_sext_v8i8_v8f32(<8 x float> %val, ptr %base, <8 x
; RV64ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV64ZVE32F-NEXT: vse32.v v12, (a2)
; RV64ZVE32F-NEXT: .LBB69_9: # %else10
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v11, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB69_15
; RV64ZVE32F-NEXT: # %bb.10: # %else12
@@ -7296,8 +7295,8 @@ define void @mscatter_baseidx_zext_v8i8_v8f32(<8 x float> %val, ptr %base, <8 x
; RV64ZVE32F-NEXT: .LBB70_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v11, v10, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v10, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB70_12
; RV64ZVE32F-NEXT: # %bb.5: # %else4
@@ -7321,8 +7320,8 @@ define void @mscatter_baseidx_zext_v8i8_v8f32(<8 x float> %val, ptr %base, <8 x
; RV64ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV64ZVE32F-NEXT: vse32.v v12, (a2)
; RV64ZVE32F-NEXT: .LBB70_9: # %else10
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v11, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB70_15
; RV64ZVE32F-NEXT: # %bb.10: # %else12
@@ -7440,8 +7439,8 @@ define void @mscatter_baseidx_v8i16_v8f32(<8 x float> %val, ptr %base, <8 x i16>
; RV64ZVE32F-NEXT: .LBB71_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e16, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v11, v10, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v10, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB71_12
; RV64ZVE32F-NEXT: # %bb.5: # %else4
@@ -7464,8 +7463,8 @@ define void @mscatter_baseidx_v8i16_v8f32(<8 x float> %val, ptr %base, <8 x i16>
; RV64ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV64ZVE32F-NEXT: vse32.v v12, (a2)
; RV64ZVE32F-NEXT: .LBB71_9: # %else10
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v11, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB71_15
; RV64ZVE32F-NEXT: # %bb.10: # %else12
@@ -7577,8 +7576,8 @@ define void @mscatter_baseidx_sext_v8i16_v8f32(<8 x float> %val, ptr %base, <8 x
; RV64ZVE32F-NEXT: .LBB72_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e16, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v11, v10, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v10, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB72_12
; RV64ZVE32F-NEXT: # %bb.5: # %else4
@@ -7601,8 +7600,8 @@ define void @mscatter_baseidx_sext_v8i16_v8f32(<8 x float> %val, ptr %base, <8 x
; RV64ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV64ZVE32F-NEXT: vse32.v v12, (a2)
; RV64ZVE32F-NEXT: .LBB72_9: # %else10
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v11, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB72_15
; RV64ZVE32F-NEXT: # %bb.10: # %else12
@@ -7718,8 +7717,8 @@ define void @mscatter_baseidx_zext_v8i16_v8f32(<8 x float> %val, ptr %base, <8 x
; RV64ZVE32F-NEXT: .LBB73_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e16, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v11, v10, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: andi a3, a2, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v10, 2
; RV64ZVE32F-NEXT: bnez a3, .LBB73_12
; RV64ZVE32F-NEXT: # %bb.5: # %else4
@@ -7743,8 +7742,8 @@ define void @mscatter_baseidx_zext_v8i16_v8f32(<8 x float> %val, ptr %base, <8 x
; RV64ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV64ZVE32F-NEXT: vse32.v v12, (a3)
; RV64ZVE32F-NEXT: .LBB73_9: # %else10
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: andi a3, a2, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v11, 2
; RV64ZVE32F-NEXT: bnez a3, .LBB73_15
; RV64ZVE32F-NEXT: # %bb.10: # %else12
@@ -7859,8 +7858,8 @@ define void @mscatter_baseidx_v8f32(<8 x float> %val, ptr %base, <8 x i32> %idxs
; RV64ZVE32F-NEXT: .LBB74_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e32, m2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v12, v10, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e32, m1, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e32, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v10, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB74_12
; RV64ZVE32F-NEXT: # %bb.5: # %else4
@@ -7883,8 +7882,8 @@ define void @mscatter_baseidx_v8f32(<8 x float> %val, ptr %base, <8 x i32> %idxs
; RV64ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV64ZVE32F-NEXT: vse32.v v10, (a2)
; RV64ZVE32F-NEXT: .LBB74_9: # %else10
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e32, m1, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e32, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v12, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB74_15
; RV64ZVE32F-NEXT: # %bb.10: # %else12
@@ -8394,81 +8393,81 @@ define void @mscatter_baseidx_v8i8_v8f64(<8 x double> %val, ptr %base, <8 x i8>
; RV32ZVE32F-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vsext.vf4 v10, v8
; RV32ZVE32F-NEXT: vsll.vi v8, v10, 3
-; RV32ZVE32F-NEXT: vadd.vx v8, v8, a0
; RV32ZVE32F-NEXT: vsetvli zero, zero, e8, mf2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a0, v0
-; RV32ZVE32F-NEXT: andi a1, a0, 1
-; RV32ZVE32F-NEXT: bnez a1, .LBB81_9
+; RV32ZVE32F-NEXT: vmv.x.s a1, v0
+; RV32ZVE32F-NEXT: andi a2, a1, 1
+; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
+; RV32ZVE32F-NEXT: vadd.vx v8, v8, a0
+; RV32ZVE32F-NEXT: bnez a2, .LBB81_9
; RV32ZVE32F-NEXT: # %bb.1: # %else
-; RV32ZVE32F-NEXT: andi a1, a0, 2
-; RV32ZVE32F-NEXT: bnez a1, .LBB81_10
+; RV32ZVE32F-NEXT: andi a0, a1, 2
+; RV32ZVE32F-NEXT: bnez a0, .LBB81_10
; RV32ZVE32F-NEXT: .LBB81_2: # %else2
-; RV32ZVE32F-NEXT: andi a1, a0, 4
-; RV32ZVE32F-NEXT: bnez a1, .LBB81_11
+; RV32ZVE32F-NEXT: andi a0, a1, 4
+; RV32ZVE32F-NEXT: bnez a0, .LBB81_11
; RV32ZVE32F-NEXT: .LBB81_3: # %else4
-; RV32ZVE32F-NEXT: andi a1, a0, 8
-; RV32ZVE32F-NEXT: bnez a1, .LBB81_12
+; RV32ZVE32F-NEXT: andi a0, a1, 8
+; RV32ZVE32F-NEXT: bnez a0, .LBB81_12
; RV32ZVE32F-NEXT: .LBB81_4: # %else6
-; RV32ZVE32F-NEXT: andi a1, a0, 16
-; RV32ZVE32F-NEXT: bnez a1, .LBB81_13
+; RV32ZVE32F-NEXT: andi a0, a1, 16
+; RV32ZVE32F-NEXT: bnez a0, .LBB81_13
; RV32ZVE32F-NEXT: .LBB81_5: # %else8
-; RV32ZVE32F-NEXT: andi a1, a0, 32
-; RV32ZVE32F-NEXT: bnez a1, .LBB81_14
+; RV32ZVE32F-NEXT: andi a0, a1, 32
+; RV32ZVE32F-NEXT: bnez a0, .LBB81_14
; RV32ZVE32F-NEXT: .LBB81_6: # %else10
-; RV32ZVE32F-NEXT: andi a1, a0, 64
-; RV32ZVE32F-NEXT: bnez a1, .LBB81_15
+; RV32ZVE32F-NEXT: andi a0, a1, 64
+; RV32ZVE32F-NEXT: bnez a0, .LBB81_15
; RV32ZVE32F-NEXT: .LBB81_7: # %else12
-; RV32ZVE32F-NEXT: andi a0, a0, -128
+; RV32ZVE32F-NEXT: andi a0, a1, -128
; RV32ZVE32F-NEXT: bnez a0, .LBB81_16
; RV32ZVE32F-NEXT: .LBB81_8: # %else14
; RV32ZVE32F-NEXT: ret
; RV32ZVE32F-NEXT: .LBB81_9: # %cond.store
-; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a1, v8
-; RV32ZVE32F-NEXT: fsd fa0, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 2
-; RV32ZVE32F-NEXT: beqz a1, .LBB81_2
+; RV32ZVE32F-NEXT: vmv.x.s a0, v8
+; RV32ZVE32F-NEXT: fsd fa0, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 2
+; RV32ZVE32F-NEXT: beqz a0, .LBB81_2
; RV32ZVE32F-NEXT: .LBB81_10: # %cond.store1
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 1
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa1, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 4
-; RV32ZVE32F-NEXT: beqz a1, .LBB81_3
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa1, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 4
+; RV32ZVE32F-NEXT: beqz a0, .LBB81_3
; RV32ZVE32F-NEXT: .LBB81_11: # %cond.store3
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 2
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa2, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 8
-; RV32ZVE32F-NEXT: beqz a1, .LBB81_4
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa2, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 8
+; RV32ZVE32F-NEXT: beqz a0, .LBB81_4
; RV32ZVE32F-NEXT: .LBB81_12: # %cond.store5
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 3
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa3, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 16
-; RV32ZVE32F-NEXT: beqz a1, .LBB81_5
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa3, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 16
+; RV32ZVE32F-NEXT: beqz a0, .LBB81_5
; RV32ZVE32F-NEXT: .LBB81_13: # %cond.store7
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 4
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa4, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 32
-; RV32ZVE32F-NEXT: beqz a1, .LBB81_6
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa4, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 32
+; RV32ZVE32F-NEXT: beqz a0, .LBB81_6
; RV32ZVE32F-NEXT: .LBB81_14: # %cond.store9
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 5
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa5, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 64
-; RV32ZVE32F-NEXT: beqz a1, .LBB81_7
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa5, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 64
+; RV32ZVE32F-NEXT: beqz a0, .LBB81_7
; RV32ZVE32F-NEXT: .LBB81_15: # %cond.store11
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 6
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa6, 0(a1)
-; RV32ZVE32F-NEXT: andi a0, a0, -128
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa6, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, -128
; RV32ZVE32F-NEXT: beqz a0, .LBB81_8
; RV32ZVE32F-NEXT: .LBB81_16: # %cond.store13
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -8501,8 +8500,8 @@ define void @mscatter_baseidx_v8i8_v8f64(<8 x double> %val, ptr %base, <8 x i8>
; RV64ZVE32F-NEXT: .LBB81_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB81_12
; RV64ZVE32F-NEXT: # %bb.5: # %else4
@@ -8594,81 +8593,81 @@ define void @mscatter_baseidx_sext_v8i8_v8f64(<8 x double> %val, ptr %base, <8 x
; RV32ZVE32F-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vsext.vf4 v10, v8
; RV32ZVE32F-NEXT: vsll.vi v8, v10, 3
-; RV32ZVE32F-NEXT: vadd.vx v8, v8, a0
; RV32ZVE32F-NEXT: vsetvli zero, zero, e8, mf2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a0, v0
-; RV32ZVE32F-NEXT: andi a1, a0, 1
-; RV32ZVE32F-NEXT: bnez a1, .LBB82_9
-; RV32ZVE32F-NEXT: # %bb.1: # %else
-; RV32ZVE32F-NEXT: andi a1, a0, 2
-; RV32ZVE32F-NEXT: bnez a1, .LBB82_10
+; RV32ZVE32F-NEXT: vmv.x.s a1, v0
+; RV32ZVE32F-NEXT: andi a2, a1, 1
+; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
+; RV32ZVE32F-NEXT: vadd.vx v8, v8, a0
+; RV32ZVE32F-NEXT: bnez a2, .LBB82_9
+; RV32ZVE32F-NEXT: # %bb.1: # %else
+; RV32ZVE32F-NEXT: andi a0, a1, 2
+; RV32ZVE32F-NEXT: bnez a0, .LBB82_10
; RV32ZVE32F-NEXT: .LBB82_2: # %else2
-; RV32ZVE32F-NEXT: andi a1, a0, 4
-; RV32ZVE32F-NEXT: bnez a1, .LBB82_11
+; RV32ZVE32F-NEXT: andi a0, a1, 4
+; RV32ZVE32F-NEXT: bnez a0, .LBB82_11
; RV32ZVE32F-NEXT: .LBB82_3: # %else4
-; RV32ZVE32F-NEXT: andi a1, a0, 8
-; RV32ZVE32F-NEXT: bnez a1, .LBB82_12
+; RV32ZVE32F-NEXT: andi a0, a1, 8
+; RV32ZVE32F-NEXT: bnez a0, .LBB82_12
; RV32ZVE32F-NEXT: .LBB82_4: # %else6
-; RV32ZVE32F-NEXT: andi a1, a0, 16
-; RV32ZVE32F-NEXT: bnez a1, .LBB82_13
+; RV32ZVE32F-NEXT: andi a0, a1, 16
+; RV32ZVE32F-NEXT: bnez a0, .LBB82_13
; RV32ZVE32F-NEXT: .LBB82_5: # %else8
-; RV32ZVE32F-NEXT: andi a1, a0, 32
-; RV32ZVE32F-NEXT: bnez a1, .LBB82_14
+; RV32ZVE32F-NEXT: andi a0, a1, 32
+; RV32ZVE32F-NEXT: bnez a0, .LBB82_14
; RV32ZVE32F-NEXT: .LBB82_6: # %else10
-; RV32ZVE32F-NEXT: andi a1, a0, 64
-; RV32ZVE32F-NEXT: bnez a1, .LBB82_15
+; RV32ZVE32F-NEXT: andi a0, a1, 64
+; RV32ZVE32F-NEXT: bnez a0, .LBB82_15
; RV32ZVE32F-NEXT: .LBB82_7: # %else12
-; RV32ZVE32F-NEXT: andi a0, a0, -128
+; RV32ZVE32F-NEXT: andi a0, a1, -128
; RV32ZVE32F-NEXT: bnez a0, .LBB82_16
; RV32ZVE32F-NEXT: .LBB82_8: # %else14
; RV32ZVE32F-NEXT: ret
; RV32ZVE32F-NEXT: .LBB82_9: # %cond.store
-; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a1, v8
-; RV32ZVE32F-NEXT: fsd fa0, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 2
-; RV32ZVE32F-NEXT: beqz a1, .LBB82_2
+; RV32ZVE32F-NEXT: vmv.x.s a0, v8
+; RV32ZVE32F-NEXT: fsd fa0, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 2
+; RV32ZVE32F-NEXT: beqz a0, .LBB82_2
; RV32ZVE32F-NEXT: .LBB82_10: # %cond.store1
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 1
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa1, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 4
-; RV32ZVE32F-NEXT: beqz a1, .LBB82_3
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa1, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 4
+; RV32ZVE32F-NEXT: beqz a0, .LBB82_3
; RV32ZVE32F-NEXT: .LBB82_11: # %cond.store3
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 2
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa2, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 8
-; RV32ZVE32F-NEXT: beqz a1, .LBB82_4
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa2, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 8
+; RV32ZVE32F-NEXT: beqz a0, .LBB82_4
; RV32ZVE32F-NEXT: .LBB82_12: # %cond.store5
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 3
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa3, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 16
-; RV32ZVE32F-NEXT: beqz a1, .LBB82_5
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa3, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 16
+; RV32ZVE32F-NEXT: beqz a0, .LBB82_5
; RV32ZVE32F-NEXT: .LBB82_13: # %cond.store7
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 4
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa4, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 32
-; RV32ZVE32F-NEXT: beqz a1, .LBB82_6
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa4, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 32
+; RV32ZVE32F-NEXT: beqz a0, .LBB82_6
; RV32ZVE32F-NEXT: .LBB82_14: # %cond.store9
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 5
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa5, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 64
-; RV32ZVE32F-NEXT: beqz a1, .LBB82_7
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa5, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 64
+; RV32ZVE32F-NEXT: beqz a0, .LBB82_7
; RV32ZVE32F-NEXT: .LBB82_15: # %cond.store11
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 6
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa6, 0(a1)
-; RV32ZVE32F-NEXT: andi a0, a0, -128
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa6, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, -128
; RV32ZVE32F-NEXT: beqz a0, .LBB82_8
; RV32ZVE32F-NEXT: .LBB82_16: # %cond.store13
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -8701,8 +8700,8 @@ define void @mscatter_baseidx_sext_v8i8_v8f64(<8 x double> %val, ptr %base, <8 x
; RV64ZVE32F-NEXT: .LBB82_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB82_12
; RV64ZVE32F-NEXT: # %bb.5: # %else4
@@ -8796,81 +8795,81 @@ define void @mscatter_baseidx_zext_v8i8_v8f64(<8 x double> %val, ptr %base, <8 x
; RV32ZVE32F-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vzext.vf4 v10, v8
; RV32ZVE32F-NEXT: vsll.vi v8, v10, 3
-; RV32ZVE32F-NEXT: vadd.vx v8, v8, a0
; RV32ZVE32F-NEXT: vsetvli zero, zero, e8, mf2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a0, v0
-; RV32ZVE32F-NEXT: andi a1, a0, 1
-; RV32ZVE32F-NEXT: bnez a1, .LBB83_9
+; RV32ZVE32F-NEXT: vmv.x.s a1, v0
+; RV32ZVE32F-NEXT: andi a2, a1, 1
+; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
+; RV32ZVE32F-NEXT: vadd.vx v8, v8, a0
+; RV32ZVE32F-NEXT: bnez a2, .LBB83_9
; RV32ZVE32F-NEXT: # %bb.1: # %else
-; RV32ZVE32F-NEXT: andi a1, a0, 2
-; RV32ZVE32F-NEXT: bnez a1, .LBB83_10
+; RV32ZVE32F-NEXT: andi a0, a1, 2
+; RV32ZVE32F-NEXT: bnez a0, .LBB83_10
; RV32ZVE32F-NEXT: .LBB83_2: # %else2
-; RV32ZVE32F-NEXT: andi a1, a0, 4
-; RV32ZVE32F-NEXT: bnez a1, .LBB83_11
+; RV32ZVE32F-NEXT: andi a0, a1, 4
+; RV32ZVE32F-NEXT: bnez a0, .LBB83_11
; RV32ZVE32F-NEXT: .LBB83_3: # %else4
-; RV32ZVE32F-NEXT: andi a1, a0, 8
-; RV32ZVE32F-NEXT: bnez a1, .LBB83_12
+; RV32ZVE32F-NEXT: andi a0, a1, 8
+; RV32ZVE32F-NEXT: bnez a0, .LBB83_12
; RV32ZVE32F-NEXT: .LBB83_4: # %else6
-; RV32ZVE32F-NEXT: andi a1, a0, 16
-; RV32ZVE32F-NEXT: bnez a1, .LBB83_13
+; RV32ZVE32F-NEXT: andi a0, a1, 16
+; RV32ZVE32F-NEXT: bnez a0, .LBB83_13
; RV32ZVE32F-NEXT: .LBB83_5: # %else8
-; RV32ZVE32F-NEXT: andi a1, a0, 32
-; RV32ZVE32F-NEXT: bnez a1, .LBB83_14
+; RV32ZVE32F-NEXT: andi a0, a1, 32
+; RV32ZVE32F-NEXT: bnez a0, .LBB83_14
; RV32ZVE32F-NEXT: .LBB83_6: # %else10
-; RV32ZVE32F-NEXT: andi a1, a0, 64
-; RV32ZVE32F-NEXT: bnez a1, .LBB83_15
+; RV32ZVE32F-NEXT: andi a0, a1, 64
+; RV32ZVE32F-NEXT: bnez a0, .LBB83_15
; RV32ZVE32F-NEXT: .LBB83_7: # %else12
-; RV32ZVE32F-NEXT: andi a0, a0, -128
+; RV32ZVE32F-NEXT: andi a0, a1, -128
; RV32ZVE32F-NEXT: bnez a0, .LBB83_16
; RV32ZVE32F-NEXT: .LBB83_8: # %else14
; RV32ZVE32F-NEXT: ret
; RV32ZVE32F-NEXT: .LBB83_9: # %cond.store
-; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a1, v8
-; RV32ZVE32F-NEXT: fsd fa0, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 2
-; RV32ZVE32F-NEXT: beqz a1, .LBB83_2
+; RV32ZVE32F-NEXT: vmv.x.s a0, v8
+; RV32ZVE32F-NEXT: fsd fa0, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 2
+; RV32ZVE32F-NEXT: beqz a0, .LBB83_2
; RV32ZVE32F-NEXT: .LBB83_10: # %cond.store1
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 1
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa1, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 4
-; RV32ZVE32F-NEXT: beqz a1, .LBB83_3
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa1, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 4
+; RV32ZVE32F-NEXT: beqz a0, .LBB83_3
; RV32ZVE32F-NEXT: .LBB83_11: # %cond.store3
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 2
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa2, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 8
-; RV32ZVE32F-NEXT: beqz a1, .LBB83_4
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa2, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 8
+; RV32ZVE32F-NEXT: beqz a0, .LBB83_4
; RV32ZVE32F-NEXT: .LBB83_12: # %cond.store5
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 3
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa3, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 16
-; RV32ZVE32F-NEXT: beqz a1, .LBB83_5
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa3, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 16
+; RV32ZVE32F-NEXT: beqz a0, .LBB83_5
; RV32ZVE32F-NEXT: .LBB83_13: # %cond.store7
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 4
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa4, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 32
-; RV32ZVE32F-NEXT: beqz a1, .LBB83_6
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa4, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 32
+; RV32ZVE32F-NEXT: beqz a0, .LBB83_6
; RV32ZVE32F-NEXT: .LBB83_14: # %cond.store9
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 5
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa5, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 64
-; RV32ZVE32F-NEXT: beqz a1, .LBB83_7
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa5, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 64
+; RV32ZVE32F-NEXT: beqz a0, .LBB83_7
; RV32ZVE32F-NEXT: .LBB83_15: # %cond.store11
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 6
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa6, 0(a1)
-; RV32ZVE32F-NEXT: andi a0, a0, -128
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa6, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, -128
; RV32ZVE32F-NEXT: beqz a0, .LBB83_8
; RV32ZVE32F-NEXT: .LBB83_16: # %cond.store13
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -8905,8 +8904,8 @@ define void @mscatter_baseidx_zext_v8i8_v8f64(<8 x double> %val, ptr %base, <8 x
; RV64ZVE32F-NEXT: .LBB83_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB83_12
; RV64ZVE32F-NEXT: # %bb.5: # %else4
@@ -9005,81 +9004,81 @@ define void @mscatter_baseidx_v8i16_v8f64(<8 x double> %val, ptr %base, <8 x i16
; RV32ZVE32F-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vsext.vf2 v10, v8
; RV32ZVE32F-NEXT: vsll.vi v8, v10, 3
-; RV32ZVE32F-NEXT: vadd.vx v8, v8, a0
; RV32ZVE32F-NEXT: vsetvli zero, zero, e8, mf2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a0, v0
-; RV32ZVE32F-NEXT: andi a1, a0, 1
-; RV32ZVE32F-NEXT: bnez a1, .LBB84_9
+; RV32ZVE32F-NEXT: vmv.x.s a1, v0
+; RV32ZVE32F-NEXT: andi a2, a1, 1
+; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
+; RV32ZVE32F-NEXT: vadd.vx v8, v8, a0
+; RV32ZVE32F-NEXT: bnez a2, .LBB84_9
; RV32ZVE32F-NEXT: # %bb.1: # %else
-; RV32ZVE32F-NEXT: andi a1, a0, 2
-; RV32ZVE32F-NEXT: bnez a1, .LBB84_10
+; RV32ZVE32F-NEXT: andi a0, a1, 2
+; RV32ZVE32F-NEXT: bnez a0, .LBB84_10
; RV32ZVE32F-NEXT: .LBB84_2: # %else2
-; RV32ZVE32F-NEXT: andi a1, a0, 4
-; RV32ZVE32F-NEXT: bnez a1, .LBB84_11
+; RV32ZVE32F-NEXT: andi a0, a1, 4
+; RV32ZVE32F-NEXT: bnez a0, .LBB84_11
; RV32ZVE32F-NEXT: .LBB84_3: # %else4
-; RV32ZVE32F-NEXT: andi a1, a0, 8
-; RV32ZVE32F-NEXT: bnez a1, .LBB84_12
+; RV32ZVE32F-NEXT: andi a0, a1, 8
+; RV32ZVE32F-NEXT: bnez a0, .LBB84_12
; RV32ZVE32F-NEXT: .LBB84_4: # %else6
-; RV32ZVE32F-NEXT: andi a1, a0, 16
-; RV32ZVE32F-NEXT: bnez a1, .LBB84_13
+; RV32ZVE32F-NEXT: andi a0, a1, 16
+; RV32ZVE32F-NEXT: bnez a0, .LBB84_13
; RV32ZVE32F-NEXT: .LBB84_5: # %else8
-; RV32ZVE32F-NEXT: andi a1, a0, 32
-; RV32ZVE32F-NEXT: bnez a1, .LBB84_14
+; RV32ZVE32F-NEXT: andi a0, a1, 32
+; RV32ZVE32F-NEXT: bnez a0, .LBB84_14
; RV32ZVE32F-NEXT: .LBB84_6: # %else10
-; RV32ZVE32F-NEXT: andi a1, a0, 64
-; RV32ZVE32F-NEXT: bnez a1, .LBB84_15
+; RV32ZVE32F-NEXT: andi a0, a1, 64
+; RV32ZVE32F-NEXT: bnez a0, .LBB84_15
; RV32ZVE32F-NEXT: .LBB84_7: # %else12
-; RV32ZVE32F-NEXT: andi a0, a0, -128
+; RV32ZVE32F-NEXT: andi a0, a1, -128
; RV32ZVE32F-NEXT: bnez a0, .LBB84_16
; RV32ZVE32F-NEXT: .LBB84_8: # %else14
; RV32ZVE32F-NEXT: ret
; RV32ZVE32F-NEXT: .LBB84_9: # %cond.store
-; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a1, v8
-; RV32ZVE32F-NEXT: fsd fa0, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 2
-; RV32ZVE32F-NEXT: beqz a1, .LBB84_2
+; RV32ZVE32F-NEXT: vmv.x.s a0, v8
+; RV32ZVE32F-NEXT: fsd fa0, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 2
+; RV32ZVE32F-NEXT: beqz a0, .LBB84_2
; RV32ZVE32F-NEXT: .LBB84_10: # %cond.store1
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 1
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa1, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 4
-; RV32ZVE32F-NEXT: beqz a1, .LBB84_3
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa1, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 4
+; RV32ZVE32F-NEXT: beqz a0, .LBB84_3
; RV32ZVE32F-NEXT: .LBB84_11: # %cond.store3
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 2
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa2, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 8
-; RV32ZVE32F-NEXT: beqz a1, .LBB84_4
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa2, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 8
+; RV32ZVE32F-NEXT: beqz a0, .LBB84_4
; RV32ZVE32F-NEXT: .LBB84_12: # %cond.store5
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 3
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa3, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 16
-; RV32ZVE32F-NEXT: beqz a1, .LBB84_5
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa3, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 16
+; RV32ZVE32F-NEXT: beqz a0, .LBB84_5
; RV32ZVE32F-NEXT: .LBB84_13: # %cond.store7
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 4
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa4, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 32
-; RV32ZVE32F-NEXT: beqz a1, .LBB84_6
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa4, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 32
+; RV32ZVE32F-NEXT: beqz a0, .LBB84_6
; RV32ZVE32F-NEXT: .LBB84_14: # %cond.store9
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 5
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa5, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 64
-; RV32ZVE32F-NEXT: beqz a1, .LBB84_7
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa5, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 64
+; RV32ZVE32F-NEXT: beqz a0, .LBB84_7
; RV32ZVE32F-NEXT: .LBB84_15: # %cond.store11
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 6
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa6, 0(a1)
-; RV32ZVE32F-NEXT: andi a0, a0, -128
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa6, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, -128
; RV32ZVE32F-NEXT: beqz a0, .LBB84_8
; RV32ZVE32F-NEXT: .LBB84_16: # %cond.store13
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -9113,8 +9112,8 @@ define void @mscatter_baseidx_v8i16_v8f64(<8 x double> %val, ptr %base, <8 x i16
; RV64ZVE32F-NEXT: .LBB84_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e16, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB84_12
; RV64ZVE32F-NEXT: # %bb.5: # %else4
@@ -9206,81 +9205,81 @@ define void @mscatter_baseidx_sext_v8i16_v8f64(<8 x double> %val, ptr %base, <8
; RV32ZVE32F-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vsext.vf2 v10, v8
; RV32ZVE32F-NEXT: vsll.vi v8, v10, 3
-; RV32ZVE32F-NEXT: vadd.vx v8, v8, a0
; RV32ZVE32F-NEXT: vsetvli zero, zero, e8, mf2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a0, v0
-; RV32ZVE32F-NEXT: andi a1, a0, 1
-; RV32ZVE32F-NEXT: bnez a1, .LBB85_9
+; RV32ZVE32F-NEXT: vmv.x.s a1, v0
+; RV32ZVE32F-NEXT: andi a2, a1, 1
+; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
+; RV32ZVE32F-NEXT: vadd.vx v8, v8, a0
+; RV32ZVE32F-NEXT: bnez a2, .LBB85_9
; RV32ZVE32F-NEXT: # %bb.1: # %else
-; RV32ZVE32F-NEXT: andi a1, a0, 2
-; RV32ZVE32F-NEXT: bnez a1, .LBB85_10
+; RV32ZVE32F-NEXT: andi a0, a1, 2
+; RV32ZVE32F-NEXT: bnez a0, .LBB85_10
; RV32ZVE32F-NEXT: .LBB85_2: # %else2
-; RV32ZVE32F-NEXT: andi a1, a0, 4
-; RV32ZVE32F-NEXT: bnez a1, .LBB85_11
+; RV32ZVE32F-NEXT: andi a0, a1, 4
+; RV32ZVE32F-NEXT: bnez a0, .LBB85_11
; RV32ZVE32F-NEXT: .LBB85_3: # %else4
-; RV32ZVE32F-NEXT: andi a1, a0, 8
-; RV32ZVE32F-NEXT: bnez a1, .LBB85_12
+; RV32ZVE32F-NEXT: andi a0, a1, 8
+; RV32ZVE32F-NEXT: bnez a0, .LBB85_12
; RV32ZVE32F-NEXT: .LBB85_4: # %else6
-; RV32ZVE32F-NEXT: andi a1, a0, 16
-; RV32ZVE32F-NEXT: bnez a1, .LBB85_13
+; RV32ZVE32F-NEXT: andi a0, a1, 16
+; RV32ZVE32F-NEXT: bnez a0, .LBB85_13
; RV32ZVE32F-NEXT: .LBB85_5: # %else8
-; RV32ZVE32F-NEXT: andi a1, a0, 32
-; RV32ZVE32F-NEXT: bnez a1, .LBB85_14
+; RV32ZVE32F-NEXT: andi a0, a1, 32
+; RV32ZVE32F-NEXT: bnez a0, .LBB85_14
; RV32ZVE32F-NEXT: .LBB85_6: # %else10
-; RV32ZVE32F-NEXT: andi a1, a0, 64
-; RV32ZVE32F-NEXT: bnez a1, .LBB85_15
+; RV32ZVE32F-NEXT: andi a0, a1, 64
+; RV32ZVE32F-NEXT: bnez a0, .LBB85_15
; RV32ZVE32F-NEXT: .LBB85_7: # %else12
-; RV32ZVE32F-NEXT: andi a0, a0, -128
+; RV32ZVE32F-NEXT: andi a0, a1, -128
; RV32ZVE32F-NEXT: bnez a0, .LBB85_16
; RV32ZVE32F-NEXT: .LBB85_8: # %else14
; RV32ZVE32F-NEXT: ret
; RV32ZVE32F-NEXT: .LBB85_9: # %cond.store
-; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a1, v8
-; RV32ZVE32F-NEXT: fsd fa0, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 2
-; RV32ZVE32F-NEXT: beqz a1, .LBB85_2
+; RV32ZVE32F-NEXT: vmv.x.s a0, v8
+; RV32ZVE32F-NEXT: fsd fa0, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 2
+; RV32ZVE32F-NEXT: beqz a0, .LBB85_2
; RV32ZVE32F-NEXT: .LBB85_10: # %cond.store1
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 1
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa1, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 4
-; RV32ZVE32F-NEXT: beqz a1, .LBB85_3
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa1, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 4
+; RV32ZVE32F-NEXT: beqz a0, .LBB85_3
; RV32ZVE32F-NEXT: .LBB85_11: # %cond.store3
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 2
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa2, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 8
-; RV32ZVE32F-NEXT: beqz a1, .LBB85_4
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa2, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 8
+; RV32ZVE32F-NEXT: beqz a0, .LBB85_4
; RV32ZVE32F-NEXT: .LBB85_12: # %cond.store5
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 3
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa3, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 16
-; RV32ZVE32F-NEXT: beqz a1, .LBB85_5
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa3, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 16
+; RV32ZVE32F-NEXT: beqz a0, .LBB85_5
; RV32ZVE32F-NEXT: .LBB85_13: # %cond.store7
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 4
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa4, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 32
-; RV32ZVE32F-NEXT: beqz a1, .LBB85_6
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa4, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 32
+; RV32ZVE32F-NEXT: beqz a0, .LBB85_6
; RV32ZVE32F-NEXT: .LBB85_14: # %cond.store9
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 5
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa5, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 64
-; RV32ZVE32F-NEXT: beqz a1, .LBB85_7
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa5, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 64
+; RV32ZVE32F-NEXT: beqz a0, .LBB85_7
; RV32ZVE32F-NEXT: .LBB85_15: # %cond.store11
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 6
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa6, 0(a1)
-; RV32ZVE32F-NEXT: andi a0, a0, -128
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa6, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, -128
; RV32ZVE32F-NEXT: beqz a0, .LBB85_8
; RV32ZVE32F-NEXT: .LBB85_16: # %cond.store13
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -9314,8 +9313,8 @@ define void @mscatter_baseidx_sext_v8i16_v8f64(<8 x double> %val, ptr %base, <8
; RV64ZVE32F-NEXT: .LBB85_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e16, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB85_12
; RV64ZVE32F-NEXT: # %bb.5: # %else4
@@ -9409,81 +9408,81 @@ define void @mscatter_baseidx_zext_v8i16_v8f64(<8 x double> %val, ptr %base, <8
; RV32ZVE32F-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vzext.vf2 v10, v8
; RV32ZVE32F-NEXT: vsll.vi v8, v10, 3
-; RV32ZVE32F-NEXT: vadd.vx v8, v8, a0
; RV32ZVE32F-NEXT: vsetvli zero, zero, e8, mf2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a0, v0
-; RV32ZVE32F-NEXT: andi a1, a0, 1
-; RV32ZVE32F-NEXT: bnez a1, .LBB86_9
+; RV32ZVE32F-NEXT: vmv.x.s a1, v0
+; RV32ZVE32F-NEXT: andi a2, a1, 1
+; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
+; RV32ZVE32F-NEXT: vadd.vx v8, v8, a0
+; RV32ZVE32F-NEXT: bnez a2, .LBB86_9
; RV32ZVE32F-NEXT: # %bb.1: # %else
-; RV32ZVE32F-NEXT: andi a1, a0, 2
-; RV32ZVE32F-NEXT: bnez a1, .LBB86_10
+; RV32ZVE32F-NEXT: andi a0, a1, 2
+; RV32ZVE32F-NEXT: bnez a0, .LBB86_10
; RV32ZVE32F-NEXT: .LBB86_2: # %else2
-; RV32ZVE32F-NEXT: andi a1, a0, 4
-; RV32ZVE32F-NEXT: bnez a1, .LBB86_11
+; RV32ZVE32F-NEXT: andi a0, a1, 4
+; RV32ZVE32F-NEXT: bnez a0, .LBB86_11
; RV32ZVE32F-NEXT: .LBB86_3: # %else4
-; RV32ZVE32F-NEXT: andi a1, a0, 8
-; RV32ZVE32F-NEXT: bnez a1, .LBB86_12
+; RV32ZVE32F-NEXT: andi a0, a1, 8
+; RV32ZVE32F-NEXT: bnez a0, .LBB86_12
; RV32ZVE32F-NEXT: .LBB86_4: # %else6
-; RV32ZVE32F-NEXT: andi a1, a0, 16
-; RV32ZVE32F-NEXT: bnez a1, .LBB86_13
+; RV32ZVE32F-NEXT: andi a0, a1, 16
+; RV32ZVE32F-NEXT: bnez a0, .LBB86_13
; RV32ZVE32F-NEXT: .LBB86_5: # %else8
-; RV32ZVE32F-NEXT: andi a1, a0, 32
-; RV32ZVE32F-NEXT: bnez a1, .LBB86_14
+; RV32ZVE32F-NEXT: andi a0, a1, 32
+; RV32ZVE32F-NEXT: bnez a0, .LBB86_14
; RV32ZVE32F-NEXT: .LBB86_6: # %else10
-; RV32ZVE32F-NEXT: andi a1, a0, 64
-; RV32ZVE32F-NEXT: bnez a1, .LBB86_15
+; RV32ZVE32F-NEXT: andi a0, a1, 64
+; RV32ZVE32F-NEXT: bnez a0, .LBB86_15
; RV32ZVE32F-NEXT: .LBB86_7: # %else12
-; RV32ZVE32F-NEXT: andi a0, a0, -128
+; RV32ZVE32F-NEXT: andi a0, a1, -128
; RV32ZVE32F-NEXT: bnez a0, .LBB86_16
; RV32ZVE32F-NEXT: .LBB86_8: # %else14
; RV32ZVE32F-NEXT: ret
; RV32ZVE32F-NEXT: .LBB86_9: # %cond.store
-; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a1, v8
-; RV32ZVE32F-NEXT: fsd fa0, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 2
-; RV32ZVE32F-NEXT: beqz a1, .LBB86_2
+; RV32ZVE32F-NEXT: vmv.x.s a0, v8
+; RV32ZVE32F-NEXT: fsd fa0, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 2
+; RV32ZVE32F-NEXT: beqz a0, .LBB86_2
; RV32ZVE32F-NEXT: .LBB86_10: # %cond.store1
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 1
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa1, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 4
-; RV32ZVE32F-NEXT: beqz a1, .LBB86_3
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa1, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 4
+; RV32ZVE32F-NEXT: beqz a0, .LBB86_3
; RV32ZVE32F-NEXT: .LBB86_11: # %cond.store3
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 2
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa2, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 8
-; RV32ZVE32F-NEXT: beqz a1, .LBB86_4
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa2, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 8
+; RV32ZVE32F-NEXT: beqz a0, .LBB86_4
; RV32ZVE32F-NEXT: .LBB86_12: # %cond.store5
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 3
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa3, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 16
-; RV32ZVE32F-NEXT: beqz a1, .LBB86_5
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa3, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 16
+; RV32ZVE32F-NEXT: beqz a0, .LBB86_5
; RV32ZVE32F-NEXT: .LBB86_13: # %cond.store7
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 4
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa4, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 32
-; RV32ZVE32F-NEXT: beqz a1, .LBB86_6
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa4, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 32
+; RV32ZVE32F-NEXT: beqz a0, .LBB86_6
; RV32ZVE32F-NEXT: .LBB86_14: # %cond.store9
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 5
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa5, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 64
-; RV32ZVE32F-NEXT: beqz a1, .LBB86_7
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa5, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 64
+; RV32ZVE32F-NEXT: beqz a0, .LBB86_7
; RV32ZVE32F-NEXT: .LBB86_15: # %cond.store11
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 6
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa6, 0(a1)
-; RV32ZVE32F-NEXT: andi a0, a0, -128
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa6, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, -128
; RV32ZVE32F-NEXT: beqz a0, .LBB86_8
; RV32ZVE32F-NEXT: .LBB86_16: # %cond.store13
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -9521,8 +9520,8 @@ define void @mscatter_baseidx_zext_v8i16_v8f64(<8 x double> %val, ptr %base, <8
; RV64ZVE32F-NEXT: .LBB86_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e16, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: andi a3, a2, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e16, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a3, .LBB86_12
; RV64ZVE32F-NEXT: # %bb.5: # %else4
@@ -9619,81 +9618,81 @@ define void @mscatter_baseidx_v8i32_v8f64(<8 x double> %val, ptr %base, <8 x i32
; RV32ZVE32F: # %bb.0:
; RV32ZVE32F-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vsll.vi v8, v8, 3
-; RV32ZVE32F-NEXT: vadd.vx v8, v8, a0
; RV32ZVE32F-NEXT: vsetvli zero, zero, e8, mf2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a0, v0
-; RV32ZVE32F-NEXT: andi a1, a0, 1
-; RV32ZVE32F-NEXT: bnez a1, .LBB87_9
+; RV32ZVE32F-NEXT: vmv.x.s a1, v0
+; RV32ZVE32F-NEXT: andi a2, a1, 1
+; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
+; RV32ZVE32F-NEXT: vadd.vx v8, v8, a0
+; RV32ZVE32F-NEXT: bnez a2, .LBB87_9
; RV32ZVE32F-NEXT: # %bb.1: # %else
-; RV32ZVE32F-NEXT: andi a1, a0, 2
-; RV32ZVE32F-NEXT: bnez a1, .LBB87_10
+; RV32ZVE32F-NEXT: andi a0, a1, 2
+; RV32ZVE32F-NEXT: bnez a0, .LBB87_10
; RV32ZVE32F-NEXT: .LBB87_2: # %else2
-; RV32ZVE32F-NEXT: andi a1, a0, 4
-; RV32ZVE32F-NEXT: bnez a1, .LBB87_11
+; RV32ZVE32F-NEXT: andi a0, a1, 4
+; RV32ZVE32F-NEXT: bnez a0, .LBB87_11
; RV32ZVE32F-NEXT: .LBB87_3: # %else4
-; RV32ZVE32F-NEXT: andi a1, a0, 8
-; RV32ZVE32F-NEXT: bnez a1, .LBB87_12
+; RV32ZVE32F-NEXT: andi a0, a1, 8
+; RV32ZVE32F-NEXT: bnez a0, .LBB87_12
; RV32ZVE32F-NEXT: .LBB87_4: # %else6
-; RV32ZVE32F-NEXT: andi a1, a0, 16
-; RV32ZVE32F-NEXT: bnez a1, .LBB87_13
+; RV32ZVE32F-NEXT: andi a0, a1, 16
+; RV32ZVE32F-NEXT: bnez a0, .LBB87_13
; RV32ZVE32F-NEXT: .LBB87_5: # %else8
-; RV32ZVE32F-NEXT: andi a1, a0, 32
-; RV32ZVE32F-NEXT: bnez a1, .LBB87_14
+; RV32ZVE32F-NEXT: andi a0, a1, 32
+; RV32ZVE32F-NEXT: bnez a0, .LBB87_14
; RV32ZVE32F-NEXT: .LBB87_6: # %else10
-; RV32ZVE32F-NEXT: andi a1, a0, 64
-; RV32ZVE32F-NEXT: bnez a1, .LBB87_15
+; RV32ZVE32F-NEXT: andi a0, a1, 64
+; RV32ZVE32F-NEXT: bnez a0, .LBB87_15
; RV32ZVE32F-NEXT: .LBB87_7: # %else12
-; RV32ZVE32F-NEXT: andi a0, a0, -128
+; RV32ZVE32F-NEXT: andi a0, a1, -128
; RV32ZVE32F-NEXT: bnez a0, .LBB87_16
; RV32ZVE32F-NEXT: .LBB87_8: # %else14
; RV32ZVE32F-NEXT: ret
; RV32ZVE32F-NEXT: .LBB87_9: # %cond.store
-; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a1, v8
-; RV32ZVE32F-NEXT: fsd fa0, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 2
-; RV32ZVE32F-NEXT: beqz a1, .LBB87_2
+; RV32ZVE32F-NEXT: vmv.x.s a0, v8
+; RV32ZVE32F-NEXT: fsd fa0, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 2
+; RV32ZVE32F-NEXT: beqz a0, .LBB87_2
; RV32ZVE32F-NEXT: .LBB87_10: # %cond.store1
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 1
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa1, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 4
-; RV32ZVE32F-NEXT: beqz a1, .LBB87_3
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa1, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 4
+; RV32ZVE32F-NEXT: beqz a0, .LBB87_3
; RV32ZVE32F-NEXT: .LBB87_11: # %cond.store3
-; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
-; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 2
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa2, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 8
-; RV32ZVE32F-NEXT: beqz a1, .LBB87_4
+; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
+; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 2
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa2, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 8
+; RV32ZVE32F-NEXT: beqz a0, .LBB87_4
; RV32ZVE32F-NEXT: .LBB87_12: # %cond.store5
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 3
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa3, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 16
-; RV32ZVE32F-NEXT: beqz a1, .LBB87_5
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa3, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 16
+; RV32ZVE32F-NEXT: beqz a0, .LBB87_5
; RV32ZVE32F-NEXT: .LBB87_13: # %cond.store7
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 4
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa4, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 32
-; RV32ZVE32F-NEXT: beqz a1, .LBB87_6
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa4, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 32
+; RV32ZVE32F-NEXT: beqz a0, .LBB87_6
; RV32ZVE32F-NEXT: .LBB87_14: # %cond.store9
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 5
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa5, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 64
-; RV32ZVE32F-NEXT: beqz a1, .LBB87_7
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa5, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 64
+; RV32ZVE32F-NEXT: beqz a0, .LBB87_7
; RV32ZVE32F-NEXT: .LBB87_15: # %cond.store11
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 6
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa6, 0(a1)
-; RV32ZVE32F-NEXT: andi a0, a0, -128
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa6, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, -128
; RV32ZVE32F-NEXT: beqz a0, .LBB87_8
; RV32ZVE32F-NEXT: .LBB87_16: # %cond.store13
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -9727,8 +9726,8 @@ define void @mscatter_baseidx_v8i32_v8f64(<8 x double> %val, ptr %base, <8 x i32
; RV64ZVE32F-NEXT: .LBB87_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e32, m2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e32, m1, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e32, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB87_12
; RV64ZVE32F-NEXT: # %bb.5: # %else4
@@ -9818,81 +9817,81 @@ define void @mscatter_baseidx_sext_v8i32_v8f64(<8 x double> %val, ptr %base, <8
; RV32ZVE32F: # %bb.0:
; RV32ZVE32F-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vsll.vi v8, v8, 3
-; RV32ZVE32F-NEXT: vadd.vx v8, v8, a0
; RV32ZVE32F-NEXT: vsetvli zero, zero, e8, mf2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a0, v0
-; RV32ZVE32F-NEXT: andi a1, a0, 1
-; RV32ZVE32F-NEXT: bnez a1, .LBB88_9
+; RV32ZVE32F-NEXT: vmv.x.s a1, v0
+; RV32ZVE32F-NEXT: andi a2, a1, 1
+; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
+; RV32ZVE32F-NEXT: vadd.vx v8, v8, a0
+; RV32ZVE32F-NEXT: bnez a2, .LBB88_9
; RV32ZVE32F-NEXT: # %bb.1: # %else
-; RV32ZVE32F-NEXT: andi a1, a0, 2
-; RV32ZVE32F-NEXT: bnez a1, .LBB88_10
+; RV32ZVE32F-NEXT: andi a0, a1, 2
+; RV32ZVE32F-NEXT: bnez a0, .LBB88_10
; RV32ZVE32F-NEXT: .LBB88_2: # %else2
-; RV32ZVE32F-NEXT: andi a1, a0, 4
-; RV32ZVE32F-NEXT: bnez a1, .LBB88_11
+; RV32ZVE32F-NEXT: andi a0, a1, 4
+; RV32ZVE32F-NEXT: bnez a0, .LBB88_11
; RV32ZVE32F-NEXT: .LBB88_3: # %else4
-; RV32ZVE32F-NEXT: andi a1, a0, 8
-; RV32ZVE32F-NEXT: bnez a1, .LBB88_12
+; RV32ZVE32F-NEXT: andi a0, a1, 8
+; RV32ZVE32F-NEXT: bnez a0, .LBB88_12
; RV32ZVE32F-NEXT: .LBB88_4: # %else6
-; RV32ZVE32F-NEXT: andi a1, a0, 16
-; RV32ZVE32F-NEXT: bnez a1, .LBB88_13
+; RV32ZVE32F-NEXT: andi a0, a1, 16
+; RV32ZVE32F-NEXT: bnez a0, .LBB88_13
; RV32ZVE32F-NEXT: .LBB88_5: # %else8
-; RV32ZVE32F-NEXT: andi a1, a0, 32
-; RV32ZVE32F-NEXT: bnez a1, .LBB88_14
+; RV32ZVE32F-NEXT: andi a0, a1, 32
+; RV32ZVE32F-NEXT: bnez a0, .LBB88_14
; RV32ZVE32F-NEXT: .LBB88_6: # %else10
-; RV32ZVE32F-NEXT: andi a1, a0, 64
-; RV32ZVE32F-NEXT: bnez a1, .LBB88_15
+; RV32ZVE32F-NEXT: andi a0, a1, 64
+; RV32ZVE32F-NEXT: bnez a0, .LBB88_15
; RV32ZVE32F-NEXT: .LBB88_7: # %else12
-; RV32ZVE32F-NEXT: andi a0, a0, -128
+; RV32ZVE32F-NEXT: andi a0, a1, -128
; RV32ZVE32F-NEXT: bnez a0, .LBB88_16
; RV32ZVE32F-NEXT: .LBB88_8: # %else14
; RV32ZVE32F-NEXT: ret
; RV32ZVE32F-NEXT: .LBB88_9: # %cond.store
-; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a1, v8
-; RV32ZVE32F-NEXT: fsd fa0, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 2
-; RV32ZVE32F-NEXT: beqz a1, .LBB88_2
+; RV32ZVE32F-NEXT: vmv.x.s a0, v8
+; RV32ZVE32F-NEXT: fsd fa0, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 2
+; RV32ZVE32F-NEXT: beqz a0, .LBB88_2
; RV32ZVE32F-NEXT: .LBB88_10: # %cond.store1
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 1
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa1, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 4
-; RV32ZVE32F-NEXT: beqz a1, .LBB88_3
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa1, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 4
+; RV32ZVE32F-NEXT: beqz a0, .LBB88_3
; RV32ZVE32F-NEXT: .LBB88_11: # %cond.store3
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 2
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa2, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 8
-; RV32ZVE32F-NEXT: beqz a1, .LBB88_4
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa2, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 8
+; RV32ZVE32F-NEXT: beqz a0, .LBB88_4
; RV32ZVE32F-NEXT: .LBB88_12: # %cond.store5
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 3
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa3, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 16
-; RV32ZVE32F-NEXT: beqz a1, .LBB88_5
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa3, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 16
+; RV32ZVE32F-NEXT: beqz a0, .LBB88_5
; RV32ZVE32F-NEXT: .LBB88_13: # %cond.store7
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 4
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa4, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 32
-; RV32ZVE32F-NEXT: beqz a1, .LBB88_6
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa4, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 32
+; RV32ZVE32F-NEXT: beqz a0, .LBB88_6
; RV32ZVE32F-NEXT: .LBB88_14: # %cond.store9
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 5
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa5, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 64
-; RV32ZVE32F-NEXT: beqz a1, .LBB88_7
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa5, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 64
+; RV32ZVE32F-NEXT: beqz a0, .LBB88_7
; RV32ZVE32F-NEXT: .LBB88_15: # %cond.store11
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 6
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa6, 0(a1)
-; RV32ZVE32F-NEXT: andi a0, a0, -128
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa6, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, -128
; RV32ZVE32F-NEXT: beqz a0, .LBB88_8
; RV32ZVE32F-NEXT: .LBB88_16: # %cond.store13
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -9926,8 +9925,8 @@ define void @mscatter_baseidx_sext_v8i32_v8f64(<8 x double> %val, ptr %base, <8
; RV64ZVE32F-NEXT: .LBB88_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e32, m2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e32, m1, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e32, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB88_12
; RV64ZVE32F-NEXT: # %bb.5: # %else4
@@ -10018,81 +10017,81 @@ define void @mscatter_baseidx_zext_v8i32_v8f64(<8 x double> %val, ptr %base, <8
; RV32ZVE32F: # %bb.0:
; RV32ZVE32F-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vsll.vi v8, v8, 3
-; RV32ZVE32F-NEXT: vadd.vx v8, v8, a0
; RV32ZVE32F-NEXT: vsetvli zero, zero, e8, mf2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a0, v0
-; RV32ZVE32F-NEXT: andi a1, a0, 1
-; RV32ZVE32F-NEXT: bnez a1, .LBB89_9
+; RV32ZVE32F-NEXT: vmv.x.s a1, v0
+; RV32ZVE32F-NEXT: andi a2, a1, 1
+; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
+; RV32ZVE32F-NEXT: vadd.vx v8, v8, a0
+; RV32ZVE32F-NEXT: bnez a2, .LBB89_9
; RV32ZVE32F-NEXT: # %bb.1: # %else
-; RV32ZVE32F-NEXT: andi a1, a0, 2
-; RV32ZVE32F-NEXT: bnez a1, .LBB89_10
+; RV32ZVE32F-NEXT: andi a0, a1, 2
+; RV32ZVE32F-NEXT: bnez a0, .LBB89_10
; RV32ZVE32F-NEXT: .LBB89_2: # %else2
-; RV32ZVE32F-NEXT: andi a1, a0, 4
-; RV32ZVE32F-NEXT: bnez a1, .LBB89_11
+; RV32ZVE32F-NEXT: andi a0, a1, 4
+; RV32ZVE32F-NEXT: bnez a0, .LBB89_11
; RV32ZVE32F-NEXT: .LBB89_3: # %else4
-; RV32ZVE32F-NEXT: andi a1, a0, 8
-; RV32ZVE32F-NEXT: bnez a1, .LBB89_12
+; RV32ZVE32F-NEXT: andi a0, a1, 8
+; RV32ZVE32F-NEXT: bnez a0, .LBB89_12
; RV32ZVE32F-NEXT: .LBB89_4: # %else6
-; RV32ZVE32F-NEXT: andi a1, a0, 16
-; RV32ZVE32F-NEXT: bnez a1, .LBB89_13
+; RV32ZVE32F-NEXT: andi a0, a1, 16
+; RV32ZVE32F-NEXT: bnez a0, .LBB89_13
; RV32ZVE32F-NEXT: .LBB89_5: # %else8
-; RV32ZVE32F-NEXT: andi a1, a0, 32
-; RV32ZVE32F-NEXT: bnez a1, .LBB89_14
+; RV32ZVE32F-NEXT: andi a0, a1, 32
+; RV32ZVE32F-NEXT: bnez a0, .LBB89_14
; RV32ZVE32F-NEXT: .LBB89_6: # %else10
-; RV32ZVE32F-NEXT: andi a1, a0, 64
-; RV32ZVE32F-NEXT: bnez a1, .LBB89_15
+; RV32ZVE32F-NEXT: andi a0, a1, 64
+; RV32ZVE32F-NEXT: bnez a0, .LBB89_15
; RV32ZVE32F-NEXT: .LBB89_7: # %else12
-; RV32ZVE32F-NEXT: andi a0, a0, -128
+; RV32ZVE32F-NEXT: andi a0, a1, -128
; RV32ZVE32F-NEXT: bnez a0, .LBB89_16
; RV32ZVE32F-NEXT: .LBB89_8: # %else14
; RV32ZVE32F-NEXT: ret
; RV32ZVE32F-NEXT: .LBB89_9: # %cond.store
-; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a1, v8
-; RV32ZVE32F-NEXT: fsd fa0, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 2
-; RV32ZVE32F-NEXT: beqz a1, .LBB89_2
+; RV32ZVE32F-NEXT: vmv.x.s a0, v8
+; RV32ZVE32F-NEXT: fsd fa0, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 2
+; RV32ZVE32F-NEXT: beqz a0, .LBB89_2
; RV32ZVE32F-NEXT: .LBB89_10: # %cond.store1
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 1
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa1, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 4
-; RV32ZVE32F-NEXT: beqz a1, .LBB89_3
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa1, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 4
+; RV32ZVE32F-NEXT: beqz a0, .LBB89_3
; RV32ZVE32F-NEXT: .LBB89_11: # %cond.store3
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 2
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa2, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 8
-; RV32ZVE32F-NEXT: beqz a1, .LBB89_4
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa2, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 8
+; RV32ZVE32F-NEXT: beqz a0, .LBB89_4
; RV32ZVE32F-NEXT: .LBB89_12: # %cond.store5
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 3
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa3, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 16
-; RV32ZVE32F-NEXT: beqz a1, .LBB89_5
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa3, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 16
+; RV32ZVE32F-NEXT: beqz a0, .LBB89_5
; RV32ZVE32F-NEXT: .LBB89_13: # %cond.store7
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 4
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa4, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 32
-; RV32ZVE32F-NEXT: beqz a1, .LBB89_6
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa4, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 32
+; RV32ZVE32F-NEXT: beqz a0, .LBB89_6
; RV32ZVE32F-NEXT: .LBB89_14: # %cond.store9
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 5
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa5, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 64
-; RV32ZVE32F-NEXT: beqz a1, .LBB89_7
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa5, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 64
+; RV32ZVE32F-NEXT: beqz a0, .LBB89_7
; RV32ZVE32F-NEXT: .LBB89_15: # %cond.store11
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 6
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa6, 0(a1)
-; RV32ZVE32F-NEXT: andi a0, a0, -128
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa6, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, -128
; RV32ZVE32F-NEXT: beqz a0, .LBB89_8
; RV32ZVE32F-NEXT: .LBB89_16: # %cond.store13
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -10128,8 +10127,8 @@ define void @mscatter_baseidx_zext_v8i32_v8f64(<8 x double> %val, ptr %base, <8
; RV64ZVE32F-NEXT: .LBB89_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e32, m2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v8, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e32, m1, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e32, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v8, v8, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB89_12
; RV64ZVE32F-NEXT: # %bb.5: # %else4
@@ -10227,95 +10226,95 @@ define void @mscatter_baseidx_v8f64(<8 x double> %val, ptr %base, <8 x i64> %idx
; RV32ZVE32F-NEXT: lw a2, 56(a1)
; RV32ZVE32F-NEXT: lw a3, 48(a1)
; RV32ZVE32F-NEXT: lw a4, 40(a1)
-; RV32ZVE32F-NEXT: lw a5, 32(a1)
-; RV32ZVE32F-NEXT: lw a6, 24(a1)
-; RV32ZVE32F-NEXT: lw a7, 16(a1)
-; RV32ZVE32F-NEXT: lw t0, 8(a1)
+; RV32ZVE32F-NEXT: lw a5, 8(a1)
; RV32ZVE32F-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vlse32.v v8, (a1), zero
-; RV32ZVE32F-NEXT: vslide1down.vx v8, v8, t0
-; RV32ZVE32F-NEXT: vslide1down.vx v8, v8, a7
-; RV32ZVE32F-NEXT: vslide1down.vx v8, v8, a6
+; RV32ZVE32F-NEXT: lw a6, 16(a1)
+; RV32ZVE32F-NEXT: lw a7, 24(a1)
+; RV32ZVE32F-NEXT: lw a1, 32(a1)
; RV32ZVE32F-NEXT: vslide1down.vx v8, v8, a5
+; RV32ZVE32F-NEXT: vslide1down.vx v8, v8, a6
+; RV32ZVE32F-NEXT: vslide1down.vx v8, v8, a7
+; RV32ZVE32F-NEXT: vslide1down.vx v8, v8, a1
; RV32ZVE32F-NEXT: vslide1down.vx v8, v8, a4
; RV32ZVE32F-NEXT: vslide1down.vx v8, v8, a3
; RV32ZVE32F-NEXT: vslide1down.vx v8, v8, a2
; RV32ZVE32F-NEXT: vsll.vi v8, v8, 3
-; RV32ZVE32F-NEXT: vadd.vx v8, v8, a0
; RV32ZVE32F-NEXT: vsetvli zero, zero, e8, mf2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a0, v0
-; RV32ZVE32F-NEXT: andi a1, a0, 1
-; RV32ZVE32F-NEXT: bnez a1, .LBB90_9
+; RV32ZVE32F-NEXT: vmv.x.s a1, v0
+; RV32ZVE32F-NEXT: andi a2, a1, 1
+; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
+; RV32ZVE32F-NEXT: vadd.vx v8, v8, a0
+; RV32ZVE32F-NEXT: bnez a2, .LBB90_9
; RV32ZVE32F-NEXT: # %bb.1: # %else
-; RV32ZVE32F-NEXT: andi a1, a0, 2
-; RV32ZVE32F-NEXT: bnez a1, .LBB90_10
+; RV32ZVE32F-NEXT: andi a0, a1, 2
+; RV32ZVE32F-NEXT: bnez a0, .LBB90_10
; RV32ZVE32F-NEXT: .LBB90_2: # %else2
-; RV32ZVE32F-NEXT: andi a1, a0, 4
-; RV32ZVE32F-NEXT: bnez a1, .LBB90_11
+; RV32ZVE32F-NEXT: andi a0, a1, 4
+; RV32ZVE32F-NEXT: bnez a0, .LBB90_11
; RV32ZVE32F-NEXT: .LBB90_3: # %else4
-; RV32ZVE32F-NEXT: andi a1, a0, 8
-; RV32ZVE32F-NEXT: bnez a1, .LBB90_12
+; RV32ZVE32F-NEXT: andi a0, a1, 8
+; RV32ZVE32F-NEXT: bnez a0, .LBB90_12
; RV32ZVE32F-NEXT: .LBB90_4: # %else6
-; RV32ZVE32F-NEXT: andi a1, a0, 16
-; RV32ZVE32F-NEXT: bnez a1, .LBB90_13
+; RV32ZVE32F-NEXT: andi a0, a1, 16
+; RV32ZVE32F-NEXT: bnez a0, .LBB90_13
; RV32ZVE32F-NEXT: .LBB90_5: # %else8
-; RV32ZVE32F-NEXT: andi a1, a0, 32
-; RV32ZVE32F-NEXT: bnez a1, .LBB90_14
+; RV32ZVE32F-NEXT: andi a0, a1, 32
+; RV32ZVE32F-NEXT: bnez a0, .LBB90_14
; RV32ZVE32F-NEXT: .LBB90_6: # %else10
-; RV32ZVE32F-NEXT: andi a1, a0, 64
-; RV32ZVE32F-NEXT: bnez a1, .LBB90_15
+; RV32ZVE32F-NEXT: andi a0, a1, 64
+; RV32ZVE32F-NEXT: bnez a0, .LBB90_15
; RV32ZVE32F-NEXT: .LBB90_7: # %else12
-; RV32ZVE32F-NEXT: andi a0, a0, -128
+; RV32ZVE32F-NEXT: andi a0, a1, -128
; RV32ZVE32F-NEXT: bnez a0, .LBB90_16
; RV32ZVE32F-NEXT: .LBB90_8: # %else14
; RV32ZVE32F-NEXT: ret
; RV32ZVE32F-NEXT: .LBB90_9: # %cond.store
-; RV32ZVE32F-NEXT: vsetvli zero, zero, e32, m2, ta, ma
-; RV32ZVE32F-NEXT: vmv.x.s a1, v8
-; RV32ZVE32F-NEXT: fsd fa0, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 2
-; RV32ZVE32F-NEXT: beqz a1, .LBB90_2
+; RV32ZVE32F-NEXT: vmv.x.s a0, v8
+; RV32ZVE32F-NEXT: fsd fa0, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 2
+; RV32ZVE32F-NEXT: beqz a0, .LBB90_2
; RV32ZVE32F-NEXT: .LBB90_10: # %cond.store1
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 1
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa1, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 4
-; RV32ZVE32F-NEXT: beqz a1, .LBB90_3
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa1, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 4
+; RV32ZVE32F-NEXT: beqz a0, .LBB90_3
; RV32ZVE32F-NEXT: .LBB90_11: # %cond.store3
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 2
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa2, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 8
-; RV32ZVE32F-NEXT: beqz a1, .LBB90_4
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa2, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 8
+; RV32ZVE32F-NEXT: beqz a0, .LBB90_4
; RV32ZVE32F-NEXT: .LBB90_12: # %cond.store5
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 3
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa3, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 16
-; RV32ZVE32F-NEXT: beqz a1, .LBB90_5
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa3, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 16
+; RV32ZVE32F-NEXT: beqz a0, .LBB90_5
; RV32ZVE32F-NEXT: .LBB90_13: # %cond.store7
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 4
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa4, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 32
-; RV32ZVE32F-NEXT: beqz a1, .LBB90_6
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa4, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 32
+; RV32ZVE32F-NEXT: beqz a0, .LBB90_6
; RV32ZVE32F-NEXT: .LBB90_14: # %cond.store9
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 5
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa5, 0(a1)
-; RV32ZVE32F-NEXT: andi a1, a0, 64
-; RV32ZVE32F-NEXT: beqz a1, .LBB90_7
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa5, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, 64
+; RV32ZVE32F-NEXT: beqz a0, .LBB90_7
; RV32ZVE32F-NEXT: .LBB90_15: # %cond.store11
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
; RV32ZVE32F-NEXT: vslidedown.vi v10, v8, 6
-; RV32ZVE32F-NEXT: vmv.x.s a1, v10
-; RV32ZVE32F-NEXT: fsd fa6, 0(a1)
-; RV32ZVE32F-NEXT: andi a0, a0, -128
+; RV32ZVE32F-NEXT: vmv.x.s a0, v10
+; RV32ZVE32F-NEXT: fsd fa6, 0(a0)
+; RV32ZVE32F-NEXT: andi a0, a1, -128
; RV32ZVE32F-NEXT: beqz a0, .LBB90_8
; RV32ZVE32F-NEXT: .LBB90_16: # %cond.store13
; RV32ZVE32F-NEXT: vsetivli zero, 1, e32, m2, ta, ma
@@ -10457,8 +10456,8 @@ define void @mscatter_baseidx_v16i8(<16 x i8> %val, ptr %base, <16 x i8> %idxs,
; RV64ZVE32F-NEXT: .LBB91_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v9, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v11, v9, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB91_25
; RV64ZVE32F-NEXT: # %bb.5: # %else4
@@ -10474,8 +10473,8 @@ define void @mscatter_baseidx_v16i8(<16 x i8> %val, ptr %base, <16 x i8> %idxs,
; RV64ZVE32F-NEXT: vslidedown.vi v11, v8, 4
; RV64ZVE32F-NEXT: vse8.v v11, (a2)
; RV64ZVE32F-NEXT: .LBB91_8: # %else8
-; RV64ZVE32F-NEXT: vsetivli zero, 8, e8, m1, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 32
+; RV64ZVE32F-NEXT: vsetivli zero, 8, e8, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v9, 8
; RV64ZVE32F-NEXT: beqz a2, .LBB91_10
; RV64ZVE32F-NEXT: # %bb.9: # %cond.store9
@@ -10487,8 +10486,8 @@ define void @mscatter_baseidx_v16i8(<16 x i8> %val, ptr %base, <16 x i8> %idxs,
; RV64ZVE32F-NEXT: vslidedown.vi v11, v8, 5
; RV64ZVE32F-NEXT: vse8.v v11, (a2)
; RV64ZVE32F-NEXT: .LBB91_10: # %else10
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v10, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB91_27
; RV64ZVE32F-NEXT: # %bb.11: # %else12
@@ -10511,8 +10510,8 @@ define void @mscatter_baseidx_v16i8(<16 x i8> %val, ptr %base, <16 x i8> %idxs,
; RV64ZVE32F-NEXT: .LBB91_15: # %else18
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v9, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 1024
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v9, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB91_30
; RV64ZVE32F-NEXT: # %bb.16: # %else20
@@ -10533,8 +10532,8 @@ define void @mscatter_baseidx_v16i8(<16 x i8> %val, ptr %base, <16 x i8> %idxs,
; RV64ZVE32F-NEXT: vslidedown.vi v9, v8, 13
; RV64ZVE32F-NEXT: vse8.v v9, (a2)
; RV64ZVE32F-NEXT: .LBB91_20: # %else26
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: slli a2, a1, 49
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v9, v10, 2
; RV64ZVE32F-NEXT: bgez a2, .LBB91_22
; RV64ZVE32F-NEXT: # %bb.21: # %cond.store27
@@ -10656,11 +10655,11 @@ define void @mscatter_baseidx_v32i8(<32 x i8> %val, ptr %base, <32 x i8> %idxs,
; RV64-NEXT: vsetivli zero, 16, e8, m2, ta, ma
; RV64-NEXT: vslidedown.vi v8, v8, 16
; RV64-NEXT: vslidedown.vi v10, v10, 16
-; RV64-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV64-NEXT: vsext.vf8 v16, v10
; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64-NEXT: vslidedown.vi v0, v0, 2
-; RV64-NEXT: vsetivli zero, 16, e8, m1, ta, ma
+; RV64-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; RV64-NEXT: vsext.vf8 v16, v10
+; RV64-NEXT: vsetvli zero, zero, e8, m1, ta, ma
; RV64-NEXT: vsoxei64.v v8, (a0), v16, v0.t
; RV64-NEXT: ret
;
@@ -10689,8 +10688,8 @@ define void @mscatter_baseidx_v32i8(<32 x i8> %val, ptr %base, <32 x i8> %idxs,
; RV64ZVE32F-NEXT: .LBB92_4: # %else2
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v13, v10, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 4
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v12, v10, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB92_49
; RV64ZVE32F-NEXT: # %bb.5: # %else4
@@ -10706,8 +10705,8 @@ define void @mscatter_baseidx_v32i8(<32 x i8> %val, ptr %base, <32 x i8> %idxs,
; RV64ZVE32F-NEXT: vslidedown.vi v12, v8, 4
; RV64ZVE32F-NEXT: vse8.v v12, (a2)
; RV64ZVE32F-NEXT: .LBB92_8: # %else8
-; RV64ZVE32F-NEXT: vsetivli zero, 8, e8, m1, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 32
+; RV64ZVE32F-NEXT: vsetivli zero, 8, e8, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v12, v10, 8
; RV64ZVE32F-NEXT: beqz a2, .LBB92_10
; RV64ZVE32F-NEXT: # %bb.9: # %cond.store9
@@ -10719,8 +10718,8 @@ define void @mscatter_baseidx_v32i8(<32 x i8> %val, ptr %base, <32 x i8> %idxs,
; RV64ZVE32F-NEXT: vslidedown.vi v14, v8, 5
; RV64ZVE32F-NEXT: vse8.v v14, (a2)
; RV64ZVE32F-NEXT: .LBB92_10: # %else10
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 64
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v13, v13, 2
; RV64ZVE32F-NEXT: bnez a2, .LBB92_51
; RV64ZVE32F-NEXT: # %bb.11: # %else12
@@ -10743,8 +10742,8 @@ define void @mscatter_baseidx_v32i8(<32 x i8> %val, ptr %base, <32 x i8> %idxs,
; RV64ZVE32F-NEXT: .LBB92_15: # %else18
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v13, v12, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: andi a2, a1, 1024
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v12, v12, 2
; RV64ZVE32F-NEXT: beqz a2, .LBB92_17
; RV64ZVE32F-NEXT: # %bb.16: # %cond.store19
@@ -10765,8 +10764,8 @@ define void @mscatter_baseidx_v32i8(<32 x i8> %val, ptr %base, <32 x i8> %idxs,
; RV64ZVE32F-NEXT: vslidedown.vi v12, v8, 11
; RV64ZVE32F-NEXT: vse8.v v12, (a2)
; RV64ZVE32F-NEXT: .LBB92_19: # %else22
-; RV64ZVE32F-NEXT: vsetivli zero, 16, e8, m2, ta, ma
; RV64ZVE32F-NEXT: slli a2, a1, 51
+; RV64ZVE32F-NEXT: vsetivli zero, 16, e8, m2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v10, 16
; RV64ZVE32F-NEXT: bgez a2, .LBB92_21
; RV64ZVE32F-NEXT: # %bb.20: # %cond.store23
@@ -10787,8 +10786,8 @@ define void @mscatter_baseidx_v32i8(<32 x i8> %val, ptr %base, <32 x i8> %idxs,
; RV64ZVE32F-NEXT: vslidedown.vi v11, v8, 13
; RV64ZVE32F-NEXT: vse8.v v11, (a2)
; RV64ZVE32F-NEXT: .LBB92_23: # %else26
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: slli a2, a1, 49
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v11, v13, 2
; RV64ZVE32F-NEXT: bltz a2, .LBB92_54
; RV64ZVE32F-NEXT: # %bb.24: # %else28
@@ -10812,8 +10811,8 @@ define void @mscatter_baseidx_v32i8(<32 x i8> %val, ptr %base, <32 x i8> %idxs,
; RV64ZVE32F-NEXT: .LBB92_28: # %else34
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v11, v10, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: slli a2, a1, 45
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v12, v10, 2
; RV64ZVE32F-NEXT: bltz a2, .LBB92_57
; RV64ZVE32F-NEXT: # %bb.29: # %else36
@@ -10830,8 +10829,8 @@ define void @mscatter_baseidx_v32i8(<32 x i8> %val, ptr %base, <32 x i8> %idxs,
; RV64ZVE32F-NEXT: vsetivli zero, 1, e8, m1, ta, ma
; RV64ZVE32F-NEXT: vse8.v v12, (a2)
; RV64ZVE32F-NEXT: .LBB92_32: # %else40
-; RV64ZVE32F-NEXT: vsetivli zero, 8, e8, m1, ta, ma
; RV64ZVE32F-NEXT: slli a2, a1, 42
+; RV64ZVE32F-NEXT: vsetivli zero, 8, e8, m1, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v10, 8
; RV64ZVE32F-NEXT: bgez a2, .LBB92_34
; RV64ZVE32F-NEXT: # %bb.33: # %cond.store41
@@ -10844,8 +10843,8 @@ define void @mscatter_baseidx_v32i8(<32 x i8> %val, ptr %base, <32 x i8> %idxs,
; RV64ZVE32F-NEXT: vsetivli zero, 1, e8, m1, ta, ma
; RV64ZVE32F-NEXT: vse8.v v12, (a2)
; RV64ZVE32F-NEXT: .LBB92_34: # %else42
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: slli a2, a1, 41
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v11, v11, 2
; RV64ZVE32F-NEXT: bltz a2, .LBB92_59
; RV64ZVE32F-NEXT: # %bb.35: # %else44
@@ -10869,8 +10868,8 @@ define void @mscatter_baseidx_v32i8(<32 x i8> %val, ptr %base, <32 x i8> %idxs,
; RV64ZVE32F-NEXT: .LBB92_39: # %else50
; RV64ZVE32F-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v11, v10, 4
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: slli a2, a1, 37
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v10, 2
; RV64ZVE32F-NEXT: bltz a2, .LBB92_62
; RV64ZVE32F-NEXT: # %bb.40: # %else52
@@ -10892,8 +10891,8 @@ define void @mscatter_baseidx_v32i8(<32 x i8> %val, ptr %base, <32 x i8> %idxs,
; RV64ZVE32F-NEXT: vsetivli zero, 1, e8, m1, ta, ma
; RV64ZVE32F-NEXT: vse8.v v12, (a2)
; RV64ZVE32F-NEXT: .LBB92_44: # %else58
-; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: slli a2, a1, 33
+; RV64ZVE32F-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64ZVE32F-NEXT: vslidedown.vi v10, v11, 2
; RV64ZVE32F-NEXT: bgez a2, .LBB92_46
; RV64ZVE32F-NEXT: # %bb.45: # %cond.store59
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-masked-store-int.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-masked-store-int.ll
index b3011d0f01cab..86c28247e97ef 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-masked-store-int.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-masked-store-int.ll
@@ -401,54 +401,41 @@ define void @masked_store_v32i64(ptr %val_ptr, ptr %a, ptr %m_ptr) nounwind {
; RV32: # %bb.0:
; RV32-NEXT: addi sp, sp, -16
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: li a4, 18
-; RV32-NEXT: mul a3, a3, a4
+; RV32-NEXT: slli a3, a3, 4
; RV32-NEXT: sub sp, sp, a3
; RV32-NEXT: addi a3, a2, 128
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; RV32-NEXT: vle64.v v24, (a2)
; RV32-NEXT: vle64.v v8, (a3)
-; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: slli a4, a3, 3
-; RV32-NEXT: add a3, a4, a3
-; RV32-NEXT: add a3, sp, a3
-; RV32-NEXT: addi a3, a3, 16
-; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: vle64.v v0, (a2)
-; RV32-NEXT: li a2, 32
-; RV32-NEXT: vsetvli zero, a2, e32, m8, ta, ma
-; RV32-NEXT: vmv.v.i v24, 0
-; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vmseq.vv v8, v0, v24
; RV32-NEXT: csrr a2, vlenb
; RV32-NEXT: slli a2, a2, 3
; RV32-NEXT: add a2, sp, a2
; RV32-NEXT: addi a2, a2, 16
-; RV32-NEXT: vs1r.v v8, (a2) # Unknown-size Folded Spill
+; RV32-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
+; RV32-NEXT: li a2, 32
+; RV32-NEXT: vsetvli zero, a2, e32, m8, ta, ma
+; RV32-NEXT: vmv.v.i v8, 0
+; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; RV32-NEXT: vmseq.vv v7, v24, v8
; RV32-NEXT: addi a2, a0, 128
-; RV32-NEXT: vle64.v v8, (a2)
+; RV32-NEXT: vle64.v v24, (a2)
; RV32-NEXT: vle64.v v16, (a0)
; RV32-NEXT: addi a0, sp, 16
; RV32-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a2, a0, 3
-; RV32-NEXT: add a0, a2, a0
+; RV32-NEXT: slli a0, a0, 3
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 16
; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; RV32-NEXT: vmseq.vv v0, v16, v24
+; RV32-NEXT: vmseq.vv v0, v16, v8
; RV32-NEXT: addi a0, a1, 128
-; RV32-NEXT: vse64.v v8, (a0), v0.t
-; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 3
-; RV32-NEXT: add a0, sp, a0
-; RV32-NEXT: addi a0, a0, 16
-; RV32-NEXT: vl1r.v v0, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vse64.v v24, (a0), v0.t
+; RV32-NEXT: vmv1r.v v0, v7
; RV32-NEXT: addi a0, sp, 16
; RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; RV32-NEXT: vse64.v v8, (a1), v0.t
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: li a1, 18
-; RV32-NEXT: mul a0, a0, a1
+; RV32-NEXT: slli a0, a0, 4
; RV32-NEXT: add sp, sp, a0
; RV32-NEXT: addi sp, sp, 16
; RV32-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-nearbyint-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-nearbyint-vp.ll
index 19f3d3ce19fa4..93b4f7d2a9c9f 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-nearbyint-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-nearbyint-vp.ll
@@ -19,9 +19,9 @@ define <2 x half> @vp_nearbyint_v2f16(<2 x half> %va, <2 x i1> %m, i32 zeroext %
; CHECK-NEXT: vsetvli zero, zero, e16, mf4, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v9, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v9, v9, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e16, mf4, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <2 x half> @llvm.vp.nearbyint.v2f16(<2 x half> %va, <2 x i1> %m, i32 %evl)
ret <2 x half> %v
@@ -38,9 +38,9 @@ define <2 x half> @vp_nearbyint_v2f16_unmasked(<2 x half> %va, i32 zeroext %evl)
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v9, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v9, v9, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e16, mf4, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <2 x half> @llvm.vp.nearbyint.v2f16(<2 x half> %va, <2 x i1> splat (i1 true), i32 %evl)
ret <2 x half> %v
@@ -61,9 +61,9 @@ define <4 x half> @vp_nearbyint_v4f16(<4 x half> %va, <4 x i1> %m, i32 zeroext %
; CHECK-NEXT: vsetvli zero, zero, e16, mf2, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v9, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v9, v9, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e16, mf2, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <4 x half> @llvm.vp.nearbyint.v4f16(<4 x half> %va, <4 x i1> %m, i32 %evl)
ret <4 x half> %v
@@ -80,9 +80,9 @@ define <4 x half> @vp_nearbyint_v4f16_unmasked(<4 x half> %va, i32 zeroext %evl)
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v9, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v9, v9, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e16, mf2, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <4 x half> @llvm.vp.nearbyint.v4f16(<4 x half> %va, <4 x i1> splat (i1 true), i32 %evl)
ret <4 x half> %v
@@ -103,9 +103,9 @@ define <8 x half> @vp_nearbyint_v8f16(<8 x half> %va, <8 x i1> %m, i32 zeroext %
; CHECK-NEXT: vsetvli zero, zero, e16, m1, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v9, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v9, v9, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e16, m1, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <8 x half> @llvm.vp.nearbyint.v8f16(<8 x half> %va, <8 x i1> %m, i32 %evl)
ret <8 x half> %v
@@ -122,9 +122,9 @@ define <8 x half> @vp_nearbyint_v8f16_unmasked(<8 x half> %va, i32 zeroext %evl)
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v9, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v9, v9, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e16, m1, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <8 x half> @llvm.vp.nearbyint.v8f16(<8 x half> %va, <8 x i1> splat (i1 true), i32 %evl)
ret <8 x half> %v
@@ -135,21 +135,21 @@ declare <16 x half> @llvm.vp.nearbyint.v16f16(<16 x half>, <16 x i1>, i32)
define <16 x half> @vp_nearbyint_v16f16(<16 x half> %va, <16 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_nearbyint_v16f16:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v10, v0
; CHECK-NEXT: lui a1, %hi(.LCPI6_0)
; CHECK-NEXT: flh fa5, %lo(.LCPI6_0)(a1)
+; CHECK-NEXT: vmv1r.v v10, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, ma
; CHECK-NEXT: vfabs.v v12, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, mu
; CHECK-NEXT: vmflt.vf v10, v12, fa5, v0.t
; CHECK-NEXT: frflags a0
-; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v12, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <16 x half> @llvm.vp.nearbyint.v16f16(<16 x half> %va, <16 x i1> %m, i32 %evl)
ret <16 x half> %v
@@ -166,9 +166,9 @@ define <16 x half> @vp_nearbyint_v16f16_unmasked(<16 x half> %va, i32 zeroext %e
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v10, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v10, v10, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v10, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <16 x half> @llvm.vp.nearbyint.v16f16(<16 x half> %va, <16 x i1> splat (i1 true), i32 %evl)
ret <16 x half> %v
@@ -189,9 +189,9 @@ define <2 x float> @vp_nearbyint_v2f32(<2 x float> %va, <2 x i1> %m, i32 zeroext
; CHECK-NEXT: vsetvli zero, zero, e32, mf2, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v9, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v9, v9, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e32, mf2, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <2 x float> @llvm.vp.nearbyint.v2f32(<2 x float> %va, <2 x i1> %m, i32 %evl)
ret <2 x float> %v
@@ -208,9 +208,9 @@ define <2 x float> @vp_nearbyint_v2f32_unmasked(<2 x float> %va, i32 zeroext %ev
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v9, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v9, v9, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e32, mf2, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <2 x float> @llvm.vp.nearbyint.v2f32(<2 x float> %va, <2 x i1> splat (i1 true), i32 %evl)
ret <2 x float> %v
@@ -231,9 +231,9 @@ define <4 x float> @vp_nearbyint_v4f32(<4 x float> %va, <4 x i1> %m, i32 zeroext
; CHECK-NEXT: vsetvli zero, zero, e32, m1, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v9, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v9, v9, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e32, m1, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <4 x float> @llvm.vp.nearbyint.v4f32(<4 x float> %va, <4 x i1> %m, i32 %evl)
ret <4 x float> %v
@@ -250,9 +250,9 @@ define <4 x float> @vp_nearbyint_v4f32_unmasked(<4 x float> %va, i32 zeroext %ev
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v9, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v9, v9, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e32, m1, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <4 x float> @llvm.vp.nearbyint.v4f32(<4 x float> %va, <4 x i1> splat (i1 true), i32 %evl)
ret <4 x float> %v
@@ -271,13 +271,13 @@ define <8 x float> @vp_nearbyint_v8f32(<8 x float> %va, <8 x i1> %m, i32 zeroext
; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, mu
; CHECK-NEXT: vmflt.vf v10, v12, fa5, v0.t
; CHECK-NEXT: frflags a0
-; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v12, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <8 x float> @llvm.vp.nearbyint.v8f32(<8 x float> %va, <8 x i1> %m, i32 %evl)
ret <8 x float> %v
@@ -294,9 +294,9 @@ define <8 x float> @vp_nearbyint_v8f32_unmasked(<8 x float> %va, i32 zeroext %ev
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v10, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v10, v10, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v10, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <8 x float> @llvm.vp.nearbyint.v8f32(<8 x float> %va, <8 x i1> splat (i1 true), i32 %evl)
ret <8 x float> %v
@@ -315,13 +315,13 @@ define <16 x float> @vp_nearbyint_v16f32(<16 x float> %va, <16 x i1> %m, i32 zer
; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, mu
; CHECK-NEXT: vmflt.vf v12, v16, fa5, v0.t
; CHECK-NEXT: frflags a0
-; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v16, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <16 x float> @llvm.vp.nearbyint.v16f32(<16 x float> %va, <16 x i1> %m, i32 %evl)
ret <16 x float> %v
@@ -338,9 +338,9 @@ define <16 x float> @vp_nearbyint_v16f32_unmasked(<16 x float> %va, i32 zeroext
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v12, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <16 x float> @llvm.vp.nearbyint.v16f32(<16 x float> %va, <16 x i1> splat (i1 true), i32 %evl)
ret <16 x float> %v
@@ -361,9 +361,9 @@ define <2 x double> @vp_nearbyint_v2f64(<2 x double> %va, <2 x i1> %m, i32 zeroe
; CHECK-NEXT: vsetvli zero, zero, e64, m1, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v9, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v9, v9, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e64, m1, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <2 x double> @llvm.vp.nearbyint.v2f64(<2 x double> %va, <2 x i1> %m, i32 %evl)
ret <2 x double> %v
@@ -380,9 +380,9 @@ define <2 x double> @vp_nearbyint_v2f64_unmasked(<2 x double> %va, i32 zeroext %
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v9, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v9, v9, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e64, m1, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <2 x double> @llvm.vp.nearbyint.v2f64(<2 x double> %va, <2 x i1> splat (i1 true), i32 %evl)
ret <2 x double> %v
@@ -393,21 +393,21 @@ declare <4 x double> @llvm.vp.nearbyint.v4f64(<4 x double>, <4 x i1>, i32)
define <4 x double> @vp_nearbyint_v4f64(<4 x double> %va, <4 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_nearbyint_v4f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v10, v0
; CHECK-NEXT: lui a1, %hi(.LCPI18_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI18_0)(a1)
+; CHECK-NEXT: vmv1r.v v10, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, ma
; CHECK-NEXT: vfabs.v v12, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, mu
; CHECK-NEXT: vmflt.vf v10, v12, fa5, v0.t
; CHECK-NEXT: frflags a0
-; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v12, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <4 x double> @llvm.vp.nearbyint.v4f64(<4 x double> %va, <4 x i1> %m, i32 %evl)
ret <4 x double> %v
@@ -424,9 +424,9 @@ define <4 x double> @vp_nearbyint_v4f64_unmasked(<4 x double> %va, i32 zeroext %
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v10, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v10, v10, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v10, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <4 x double> @llvm.vp.nearbyint.v4f64(<4 x double> %va, <4 x i1> splat (i1 true), i32 %evl)
ret <4 x double> %v
@@ -437,21 +437,21 @@ declare <8 x double> @llvm.vp.nearbyint.v8f64(<8 x double>, <8 x i1>, i32)
define <8 x double> @vp_nearbyint_v8f64(<8 x double> %va, <8 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_nearbyint_v8f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v12, v0
; CHECK-NEXT: lui a1, %hi(.LCPI20_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI20_0)(a1)
+; CHECK-NEXT: vmv1r.v v12, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, ma
; CHECK-NEXT: vfabs.v v16, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, mu
; CHECK-NEXT: vmflt.vf v12, v16, fa5, v0.t
; CHECK-NEXT: frflags a0
-; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v16, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <8 x double> @llvm.vp.nearbyint.v8f64(<8 x double> %va, <8 x i1> %m, i32 %evl)
ret <8 x double> %v
@@ -468,9 +468,9 @@ define <8 x double> @vp_nearbyint_v8f64_unmasked(<8 x double> %va, i32 zeroext %
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v12, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <8 x double> @llvm.vp.nearbyint.v8f64(<8 x double> %va, <8 x i1> splat (i1 true), i32 %evl)
ret <8 x double> %v
@@ -481,21 +481,21 @@ declare <15 x double> @llvm.vp.nearbyint.v15f64(<15 x double>, <15 x i1>, i32)
define <15 x double> @vp_nearbyint_v15f64(<15 x double> %va, <15 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_nearbyint_v15f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: lui a1, %hi(.LCPI22_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI22_0)(a1)
+; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
; CHECK-NEXT: vmflt.vf v16, v24, fa5, v0.t
; CHECK-NEXT: frflags a0
-; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v24, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <15 x double> @llvm.vp.nearbyint.v15f64(<15 x double> %va, <15 x i1> %m, i32 %evl)
ret <15 x double> %v
@@ -512,9 +512,9 @@ define <15 x double> @vp_nearbyint_v15f64_unmasked(<15 x double> %va, i32 zeroex
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v16, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <15 x double> @llvm.vp.nearbyint.v15f64(<15 x double> %va, <15 x i1> splat (i1 true), i32 %evl)
ret <15 x double> %v
@@ -525,21 +525,21 @@ declare <16 x double> @llvm.vp.nearbyint.v16f64(<16 x double>, <16 x i1>, i32)
define <16 x double> @vp_nearbyint_v16f64(<16 x double> %va, <16 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_nearbyint_v16f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: lui a1, %hi(.LCPI24_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI24_0)(a1)
+; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
; CHECK-NEXT: vmflt.vf v16, v24, fa5, v0.t
; CHECK-NEXT: frflags a0
-; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v24, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <16 x double> @llvm.vp.nearbyint.v16f64(<16 x double> %va, <16 x i1> %m, i32 %evl)
ret <16 x double> %v
@@ -556,9 +556,9 @@ define <16 x double> @vp_nearbyint_v16f64_unmasked(<16 x double> %va, i32 zeroex
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v16, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <16 x double> @llvm.vp.nearbyint.v16f64(<16 x double> %va, <16 x i1> splat (i1 true), i32 %evl)
ret <16 x double> %v
@@ -569,17 +569,9 @@ declare <32 x double> @llvm.vp.nearbyint.v32f64(<32 x double>, <32 x i1>, i32)
define <32 x double> @vp_nearbyint_v32f64(<32 x double> %va, <32 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_nearbyint_v32f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: addi sp, sp, -16
-; CHECK-NEXT: .cfi_def_cfa_offset 16
-; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a1, a1, 3
-; CHECK-NEXT: sub sp, sp, a1
-; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
-; CHECK-NEXT: vmv1r.v v25, v0
-; CHECK-NEXT: addi a1, sp, 16
-; CHECK-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
-; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
+; CHECK-NEXT: vmv1r.v v6, v0
; CHECK-NEXT: li a2, 16
+; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; CHECK-NEXT: vslidedown.vi v7, v0, 2
; CHECK-NEXT: mv a1, a0
; CHECK-NEXT: bltu a0, a2, .LBB26_2
@@ -588,43 +580,36 @@ define <32 x double> @vp_nearbyint_v32f64(<32 x double> %va, <32 x i1> %m, i32 z
; CHECK-NEXT: .LBB26_2:
; CHECK-NEXT: lui a2, %hi(.LCPI26_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI26_0)(a2)
+; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v25
-; CHECK-NEXT: vfabs.v v16, v8, v0.t
+; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vmflt.vf v25, v16, fa5, v0.t
+; CHECK-NEXT: vmflt.vf v6, v24, fa5, v0.t
; CHECK-NEXT: frflags a1
+; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v25
-; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
-; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
+; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
+; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
; CHECK-NEXT: fsflags a1
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vfsgnj.vv v8, v16, v8, v0.t
+; CHECK-NEXT: vfsgnj.vv v8, v24, v8, v0.t
; CHECK-NEXT: addi a1, a0, -16
; CHECK-NEXT: sltu a0, a0, a1
; CHECK-NEXT: addi a0, a0, -1
; CHECK-NEXT: and a0, a0, a1
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v7
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vfabs.v v16, v24, v0.t
+; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT: vfabs.v v24, v16, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vmflt.vf v7, v16, fa5, v0.t
+; CHECK-NEXT: vmflt.vf v7, v24, fa5, v0.t
; CHECK-NEXT: frflags a0
-; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v7
-; CHECK-NEXT: vfcvt.x.f.v v16, v24, v0.t
-; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
-; CHECK-NEXT: fsflags a0
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
+; CHECK-NEXT: vfcvt.x.f.v v24, v16, v0.t
+; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vfsgnj.vv v24, v16, v24, v0.t
-; CHECK-NEXT: vmv.v.v v16, v24
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add sp, sp, a0
-; CHECK-NEXT: addi sp, sp, 16
+; CHECK-NEXT: vfsgnj.vv v16, v24, v16, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <32 x double> @llvm.vp.nearbyint.v32f64(<32 x double> %va, <32 x i1> %m, i32 %evl)
ret <32 x double> %v
@@ -660,9 +645,9 @@ define <32 x double> @vp_nearbyint_v32f64_unmasked(<32 x double> %va, i32 zeroex
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v24, v16, v0.t
; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
; CHECK-NEXT: vfsgnj.vv v16, v24, v16, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <32 x double> @llvm.vp.nearbyint.v32f64(<32 x double> %va, <32 x i1> splat (i1 true), i32 %evl)
ret <32 x double> %v
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-reduction-formation.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-reduction-formation.ll
index 03624113a8262..c0bd49cc9c5cb 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-reduction-formation.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-reduction-formation.ll
@@ -163,12 +163,12 @@ define i32 @reduce_sum_16xi32_prefix5(ptr %p) {
; CHECK-NEXT: li a1, -32
; CHECK-NEXT: vsetivli zero, 8, e8, mf2, ta, ma
; CHECK-NEXT: vmv.s.x v0, a1
-; CHECK-NEXT: vmv.v.i v8, -1
-; CHECK-NEXT: vmerge.vim v8, v8, 0, v0
+; CHECK-NEXT: vle32.v v8, (a0)
+; CHECK-NEXT: vmv.v.i v10, -1
+; CHECK-NEXT: vmerge.vim v10, v10, 0, v0
; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, ma
-; CHECK-NEXT: vle32.v v10, (a0)
-; CHECK-NEXT: vsext.vf4 v12, v8
-; CHECK-NEXT: vand.vv v8, v10, v12
+; CHECK-NEXT: vsext.vf4 v12, v10
+; CHECK-NEXT: vand.vv v8, v8, v12
; CHECK-NEXT: vmv.s.x v10, zero
; CHECK-NEXT: vredsum.vs v8, v8, v10
; CHECK-NEXT: vmv.x.s a0, v8
@@ -192,12 +192,12 @@ define i32 @reduce_sum_16xi32_prefix6(ptr %p) {
; CHECK-NEXT: li a1, 192
; CHECK-NEXT: vsetivli zero, 8, e8, mf2, ta, ma
; CHECK-NEXT: vmv.s.x v0, a1
-; CHECK-NEXT: vmv.v.i v8, -1
-; CHECK-NEXT: vmerge.vim v8, v8, 0, v0
+; CHECK-NEXT: vle32.v v8, (a0)
+; CHECK-NEXT: vmv.v.i v10, -1
+; CHECK-NEXT: vmerge.vim v10, v10, 0, v0
; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, ma
-; CHECK-NEXT: vle32.v v10, (a0)
-; CHECK-NEXT: vsext.vf4 v12, v8
-; CHECK-NEXT: vand.vv v8, v10, v12
+; CHECK-NEXT: vsext.vf4 v12, v10
+; CHECK-NEXT: vand.vv v8, v8, v12
; CHECK-NEXT: vmv.s.x v10, zero
; CHECK-NEXT: vredsum.vs v8, v8, v10
; CHECK-NEXT: vmv.x.s a0, v8
@@ -221,10 +221,10 @@ define i32 @reduce_sum_16xi32_prefix7(ptr %p) {
; CHECK-LABEL: reduce_sum_16xi32_prefix7:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetivli zero, 8, e32, m2, ta, ma
-; CHECK-NEXT: vmv.s.x v8, zero
-; CHECK-NEXT: vle32.v v10, (a0)
-; CHECK-NEXT: vslideup.vi v10, v8, 7
-; CHECK-NEXT: vredsum.vs v8, v10, v8
+; CHECK-NEXT: vle32.v v8, (a0)
+; CHECK-NEXT: vmv.s.x v10, zero
+; CHECK-NEXT: vslideup.vi v8, v10, 7
+; CHECK-NEXT: vredsum.vs v8, v8, v10
; CHECK-NEXT: vmv.x.s a0, v8
; CHECK-NEXT: ret
%v = load <16 x i32>, ptr %p, align 256
@@ -248,9 +248,9 @@ define i32 @reduce_sum_16xi32_prefix8(ptr %p) {
; CHECK-LABEL: reduce_sum_16xi32_prefix8:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetivli zero, 8, e32, m2, ta, ma
-; CHECK-NEXT: vmv.s.x v8, zero
-; CHECK-NEXT: vle32.v v10, (a0)
-; CHECK-NEXT: vredsum.vs v8, v10, v8
+; CHECK-NEXT: vle32.v v8, (a0)
+; CHECK-NEXT: vmv.s.x v10, zero
+; CHECK-NEXT: vredsum.vs v8, v8, v10
; CHECK-NEXT: vmv.x.s a0, v8
; CHECK-NEXT: ret
%v = load <16 x i32>, ptr %p, align 256
@@ -535,12 +535,12 @@ define i32 @reduce_xor_16xi32_prefix5(ptr %p) {
; CHECK-NEXT: li a1, -32
; CHECK-NEXT: vsetivli zero, 8, e8, mf2, ta, ma
; CHECK-NEXT: vmv.s.x v0, a1
-; CHECK-NEXT: vmv.v.i v8, -1
-; CHECK-NEXT: vmerge.vim v8, v8, 0, v0
+; CHECK-NEXT: vle32.v v8, (a0)
+; CHECK-NEXT: vmv.v.i v10, -1
+; CHECK-NEXT: vmerge.vim v10, v10, 0, v0
; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, ma
-; CHECK-NEXT: vle32.v v10, (a0)
-; CHECK-NEXT: vsext.vf4 v12, v8
-; CHECK-NEXT: vand.vv v8, v10, v12
+; CHECK-NEXT: vsext.vf4 v12, v10
+; CHECK-NEXT: vand.vv v8, v8, v12
; CHECK-NEXT: vmv.s.x v10, zero
; CHECK-NEXT: vredxor.vs v8, v8, v10
; CHECK-NEXT: vmv.x.s a0, v8
@@ -576,17 +576,17 @@ define i32 @reduce_and_16xi32_prefix2(ptr %p) {
define i32 @reduce_and_16xi32_prefix5(ptr %p) {
; CHECK-LABEL: reduce_and_16xi32_prefix5:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetivli zero, 8, e32, m1, ta, ma
-; CHECK-NEXT: vmv.v.i v8, -1
; CHECK-NEXT: vsetivli zero, 8, e32, m2, ta, ma
-; CHECK-NEXT: vle32.v v10, (a0)
+; CHECK-NEXT: vle32.v v8, (a0)
+; CHECK-NEXT: vsetivli zero, 8, e32, m1, ta, ma
+; CHECK-NEXT: vmv.v.i v10, -1
; CHECK-NEXT: vsetivli zero, 6, e32, m2, tu, ma
-; CHECK-NEXT: vslideup.vi v10, v8, 5
+; CHECK-NEXT: vslideup.vi v8, v10, 5
; CHECK-NEXT: vsetivli zero, 7, e32, m2, tu, ma
-; CHECK-NEXT: vslideup.vi v10, v8, 6
+; CHECK-NEXT: vslideup.vi v8, v10, 6
; CHECK-NEXT: vsetivli zero, 8, e32, m2, ta, ma
-; CHECK-NEXT: vslideup.vi v10, v8, 7
-; CHECK-NEXT: vredand.vs v8, v10, v10
+; CHECK-NEXT: vslideup.vi v8, v10, 7
+; CHECK-NEXT: vredand.vs v8, v8, v8
; CHECK-NEXT: vmv.x.s a0, v8
; CHECK-NEXT: ret
%v = load <16 x i32>, ptr %p, align 256
@@ -623,12 +623,12 @@ define i32 @reduce_or_16xi32_prefix5(ptr %p) {
; CHECK-NEXT: li a1, -32
; CHECK-NEXT: vsetivli zero, 8, e8, mf2, ta, ma
; CHECK-NEXT: vmv.s.x v0, a1
-; CHECK-NEXT: vmv.v.i v8, -1
-; CHECK-NEXT: vmerge.vim v8, v8, 0, v0
+; CHECK-NEXT: vle32.v v8, (a0)
+; CHECK-NEXT: vmv.v.i v10, -1
+; CHECK-NEXT: vmerge.vim v10, v10, 0, v0
; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, ma
-; CHECK-NEXT: vle32.v v10, (a0)
-; CHECK-NEXT: vsext.vf4 v12, v8
-; CHECK-NEXT: vand.vv v8, v10, v12
+; CHECK-NEXT: vsext.vf4 v12, v10
+; CHECK-NEXT: vand.vv v8, v8, v12
; CHECK-NEXT: vredor.vs v8, v8, v8
; CHECK-NEXT: vmv.x.s a0, v8
; CHECK-NEXT: ret
@@ -668,17 +668,17 @@ define i32 @reduce_smax_16xi32_prefix2(ptr %p) {
define i32 @reduce_smax_16xi32_prefix5(ptr %p) {
; CHECK-LABEL: reduce_smax_16xi32_prefix5:
; CHECK: # %bb.0:
-; CHECK-NEXT: lui a1, 524288
; CHECK-NEXT: vsetivli zero, 8, e32, m2, ta, ma
-; CHECK-NEXT: vmv.s.x v8, a1
-; CHECK-NEXT: vle32.v v10, (a0)
+; CHECK-NEXT: vle32.v v8, (a0)
+; CHECK-NEXT: lui a0, 524288
+; CHECK-NEXT: vmv.s.x v10, a0
; CHECK-NEXT: vsetivli zero, 6, e32, m2, tu, ma
-; CHECK-NEXT: vslideup.vi v10, v8, 5
+; CHECK-NEXT: vslideup.vi v8, v10, 5
; CHECK-NEXT: vsetivli zero, 7, e32, m2, tu, ma
-; CHECK-NEXT: vslideup.vi v10, v8, 6
+; CHECK-NEXT: vslideup.vi v8, v10, 6
; CHECK-NEXT: vsetivli zero, 8, e32, m2, ta, ma
-; CHECK-NEXT: vslideup.vi v10, v8, 7
-; CHECK-NEXT: vredmax.vs v8, v10, v10
+; CHECK-NEXT: vslideup.vi v8, v10, 7
+; CHECK-NEXT: vredmax.vs v8, v8, v8
; CHECK-NEXT: vmv.x.s a0, v8
; CHECK-NEXT: ret
%v = load <16 x i32>, ptr %p, align 256
@@ -713,17 +713,17 @@ define i32 @reduce_smin_16xi32_prefix5(ptr %p) {
; CHECK-LABEL: reduce_smin_16xi32_prefix5:
; CHECK: # %bb.0:
; CHECK-NEXT: lui a1, 524288
-; CHECK-NEXT: addi a1, a1, -1
; CHECK-NEXT: vsetivli zero, 8, e32, m2, ta, ma
-; CHECK-NEXT: vmv.s.x v8, a1
-; CHECK-NEXT: vle32.v v10, (a0)
+; CHECK-NEXT: vle32.v v8, (a0)
+; CHECK-NEXT: addi a1, a1, -1
+; CHECK-NEXT: vmv.s.x v10, a1
; CHECK-NEXT: vsetivli zero, 6, e32, m2, tu, ma
-; CHECK-NEXT: vslideup.vi v10, v8, 5
+; CHECK-NEXT: vslideup.vi v8, v10, 5
; CHECK-NEXT: vsetivli zero, 7, e32, m2, tu, ma
-; CHECK-NEXT: vslideup.vi v10, v8, 6
+; CHECK-NEXT: vslideup.vi v8, v10, 6
; CHECK-NEXT: vsetivli zero, 8, e32, m2, ta, ma
-; CHECK-NEXT: vslideup.vi v10, v8, 7
-; CHECK-NEXT: vredmin.vs v8, v10, v10
+; CHECK-NEXT: vslideup.vi v8, v10, 7
+; CHECK-NEXT: vredmin.vs v8, v8, v8
; CHECK-NEXT: vmv.x.s a0, v8
; CHECK-NEXT: ret
%v = load <16 x i32>, ptr %p, align 256
@@ -760,12 +760,12 @@ define i32 @reduce_umax_16xi32_prefix5(ptr %p) {
; CHECK-NEXT: li a1, -32
; CHECK-NEXT: vsetivli zero, 8, e8, mf2, ta, ma
; CHECK-NEXT: vmv.s.x v0, a1
-; CHECK-NEXT: vmv.v.i v8, -1
-; CHECK-NEXT: vmerge.vim v8, v8, 0, v0
+; CHECK-NEXT: vle32.v v8, (a0)
+; CHECK-NEXT: vmv.v.i v10, -1
+; CHECK-NEXT: vmerge.vim v10, v10, 0, v0
; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, ma
-; CHECK-NEXT: vle32.v v10, (a0)
-; CHECK-NEXT: vsext.vf4 v12, v8
-; CHECK-NEXT: vand.vv v8, v10, v12
+; CHECK-NEXT: vsext.vf4 v12, v10
+; CHECK-NEXT: vand.vv v8, v8, v12
; CHECK-NEXT: vredmaxu.vs v8, v8, v8
; CHECK-NEXT: vmv.x.s a0, v8
; CHECK-NEXT: ret
@@ -800,17 +800,17 @@ define i32 @reduce_umin_16xi32_prefix2(ptr %p) {
define i32 @reduce_umin_16xi32_prefix5(ptr %p) {
; CHECK-LABEL: reduce_umin_16xi32_prefix5:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetivli zero, 8, e32, m1, ta, ma
-; CHECK-NEXT: vmv.v.i v8, -1
; CHECK-NEXT: vsetivli zero, 8, e32, m2, ta, ma
-; CHECK-NEXT: vle32.v v10, (a0)
+; CHECK-NEXT: vle32.v v8, (a0)
+; CHECK-NEXT: vsetivli zero, 8, e32, m1, ta, ma
+; CHECK-NEXT: vmv.v.i v10, -1
; CHECK-NEXT: vsetivli zero, 6, e32, m2, tu, ma
-; CHECK-NEXT: vslideup.vi v10, v8, 5
+; CHECK-NEXT: vslideup.vi v8, v10, 5
; CHECK-NEXT: vsetivli zero, 7, e32, m2, tu, ma
-; CHECK-NEXT: vslideup.vi v10, v8, 6
+; CHECK-NEXT: vslideup.vi v8, v10, 6
; CHECK-NEXT: vsetivli zero, 8, e32, m2, ta, ma
-; CHECK-NEXT: vslideup.vi v10, v8, 7
-; CHECK-NEXT: vredminu.vs v8, v10, v10
+; CHECK-NEXT: vslideup.vi v8, v10, 7
+; CHECK-NEXT: vredminu.vs v8, v8, v8
; CHECK-NEXT: vmv.x.s a0, v8
; CHECK-NEXT: ret
%v = load <16 x i32>, ptr %p, align 256
@@ -830,9 +830,9 @@ define float @reduce_fadd_16xf32_prefix2(ptr %p) {
; CHECK-LABEL: reduce_fadd_16xf32_prefix2:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
-; CHECK-NEXT: vmv.s.x v8, zero
-; CHECK-NEXT: vle32.v v9, (a0)
-; CHECK-NEXT: vfredusum.vs v8, v9, v8
+; CHECK-NEXT: vle32.v v8, (a0)
+; CHECK-NEXT: vmv.s.x v9, zero
+; CHECK-NEXT: vfredusum.vs v8, v8, v9
; CHECK-NEXT: vfmv.f.s fa0, v8
; CHECK-NEXT: ret
%v = load <16 x float>, ptr %p, align 256
@@ -845,17 +845,17 @@ define float @reduce_fadd_16xf32_prefix2(ptr %p) {
define float @reduce_fadd_16xi32_prefix5(ptr %p) {
; CHECK-LABEL: reduce_fadd_16xi32_prefix5:
; CHECK: # %bb.0:
-; CHECK-NEXT: lui a1, 524288
; CHECK-NEXT: vsetivli zero, 8, e32, m2, ta, ma
-; CHECK-NEXT: vmv.s.x v8, a1
-; CHECK-NEXT: vle32.v v10, (a0)
+; CHECK-NEXT: vle32.v v8, (a0)
+; CHECK-NEXT: lui a0, 524288
+; CHECK-NEXT: vmv.s.x v10, a0
; CHECK-NEXT: vsetivli zero, 6, e32, m2, tu, ma
-; CHECK-NEXT: vslideup.vi v10, v8, 5
+; CHECK-NEXT: vslideup.vi v8, v10, 5
; CHECK-NEXT: vsetivli zero, 7, e32, m2, tu, ma
-; CHECK-NEXT: vslideup.vi v10, v8, 6
+; CHECK-NEXT: vslideup.vi v8, v10, 6
; CHECK-NEXT: vsetivli zero, 8, e32, m2, ta, ma
-; CHECK-NEXT: vslideup.vi v10, v8, 7
-; CHECK-NEXT: vfredusum.vs v8, v10, v8
+; CHECK-NEXT: vslideup.vi v8, v10, 7
+; CHECK-NEXT: vfredusum.vs v8, v8, v10
; CHECK-NEXT: vfmv.f.s fa0, v8
; CHECK-NEXT: ret
%v = load <16 x float>, ptr %p, align 256
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-reduction-fp-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-reduction-fp-vp.ll
index 9df160bf30f00..7adaaa05f9dd9 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-reduction-fp-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-reduction-fp-vp.ll
@@ -177,8 +177,8 @@ declare float @llvm.vp.reduce.fadd.v64f32(float, <64 x float>, <64 x i1>, i32)
define float @vpreduce_fadd_v64f32(float %s, <64 x float> %v, <64 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vpreduce_fadd_v64f32:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; CHECK-NEXT: li a2, 32
+; CHECK-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; CHECK-NEXT: vslidedown.vi v24, v0, 4
; CHECK-NEXT: mv a1, a0
; CHECK-NEXT: bltu a0, a2, .LBB8_2
@@ -193,8 +193,8 @@ define float @vpreduce_fadd_v64f32(float %s, <64 x float> %v, <64 x i1> %m, i32
; CHECK-NEXT: sltu a0, a0, a1
; CHECK-NEXT: addi a0, a0, -1
; CHECK-NEXT: and a0, a0, a1
-; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; CHECK-NEXT: vfredusum.vs v25, v16, v25, v0.t
; CHECK-NEXT: vfmv.f.s fa0, v25
; CHECK-NEXT: ret
@@ -205,8 +205,8 @@ define float @vpreduce_fadd_v64f32(float %s, <64 x float> %v, <64 x i1> %m, i32
define float @vpreduce_ord_fadd_v64f32(float %s, <64 x float> %v, <64 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vpreduce_ord_fadd_v64f32:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; CHECK-NEXT: li a2, 32
+; CHECK-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; CHECK-NEXT: vslidedown.vi v24, v0, 4
; CHECK-NEXT: mv a1, a0
; CHECK-NEXT: bltu a0, a2, .LBB9_2
@@ -221,8 +221,8 @@ define float @vpreduce_ord_fadd_v64f32(float %s, <64 x float> %v, <64 x i1> %m,
; CHECK-NEXT: sltu a0, a0, a1
; CHECK-NEXT: addi a0, a0, -1
; CHECK-NEXT: and a0, a0, a1
-; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; CHECK-NEXT: vfredosum.vs v25, v16, v25, v0.t
; CHECK-NEXT: vfmv.f.s fa0, v25
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-reduction-fp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-reduction-fp.ll
index 7dcfb247d37cb..a6763fa22822e 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-reduction-fp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-reduction-fp.ll
@@ -1853,9 +1853,9 @@ define float @vreduce_fminimum_v128f32(ptr %x) {
; CHECK-NEXT: sub sp, sp, a1
; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x18, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 24 * vlenb
; CHECK-NEXT: li a1, 32
+; CHECK-NEXT: addi a2, a0, 128
; CHECK-NEXT: vsetvli zero, a1, e32, m8, ta, ma
-; CHECK-NEXT: addi a1, a0, 128
-; CHECK-NEXT: vle32.v v16, (a1)
+; CHECK-NEXT: vle32.v v16, (a2)
; CHECK-NEXT: addi a1, a0, 384
; CHECK-NEXT: vle32.v v8, (a1)
; CHECK-NEXT: addi a1, a0, 256
@@ -2188,8 +2188,8 @@ define double @vreduce_fminimum_v64f64(ptr %x) {
; CHECK-NEXT: add a1, a1, a2
; CHECK-NEXT: sub sp, sp, a1
; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x18, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 24 * vlenb
-; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; CHECK-NEXT: addi a1, a0, 128
+; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; CHECK-NEXT: vle64.v v16, (a1)
; CHECK-NEXT: addi a1, a0, 384
; CHECK-NEXT: vle64.v v8, (a1)
@@ -2286,9 +2286,9 @@ define double @vreduce_fminimum_v64f64_nonans(ptr %x) {
; CHECK: # %bb.0:
; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; CHECK-NEXT: vle64.v v8, (a0)
-; CHECK-NEXT: addi a1, a0, 384
-; CHECK-NEXT: vle64.v v16, (a1)
; CHECK-NEXT: addi a1, a0, 256
+; CHECK-NEXT: addi a2, a0, 384
+; CHECK-NEXT: vle64.v v16, (a2)
; CHECK-NEXT: addi a0, a0, 128
; CHECK-NEXT: vle64.v v24, (a0)
; CHECK-NEXT: vle64.v v0, (a1)
@@ -2563,9 +2563,9 @@ define float @vreduce_fmaximum_v128f32(ptr %x) {
; CHECK-NEXT: sub sp, sp, a1
; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x18, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 24 * vlenb
; CHECK-NEXT: li a1, 32
+; CHECK-NEXT: addi a2, a0, 128
; CHECK-NEXT: vsetvli zero, a1, e32, m8, ta, ma
-; CHECK-NEXT: addi a1, a0, 128
-; CHECK-NEXT: vle32.v v16, (a1)
+; CHECK-NEXT: vle32.v v16, (a2)
; CHECK-NEXT: addi a1, a0, 384
; CHECK-NEXT: vle32.v v8, (a1)
; CHECK-NEXT: addi a1, a0, 256
@@ -2898,8 +2898,8 @@ define double @vreduce_fmaximum_v64f64(ptr %x) {
; CHECK-NEXT: add a1, a1, a2
; CHECK-NEXT: sub sp, sp, a1
; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x18, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 24 * vlenb
-; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; CHECK-NEXT: addi a1, a0, 128
+; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; CHECK-NEXT: vle64.v v16, (a1)
; CHECK-NEXT: addi a1, a0, 384
; CHECK-NEXT: vle64.v v8, (a1)
@@ -2996,9 +2996,9 @@ define double @vreduce_fmaximum_v64f64_nonans(ptr %x) {
; CHECK: # %bb.0:
; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; CHECK-NEXT: vle64.v v8, (a0)
-; CHECK-NEXT: addi a1, a0, 384
-; CHECK-NEXT: vle64.v v16, (a1)
; CHECK-NEXT: addi a1, a0, 256
+; CHECK-NEXT: addi a2, a0, 384
+; CHECK-NEXT: vle64.v v16, (a2)
; CHECK-NEXT: addi a0, a0, 128
; CHECK-NEXT: vle64.v v24, (a0)
; CHECK-NEXT: vle64.v v0, (a1)
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-reduction-int-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-reduction-int-vp.ll
index 02a989a969960..016f95bfef7e7 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-reduction-int-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-reduction-int-vp.ll
@@ -801,8 +801,8 @@ declare i32 @llvm.vp.reduce.xor.v64i32(i32, <64 x i32>, <64 x i1>, i32)
define signext i32 @vpreduce_xor_v64i32(i32 signext %s, <64 x i32> %v, <64 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vpreduce_xor_v64i32:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; CHECK-NEXT: li a3, 32
+; CHECK-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; CHECK-NEXT: vslidedown.vi v24, v0, 4
; CHECK-NEXT: mv a2, a1
; CHECK-NEXT: bltu a1, a3, .LBB49_2
@@ -817,8 +817,8 @@ define signext i32 @vpreduce_xor_v64i32(i32 signext %s, <64 x i32> %v, <64 x i1>
; CHECK-NEXT: sltu a1, a1, a0
; CHECK-NEXT: addi a1, a1, -1
; CHECK-NEXT: and a0, a1, a0
-; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; CHECK-NEXT: vredxor.vs v25, v16, v25, v0.t
; CHECK-NEXT: vmv.x.s a0, v25
; CHECK-NEXT: ret
@@ -1750,9 +1750,9 @@ define signext i8 @vpreduce_mul_v64i8(i8 signext %s, <64 x i8> %v, <64 x i1> %m,
; RV32-NEXT: sw ra, 12(sp) # 4-byte Folded Spill
; RV32-NEXT: .cfi_offset ra, -4
; RV32-NEXT: li a3, 32
-; RV32-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; RV32-NEXT: lui a2, %hi(.LCPI72_0)
; RV32-NEXT: addi a2, a2, %lo(.LCPI72_0)
+; RV32-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; RV32-NEXT: vle8.v v12, (a2)
; RV32-NEXT: mv a2, a0
; RV32-NEXT: vid.v v16
@@ -1794,9 +1794,9 @@ define signext i8 @vpreduce_mul_v64i8(i8 signext %s, <64 x i8> %v, <64 x i1> %m,
; RV64-NEXT: sd ra, 8(sp) # 8-byte Folded Spill
; RV64-NEXT: .cfi_offset ra, -8
; RV64-NEXT: li a3, 32
-; RV64-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; RV64-NEXT: lui a2, %hi(.LCPI72_0)
; RV64-NEXT: addi a2, a2, %lo(.LCPI72_0)
+; RV64-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; RV64-NEXT: vle8.v v12, (a2)
; RV64-NEXT: mv a2, a0
; RV64-NEXT: vid.v v16
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-reduction-int.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-reduction-int.ll
index 6c75c9b9c2949..28ce6a12c4c89 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-reduction-int.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-reduction-int.ll
@@ -1540,22 +1540,21 @@ define i64 @vwreduce_add_v64i64(ptr %x) {
; RV32-NEXT: vslidedown.vi v24, v8, 16
; RV32-NEXT: vslidedown.vi v0, v16, 16
; RV32-NEXT: vsetivli zero, 16, e32, m4, ta, ma
-; RV32-NEXT: vmv4r.v v8, v0
-; RV32-NEXT: vwadd.vv v0, v24, v8
+; RV32-NEXT: vwadd.vv v8, v24, v0
; RV32-NEXT: csrr a0, vlenb
; RV32-NEXT: slli a0, a0, 3
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 16
-; RV32-NEXT: vs8r.v v0, (a0) # Unknown-size Folded Spill
+; RV32-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; RV32-NEXT: addi a0, sp, 16
; RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; RV32-NEXT: vwadd.vv v0, v8, v16
-; RV32-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; RV32-NEXT: csrr a0, vlenb
; RV32-NEXT: slli a0, a0, 3
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 16
; RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; RV32-NEXT: vadd.vv v8, v0, v8
; RV32-NEXT: vmv.s.x v16, zero
; RV32-NEXT: vredsum.vs v8, v8, v16
@@ -1588,22 +1587,21 @@ define i64 @vwreduce_add_v64i64(ptr %x) {
; RV64-NEXT: vslidedown.vi v24, v8, 16
; RV64-NEXT: vslidedown.vi v0, v16, 16
; RV64-NEXT: vsetivli zero, 16, e32, m4, ta, ma
-; RV64-NEXT: vmv4r.v v8, v0
-; RV64-NEXT: vwadd.vv v0, v24, v8
+; RV64-NEXT: vwadd.vv v8, v24, v0
; RV64-NEXT: csrr a0, vlenb
; RV64-NEXT: slli a0, a0, 3
; RV64-NEXT: add a0, sp, a0
; RV64-NEXT: addi a0, a0, 16
-; RV64-NEXT: vs8r.v v0, (a0) # Unknown-size Folded Spill
+; RV64-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; RV64-NEXT: addi a0, sp, 16
; RV64-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; RV64-NEXT: vwadd.vv v0, v8, v16
-; RV64-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; RV64-NEXT: csrr a0, vlenb
; RV64-NEXT: slli a0, a0, 3
; RV64-NEXT: add a0, sp, a0
; RV64-NEXT: addi a0, a0, 16
; RV64-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; RV64-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; RV64-NEXT: vadd.vv v8, v0, v8
; RV64-NEXT: vmv.s.x v16, zero
; RV64-NEXT: vredsum.vs v8, v8, v16
@@ -1639,22 +1637,21 @@ define i64 @vwreduce_uadd_v64i64(ptr %x) {
; RV32-NEXT: vslidedown.vi v24, v8, 16
; RV32-NEXT: vslidedown.vi v0, v16, 16
; RV32-NEXT: vsetivli zero, 16, e32, m4, ta, ma
-; RV32-NEXT: vmv4r.v v8, v0
-; RV32-NEXT: vwaddu.vv v0, v24, v8
+; RV32-NEXT: vwaddu.vv v8, v24, v0
; RV32-NEXT: csrr a0, vlenb
; RV32-NEXT: slli a0, a0, 3
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 16
-; RV32-NEXT: vs8r.v v0, (a0) # Unknown-size Folded Spill
+; RV32-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; RV32-NEXT: addi a0, sp, 16
; RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; RV32-NEXT: vwaddu.vv v0, v8, v16
-; RV32-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; RV32-NEXT: csrr a0, vlenb
; RV32-NEXT: slli a0, a0, 3
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 16
; RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; RV32-NEXT: vadd.vv v8, v0, v8
; RV32-NEXT: vmv.s.x v16, zero
; RV32-NEXT: vredsum.vs v8, v8, v16
@@ -1687,22 +1684,21 @@ define i64 @vwreduce_uadd_v64i64(ptr %x) {
; RV64-NEXT: vslidedown.vi v24, v8, 16
; RV64-NEXT: vslidedown.vi v0, v16, 16
; RV64-NEXT: vsetivli zero, 16, e32, m4, ta, ma
-; RV64-NEXT: vmv4r.v v8, v0
-; RV64-NEXT: vwaddu.vv v0, v24, v8
+; RV64-NEXT: vwaddu.vv v8, v24, v0
; RV64-NEXT: csrr a0, vlenb
; RV64-NEXT: slli a0, a0, 3
; RV64-NEXT: add a0, sp, a0
; RV64-NEXT: addi a0, a0, 16
-; RV64-NEXT: vs8r.v v0, (a0) # Unknown-size Folded Spill
+; RV64-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; RV64-NEXT: addi a0, sp, 16
; RV64-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; RV64-NEXT: vwaddu.vv v0, v8, v16
-; RV64-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; RV64-NEXT: csrr a0, vlenb
; RV64-NEXT: slli a0, a0, 3
; RV64-NEXT: add a0, sp, a0
; RV64-NEXT: addi a0, a0, 16
; RV64-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; RV64-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; RV64-NEXT: vadd.vv v8, v0, v8
; RV64-NEXT: vmv.s.x v16, zero
; RV64-NEXT: vredsum.vs v8, v8, v16
@@ -2286,9 +2282,9 @@ define i64 @vreduce_and_v64i64(ptr %x) nounwind {
; RV64: # %bb.0:
; RV64-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV64-NEXT: vle64.v v8, (a0)
-; RV64-NEXT: addi a1, a0, 384
-; RV64-NEXT: vle64.v v16, (a1)
; RV64-NEXT: addi a1, a0, 256
+; RV64-NEXT: addi a2, a0, 384
+; RV64-NEXT: vle64.v v16, (a2)
; RV64-NEXT: addi a0, a0, 128
; RV64-NEXT: vle64.v v24, (a0)
; RV64-NEXT: vle64.v v0, (a1)
@@ -2871,9 +2867,9 @@ define i64 @vreduce_or_v64i64(ptr %x) nounwind {
; RV64: # %bb.0:
; RV64-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV64-NEXT: vle64.v v8, (a0)
-; RV64-NEXT: addi a1, a0, 384
-; RV64-NEXT: vle64.v v16, (a1)
; RV64-NEXT: addi a1, a0, 256
+; RV64-NEXT: addi a2, a0, 384
+; RV64-NEXT: vle64.v v16, (a2)
; RV64-NEXT: addi a0, a0, 128
; RV64-NEXT: vle64.v v24, (a0)
; RV64-NEXT: vle64.v v0, (a1)
@@ -4074,9 +4070,9 @@ define i64 @vreduce_smin_v64i64(ptr %x) nounwind {
; RV64: # %bb.0:
; RV64-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV64-NEXT: vle64.v v8, (a0)
-; RV64-NEXT: addi a1, a0, 384
-; RV64-NEXT: vle64.v v16, (a1)
; RV64-NEXT: addi a1, a0, 256
+; RV64-NEXT: addi a2, a0, 384
+; RV64-NEXT: vle64.v v16, (a2)
; RV64-NEXT: addi a0, a0, 128
; RV64-NEXT: vle64.v v24, (a0)
; RV64-NEXT: vle64.v v0, (a1)
@@ -4659,9 +4655,9 @@ define i64 @vreduce_smax_v64i64(ptr %x) nounwind {
; RV64: # %bb.0:
; RV64-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV64-NEXT: vle64.v v8, (a0)
-; RV64-NEXT: addi a1, a0, 384
-; RV64-NEXT: vle64.v v16, (a1)
; RV64-NEXT: addi a1, a0, 256
+; RV64-NEXT: addi a2, a0, 384
+; RV64-NEXT: vle64.v v16, (a2)
; RV64-NEXT: addi a0, a0, 128
; RV64-NEXT: vle64.v v24, (a0)
; RV64-NEXT: vle64.v v0, (a1)
@@ -5244,9 +5240,9 @@ define i64 @vreduce_umin_v64i64(ptr %x) nounwind {
; RV64: # %bb.0:
; RV64-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV64-NEXT: vle64.v v8, (a0)
-; RV64-NEXT: addi a1, a0, 384
-; RV64-NEXT: vle64.v v16, (a1)
; RV64-NEXT: addi a1, a0, 256
+; RV64-NEXT: addi a2, a0, 384
+; RV64-NEXT: vle64.v v16, (a2)
; RV64-NEXT: addi a0, a0, 128
; RV64-NEXT: vle64.v v24, (a0)
; RV64-NEXT: vle64.v v0, (a1)
@@ -5829,9 +5825,9 @@ define i64 @vreduce_umax_v64i64(ptr %x) nounwind {
; RV64: # %bb.0:
; RV64-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV64-NEXT: vle64.v v8, (a0)
-; RV64-NEXT: addi a1, a0, 384
-; RV64-NEXT: vle64.v v16, (a1)
; RV64-NEXT: addi a1, a0, 256
+; RV64-NEXT: addi a2, a0, 384
+; RV64-NEXT: vle64.v v16, (a2)
; RV64-NEXT: addi a0, a0, 128
; RV64-NEXT: vle64.v v24, (a0)
; RV64-NEXT: vle64.v v0, (a1)
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-reduction-mask-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-reduction-mask-vp.ll
index a1f010f98ab40..dc0f4e7430555 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-reduction-mask-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-reduction-mask-vp.ll
@@ -24,8 +24,8 @@ define zeroext i1 @vpreduce_or_v1i1(i1 zeroext %s, <1 x i1> %v, <1 x i1> %m, i32
; CHECK-LABEL: vpreduce_or_v1i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: snez a1, a1
; CHECK-NEXT: or a0, a1, a0
@@ -40,8 +40,8 @@ define zeroext i1 @vpreduce_xor_v1i1(i1 zeroext %s, <1 x i1> %v, <1 x i1> %m, i3
; CHECK-LABEL: vpreduce_xor_v1i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: andi a1, a1, 1
; CHECK-NEXT: xor a0, a1, a0
@@ -72,8 +72,8 @@ define zeroext i1 @vpreduce_or_v2i1(i1 zeroext %s, <2 x i1> %v, <2 x i1> %m, i32
; CHECK-LABEL: vpreduce_or_v2i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: snez a1, a1
; CHECK-NEXT: or a0, a1, a0
@@ -88,8 +88,8 @@ define zeroext i1 @vpreduce_xor_v2i1(i1 zeroext %s, <2 x i1> %v, <2 x i1> %m, i3
; CHECK-LABEL: vpreduce_xor_v2i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: andi a1, a1, 1
; CHECK-NEXT: xor a0, a1, a0
@@ -120,8 +120,8 @@ define zeroext i1 @vpreduce_or_v4i1(i1 zeroext %s, <4 x i1> %v, <4 x i1> %m, i32
; CHECK-LABEL: vpreduce_or_v4i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, mf4, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: snez a1, a1
; CHECK-NEXT: or a0, a1, a0
@@ -136,8 +136,8 @@ define zeroext i1 @vpreduce_xor_v4i1(i1 zeroext %s, <4 x i1> %v, <4 x i1> %m, i3
; CHECK-LABEL: vpreduce_xor_v4i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, mf4, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: andi a1, a1, 1
; CHECK-NEXT: xor a0, a1, a0
@@ -168,8 +168,8 @@ define zeroext i1 @vpreduce_or_v8i1(i1 zeroext %s, <8 x i1> %v, <8 x i1> %m, i32
; CHECK-LABEL: vpreduce_or_v8i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, mf2, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: snez a1, a1
; CHECK-NEXT: or a0, a1, a0
@@ -184,8 +184,8 @@ define zeroext i1 @vpreduce_xor_v8i1(i1 zeroext %s, <8 x i1> %v, <8 x i1> %m, i3
; CHECK-LABEL: vpreduce_xor_v8i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, mf2, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: andi a1, a1, 1
; CHECK-NEXT: xor a0, a1, a0
@@ -264,8 +264,8 @@ define zeroext i1 @vpreduce_or_v16i1(i1 zeroext %s, <16 x i1> %v, <16 x i1> %m,
; CHECK-LABEL: vpreduce_or_v16i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m1, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, m1, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: snez a1, a1
; CHECK-NEXT: or a0, a1, a0
@@ -280,8 +280,8 @@ define zeroext i1 @vpreduce_xor_v16i1(i1 zeroext %s, <16 x i1> %v, <16 x i1> %m,
; CHECK-LABEL: vpreduce_xor_v16i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m1, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, m1, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: andi a1, a1, 1
; CHECK-NEXT: xor a0, a1, a0
@@ -296,8 +296,8 @@ define zeroext i1 @vpreduce_add_v1i1(i1 zeroext %s, <1 x i1> %v, <1 x i1> %m, i3
; CHECK-LABEL: vpreduce_add_v1i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: andi a1, a1, 1
; CHECK-NEXT: xor a0, a1, a0
@@ -312,8 +312,8 @@ define zeroext i1 @vpreduce_add_v2i1(i1 zeroext %s, <2 x i1> %v, <2 x i1> %m, i3
; CHECK-LABEL: vpreduce_add_v2i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: andi a1, a1, 1
; CHECK-NEXT: xor a0, a1, a0
@@ -328,8 +328,8 @@ define zeroext i1 @vpreduce_add_v4i1(i1 zeroext %s, <4 x i1> %v, <4 x i1> %m, i3
; CHECK-LABEL: vpreduce_add_v4i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, mf4, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: andi a1, a1, 1
; CHECK-NEXT: xor a0, a1, a0
@@ -344,8 +344,8 @@ define zeroext i1 @vpreduce_add_v8i1(i1 zeroext %s, <8 x i1> %v, <8 x i1> %m, i3
; CHECK-LABEL: vpreduce_add_v8i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, mf2, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: andi a1, a1, 1
; CHECK-NEXT: xor a0, a1, a0
@@ -360,8 +360,8 @@ define zeroext i1 @vpreduce_add_v16i1(i1 zeroext %s, <16 x i1> %v, <16 x i1> %m,
; CHECK-LABEL: vpreduce_add_v16i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m1, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, m1, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: andi a1, a1, 1
; CHECK-NEXT: xor a0, a1, a0
@@ -488,8 +488,8 @@ define zeroext i1 @vpreduce_smin_v1i1(i1 zeroext %s, <1 x i1> %v, <1 x i1> %m, i
; CHECK-LABEL: vpreduce_smin_v1i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: snez a1, a1
; CHECK-NEXT: or a0, a1, a0
@@ -504,8 +504,8 @@ define zeroext i1 @vpreduce_smin_v2i1(i1 zeroext %s, <2 x i1> %v, <2 x i1> %m, i
; CHECK-LABEL: vpreduce_smin_v2i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: snez a1, a1
; CHECK-NEXT: or a0, a1, a0
@@ -520,8 +520,8 @@ define zeroext i1 @vpreduce_smin_v4i1(i1 zeroext %s, <4 x i1> %v, <4 x i1> %m, i
; CHECK-LABEL: vpreduce_smin_v4i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, mf4, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: snez a1, a1
; CHECK-NEXT: or a0, a1, a0
@@ -536,8 +536,8 @@ define zeroext i1 @vpreduce_smin_v8i1(i1 zeroext %s, <8 x i1> %v, <8 x i1> %m, i
; CHECK-LABEL: vpreduce_smin_v8i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, mf2, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: snez a1, a1
; CHECK-NEXT: or a0, a1, a0
@@ -552,8 +552,8 @@ define zeroext i1 @vpreduce_smin_v16i1(i1 zeroext %s, <16 x i1> %v, <16 x i1> %m
; CHECK-LABEL: vpreduce_smin_v16i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m1, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, m1, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: snez a1, a1
; CHECK-NEXT: or a0, a1, a0
@@ -568,8 +568,8 @@ define zeroext i1 @vpreduce_smin_v32i1(i1 zeroext %s, <32 x i1> %v, <32 x i1> %m
; CHECK-LABEL: vpreduce_smin_v32i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, m2, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: snez a1, a1
; CHECK-NEXT: or a0, a1, a0
@@ -584,8 +584,8 @@ define zeroext i1 @vpreduce_smin_v64i1(i1 zeroext %s, <64 x i1> %v, <64 x i1> %m
; CHECK-LABEL: vpreduce_smin_v64i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, m4, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: snez a1, a1
; CHECK-NEXT: or a0, a1, a0
@@ -600,8 +600,8 @@ define zeroext i1 @vpreduce_umax_v1i1(i1 zeroext %s, <1 x i1> %v, <1 x i1> %m, i
; CHECK-LABEL: vpreduce_umax_v1i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: snez a1, a1
; CHECK-NEXT: or a0, a1, a0
@@ -616,8 +616,8 @@ define zeroext i1 @vpreduce_umax_v2i1(i1 zeroext %s, <2 x i1> %v, <2 x i1> %m, i
; CHECK-LABEL: vpreduce_umax_v2i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: snez a1, a1
; CHECK-NEXT: or a0, a1, a0
@@ -632,8 +632,8 @@ define zeroext i1 @vpreduce_umax_v4i1(i1 zeroext %s, <4 x i1> %v, <4 x i1> %m, i
; CHECK-LABEL: vpreduce_umax_v4i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, mf4, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: snez a1, a1
; CHECK-NEXT: or a0, a1, a0
@@ -648,8 +648,8 @@ define zeroext i1 @vpreduce_umax_v8i1(i1 zeroext %s, <8 x i1> %v, <8 x i1> %m, i
; CHECK-LABEL: vpreduce_umax_v8i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, mf2, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: snez a1, a1
; CHECK-NEXT: or a0, a1, a0
@@ -664,8 +664,8 @@ define zeroext i1 @vpreduce_umax_v16i1(i1 zeroext %s, <16 x i1> %v, <16 x i1> %m
; CHECK-LABEL: vpreduce_umax_v16i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m1, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, m1, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: snez a1, a1
; CHECK-NEXT: or a0, a1, a0
@@ -680,8 +680,8 @@ define zeroext i1 @vpreduce_umax_v32i1(i1 zeroext %s, <32 x i1> %v, <32 x i1> %m
; CHECK-LABEL: vpreduce_umax_v32i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, m2, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: snez a1, a1
; CHECK-NEXT: or a0, a1, a0
@@ -696,8 +696,8 @@ define zeroext i1 @vpreduce_umax_v64i1(i1 zeroext %s, <64 x i1> %v, <64 x i1> %m
; CHECK-LABEL: vpreduce_umax_v64i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, m4, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: snez a1, a1
; CHECK-NEXT: or a0, a1, a0
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-rint-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-rint-vp.ll
index 920d0d5fe7ba7..1f856d04ca89f 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-rint-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-rint-vp.ll
@@ -123,15 +123,15 @@ declare <16 x half> @llvm.vp.rint.v16f16(<16 x half>, <16 x i1>, i32)
define <16 x half> @vp_rint_v16f16(<16 x half> %va, <16 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_rint_v16f16:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v10, v0
; CHECK-NEXT: lui a1, %hi(.LCPI6_0)
; CHECK-NEXT: flh fa5, %lo(.LCPI6_0)(a1)
+; CHECK-NEXT: vmv1r.v v10, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, ma
; CHECK-NEXT: vfabs.v v12, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, mu
; CHECK-NEXT: vmflt.vf v10, v12, fa5, v0.t
-; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, mu
@@ -246,8 +246,8 @@ define <8 x float> @vp_rint_v8f32(<8 x float> %va, <8 x i1> %m, i32 zeroext %evl
; CHECK-NEXT: fmv.w.x fa5, a0
; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, mu
; CHECK-NEXT: vmflt.vf v10, v12, fa5, v0.t
-; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, mu
@@ -286,8 +286,8 @@ define <16 x float> @vp_rint_v16f32(<16 x float> %va, <16 x i1> %m, i32 zeroext
; CHECK-NEXT: fmv.w.x fa5, a0
; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, mu
; CHECK-NEXT: vmflt.vf v12, v16, fa5, v0.t
-; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, mu
@@ -357,15 +357,15 @@ declare <4 x double> @llvm.vp.rint.v4f64(<4 x double>, <4 x i1>, i32)
define <4 x double> @vp_rint_v4f64(<4 x double> %va, <4 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_rint_v4f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v10, v0
; CHECK-NEXT: lui a1, %hi(.LCPI18_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI18_0)(a1)
+; CHECK-NEXT: vmv1r.v v10, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, ma
; CHECK-NEXT: vfabs.v v12, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, mu
; CHECK-NEXT: vmflt.vf v10, v12, fa5, v0.t
-; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, mu
@@ -397,15 +397,15 @@ declare <8 x double> @llvm.vp.rint.v8f64(<8 x double>, <8 x i1>, i32)
define <8 x double> @vp_rint_v8f64(<8 x double> %va, <8 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_rint_v8f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v12, v0
; CHECK-NEXT: lui a1, %hi(.LCPI20_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI20_0)(a1)
+; CHECK-NEXT: vmv1r.v v12, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, ma
; CHECK-NEXT: vfabs.v v16, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, mu
; CHECK-NEXT: vmflt.vf v12, v16, fa5, v0.t
-; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, mu
@@ -437,15 +437,15 @@ declare <15 x double> @llvm.vp.rint.v15f64(<15 x double>, <15 x i1>, i32)
define <15 x double> @vp_rint_v15f64(<15 x double> %va, <15 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_rint_v15f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: lui a1, %hi(.LCPI22_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI22_0)(a1)
+; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
; CHECK-NEXT: vmflt.vf v16, v24, fa5, v0.t
-; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
@@ -477,15 +477,15 @@ declare <16 x double> @llvm.vp.rint.v16f64(<16 x double>, <16 x i1>, i32)
define <16 x double> @vp_rint_v16f64(<16 x double> %va, <16 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_rint_v16f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: lui a1, %hi(.LCPI24_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI24_0)(a1)
+; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
; CHECK-NEXT: vmflt.vf v16, v24, fa5, v0.t
-; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
@@ -517,65 +517,54 @@ declare <32 x double> @llvm.vp.rint.v32f64(<32 x double>, <32 x i1>, i32)
define <32 x double> @vp_rint_v32f64(<32 x double> %va, <32 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_rint_v32f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: addi sp, sp, -16
-; CHECK-NEXT: .cfi_def_cfa_offset 16
-; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a1, a1, 4
-; CHECK-NEXT: sub sp, sp, a1
-; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
-; CHECK-NEXT: vmv1r.v v25, v0
-; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a1, a1, 3
-; CHECK-NEXT: add a1, sp, a1
-; CHECK-NEXT: addi a1, a1, 16
-; CHECK-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
-; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
+; CHECK-NEXT: vmv1r.v v6, v0
; CHECK-NEXT: li a2, 16
-; CHECK-NEXT: vslidedown.vi v24, v0, 2
+; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
+; CHECK-NEXT: vslidedown.vi v7, v0, 2
; CHECK-NEXT: mv a1, a0
; CHECK-NEXT: bltu a0, a2, .LBB26_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: li a1, 16
; CHECK-NEXT: .LBB26_2:
+; CHECK-NEXT: addi sp, sp, -16
+; CHECK-NEXT: .cfi_def_cfa_offset 16
+; CHECK-NEXT: csrr a2, vlenb
+; CHECK-NEXT: slli a2, a2, 3
+; CHECK-NEXT: sub sp, sp, a2
+; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
; CHECK-NEXT: lui a2, %hi(.LCPI26_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI26_0)(a2)
+; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v25
-; CHECK-NEXT: vfabs.v v16, v8, v0.t
+; CHECK-NEXT: vfabs.v v24, v8, v0.t
+; CHECK-NEXT: addi a1, sp, 16
+; CHECK-NEXT: vs8r.v v24, (a1) # Unknown-size Folded Spill
+; CHECK-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vmflt.vf v25, v16, fa5, v0.t
+; CHECK-NEXT: vmflt.vf v6, v24, fa5, v0.t
+; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v25
-; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
-; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
+; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
+; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vfsgnj.vv v8, v16, v8, v0.t
-; CHECK-NEXT: addi a1, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
+; CHECK-NEXT: vfsgnj.vv v8, v24, v8, v0.t
; CHECK-NEXT: addi a1, a0, -16
; CHECK-NEXT: sltu a0, a0, a1
; CHECK-NEXT: addi a0, a0, -1
; CHECK-NEXT: and a0, a0, a1
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v24
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vfabs.v v8, v16, v0.t
+; CHECK-NEXT: vfabs.v v24, v16, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vmflt.vf v24, v8, fa5, v0.t
+; CHECK-NEXT: vmflt.vf v7, v24, fa5, v0.t
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v24
-; CHECK-NEXT: vfcvt.x.f.v v8, v16, v0.t
-; CHECK-NEXT: vfcvt.f.x.v v8, v8, v0.t
+; CHECK-NEXT: vfcvt.x.f.v v24, v16, v0.t
+; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vfsgnj.vv v16, v8, v16, v0.t
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vfsgnj.vv v16, v24, v16, v0.t
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 4
+; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add sp, sp, a0
; CHECK-NEXT: addi sp, sp, 16
; CHECK-NEXT: ret
@@ -597,17 +586,20 @@ define <32 x double> @vp_rint_v32f64_unmasked(<32 x double> %va, i32 zeroext %ev
; CHECK-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; CHECK-NEXT: vfabs.v v24, v8
; CHECK-NEXT: vmflt.vf v0, v24, fa5
+; CHECK-NEXT: addi a2, a0, -16
+; CHECK-NEXT: sltu a0, a0, a2
+; CHECK-NEXT: addi a0, a0, -1
+; CHECK-NEXT: and a0, a0, a2
+; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT: vfabs.v v24, v16
+; CHECK-NEXT: vmflt.vf v7, v24, fa5
+; CHECK-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v24, v8, v0.t
-; CHECK-NEXT: addi a1, a0, -16
-; CHECK-NEXT: sltu a0, a0, a1
-; CHECK-NEXT: addi a0, a0, -1
-; CHECK-NEXT: and a0, a0, a1
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: vfabs.v v24, v16
-; CHECK-NEXT: vmflt.vf v0, v24, fa5
; CHECK-NEXT: vfcvt.x.f.v v24, v16, v0.t
; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-round-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-round-vp.ll
index 716cf7b0f46fa..0f587232680df 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-round-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-round-vp.ll
@@ -204,8 +204,8 @@ define <8 x half> @vp_round_v8f16(<8 x half> %va, <8 x i1> %m, i32 zeroext %evl)
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m2, ta, mu
; ZVFHMIN-NEXT: vmflt.vf v9, v12, fa5, v0.t
; ZVFHMIN-NEXT: fsrmi a0, 4
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v9
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; ZVFHMIN-NEXT: vfcvt.x.f.v v12, v10, v0.t
; ZVFHMIN-NEXT: fsrm a0
; ZVFHMIN-NEXT: vfcvt.f.x.v v12, v12, v0.t
@@ -261,16 +261,16 @@ declare <16 x half> @llvm.vp.round.v16f16(<16 x half>, <16 x i1>, i32)
define <16 x half> @vp_round_v16f16(<16 x half> %va, <16 x i1> %m, i32 zeroext %evl) {
; ZVFH-LABEL: vp_round_v16f16:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vmv1r.v v10, v0
; ZVFH-NEXT: lui a1, %hi(.LCPI6_0)
; ZVFH-NEXT: flh fa5, %lo(.LCPI6_0)(a1)
+; ZVFH-NEXT: vmv1r.v v10, v0
; ZVFH-NEXT: vsetvli zero, a0, e16, m2, ta, ma
; ZVFH-NEXT: vfabs.v v12, v8, v0.t
; ZVFH-NEXT: vsetvli zero, zero, e16, m2, ta, mu
; ZVFH-NEXT: vmflt.vf v10, v12, fa5, v0.t
; ZVFH-NEXT: fsrmi a0, 4
-; ZVFH-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; ZVFH-NEXT: vmv1r.v v0, v10
+; ZVFH-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; ZVFH-NEXT: vfcvt.x.f.v v12, v8, v0.t
; ZVFH-NEXT: fsrm a0
; ZVFH-NEXT: vfcvt.f.x.v v12, v12, v0.t
@@ -290,8 +290,8 @@ define <16 x half> @vp_round_v16f16(<16 x half> %va, <16 x i1> %m, i32 zeroext %
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m4, ta, mu
; ZVFHMIN-NEXT: vmflt.vf v10, v16, fa5, v0.t
; ZVFHMIN-NEXT: fsrmi a0, 4
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v10
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; ZVFHMIN-NEXT: vfcvt.x.f.v v16, v12, v0.t
; ZVFHMIN-NEXT: fsrm a0
; ZVFHMIN-NEXT: vfcvt.f.x.v v16, v16, v0.t
@@ -439,8 +439,8 @@ define <8 x float> @vp_round_v8f32(<8 x float> %va, <8 x i1> %m, i32 zeroext %ev
; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, mu
; CHECK-NEXT: vmflt.vf v10, v12, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 4
-; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
@@ -483,8 +483,8 @@ define <16 x float> @vp_round_v16f32(<16 x float> %va, <16 x i1> %m, i32 zeroext
; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, mu
; CHECK-NEXT: vmflt.vf v12, v16, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 4
-; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
@@ -561,16 +561,16 @@ declare <4 x double> @llvm.vp.round.v4f64(<4 x double>, <4 x i1>, i32)
define <4 x double> @vp_round_v4f64(<4 x double> %va, <4 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_round_v4f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v10, v0
; CHECK-NEXT: lui a1, %hi(.LCPI18_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI18_0)(a1)
+; CHECK-NEXT: vmv1r.v v10, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, ma
; CHECK-NEXT: vfabs.v v12, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, mu
; CHECK-NEXT: vmflt.vf v10, v12, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 4
-; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
@@ -605,16 +605,16 @@ declare <8 x double> @llvm.vp.round.v8f64(<8 x double>, <8 x i1>, i32)
define <8 x double> @vp_round_v8f64(<8 x double> %va, <8 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_round_v8f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v12, v0
; CHECK-NEXT: lui a1, %hi(.LCPI20_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI20_0)(a1)
+; CHECK-NEXT: vmv1r.v v12, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, ma
; CHECK-NEXT: vfabs.v v16, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, mu
; CHECK-NEXT: vmflt.vf v12, v16, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 4
-; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
@@ -649,16 +649,16 @@ declare <15 x double> @llvm.vp.round.v15f64(<15 x double>, <15 x i1>, i32)
define <15 x double> @vp_round_v15f64(<15 x double> %va, <15 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_round_v15f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: lui a1, %hi(.LCPI22_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI22_0)(a1)
+; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
; CHECK-NEXT: vmflt.vf v16, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 4
-; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
@@ -693,16 +693,16 @@ declare <16 x double> @llvm.vp.round.v16f64(<16 x double>, <16 x i1>, i32)
define <16 x double> @vp_round_v16f64(<16 x double> %va, <16 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_round_v16f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: lui a1, %hi(.LCPI24_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI24_0)(a1)
+; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
; CHECK-NEXT: vmflt.vf v16, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 4
-; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
@@ -737,69 +737,59 @@ declare <32 x double> @llvm.vp.round.v32f64(<32 x double>, <32 x i1>, i32)
define <32 x double> @vp_round_v32f64(<32 x double> %va, <32 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_round_v32f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: addi sp, sp, -16
-; CHECK-NEXT: .cfi_def_cfa_offset 16
-; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a1, a1, 4
-; CHECK-NEXT: sub sp, sp, a1
-; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
-; CHECK-NEXT: vmv1r.v v25, v0
-; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a1, a1, 3
-; CHECK-NEXT: add a1, sp, a1
-; CHECK-NEXT: addi a1, a1, 16
-; CHECK-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
-; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
+; CHECK-NEXT: vmv1r.v v6, v0
; CHECK-NEXT: li a2, 16
-; CHECK-NEXT: vslidedown.vi v24, v0, 2
+; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
+; CHECK-NEXT: vslidedown.vi v7, v0, 2
; CHECK-NEXT: mv a1, a0
; CHECK-NEXT: bltu a0, a2, .LBB26_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: li a1, 16
; CHECK-NEXT: .LBB26_2:
+; CHECK-NEXT: addi sp, sp, -16
+; CHECK-NEXT: .cfi_def_cfa_offset 16
+; CHECK-NEXT: csrr a2, vlenb
+; CHECK-NEXT: slli a2, a2, 3
+; CHECK-NEXT: sub sp, sp, a2
+; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
; CHECK-NEXT: lui a2, %hi(.LCPI26_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI26_0)(a2)
+; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v25
-; CHECK-NEXT: vfabs.v v16, v8, v0.t
+; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vmflt.vf v25, v16, fa5, v0.t
+; CHECK-NEXT: vmflt.vf v6, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a1, 4
+; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v25
-; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
+; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
+; CHECK-NEXT: addi a2, sp, 16
+; CHECK-NEXT: vs8r.v v24, (a2) # Unknown-size Folded Spill
; CHECK-NEXT: fsrm a1
-; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
-; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vfsgnj.vv v8, v16, v8, v0.t
; CHECK-NEXT: addi a1, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
+; CHECK-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v24, v8, v0.t
; CHECK-NEXT: addi a1, a0, -16
; CHECK-NEXT: sltu a0, a0, a1
; CHECK-NEXT: addi a0, a0, -1
; CHECK-NEXT: and a0, a0, a1
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v24
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vfabs.v v8, v16, v0.t
+; CHECK-NEXT: vfabs.v v24, v16, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vmflt.vf v24, v8, fa5, v0.t
+; CHECK-NEXT: vmflt.vf v7, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 4
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v24
-; CHECK-NEXT: vfcvt.x.f.v v8, v16, v0.t
+; CHECK-NEXT: vfcvt.x.f.v v24, v16, v0.t
; CHECK-NEXT: fsrm a0
-; CHECK-NEXT: vfcvt.f.x.v v8, v8, v0.t
+; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vfsgnj.vv v16, v8, v16, v0.t
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vfsgnj.vv v16, v24, v16, v0.t
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 4
+; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add sp, sp, a0
; CHECK-NEXT: addi sp, sp, 16
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-roundeven-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-roundeven-vp.ll
index 603f9397dc90f..0fb7e6a7de569 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-roundeven-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-roundeven-vp.ll
@@ -204,8 +204,8 @@ define <8 x half> @vp_roundeven_v8f16(<8 x half> %va, <8 x i1> %m, i32 zeroext %
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m2, ta, mu
; ZVFHMIN-NEXT: vmflt.vf v9, v12, fa5, v0.t
; ZVFHMIN-NEXT: fsrmi a0, 0
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v9
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; ZVFHMIN-NEXT: vfcvt.x.f.v v12, v10, v0.t
; ZVFHMIN-NEXT: fsrm a0
; ZVFHMIN-NEXT: vfcvt.f.x.v v12, v12, v0.t
@@ -261,16 +261,16 @@ declare <16 x half> @llvm.vp.roundeven.v16f16(<16 x half>, <16 x i1>, i32)
define <16 x half> @vp_roundeven_v16f16(<16 x half> %va, <16 x i1> %m, i32 zeroext %evl) {
; ZVFH-LABEL: vp_roundeven_v16f16:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vmv1r.v v10, v0
; ZVFH-NEXT: lui a1, %hi(.LCPI6_0)
; ZVFH-NEXT: flh fa5, %lo(.LCPI6_0)(a1)
+; ZVFH-NEXT: vmv1r.v v10, v0
; ZVFH-NEXT: vsetvli zero, a0, e16, m2, ta, ma
; ZVFH-NEXT: vfabs.v v12, v8, v0.t
; ZVFH-NEXT: vsetvli zero, zero, e16, m2, ta, mu
; ZVFH-NEXT: vmflt.vf v10, v12, fa5, v0.t
; ZVFH-NEXT: fsrmi a0, 0
-; ZVFH-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; ZVFH-NEXT: vmv1r.v v0, v10
+; ZVFH-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; ZVFH-NEXT: vfcvt.x.f.v v12, v8, v0.t
; ZVFH-NEXT: fsrm a0
; ZVFH-NEXT: vfcvt.f.x.v v12, v12, v0.t
@@ -290,8 +290,8 @@ define <16 x half> @vp_roundeven_v16f16(<16 x half> %va, <16 x i1> %m, i32 zeroe
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m4, ta, mu
; ZVFHMIN-NEXT: vmflt.vf v10, v16, fa5, v0.t
; ZVFHMIN-NEXT: fsrmi a0, 0
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v10
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; ZVFHMIN-NEXT: vfcvt.x.f.v v16, v12, v0.t
; ZVFHMIN-NEXT: fsrm a0
; ZVFHMIN-NEXT: vfcvt.f.x.v v16, v16, v0.t
@@ -439,8 +439,8 @@ define <8 x float> @vp_roundeven_v8f32(<8 x float> %va, <8 x i1> %m, i32 zeroext
; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, mu
; CHECK-NEXT: vmflt.vf v10, v12, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 0
-; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
@@ -483,8 +483,8 @@ define <16 x float> @vp_roundeven_v16f32(<16 x float> %va, <16 x i1> %m, i32 zer
; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, mu
; CHECK-NEXT: vmflt.vf v12, v16, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 0
-; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
@@ -561,16 +561,16 @@ declare <4 x double> @llvm.vp.roundeven.v4f64(<4 x double>, <4 x i1>, i32)
define <4 x double> @vp_roundeven_v4f64(<4 x double> %va, <4 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_roundeven_v4f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v10, v0
; CHECK-NEXT: lui a1, %hi(.LCPI18_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI18_0)(a1)
+; CHECK-NEXT: vmv1r.v v10, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, ma
; CHECK-NEXT: vfabs.v v12, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, mu
; CHECK-NEXT: vmflt.vf v10, v12, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 0
-; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
@@ -605,16 +605,16 @@ declare <8 x double> @llvm.vp.roundeven.v8f64(<8 x double>, <8 x i1>, i32)
define <8 x double> @vp_roundeven_v8f64(<8 x double> %va, <8 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_roundeven_v8f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v12, v0
; CHECK-NEXT: lui a1, %hi(.LCPI20_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI20_0)(a1)
+; CHECK-NEXT: vmv1r.v v12, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, ma
; CHECK-NEXT: vfabs.v v16, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, mu
; CHECK-NEXT: vmflt.vf v12, v16, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 0
-; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
@@ -649,16 +649,16 @@ declare <15 x double> @llvm.vp.roundeven.v15f64(<15 x double>, <15 x i1>, i32)
define <15 x double> @vp_roundeven_v15f64(<15 x double> %va, <15 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_roundeven_v15f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: lui a1, %hi(.LCPI22_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI22_0)(a1)
+; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
; CHECK-NEXT: vmflt.vf v16, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 0
-; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
@@ -693,16 +693,16 @@ declare <16 x double> @llvm.vp.roundeven.v16f64(<16 x double>, <16 x i1>, i32)
define <16 x double> @vp_roundeven_v16f64(<16 x double> %va, <16 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_roundeven_v16f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: lui a1, %hi(.LCPI24_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI24_0)(a1)
+; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
; CHECK-NEXT: vmflt.vf v16, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 0
-; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
@@ -737,69 +737,59 @@ declare <32 x double> @llvm.vp.roundeven.v32f64(<32 x double>, <32 x i1>, i32)
define <32 x double> @vp_roundeven_v32f64(<32 x double> %va, <32 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_roundeven_v32f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: addi sp, sp, -16
-; CHECK-NEXT: .cfi_def_cfa_offset 16
-; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a1, a1, 4
-; CHECK-NEXT: sub sp, sp, a1
-; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
-; CHECK-NEXT: vmv1r.v v25, v0
-; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a1, a1, 3
-; CHECK-NEXT: add a1, sp, a1
-; CHECK-NEXT: addi a1, a1, 16
-; CHECK-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
-; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
+; CHECK-NEXT: vmv1r.v v6, v0
; CHECK-NEXT: li a2, 16
-; CHECK-NEXT: vslidedown.vi v24, v0, 2
+; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
+; CHECK-NEXT: vslidedown.vi v7, v0, 2
; CHECK-NEXT: mv a1, a0
; CHECK-NEXT: bltu a0, a2, .LBB26_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: li a1, 16
; CHECK-NEXT: .LBB26_2:
+; CHECK-NEXT: addi sp, sp, -16
+; CHECK-NEXT: .cfi_def_cfa_offset 16
+; CHECK-NEXT: csrr a2, vlenb
+; CHECK-NEXT: slli a2, a2, 3
+; CHECK-NEXT: sub sp, sp, a2
+; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
; CHECK-NEXT: lui a2, %hi(.LCPI26_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI26_0)(a2)
+; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v25
-; CHECK-NEXT: vfabs.v v16, v8, v0.t
+; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vmflt.vf v25, v16, fa5, v0.t
+; CHECK-NEXT: vmflt.vf v6, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a1, 0
+; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v25
-; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
+; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
+; CHECK-NEXT: addi a2, sp, 16
+; CHECK-NEXT: vs8r.v v24, (a2) # Unknown-size Folded Spill
; CHECK-NEXT: fsrm a1
-; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
-; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vfsgnj.vv v8, v16, v8, v0.t
; CHECK-NEXT: addi a1, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
+; CHECK-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v24, v8, v0.t
; CHECK-NEXT: addi a1, a0, -16
; CHECK-NEXT: sltu a0, a0, a1
; CHECK-NEXT: addi a0, a0, -1
; CHECK-NEXT: and a0, a0, a1
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v24
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vfabs.v v8, v16, v0.t
+; CHECK-NEXT: vfabs.v v24, v16, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vmflt.vf v24, v8, fa5, v0.t
+; CHECK-NEXT: vmflt.vf v7, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 0
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v24
-; CHECK-NEXT: vfcvt.x.f.v v8, v16, v0.t
+; CHECK-NEXT: vfcvt.x.f.v v24, v16, v0.t
; CHECK-NEXT: fsrm a0
-; CHECK-NEXT: vfcvt.f.x.v v8, v8, v0.t
+; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vfsgnj.vv v16, v8, v16, v0.t
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vfsgnj.vv v16, v24, v16, v0.t
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 4
+; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add sp, sp, a0
; CHECK-NEXT: addi sp, sp, 16
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-roundtozero-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-roundtozero-vp.ll
index a5adfc36887ad..927f96b644227 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-roundtozero-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-roundtozero-vp.ll
@@ -204,8 +204,8 @@ define <8 x half> @vp_roundtozero_v8f16(<8 x half> %va, <8 x i1> %m, i32 zeroext
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m2, ta, mu
; ZVFHMIN-NEXT: vmflt.vf v9, v12, fa5, v0.t
; ZVFHMIN-NEXT: fsrmi a0, 1
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v9
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; ZVFHMIN-NEXT: vfcvt.x.f.v v12, v10, v0.t
; ZVFHMIN-NEXT: fsrm a0
; ZVFHMIN-NEXT: vfcvt.f.x.v v12, v12, v0.t
@@ -261,16 +261,16 @@ declare <16 x half> @llvm.vp.roundtozero.v16f16(<16 x half>, <16 x i1>, i32)
define <16 x half> @vp_roundtozero_v16f16(<16 x half> %va, <16 x i1> %m, i32 zeroext %evl) {
; ZVFH-LABEL: vp_roundtozero_v16f16:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vmv1r.v v10, v0
; ZVFH-NEXT: lui a1, %hi(.LCPI6_0)
; ZVFH-NEXT: flh fa5, %lo(.LCPI6_0)(a1)
+; ZVFH-NEXT: vmv1r.v v10, v0
; ZVFH-NEXT: vsetvli zero, a0, e16, m2, ta, ma
; ZVFH-NEXT: vfabs.v v12, v8, v0.t
; ZVFH-NEXT: vsetvli zero, zero, e16, m2, ta, mu
; ZVFH-NEXT: vmflt.vf v10, v12, fa5, v0.t
; ZVFH-NEXT: fsrmi a0, 1
-; ZVFH-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; ZVFH-NEXT: vmv1r.v v0, v10
+; ZVFH-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; ZVFH-NEXT: vfcvt.x.f.v v12, v8, v0.t
; ZVFH-NEXT: fsrm a0
; ZVFH-NEXT: vfcvt.f.x.v v12, v12, v0.t
@@ -290,8 +290,8 @@ define <16 x half> @vp_roundtozero_v16f16(<16 x half> %va, <16 x i1> %m, i32 zer
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m4, ta, mu
; ZVFHMIN-NEXT: vmflt.vf v10, v16, fa5, v0.t
; ZVFHMIN-NEXT: fsrmi a0, 1
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v10
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; ZVFHMIN-NEXT: vfcvt.x.f.v v16, v12, v0.t
; ZVFHMIN-NEXT: fsrm a0
; ZVFHMIN-NEXT: vfcvt.f.x.v v16, v16, v0.t
@@ -439,8 +439,8 @@ define <8 x float> @vp_roundtozero_v8f32(<8 x float> %va, <8 x i1> %m, i32 zeroe
; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, mu
; CHECK-NEXT: vmflt.vf v10, v12, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 1
-; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
@@ -483,8 +483,8 @@ define <16 x float> @vp_roundtozero_v16f32(<16 x float> %va, <16 x i1> %m, i32 z
; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, mu
; CHECK-NEXT: vmflt.vf v12, v16, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 1
-; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
@@ -561,16 +561,16 @@ declare <4 x double> @llvm.vp.roundtozero.v4f64(<4 x double>, <4 x i1>, i32)
define <4 x double> @vp_roundtozero_v4f64(<4 x double> %va, <4 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_roundtozero_v4f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v10, v0
; CHECK-NEXT: lui a1, %hi(.LCPI18_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI18_0)(a1)
+; CHECK-NEXT: vmv1r.v v10, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, ma
; CHECK-NEXT: vfabs.v v12, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, mu
; CHECK-NEXT: vmflt.vf v10, v12, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 1
-; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
@@ -605,16 +605,16 @@ declare <8 x double> @llvm.vp.roundtozero.v8f64(<8 x double>, <8 x i1>, i32)
define <8 x double> @vp_roundtozero_v8f64(<8 x double> %va, <8 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_roundtozero_v8f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v12, v0
; CHECK-NEXT: lui a1, %hi(.LCPI20_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI20_0)(a1)
+; CHECK-NEXT: vmv1r.v v12, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, ma
; CHECK-NEXT: vfabs.v v16, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, mu
; CHECK-NEXT: vmflt.vf v12, v16, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 1
-; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
@@ -649,16 +649,16 @@ declare <15 x double> @llvm.vp.roundtozero.v15f64(<15 x double>, <15 x i1>, i32)
define <15 x double> @vp_roundtozero_v15f64(<15 x double> %va, <15 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_roundtozero_v15f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: lui a1, %hi(.LCPI22_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI22_0)(a1)
+; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
; CHECK-NEXT: vmflt.vf v16, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 1
-; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
@@ -693,16 +693,16 @@ declare <16 x double> @llvm.vp.roundtozero.v16f64(<16 x double>, <16 x i1>, i32)
define <16 x double> @vp_roundtozero_v16f64(<16 x double> %va, <16 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_roundtozero_v16f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: lui a1, %hi(.LCPI24_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI24_0)(a1)
+; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
; CHECK-NEXT: vmflt.vf v16, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 1
-; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
@@ -737,69 +737,59 @@ declare <32 x double> @llvm.vp.roundtozero.v32f64(<32 x double>, <32 x i1>, i32)
define <32 x double> @vp_roundtozero_v32f64(<32 x double> %va, <32 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_roundtozero_v32f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: addi sp, sp, -16
-; CHECK-NEXT: .cfi_def_cfa_offset 16
-; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a1, a1, 4
-; CHECK-NEXT: sub sp, sp, a1
-; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
-; CHECK-NEXT: vmv1r.v v25, v0
-; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a1, a1, 3
-; CHECK-NEXT: add a1, sp, a1
-; CHECK-NEXT: addi a1, a1, 16
-; CHECK-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
-; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
+; CHECK-NEXT: vmv1r.v v6, v0
; CHECK-NEXT: li a2, 16
-; CHECK-NEXT: vslidedown.vi v24, v0, 2
+; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
+; CHECK-NEXT: vslidedown.vi v7, v0, 2
; CHECK-NEXT: mv a1, a0
; CHECK-NEXT: bltu a0, a2, .LBB26_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: li a1, 16
; CHECK-NEXT: .LBB26_2:
+; CHECK-NEXT: addi sp, sp, -16
+; CHECK-NEXT: .cfi_def_cfa_offset 16
+; CHECK-NEXT: csrr a2, vlenb
+; CHECK-NEXT: slli a2, a2, 3
+; CHECK-NEXT: sub sp, sp, a2
+; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
; CHECK-NEXT: lui a2, %hi(.LCPI26_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI26_0)(a2)
+; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v25
-; CHECK-NEXT: vfabs.v v16, v8, v0.t
+; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vmflt.vf v25, v16, fa5, v0.t
+; CHECK-NEXT: vmflt.vf v6, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a1, 1
+; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v25
-; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
+; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
+; CHECK-NEXT: addi a2, sp, 16
+; CHECK-NEXT: vs8r.v v24, (a2) # Unknown-size Folded Spill
; CHECK-NEXT: fsrm a1
-; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
-; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vfsgnj.vv v8, v16, v8, v0.t
; CHECK-NEXT: addi a1, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
+; CHECK-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v24, v8, v0.t
; CHECK-NEXT: addi a1, a0, -16
; CHECK-NEXT: sltu a0, a0, a1
; CHECK-NEXT: addi a0, a0, -1
; CHECK-NEXT: and a0, a0, a1
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v24
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vfabs.v v8, v16, v0.t
+; CHECK-NEXT: vfabs.v v24, v16, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vmflt.vf v24, v8, fa5, v0.t
+; CHECK-NEXT: vmflt.vf v7, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v24
-; CHECK-NEXT: vfcvt.x.f.v v8, v16, v0.t
+; CHECK-NEXT: vfcvt.x.f.v v24, v16, v0.t
; CHECK-NEXT: fsrm a0
-; CHECK-NEXT: vfcvt.f.x.v v8, v8, v0.t
+; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vfsgnj.vv v16, v8, v16, v0.t
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vfsgnj.vv v16, v24, v16, v0.t
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 4
+; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add sp, sp, a0
; CHECK-NEXT: addi sp, sp, 16
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-sad.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-sad.ll
index a4ab67f41595d..80561be0ca2f5 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-sad.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-sad.ll
@@ -115,17 +115,17 @@ define signext i32 @sad_2block_16xi8_as_i32(ptr %a, ptr %b, i32 signext %stridea
; CHECK-NEXT: vwaddu.vv v10, v9, v8
; CHECK-NEXT: vminu.vv v8, v12, v13
; CHECK-NEXT: vmaxu.vv v9, v12, v13
-; CHECK-NEXT: vsub.vv v8, v9, v8
-; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; CHECK-NEXT: add a0, a0, a2
; CHECK-NEXT: add a1, a1, a3
-; CHECK-NEXT: vle8.v v9, (a0)
-; CHECK-NEXT: vle8.v v12, (a1)
+; CHECK-NEXT: vle8.v v12, (a0)
+; CHECK-NEXT: vle8.v v13, (a1)
+; CHECK-NEXT: vsub.vv v8, v9, v8
+; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; CHECK-NEXT: vzext.vf2 v14, v8
; CHECK-NEXT: vwaddu.vv v16, v14, v10
; CHECK-NEXT: vsetvli zero, zero, e8, m1, ta, ma
-; CHECK-NEXT: vminu.vv v8, v9, v12
-; CHECK-NEXT: vmaxu.vv v9, v9, v12
+; CHECK-NEXT: vminu.vv v8, v12, v13
+; CHECK-NEXT: vmaxu.vv v9, v12, v13
; CHECK-NEXT: vsub.vv v8, v9, v8
; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; CHECK-NEXT: vzext.vf2 v10, v8
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-setcc-fp-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-setcc-fp-vp.ll
index 4598bf67a2363..33e9cde4c30ab 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-setcc-fp-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-setcc-fp-vp.ll
@@ -1163,31 +1163,31 @@ define <128 x i1> @fcmp_oeq_vv_v128f16(<128 x half> %va, <128 x half> %vb, <128
; ZVFH-NEXT: addi a0, sp, 16
; ZVFH-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; ZVFH-NEXT: vsetivli zero, 8, e8, m1, ta, ma
-; ZVFH-NEXT: vslidedown.vi v7, v0, 8
+; ZVFH-NEXT: vslidedown.vi v6, v0, 8
; ZVFH-NEXT: mv a0, a2
; ZVFH-NEXT: bltu a2, a3, .LBB43_2
; ZVFH-NEXT: # %bb.1:
; ZVFH-NEXT: li a0, 64
; ZVFH-NEXT: .LBB43_2:
+; ZVFH-NEXT: addi a1, sp, 16
+; ZVFH-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
; ZVFH-NEXT: vsetvli zero, a0, e16, m8, ta, ma
-; ZVFH-NEXT: addi a0, sp, 16
-; ZVFH-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
-; ZVFH-NEXT: vmfeq.vv v6, v8, v24, v0.t
+; ZVFH-NEXT: vmfeq.vv v7, v8, v24, v0.t
; ZVFH-NEXT: addi a0, a2, -64
; ZVFH-NEXT: sltu a1, a2, a0
; ZVFH-NEXT: addi a1, a1, -1
; ZVFH-NEXT: and a0, a1, a0
+; ZVFH-NEXT: vmv1r.v v0, v6
+; ZVFH-NEXT: csrr a1, vlenb
+; ZVFH-NEXT: slli a1, a1, 3
+; ZVFH-NEXT: add a1, sp, a1
+; ZVFH-NEXT: addi a1, a1, 16
+; ZVFH-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
; ZVFH-NEXT: vsetvli zero, a0, e16, m8, ta, ma
-; ZVFH-NEXT: vmv1r.v v0, v7
-; ZVFH-NEXT: csrr a0, vlenb
-; ZVFH-NEXT: slli a0, a0, 3
-; ZVFH-NEXT: add a0, sp, a0
-; ZVFH-NEXT: addi a0, a0, 16
-; ZVFH-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; ZVFH-NEXT: vmfeq.vv v24, v16, v8, v0.t
+; ZVFH-NEXT: vmfeq.vv v8, v16, v24, v0.t
; ZVFH-NEXT: vsetivli zero, 16, e8, m1, ta, ma
-; ZVFH-NEXT: vslideup.vi v6, v24, 8
-; ZVFH-NEXT: vmv.v.v v0, v6
+; ZVFH-NEXT: vslideup.vi v7, v8, 8
+; ZVFH-NEXT: vmv.v.v v0, v7
; ZVFH-NEXT: csrr a0, vlenb
; ZVFH-NEXT: slli a0, a0, 4
; ZVFH-NEXT: add sp, sp, a0
@@ -2865,37 +2865,36 @@ define <32 x i1> @fcmp_oeq_vv_v32f64(<32 x double> %va, <32 x double> %vb, <32 x
; CHECK-NEXT: add a1, sp, a1
; CHECK-NEXT: addi a1, a1, 16
; CHECK-NEXT: vs8r.v v24, (a1) # Unknown-size Folded Spill
-; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
-; CHECK-NEXT: vslidedown.vi v7, v0, 2
-; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; CHECK-NEXT: vle64.v v24, (a0)
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: li a1, 16
+; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
+; CHECK-NEXT: vslidedown.vi v6, v0, 2
; CHECK-NEXT: mv a0, a2
; CHECK-NEXT: bltu a2, a1, .LBB87_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: li a0, 16
; CHECK-NEXT: .LBB87_2:
+; CHECK-NEXT: addi a1, sp, 16
+; CHECK-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vmfeq.vv v6, v8, v24, v0.t
+; CHECK-NEXT: vmfeq.vv v7, v8, v24, v0.t
; CHECK-NEXT: addi a0, a2, -16
; CHECK-NEXT: sltu a1, a2, a0
; CHECK-NEXT: addi a1, a1, -1
; CHECK-NEXT: and a0, a1, a0
+; CHECK-NEXT: vmv1r.v v0, v6
+; CHECK-NEXT: csrr a1, vlenb
+; CHECK-NEXT: slli a1, a1, 3
+; CHECK-NEXT: add a1, sp, a1
+; CHECK-NEXT: addi a1, a1, 16
+; CHECK-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v7
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vmfeq.vv v24, v16, v8, v0.t
+; CHECK-NEXT: vmfeq.vv v8, v16, v24, v0.t
; CHECK-NEXT: vsetivli zero, 4, e8, mf4, ta, ma
-; CHECK-NEXT: vslideup.vi v6, v24, 2
-; CHECK-NEXT: vmv1r.v v0, v6
+; CHECK-NEXT: vslideup.vi v7, v8, 2
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 4
; CHECK-NEXT: add sp, sp, a0
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-setcc-int-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-setcc-int-vp.ll
index 21bbca00921d6..5f3847e085055 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-setcc-int-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-setcc-int-vp.ll
@@ -611,10 +611,10 @@ define <256 x i1> @icmp_eq_vv_v256i8(<256 x i8> %va, <256 x i8> %vb, <256 x i1>
; CHECK-NEXT: vle8.v v8, (a2)
; CHECK-NEXT: addi a2, a3, -128
; CHECK-NEXT: sltu a4, a3, a2
-; CHECK-NEXT: addi a4, a4, -1
; CHECK-NEXT: vle8.v v24, (a0)
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: addi a4, a4, -1
; CHECK-NEXT: and a2, a4, a2
; CHECK-NEXT: vsetvli zero, a2, e8, m8, ta, ma
; CHECK-NEXT: vmseq.vv v6, v16, v8, v0.t
@@ -622,7 +622,6 @@ define <256 x i1> @icmp_eq_vv_v256i8(<256 x i8> %va, <256 x i8> %vb, <256 x i1>
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: li a3, 128
; CHECK-NEXT: .LBB51_2:
-; CHECK-NEXT: vsetvli zero, a3, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
@@ -631,6 +630,7 @@ define <256 x i1> @icmp_eq_vv_v256i8(<256 x i8> %va, <256 x i8> %vb, <256 x i1>
; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, a3, e8, m8, ta, ma
; CHECK-NEXT: vmseq.vv v16, v8, v24, v0.t
; CHECK-NEXT: vmv1r.v v0, v16
; CHECK-NEXT: vmv1r.v v8, v6
@@ -660,8 +660,8 @@ define <256 x i1> @icmp_eq_vx_v256i8(<256 x i8> %va, i8 %b, <256 x i1> %m, i32 z
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: li a2, 128
; CHECK-NEXT: .LBB52_2:
-; CHECK-NEXT: vsetvli zero, a2, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a2, e8, m8, ta, ma
; CHECK-NEXT: vmseq.vx v16, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v16
; CHECK-NEXT: vmv1r.v v8, v25
@@ -689,8 +689,8 @@ define <256 x i1> @icmp_eq_vx_swap_v256i8(<256 x i8> %va, i8 %b, <256 x i1> %m,
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: li a2, 128
; CHECK-NEXT: .LBB53_2:
-; CHECK-NEXT: vsetvli zero, a2, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a2, e8, m8, ta, ma
; CHECK-NEXT: vmseq.vx v16, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v16
; CHECK-NEXT: vmv1r.v v8, v25
@@ -1264,31 +1264,31 @@ define <64 x i1> @icmp_eq_vv_v64i32(<64 x i32> %va, <64 x i32> %vb, <64 x i1> %m
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
-; CHECK-NEXT: vslidedown.vi v7, v0, 4
+; CHECK-NEXT: vslidedown.vi v6, v0, 4
; CHECK-NEXT: mv a0, a2
; CHECK-NEXT: bltu a2, a3, .LBB99_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: li a0, 32
; CHECK-NEXT: .LBB99_2:
+; CHECK-NEXT: addi a1, sp, 16
+; CHECK-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vmseq.vv v6, v8, v24, v0.t
+; CHECK-NEXT: vmseq.vv v7, v8, v24, v0.t
; CHECK-NEXT: addi a0, a2, -32
; CHECK-NEXT: sltu a1, a2, a0
; CHECK-NEXT: addi a1, a1, -1
; CHECK-NEXT: and a0, a1, a0
+; CHECK-NEXT: vmv1r.v v0, v6
+; CHECK-NEXT: csrr a1, vlenb
+; CHECK-NEXT: slli a1, a1, 3
+; CHECK-NEXT: add a1, sp, a1
+; CHECK-NEXT: addi a1, a1, 16
+; CHECK-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v7
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vmseq.vv v24, v16, v8, v0.t
+; CHECK-NEXT: vmseq.vv v8, v16, v24, v0.t
; CHECK-NEXT: vsetivli zero, 8, e8, mf2, ta, ma
-; CHECK-NEXT: vslideup.vi v6, v24, 4
-; CHECK-NEXT: vmv1r.v v0, v6
+; CHECK-NEXT: vslideup.vi v7, v8, 4
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 4
; CHECK-NEXT: add sp, sp, a0
@@ -1301,8 +1301,8 @@ define <64 x i1> @icmp_eq_vv_v64i32(<64 x i32> %va, <64 x i32> %vb, <64 x i1> %m
define <64 x i1> @icmp_eq_vx_v64i32(<64 x i32> %va, i32 %b, <64 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: icmp_eq_vx_v64i32:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; CHECK-NEXT: li a3, 32
+; CHECK-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; CHECK-NEXT: vslidedown.vi v24, v0, 4
; CHECK-NEXT: mv a2, a1
; CHECK-NEXT: bltu a1, a3, .LBB100_2
@@ -1315,8 +1315,8 @@ define <64 x i1> @icmp_eq_vx_v64i32(<64 x i32> %va, i32 %b, <64 x i1> %m, i32 ze
; CHECK-NEXT: sltu a1, a1, a2
; CHECK-NEXT: addi a1, a1, -1
; CHECK-NEXT: and a1, a1, a2
-; CHECK-NEXT: vsetvli zero, a1, e32, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a1, e32, m8, ta, ma
; CHECK-NEXT: vmseq.vx v8, v16, a0, v0.t
; CHECK-NEXT: vsetivli zero, 8, e8, mf2, ta, ma
; CHECK-NEXT: vslideup.vi v25, v8, 4
@@ -1331,8 +1331,8 @@ define <64 x i1> @icmp_eq_vx_v64i32(<64 x i32> %va, i32 %b, <64 x i1> %m, i32 ze
define <64 x i1> @icmp_eq_vx_swap_v64i32(<64 x i32> %va, i32 %b, <64 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: icmp_eq_vx_swap_v64i32:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; CHECK-NEXT: li a3, 32
+; CHECK-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; CHECK-NEXT: vslidedown.vi v24, v0, 4
; CHECK-NEXT: mv a2, a1
; CHECK-NEXT: bltu a1, a3, .LBB101_2
@@ -1345,8 +1345,8 @@ define <64 x i1> @icmp_eq_vx_swap_v64i32(<64 x i32> %va, i32 %b, <64 x i1> %m, i
; CHECK-NEXT: sltu a1, a1, a2
; CHECK-NEXT: addi a1, a1, -1
; CHECK-NEXT: and a1, a1, a2
-; CHECK-NEXT: vsetvli zero, a1, e32, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a1, e32, m8, ta, ma
; CHECK-NEXT: vmseq.vx v8, v16, a0, v0.t
; CHECK-NEXT: vsetivli zero, 8, e8, mf2, ta, ma
; CHECK-NEXT: vslideup.vi v25, v8, 4
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-sext-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-sext-vp.ll
index 52596d8892411..d1980ee3b0a6f 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-sext-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-sext-vp.ll
@@ -151,8 +151,8 @@ declare <32 x i64> @llvm.vp.sext.v32i64.v32i32(<32 x i32>, <32 x i1>, i32)
define <32 x i64> @vsext_v32i64_v32i32(<32 x i32> %va, <32 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vsext_v32i64_v32i32:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; CHECK-NEXT: li a2, 16
+; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; CHECK-NEXT: vslidedown.vi v16, v0, 2
; CHECK-NEXT: mv a1, a0
; CHECK-NEXT: bltu a0, a2, .LBB12_2
@@ -167,8 +167,8 @@ define <32 x i64> @vsext_v32i64_v32i32(<32 x i32> %va, <32 x i1> %m, i32 zeroext
; CHECK-NEXT: and a0, a0, a1
; CHECK-NEXT: vsetivli zero, 16, e32, m8, ta, ma
; CHECK-NEXT: vslidedown.vi v8, v8, 16
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vsext.vf2 v16, v8, v0.t
; CHECK-NEXT: vmv8r.v v8, v24
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-shuffle-concat.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-shuffle-concat.ll
index 609b4e9824892..925366e8b1d50 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-shuffle-concat.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-shuffle-concat.ll
@@ -33,8 +33,8 @@ define <8 x i32> @concat_4xv2i32(<2 x i32> %a, <2 x i32> %b, <2 x i32> %c, <2 x
; VLS-LABEL: concat_4xv2i32:
; VLS: # %bb.0:
; VLS-NEXT: vmv1r.v v13, v10
-; VLS-NEXT: vsetivli zero, 4, e32, m1, ta, ma
; VLS-NEXT: vmv1r.v v12, v8
+; VLS-NEXT: vsetivli zero, 4, e32, m1, ta, ma
; VLS-NEXT: vslideup.vi v13, v11, 2
; VLS-NEXT: vslideup.vi v12, v9, 2
; VLS-NEXT: vmv2r.v v8, v12
@@ -147,8 +147,8 @@ define <16 x i32> @concat_8xv2i32(<2 x i32> %a, <2 x i32> %b, <2 x i32> %c, <2 x
; VLS-NEXT: vmv1r.v v19, v14
; VLS-NEXT: vmv1r.v v18, v12
; VLS-NEXT: vmv1r.v v17, v10
-; VLS-NEXT: vsetivli zero, 4, e32, m1, ta, ma
; VLS-NEXT: vmv1r.v v16, v8
+; VLS-NEXT: vsetivli zero, 4, e32, m1, ta, ma
; VLS-NEXT: vslideup.vi v19, v15, 2
; VLS-NEXT: vslideup.vi v18, v13, 2
; VLS-NEXT: vslideup.vi v17, v11, 2
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-shuffle-exact-vlen.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-shuffle-exact-vlen.ll
index 8499086994bc0..d461fa8378cff 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-shuffle-exact-vlen.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-shuffle-exact-vlen.ll
@@ -164,11 +164,10 @@ define <4 x i64> @m2_splat_into_slide_two_source_v2_lo(<4 x i64> %v1, <4 x i64>
define <4 x i64> @m2_splat_into_slide_two_source(<4 x i64> %v1, <4 x i64> %v2) vscale_range(2,2) {
; CHECK-LABEL: m2_splat_into_slide_two_source:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetivli zero, 4, e64, m2, ta, ma
-; CHECK-NEXT: vrgather.vi v12, v8, 0
; CHECK-NEXT: vsetivli zero, 1, e8, mf8, ta, ma
; CHECK-NEXT: vmv.v.i v0, 12
; CHECK-NEXT: vsetivli zero, 4, e64, m2, ta, mu
+; CHECK-NEXT: vrgather.vi v12, v8, 0
; CHECK-NEXT: vslideup.vi v12, v10, 1, v0.t
; CHECK-NEXT: vmv.v.v v8, v12
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-shuffle-reverse.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-shuffle-reverse.ll
index 47d7baade8b49..d70ed2fb0e266 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-shuffle-reverse.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-shuffle-reverse.ll
@@ -101,10 +101,10 @@ define <32 x i1> @reverse_v32i1(<32 x i1> %a) {
; NO-ZVBB-LABEL: reverse_v32i1:
; NO-ZVBB: # %bb.0:
; NO-ZVBB-NEXT: li a0, 32
+; NO-ZVBB-NEXT: lui a1, %hi(.LCPI4_0)
+; NO-ZVBB-NEXT: addi a1, a1, %lo(.LCPI4_0)
; NO-ZVBB-NEXT: vsetvli zero, a0, e8, m2, ta, ma
-; NO-ZVBB-NEXT: lui a0, %hi(.LCPI4_0)
-; NO-ZVBB-NEXT: addi a0, a0, %lo(.LCPI4_0)
-; NO-ZVBB-NEXT: vle8.v v8, (a0)
+; NO-ZVBB-NEXT: vle8.v v8, (a1)
; NO-ZVBB-NEXT: vmv.v.i v10, 0
; NO-ZVBB-NEXT: vmerge.vim v10, v10, 1, v0
; NO-ZVBB-NEXT: vrgather.vv v12, v10, v8
@@ -124,10 +124,10 @@ define <64 x i1> @reverse_v64i1(<64 x i1> %a) {
; NO-ZVBB-LABEL: reverse_v64i1:
; NO-ZVBB: # %bb.0:
; NO-ZVBB-NEXT: li a0, 64
+; NO-ZVBB-NEXT: lui a1, %hi(.LCPI5_0)
+; NO-ZVBB-NEXT: addi a1, a1, %lo(.LCPI5_0)
; NO-ZVBB-NEXT: vsetvli zero, a0, e8, m4, ta, ma
-; NO-ZVBB-NEXT: lui a0, %hi(.LCPI5_0)
-; NO-ZVBB-NEXT: addi a0, a0, %lo(.LCPI5_0)
-; NO-ZVBB-NEXT: vle8.v v8, (a0)
+; NO-ZVBB-NEXT: vle8.v v8, (a1)
; NO-ZVBB-NEXT: vmv.v.i v12, 0
; NO-ZVBB-NEXT: vmerge.vim v12, v12, 1, v0
; NO-ZVBB-NEXT: vrgather.vv v16, v12, v8
@@ -147,10 +147,10 @@ define <128 x i1> @reverse_v128i1(<128 x i1> %a) {
; CHECK-LABEL: reverse_v128i1:
; CHECK: # %bb.0:
; CHECK-NEXT: li a0, 128
+; CHECK-NEXT: lui a1, %hi(.LCPI6_0)
+; CHECK-NEXT: addi a1, a1, %lo(.LCPI6_0)
; CHECK-NEXT: vsetvli zero, a0, e8, m8, ta, ma
-; CHECK-NEXT: lui a0, %hi(.LCPI6_0)
-; CHECK-NEXT: addi a0, a0, %lo(.LCPI6_0)
-; CHECK-NEXT: vle8.v v8, (a0)
+; CHECK-NEXT: vle8.v v8, (a1)
; CHECK-NEXT: vmv.v.i v16, 0
; CHECK-NEXT: vmerge.vim v16, v16, 1, v0
; CHECK-NEXT: vrgather.vv v24, v16, v8
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-shuffle-transpose.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-shuffle-transpose.ll
index 038fead011d89..8c74d8f538525 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-shuffle-transpose.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-shuffle-transpose.ll
@@ -19,8 +19,8 @@ define <8 x i8> @trn1.v8i8(<8 x i8> %v0, <8 x i8> %v1) {
define <8 x i8> @trn2.v8i8(<8 x i8> %v0, <8 x i8> %v1) {
; CHECK-LABEL: trn2.v8i8:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetivli zero, 8, e8, mf2, ta, ma
; CHECK-NEXT: li a0, 170
+; CHECK-NEXT: vsetivli zero, 8, e8, mf2, ta, ma
; CHECK-NEXT: vmv.s.x v0, a0
; CHECK-NEXT: vslidedown.vi v8, v8, 1
; CHECK-NEXT: vmerge.vvm v8, v8, v9, v0
@@ -46,13 +46,12 @@ define <16 x i8> @trn1.v16i8(<16 x i8> %v0, <16 x i8> %v1) {
define <16 x i8> @trn2.v16i8(<16 x i8> %v0, <16 x i8> %v1) {
; CHECK-LABEL: trn2.v16i8:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetivli zero, 16, e8, m1, ta, ma
-; CHECK-NEXT: vslidedown.vi v8, v8, 1
; CHECK-NEXT: lui a0, 11
; CHECK-NEXT: addi a0, a0, -1366
-; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, ma
+; CHECK-NEXT: vsetivli zero, 1, e16, m1, ta, ma
; CHECK-NEXT: vmv.s.x v0, a0
-; CHECK-NEXT: vsetvli zero, zero, e8, m1, ta, ma
+; CHECK-NEXT: vsetivli zero, 16, e8, m1, ta, ma
+; CHECK-NEXT: vslidedown.vi v8, v8, 1
; CHECK-NEXT: vmerge.vvm v8, v8, v9, v0
; CHECK-NEXT: ret
%tmp0 = shufflevector <16 x i8> %v0, <16 x i8> %v1, <16 x i32> <i32 1, i32 17, i32 3, i32 19, i32 5, i32 21, i32 7, i32 23, i32 9, i32 25, i32 11, i32 27, i32 13, i32 29, i32 15, i32 31>
@@ -74,8 +73,9 @@ define <4 x i16> @trn1.v4i16(<4 x i16> %v0, <4 x i16> %v1) {
define <4 x i16> @trn2.v4i16(<4 x i16> %v0, <4 x i16> %v1) {
; CHECK-LABEL: trn2.v4i16:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetivli zero, 4, e16, mf2, ta, ma
+; CHECK-NEXT: vsetivli zero, 1, e8, mf8, ta, ma
; CHECK-NEXT: vmv.v.i v0, 10
+; CHECK-NEXT: vsetivli zero, 4, e16, mf2, ta, ma
; CHECK-NEXT: vslidedown.vi v8, v8, 1
; CHECK-NEXT: vmerge.vvm v8, v8, v9, v0
; CHECK-NEXT: ret
@@ -98,8 +98,8 @@ define <8 x i16> @trn1.v8i16(<8 x i16> %v0, <8 x i16> %v1) {
define <8 x i16> @trn2.v8i16(<8 x i16> %v0, <8 x i16> %v1) {
; CHECK-LABEL: trn2.v8i16:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetivli zero, 8, e16, m1, ta, ma
; CHECK-NEXT: li a0, 170
+; CHECK-NEXT: vsetivli zero, 8, e16, m1, ta, ma
; CHECK-NEXT: vmv.s.x v0, a0
; CHECK-NEXT: vslidedown.vi v8, v8, 1
; CHECK-NEXT: vmerge.vvm v8, v8, v9, v0
@@ -121,8 +121,9 @@ define <2 x i32> @trn1.v2i32(<2 x i32> %v0, <2 x i32> %v1) {
define <2 x i32> @trn2.v2i32(<2 x i32> %v0, <2 x i32> %v1) {
; CHECK-LABEL: trn2.v2i32:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
+; CHECK-NEXT: vsetivli zero, 1, e8, mf8, ta, ma
; CHECK-NEXT: vmv.v.i v0, 2
+; CHECK-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
; CHECK-NEXT: vrgather.vi v10, v8, 1
; CHECK-NEXT: vmerge.vvm v8, v10, v9, v0
; CHECK-NEXT: ret
@@ -145,8 +146,9 @@ define <4 x i32> @trn1.v4i32(<4 x i32> %v0, <4 x i32> %v1) {
define <4 x i32> @trn2.v4i32(<4 x i32> %v0, <4 x i32> %v1) {
; CHECK-LABEL: trn2.v4i32:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetivli zero, 4, e32, m1, ta, ma
+; CHECK-NEXT: vsetivli zero, 1, e8, mf8, ta, ma
; CHECK-NEXT: vmv.v.i v0, 10
+; CHECK-NEXT: vsetivli zero, 4, e32, m1, ta, ma
; CHECK-NEXT: vslidedown.vi v8, v8, 1
; CHECK-NEXT: vmerge.vvm v8, v8, v9, v0
; CHECK-NEXT: ret
@@ -167,8 +169,9 @@ define <2 x i64> @trn1.v2i64(<2 x i64> %v0, <2 x i64> %v1) {
define <2 x i64> @trn2.v2i64(<2 x i64> %v0, <2 x i64> %v1) {
; CHECK-LABEL: trn2.v2i64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetivli zero, 2, e64, m1, ta, ma
+; CHECK-NEXT: vsetivli zero, 1, e8, mf8, ta, ma
; CHECK-NEXT: vmv.v.i v0, 2
+; CHECK-NEXT: vsetivli zero, 2, e64, m1, ta, ma
; CHECK-NEXT: vrgather.vi v10, v8, 1
; CHECK-NEXT: vmerge.vvm v8, v10, v9, v0
; CHECK-NEXT: ret
@@ -189,8 +192,9 @@ define <2 x float> @trn1.v2f32(<2 x float> %v0, <2 x float> %v1) {
define <2 x float> @trn2.v2f32(<2 x float> %v0, <2 x float> %v1) {
; CHECK-LABEL: trn2.v2f32:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
+; CHECK-NEXT: vsetivli zero, 1, e8, mf8, ta, ma
; CHECK-NEXT: vmv.v.i v0, 2
+; CHECK-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
; CHECK-NEXT: vrgather.vi v10, v8, 1
; CHECK-NEXT: vmerge.vvm v8, v10, v9, v0
; CHECK-NEXT: ret
@@ -213,8 +217,9 @@ define <4 x float> @trn1.v4f32(<4 x float> %v0, <4 x float> %v1) {
define <4 x float> @trn2.v4f32(<4 x float> %v0, <4 x float> %v1) {
; CHECK-LABEL: trn2.v4f32:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetivli zero, 4, e32, m1, ta, ma
+; CHECK-NEXT: vsetivli zero, 1, e8, mf8, ta, ma
; CHECK-NEXT: vmv.v.i v0, 10
+; CHECK-NEXT: vsetivli zero, 4, e32, m1, ta, ma
; CHECK-NEXT: vslidedown.vi v8, v8, 1
; CHECK-NEXT: vmerge.vvm v8, v8, v9, v0
; CHECK-NEXT: ret
@@ -235,8 +240,9 @@ define <2 x double> @trn1.v2f64(<2 x double> %v0, <2 x double> %v1) {
define <2 x double> @trn2.v2f64(<2 x double> %v0, <2 x double> %v1) {
; CHECK-LABEL: trn2.v2f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetivli zero, 2, e64, m1, ta, ma
+; CHECK-NEXT: vsetivli zero, 1, e8, mf8, ta, ma
; CHECK-NEXT: vmv.v.i v0, 2
+; CHECK-NEXT: vsetivli zero, 2, e64, m1, ta, ma
; CHECK-NEXT: vrgather.vi v10, v8, 1
; CHECK-NEXT: vmerge.vvm v8, v10, v9, v0
; CHECK-NEXT: ret
@@ -259,8 +265,9 @@ define <4 x half> @trn1.v4f16(<4 x half> %v0, <4 x half> %v1) {
define <4 x half> @trn2.v4f16(<4 x half> %v0, <4 x half> %v1) {
; CHECK-LABEL: trn2.v4f16:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetivli zero, 4, e16, mf2, ta, ma
+; CHECK-NEXT: vsetivli zero, 1, e8, mf8, ta, ma
; CHECK-NEXT: vmv.v.i v0, 10
+; CHECK-NEXT: vsetivli zero, 4, e16, mf2, ta, ma
; CHECK-NEXT: vslidedown.vi v8, v8, 1
; CHECK-NEXT: vmerge.vvm v8, v8, v9, v0
; CHECK-NEXT: ret
@@ -283,8 +290,8 @@ define <8 x half> @trn1.v8f16(<8 x half> %v0, <8 x half> %v1) {
define <8 x half> @trn2.v8f16(<8 x half> %v0, <8 x half> %v1) {
; CHECK-LABEL: trn2.v8f16:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetivli zero, 8, e16, m1, ta, ma
; CHECK-NEXT: li a0, 170
+; CHECK-NEXT: vsetivli zero, 8, e16, m1, ta, ma
; CHECK-NEXT: vmv.s.x v0, a0
; CHECK-NEXT: vslidedown.vi v8, v8, 1
; CHECK-NEXT: vmerge.vvm v8, v8, v9, v0
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-sitofp-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-sitofp-vp.ll
index 5e93fdfc7a652..bf0eab77d0ac8 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-sitofp-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-sitofp-vp.ll
@@ -390,8 +390,8 @@ declare <32 x double> @llvm.vp.sitofp.v32f64.v32i64(<32 x i64>, <32 x i1>, i32)
define <32 x double> @vsitofp_v32f64_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vsitofp_v32f64_v32i64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; CHECK-NEXT: li a2, 16
+; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; CHECK-NEXT: vslidedown.vi v24, v0, 2
; CHECK-NEXT: mv a1, a0
; CHECK-NEXT: bltu a0, a2, .LBB25_2
@@ -404,8 +404,8 @@ define <32 x double> @vsitofp_v32f64_v32i64(<32 x i64> %va, <32 x i1> %m, i32 ze
; CHECK-NEXT: sltu a0, a0, a1
; CHECK-NEXT: addi a0, a0, -1
; CHECK-NEXT: and a0, a0, a1
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
; CHECK-NEXT: ret
%v = call <32 x double> @llvm.vp.sitofp.v32f64.v32i64(<32 x i64> %va, <32 x i1> %m, i32 %evl)
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-strided-load-combine.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-strided-load-combine.ll
index 0e6b03bf16323..0e1105848440a 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-strided-load-combine.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-strided-load-combine.ll
@@ -75,9 +75,9 @@ define void @widen_4xv4i16_unaligned(ptr %x, ptr %z) {
; CHECK-NO-MISALIGN: # %bb.0:
; CHECK-NO-MISALIGN-NEXT: vsetivli zero, 8, e16, m1, ta, ma
; CHECK-NO-MISALIGN-NEXT: vle8.v v8, (a0)
-; CHECK-NO-MISALIGN-NEXT: addi a2, a0, 16
-; CHECK-NO-MISALIGN-NEXT: vle8.v v10, (a2)
; CHECK-NO-MISALIGN-NEXT: addi a2, a0, 8
+; CHECK-NO-MISALIGN-NEXT: addi a3, a0, 16
+; CHECK-NO-MISALIGN-NEXT: vle8.v v10, (a3)
; CHECK-NO-MISALIGN-NEXT: addi a0, a0, 24
; CHECK-NO-MISALIGN-NEXT: vle8.v v9, (a0)
; CHECK-NO-MISALIGN-NEXT: vle8.v v11, (a2)
@@ -186,9 +186,9 @@ define void @strided_constant_mismatch_4xv4i16(ptr %x, ptr %z) {
; CHECK: # %bb.0:
; CHECK-NEXT: vsetivli zero, 4, e16, mf2, ta, ma
; CHECK-NEXT: vle16.v v8, (a0)
-; CHECK-NEXT: addi a2, a0, 6
-; CHECK-NEXT: vle16.v v10, (a2)
; CHECK-NEXT: addi a2, a0, 2
+; CHECK-NEXT: addi a3, a0, 6
+; CHECK-NEXT: vle16.v v10, (a3)
; CHECK-NEXT: addi a0, a0, 8
; CHECK-NEXT: vle16.v v9, (a0)
; CHECK-NEXT: vle16.v v11, (a2)
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-strided-vpload.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-strided-vpload.ll
index 6a8d2008de74d..5e64e9fbc1a2f 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-strided-vpload.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-strided-vpload.ll
@@ -480,14 +480,14 @@ define <32 x double> @strided_vpload_v32f64(ptr %ptr, i32 signext %stride, <32 x
; CHECK-NEXT: addi a5, a2, -16
; CHECK-NEXT: sltu a2, a2, a5
; CHECK-NEXT: addi a2, a2, -1
-; CHECK-NEXT: and a2, a2, a5
; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; CHECK-NEXT: vslidedown.vi v8, v9, 2
-; CHECK-NEXT: vsetvli zero, a2, e64, m8, ta, ma
+; CHECK-NEXT: and a2, a2, a5
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a2, e64, m8, ta, ma
; CHECK-NEXT: vlse64.v v16, (a4), a1, v0.t
-; CHECK-NEXT: vsetvli zero, a3, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a3, e64, m8, ta, ma
; CHECK-NEXT: vlse64.v v8, (a0), a1, v0.t
; CHECK-NEXT: ret
%load = call <32 x double> @llvm.experimental.vp.strided.load.v32f64.p0.i32(ptr %ptr, i32 %stride, <32 x i1> %m, i32 %evl)
@@ -555,13 +555,13 @@ define <33 x double> @strided_load_v33f64(ptr %ptr, i64 %stride, <33 x i1> %mask
; CHECK-RV32-NEXT: li a4, 16
; CHECK-RV32-NEXT: .LBB42_6:
; CHECK-RV32-NEXT: mul a5, a4, a2
-; CHECK-RV32-NEXT: add a5, a1, a5
; CHECK-RV32-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; CHECK-RV32-NEXT: vslidedown.vi v0, v8, 2
+; CHECK-RV32-NEXT: add a5, a1, a5
; CHECK-RV32-NEXT: vsetvli zero, a3, e64, m8, ta, ma
; CHECK-RV32-NEXT: vlse64.v v24, (a5), a2, v0.t
-; CHECK-RV32-NEXT: vsetvli zero, a4, e64, m8, ta, ma
; CHECK-RV32-NEXT: vmv1r.v v0, v8
+; CHECK-RV32-NEXT: vsetvli zero, a4, e64, m8, ta, ma
; CHECK-RV32-NEXT: vlse64.v v8, (a1), a2, v0.t
; CHECK-RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; CHECK-RV32-NEXT: vse64.v v8, (a0)
@@ -605,13 +605,13 @@ define <33 x double> @strided_load_v33f64(ptr %ptr, i64 %stride, <33 x i1> %mask
; CHECK-RV64-NEXT: li a3, 16
; CHECK-RV64-NEXT: .LBB42_6:
; CHECK-RV64-NEXT: mul a5, a3, a2
-; CHECK-RV64-NEXT: add a5, a1, a5
; CHECK-RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; CHECK-RV64-NEXT: vslidedown.vi v0, v8, 2
+; CHECK-RV64-NEXT: add a5, a1, a5
; CHECK-RV64-NEXT: vsetvli zero, a4, e64, m8, ta, ma
; CHECK-RV64-NEXT: vlse64.v v24, (a5), a2, v0.t
-; CHECK-RV64-NEXT: vsetvli zero, a3, e64, m8, ta, ma
; CHECK-RV64-NEXT: vmv1r.v v0, v8
+; CHECK-RV64-NEXT: vsetvli zero, a3, e64, m8, ta, ma
; CHECK-RV64-NEXT: vlse64.v v8, (a1), a2, v0.t
; CHECK-RV64-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; CHECK-RV64-NEXT: vse64.v v8, (a0)
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-strided-vpstore.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-strided-vpstore.ll
index dee422a4c17d1..35f123f1157f2 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-strided-vpstore.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-strided-vpstore.ll
@@ -420,9 +420,9 @@ define void @strided_store_v32f64(<32 x double> %v, ptr %ptr, i32 signext %strid
; CHECK-NEXT: addi a3, a2, -16
; CHECK-NEXT: sltu a2, a2, a3
; CHECK-NEXT: addi a2, a2, -1
-; CHECK-NEXT: and a2, a2, a3
; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; CHECK-NEXT: vslidedown.vi v0, v0, 2
+; CHECK-NEXT: and a2, a2, a3
; CHECK-NEXT: vsetvli zero, a2, e64, m8, ta, ma
; CHECK-NEXT: vsse64.v v16, (a0), a1, v0.t
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-trunc-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-trunc-vp.ll
index 9fa8ab39723f7..7513d31b54bd1 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-trunc-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-trunc-vp.ll
@@ -54,8 +54,8 @@ define <128 x i7> @vtrunc_v128i7_v128i16(<128 x i16> %a, <128 x i1> %m, i32 zero
; CHECK-LABEL: vtrunc_v128i7_v128i16:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv8r.v v24, v8
-; CHECK-NEXT: vsetivli zero, 8, e8, m1, ta, ma
; CHECK-NEXT: li a1, 64
+; CHECK-NEXT: vsetivli zero, 8, e8, m1, ta, ma
; CHECK-NEXT: vslidedown.vi v12, v0, 8
; CHECK-NEXT: mv a2, a0
; CHECK-NEXT: bltu a0, a1, .LBB4_2
@@ -68,8 +68,8 @@ define <128 x i7> @vtrunc_v128i7_v128i16(<128 x i16> %a, <128 x i1> %m, i32 zero
; CHECK-NEXT: sltu a0, a0, a2
; CHECK-NEXT: addi a0, a0, -1
; CHECK-NEXT: and a0, a0, a2
-; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, ma
; CHECK-NEXT: vnsrl.wi v24, v16, 0, v0.t
; CHECK-NEXT: li a0, 128
; CHECK-NEXT: vsetvli zero, a0, e8, m8, ta, ma
@@ -243,75 +243,67 @@ define <128 x i32> @vtrunc_v128i32_v128i64(<128 x i64> %a, <128 x i1> %m, i32 ze
; CHECK-NEXT: addi a2, a2, 16
; CHECK-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
; CHECK-NEXT: vsetivli zero, 8, e8, m1, ta, ma
-; CHECK-NEXT: vslidedown.vi v5, v0, 8
-; CHECK-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
-; CHECK-NEXT: vslidedown.vi v26, v0, 4
+; CHECK-NEXT: vslidedown.vi v25, v0, 8
; CHECK-NEXT: addi a2, a1, 512
-; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; CHECK-NEXT: vle64.v v8, (a2)
-; CHECK-NEXT: csrr a2, vlenb
-; CHECK-NEXT: li a3, 48
-; CHECK-NEXT: mul a2, a2, a3
-; CHECK-NEXT: add a2, sp, a2
-; CHECK-NEXT: addi a2, a2, 16
-; CHECK-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
; CHECK-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
-; CHECK-NEXT: vslidedown.vi v27, v5, 4
-; CHECK-NEXT: addi a2, a1, 640
+; CHECK-NEXT: vslidedown.vi v27, v25, 4
+; CHECK-NEXT: addi a3, a1, 640
; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; CHECK-NEXT: vle64.v v16, (a2)
+; CHECK-NEXT: vle64.v v8, (a3)
; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
-; CHECK-NEXT: addi a2, a7, -64
-; CHECK-NEXT: sltu a3, a7, a2
-; CHECK-NEXT: addi a3, a3, -1
-; CHECK-NEXT: and a4, a3, a2
-; CHECK-NEXT: addi a2, a4, -32
-; CHECK-NEXT: sltu a3, a4, a2
-; CHECK-NEXT: addi a3, a3, -1
-; CHECK-NEXT: and a3, a3, a2
-; CHECK-NEXT: addi a2, a3, -16
-; CHECK-NEXT: sltu a5, a3, a2
-; CHECK-NEXT: addi a5, a5, -1
-; CHECK-NEXT: and a2, a5, a2
; CHECK-NEXT: vslidedown.vi v0, v27, 2
-; CHECK-NEXT: vsetvli zero, a2, e32, m4, ta, ma
-; CHECK-NEXT: vnsrl.wi v8, v16, 0, v0.t
-; CHECK-NEXT: csrr a2, vlenb
-; CHECK-NEXT: li a5, 24
-; CHECK-NEXT: mul a2, a2, a5
-; CHECK-NEXT: add a2, sp, a2
-; CHECK-NEXT: addi a2, a2, 16
-; CHECK-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
-; CHECK-NEXT: li a2, 16
+; CHECK-NEXT: addi a3, a7, -64
+; CHECK-NEXT: sltu a4, a7, a3
+; CHECK-NEXT: addi a4, a4, -1
+; CHECK-NEXT: and a4, a4, a3
+; CHECK-NEXT: addi a3, a4, -32
+; CHECK-NEXT: sltu a5, a4, a3
+; CHECK-NEXT: addi a5, a5, -1
+; CHECK-NEXT: and a3, a5, a3
+; CHECK-NEXT: addi a5, a3, -16
+; CHECK-NEXT: sltu a6, a3, a5
+; CHECK-NEXT: addi a6, a6, -1
+; CHECK-NEXT: and a5, a6, a5
+; CHECK-NEXT: vsetvli zero, a5, e32, m4, ta, ma
+; CHECK-NEXT: vnsrl.wi v16, v8, 0, v0.t
+; CHECK-NEXT: csrr a5, vlenb
+; CHECK-NEXT: li a6, 24
+; CHECK-NEXT: mul a5, a5, a6
+; CHECK-NEXT: add a5, sp, a5
+; CHECK-NEXT: addi a5, a5, 16
+; CHECK-NEXT: vs8r.v v16, (a5) # Unknown-size Folded Spill
+; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; CHECK-NEXT: vle64.v v8, (a2)
; CHECK-NEXT: addi a5, a1, 128
+; CHECK-NEXT: li a2, 16
+; CHECK-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
+; CHECK-NEXT: vslidedown.vi v26, v7, 4
; CHECK-NEXT: bltu a3, a2, .LBB16_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: li a3, 16
; CHECK-NEXT: .LBB16_2:
-; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
-; CHECK-NEXT: vslidedown.vi v28, v26, 2
; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; CHECK-NEXT: vle64.v v8, (a5)
-; CHECK-NEXT: addi a5, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a5) # Unknown-size Folded Spill
-; CHECK-NEXT: vsetvli zero, a3, e32, m4, ta, ma
-; CHECK-NEXT: li a3, 64
-; CHECK-NEXT: vmv1r.v v0, v27
+; CHECK-NEXT: vle64.v v16, (a5)
; CHECK-NEXT: csrr a5, vlenb
; CHECK-NEXT: li a6, 48
; CHECK-NEXT: mul a5, a5, a6
; CHECK-NEXT: add a5, sp, a5
; CHECK-NEXT: addi a5, a5, 16
-; CHECK-NEXT: vl8r.v v8, (a5) # Unknown-size Folded Reload
+; CHECK-NEXT: vs8r.v v16, (a5) # Unknown-size Folded Spill
+; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
+; CHECK-NEXT: vslidedown.vi v28, v26, 2
+; CHECK-NEXT: li a5, 64
+; CHECK-NEXT: vmv1r.v v0, v27
+; CHECK-NEXT: vsetvli zero, a3, e32, m4, ta, ma
; CHECK-NEXT: vnsrl.wi v16, v8, 0, v0.t
-; CHECK-NEXT: csrr a5, vlenb
+; CHECK-NEXT: csrr a3, vlenb
; CHECK-NEXT: li a6, 56
-; CHECK-NEXT: mul a5, a5, a6
-; CHECK-NEXT: add a5, sp, a5
-; CHECK-NEXT: addi a5, a5, 16
-; CHECK-NEXT: vs8r.v v16, (a5) # Unknown-size Folded Spill
+; CHECK-NEXT: mul a3, a3, a6
+; CHECK-NEXT: add a3, sp, a3
+; CHECK-NEXT: addi a3, a3, 16
+; CHECK-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
; CHECK-NEXT: mv a6, a7
-; CHECK-NEXT: bltu a7, a3, .LBB16_4
+; CHECK-NEXT: bltu a7, a5, .LBB16_4
; CHECK-NEXT: # %bb.3:
; CHECK-NEXT: li a6, 64
; CHECK-NEXT: .LBB16_4:
@@ -332,10 +324,14 @@ define <128 x i32> @vtrunc_v128i32_v128i64(<128 x i64> %a, <128 x i1> %m, i32 ze
; CHECK-NEXT: sltu t1, a6, t0
; CHECK-NEXT: addi t1, t1, -1
; CHECK-NEXT: and t0, t1, t0
-; CHECK-NEXT: vsetvli zero, t0, e32, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v28
-; CHECK-NEXT: addi t0, sp, 16
-; CHECK-NEXT: vl8r.v v16, (t0) # Unknown-size Folded Reload
+; CHECK-NEXT: csrr t1, vlenb
+; CHECK-NEXT: li t2, 48
+; CHECK-NEXT: mul t1, t1, t2
+; CHECK-NEXT: add t1, sp, t1
+; CHECK-NEXT: addi t1, t1, 16
+; CHECK-NEXT: vl8r.v v16, (t1) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, t0, e32, m4, ta, ma
; CHECK-NEXT: vnsrl.wi v8, v16, 0, v0.t
; CHECK-NEXT: csrr t0, vlenb
; CHECK-NEXT: slli t0, t0, 4
@@ -346,19 +342,21 @@ define <128 x i32> @vtrunc_v128i32_v128i64(<128 x i64> %a, <128 x i1> %m, i32 ze
; CHECK-NEXT: # %bb.5:
; CHECK-NEXT: li a6, 16
; CHECK-NEXT: .LBB16_6:
-; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
-; CHECK-NEXT: vslidedown.vi v20, v5, 2
; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; CHECK-NEXT: vle64.v v8, (a5)
+; CHECK-NEXT: addi a5, sp, 16
+; CHECK-NEXT: vs8r.v v8, (a5) # Unknown-size Folded Spill
; CHECK-NEXT: addi a1, a1, 256
-; CHECK-NEXT: vsetvli zero, a6, e32, m4, ta, ma
+; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
+; CHECK-NEXT: vslidedown.vi v27, v25, 2
; CHECK-NEXT: vmv1r.v v0, v26
; CHECK-NEXT: csrr a5, vlenb
; CHECK-NEXT: slli a5, a5, 3
; CHECK-NEXT: add a5, sp, a5
; CHECK-NEXT: addi a5, a5, 16
-; CHECK-NEXT: vl8r.v v24, (a5) # Unknown-size Folded Reload
-; CHECK-NEXT: vnsrl.wi v16, v24, 0, v0.t
+; CHECK-NEXT: vl8r.v v8, (a5) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, a6, e32, m4, ta, ma
+; CHECK-NEXT: vnsrl.wi v16, v8, 0, v0.t
; CHECK-NEXT: csrr a5, vlenb
; CHECK-NEXT: li a6, 48
; CHECK-NEXT: mul a5, a5, a6
@@ -371,13 +369,20 @@ define <128 x i32> @vtrunc_v128i32_v128i64(<128 x i64> %a, <128 x i1> %m, i32 ze
; CHECK-NEXT: li a5, 32
; CHECK-NEXT: .LBB16_8:
; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; CHECK-NEXT: vle64.v v24, (a1)
+; CHECK-NEXT: vle64.v v8, (a1)
+; CHECK-NEXT: csrr a1, vlenb
+; CHECK-NEXT: slli a1, a1, 3
+; CHECK-NEXT: add a1, sp, a1
+; CHECK-NEXT: addi a1, a1, 16
+; CHECK-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
; CHECK-NEXT: addi a1, a5, -16
; CHECK-NEXT: sltu a5, a5, a1
; CHECK-NEXT: addi a5, a5, -1
; CHECK-NEXT: and a1, a5, a1
+; CHECK-NEXT: vmv1r.v v0, v27
+; CHECK-NEXT: addi a5, sp, 16
+; CHECK-NEXT: vl8r.v v8, (a5) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, a1, e32, m4, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v20
; CHECK-NEXT: vnsrl.wi v16, v8, 0, v0.t
; CHECK-NEXT: bltu a4, a2, .LBB16_10
; CHECK-NEXT: # %bb.9:
@@ -385,8 +390,13 @@ define <128 x i32> @vtrunc_v128i32_v128i64(<128 x i64> %a, <128 x i1> %m, i32 ze
; CHECK-NEXT: .LBB16_10:
; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; CHECK-NEXT: vslidedown.vi v6, v7, 2
+; CHECK-NEXT: vmv1r.v v0, v25
+; CHECK-NEXT: csrr a1, vlenb
+; CHECK-NEXT: slli a1, a1, 3
+; CHECK-NEXT: add a1, sp, a1
+; CHECK-NEXT: addi a1, a1, 16
+; CHECK-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, a4, e32, m4, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v5
; CHECK-NEXT: vnsrl.wi v8, v24, 0, v0.t
; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: mv a1, a7
@@ -401,13 +411,13 @@ define <128 x i32> @vtrunc_v128i32_v128i64(<128 x i64> %a, <128 x i1> %m, i32 ze
; CHECK-NEXT: addi a4, a4, 16
; CHECK-NEXT: vl8r.v v8, (a4) # Unknown-size Folded Reload
; CHECK-NEXT: vmv4r.v v24, v8
-; CHECK-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; CHECK-NEXT: csrr a4, vlenb
; CHECK-NEXT: li a5, 56
; CHECK-NEXT: mul a4, a4, a5
; CHECK-NEXT: add a4, sp, a4
; CHECK-NEXT: addi a4, a4, 16
; CHECK-NEXT: vl8r.v v8, (a4) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; CHECK-NEXT: vslideup.vi v8, v24, 16
; CHECK-NEXT: csrr a4, vlenb
; CHECK-NEXT: li a5, 56
@@ -446,19 +456,18 @@ define <128 x i32> @vtrunc_v128i32_v128i64(<128 x i64> %a, <128 x i1> %m, i32 ze
; CHECK-NEXT: sltu a1, a1, a4
; CHECK-NEXT: addi a1, a1, -1
; CHECK-NEXT: and a1, a1, a4
-; CHECK-NEXT: vsetvli zero, a1, e32, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v6
-; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a1, a1, 5
-; CHECK-NEXT: add a1, sp, a1
-; CHECK-NEXT: addi a1, a1, 16
-; CHECK-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
+; CHECK-NEXT: csrr a4, vlenb
+; CHECK-NEXT: slli a4, a4, 5
+; CHECK-NEXT: add a4, sp, a4
+; CHECK-NEXT: addi a4, a4, 16
+; CHECK-NEXT: vl8r.v v16, (a4) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, a1, e32, m4, ta, ma
; CHECK-NEXT: vnsrl.wi v8, v16, 0, v0.t
; CHECK-NEXT: bltu a7, a2, .LBB16_14
; CHECK-NEXT: # %bb.13:
; CHECK-NEXT: li a7, 16
; CHECK-NEXT: .LBB16_14:
-; CHECK-NEXT: vsetvli zero, a7, e32, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: csrr a1, vlenb
; CHECK-NEXT: li a2, 40
@@ -466,6 +475,7 @@ define <128 x i32> @vtrunc_v128i32_v128i64(<128 x i64> %a, <128 x i1> %m, i32 ze
; CHECK-NEXT: add a1, sp, a1
; CHECK-NEXT: addi a1, a1, 16
; CHECK-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, a7, e32, m4, ta, ma
; CHECK-NEXT: vnsrl.wi v16, v24, 0, v0.t
; CHECK-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; CHECK-NEXT: vslideup.vi v16, v8, 16
@@ -509,8 +519,8 @@ define <32 x i32> @vtrunc_v32i32_v32i64(<32 x i64> %a, <32 x i1> %m, i32 zeroext
; CHECK-LABEL: vtrunc_v32i32_v32i64:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv8r.v v24, v8
-; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; CHECK-NEXT: li a2, 16
+; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; CHECK-NEXT: vslidedown.vi v12, v0, 2
; CHECK-NEXT: mv a1, a0
; CHECK-NEXT: bltu a0, a2, .LBB17_2
@@ -523,8 +533,8 @@ define <32 x i32> @vtrunc_v32i32_v32i64(<32 x i64> %a, <32 x i1> %m, i32 zeroext
; CHECK-NEXT: sltu a0, a0, a1
; CHECK-NEXT: addi a0, a0, -1
; CHECK-NEXT: and a0, a0, a1
-; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, ma
; CHECK-NEXT: vnsrl.wi v24, v16, 0, v0.t
; CHECK-NEXT: li a0, 32
; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-uitofp-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-uitofp-vp.ll
index 698c48bc55650..e28d55f46abcb 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-uitofp-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-uitofp-vp.ll
@@ -390,8 +390,8 @@ declare <32 x double> @llvm.vp.uitofp.v32f64.v32i64(<32 x i64>, <32 x i1>, i32)
define <32 x double> @vuitofp_v32f64_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vuitofp_v32f64_v32i64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; CHECK-NEXT: li a2, 16
+; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; CHECK-NEXT: vslidedown.vi v24, v0, 2
; CHECK-NEXT: mv a1, a0
; CHECK-NEXT: bltu a0, a2, .LBB25_2
@@ -404,8 +404,8 @@ define <32 x double> @vuitofp_v32f64_v32i64(<32 x i64> %va, <32 x i1> %m, i32 ze
; CHECK-NEXT: sltu a0, a0, a1
; CHECK-NEXT: addi a0, a0, -1
; CHECK-NEXT: and a0, a0, a1
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfcvt.f.xu.v v16, v16, v0.t
; CHECK-NEXT: ret
%v = call <32 x double> @llvm.vp.uitofp.v32f64.v32i64(<32 x i64> %va, <32 x i1> %m, i32 %evl)
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vadd-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vadd-vp.ll
index 2c62cbd583d00..5601bd5ee7a3a 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vadd-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vadd-vp.ll
@@ -377,8 +377,8 @@ define <256 x i8> @vadd_vi_v258i8(<256 x i8> %va, <256 x i1> %m, i32 zeroext %ev
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: li a1, 128
; CHECK-NEXT: .LBB32_2:
-; CHECK-NEXT: vsetvli zero, a1, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a1, e8, m8, ta, ma
; CHECK-NEXT: vadd.vi v8, v8, -1, v0.t
; CHECK-NEXT: ret
%v = call <256 x i8> @llvm.vp.add.v258i8(<256 x i8> %va, <256 x i8> splat (i8 -1), <256 x i1> %m, i32 %evl)
@@ -416,8 +416,8 @@ define <256 x i8> @vadd_vi_v258i8_evl129(<256 x i8> %va, <256 x i1> %m) {
; CHECK-NEXT: vsetvli zero, a1, e8, m8, ta, ma
; CHECK-NEXT: vlm.v v24, (a0)
; CHECK-NEXT: vadd.vi v8, v8, -1, v0.t
-; CHECK-NEXT: vsetivli zero, 1, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetivli zero, 1, e8, m8, ta, ma
; CHECK-NEXT: vadd.vi v16, v16, -1, v0.t
; CHECK-NEXT: ret
%v = call <256 x i8> @llvm.vp.add.v258i8(<256 x i8> %va, <256 x i8> splat (i8 -1), <256 x i1> %m, i32 129)
@@ -1348,8 +1348,8 @@ declare <32 x i64> @llvm.vp.add.v32i64(<32 x i64>, <32 x i64>, <32 x i1>, i32)
define <32 x i64> @vadd_vx_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %evl) {
; RV32-LABEL: vadd_vx_v32i64:
; RV32: # %bb.0:
-; RV32-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV32-NEXT: li a2, 16
+; RV32-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV32-NEXT: vslidedown.vi v7, v0, 2
; RV32-NEXT: mv a1, a0
; RV32-NEXT: bltu a0, a2, .LBB108_2
@@ -1365,15 +1365,15 @@ define <32 x i64> @vadd_vx_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %evl
; RV32-NEXT: sltu a0, a0, a1
; RV32-NEXT: addi a0, a0, -1
; RV32-NEXT: and a0, a0, a1
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vmv1r.v v0, v7
+; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vadd.vv v16, v16, v24, v0.t
; RV32-NEXT: ret
;
; RV64-LABEL: vadd_vx_v32i64:
; RV64: # %bb.0:
-; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64-NEXT: li a2, 16
+; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64-NEXT: vslidedown.vi v24, v0, 2
; RV64-NEXT: mv a1, a0
; RV64-NEXT: bltu a0, a2, .LBB108_2
@@ -1386,8 +1386,8 @@ define <32 x i64> @vadd_vx_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %evl
; RV64-NEXT: sltu a0, a0, a1
; RV64-NEXT: addi a0, a0, -1
; RV64-NEXT: and a0, a0, a1
-; RV64-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV64-NEXT: vmv1r.v v0, v24
+; RV64-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV64-NEXT: vadd.vi v16, v16, -1, v0.t
; RV64-NEXT: ret
%v = call <32 x i64> @llvm.vp.add.v32i64(<32 x i64> %va, <32 x i64> splat (i64 -1), <32 x i1> %m, i32 %evl)
@@ -1468,8 +1468,8 @@ define <32 x i64> @vadd_vx_v32i64_evl27(<32 x i64> %va, <32 x i1> %m) {
; RV32-NEXT: vmv.v.i v24, -1
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV32-NEXT: vadd.vv v8, v8, v24, v0.t
-; RV32-NEXT: vsetivli zero, 11, e64, m8, ta, ma
; RV32-NEXT: vmv1r.v v0, v7
+; RV32-NEXT: vsetivli zero, 11, e64, m8, ta, ma
; RV32-NEXT: vadd.vv v16, v16, v24, v0.t
; RV32-NEXT: ret
;
@@ -1479,8 +1479,8 @@ define <32 x i64> @vadd_vx_v32i64_evl27(<32 x i64> %va, <32 x i1> %m) {
; RV64-NEXT: vslidedown.vi v24, v0, 2
; RV64-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV64-NEXT: vadd.vi v8, v8, -1, v0.t
-; RV64-NEXT: vsetivli zero, 11, e64, m8, ta, ma
; RV64-NEXT: vmv1r.v v0, v24
+; RV64-NEXT: vsetivli zero, 11, e64, m8, ta, ma
; RV64-NEXT: vadd.vi v16, v16, -1, v0.t
; RV64-NEXT: ret
%v = call <32 x i64> @llvm.vp.add.v32i64(<32 x i64> %va, <32 x i64> splat (i64 -1), <32 x i1> %m, i32 27)
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vand-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vand-vp.ll
index 507cf5cc6b80c..d414be76672ab 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vand-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vand-vp.ll
@@ -1140,15 +1140,16 @@ define <11 x i64> @vand_vx_v11i64(<11 x i64> %va, i64 %b, <11 x i1> %m, i32 zero
; RV32-LABEL: vand_vx_v11i64:
; RV32: # %bb.0:
; RV32-NEXT: vmv1r.v v16, v0
-; RV32-NEXT: li a3, 32
-; RV32-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; RV32-NEXT: lui a3, 341
; RV32-NEXT: addi a3, a3, 1365
+; RV32-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; RV32-NEXT: vmv.s.x v0, a3
+; RV32-NEXT: li a3, 32
+; RV32-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; RV32-NEXT: vmv.v.x v24, a1
; RV32-NEXT: vmerge.vxm v24, v24, a0, v0
-; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
; RV32-NEXT: vmv1r.v v0, v16
+; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
; RV32-NEXT: vand.vv v8, v8, v24, v0.t
; RV32-NEXT: ret
;
@@ -1167,10 +1168,11 @@ define <11 x i64> @vand_vx_v11i64_unmasked(<11 x i64> %va, i64 %b, i32 zeroext %
; RV32-LABEL: vand_vx_v11i64_unmasked:
; RV32: # %bb.0:
; RV32-NEXT: li a3, 32
+; RV32-NEXT: lui a4, 341
+; RV32-NEXT: addi a4, a4, 1365
+; RV32-NEXT: vsetivli zero, 1, e32, m1, ta, ma
+; RV32-NEXT: vmv.s.x v0, a4
; RV32-NEXT: vsetvli zero, a3, e32, m8, ta, ma
-; RV32-NEXT: lui a3, 341
-; RV32-NEXT: addi a3, a3, 1365
-; RV32-NEXT: vmv.s.x v0, a3
; RV32-NEXT: vmv.v.x v16, a1
; RV32-NEXT: vmerge.vxm v16, v16, a0, v0
; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vcopysign-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vcopysign-vp.ll
index 01b07b4081e6d..77a095303675f 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vcopysign-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vcopysign-vp.ll
@@ -298,37 +298,46 @@ define <32 x double> @vfsgnj_vv_v32f64(<32 x double> %va, <32 x double> %vb, <32
; CHECK-NEXT: addi sp, sp, -16
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a1, a1, 3
+; CHECK-NEXT: slli a1, a1, 4
; CHECK-NEXT: sub sp, sp, a1
-; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
; CHECK-NEXT: addi a1, a0, 128
; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; CHECK-NEXT: vle64.v v24, (a1)
-; CHECK-NEXT: addi a1, sp, 16
+; CHECK-NEXT: csrr a1, vlenb
+; CHECK-NEXT: slli a1, a1, 3
+; CHECK-NEXT: add a1, sp, a1
+; CHECK-NEXT: addi a1, a1, 16
; CHECK-NEXT: vs8r.v v24, (a1) # Unknown-size Folded Spill
-; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
-; CHECK-NEXT: vslidedown.vi v7, v0, 2
-; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; CHECK-NEXT: vle64.v v24, (a0)
+; CHECK-NEXT: addi a0, sp, 16
+; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: li a1, 16
+; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
+; CHECK-NEXT: vslidedown.vi v7, v0, 2
; CHECK-NEXT: mv a0, a2
; CHECK-NEXT: bltu a2, a1, .LBB26_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: li a0, 16
; CHECK-NEXT: .LBB26_2:
+; CHECK-NEXT: addi a1, sp, 16
+; CHECK-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfsgnj.vv v8, v8, v24, v0.t
; CHECK-NEXT: addi a0, a2, -16
; CHECK-NEXT: sltu a1, a2, a0
; CHECK-NEXT: addi a1, a1, -1
; CHECK-NEXT: and a0, a1, a0
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v7
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: csrr a1, vlenb
+; CHECK-NEXT: slli a1, a1, 3
+; CHECK-NEXT: add a1, sp, a1
+; CHECK-NEXT: addi a1, a1, 16
+; CHECK-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfsgnj.vv v16, v16, v24, v0.t
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
+; CHECK-NEXT: slli a0, a0, 4
; CHECK-NEXT: add sp, sp, a0
; CHECK-NEXT: addi sp, sp, 16
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfabs-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfabs-vp.ll
index f32e2bbf37946..ae3dce497c6d0 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfabs-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfabs-vp.ll
@@ -379,8 +379,8 @@ declare <32 x double> @llvm.vp.fabs.v32f64(<32 x double>, <32 x i1>, i32)
define <32 x double> @vfabs_vv_v32f64(<32 x double> %va, <32 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfabs_vv_v32f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; CHECK-NEXT: li a2, 16
+; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; CHECK-NEXT: vslidedown.vi v24, v0, 2
; CHECK-NEXT: mv a1, a0
; CHECK-NEXT: bltu a0, a2, .LBB26_2
@@ -393,8 +393,8 @@ define <32 x double> @vfabs_vv_v32f64(<32 x double> %va, <32 x i1> %m, i32 zeroe
; CHECK-NEXT: sltu a0, a0, a1
; CHECK-NEXT: addi a0, a0, -1
; CHECK-NEXT: and a0, a0, a1
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfabs.v v16, v16, v0.t
; CHECK-NEXT: ret
%v = call <32 x double> @llvm.vp.fabs.v32f64(<32 x double> %va, <32 x i1> %m, i32 %evl)
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfma-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfma-vp.ll
index 0574773fb2fd9..e2e48cee3eacc 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfma-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfma-vp.ll
@@ -862,51 +862,51 @@ define <32 x double> @vfma_vv_v32f64(<32 x double> %va, <32 x double> %b, <32 x
; CHECK-NEXT: add a1, sp, a1
; CHECK-NEXT: addi a1, a1, 16
; CHECK-NEXT: vs8r.v v24, (a1) # Unknown-size Folded Spill
-; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
-; CHECK-NEXT: vslidedown.vi v7, v0, 2
-; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; CHECK-NEXT: vle64.v v24, (a2)
; CHECK-NEXT: addi a1, sp, 16
; CHECK-NEXT: vs8r.v v24, (a1) # Unknown-size Folded Spill
; CHECK-NEXT: vle64.v v24, (a0)
-; CHECK-NEXT: li a1, 16
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: li a1, 16
; CHECK-NEXT: mv a0, a4
+; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
+; CHECK-NEXT: vslidedown.vi v7, v0, 2
; CHECK-NEXT: bltu a4, a1, .LBB50_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: li a0, 16
; CHECK-NEXT: .LBB50_2:
+; CHECK-NEXT: addi a1, sp, 16
+; CHECK-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vfmadd.vv v24, v8, v16, v0.t
+; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: addi a0, a4, -16
; CHECK-NEXT: sltu a1, a4, a0
; CHECK-NEXT: addi a1, a1, -1
; CHECK-NEXT: and a0, a1, a0
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v7
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a1, 24
-; CHECK-NEXT: mul a0, a0, a1
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 4
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: csrr a1, vlenb
+; CHECK-NEXT: li a2, 24
+; CHECK-NEXT: mul a1, a1, a2
+; CHECK-NEXT: add a1, sp, a1
+; CHECK-NEXT: addi a1, a1, 16
+; CHECK-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
+; CHECK-NEXT: csrr a1, vlenb
+; CHECK-NEXT: slli a1, a1, 4
+; CHECK-NEXT: add a1, sp, a1
+; CHECK-NEXT: addi a1, a1, 16
+; CHECK-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
+; CHECK-NEXT: csrr a1, vlenb
+; CHECK-NEXT: slli a1, a1, 3
+; CHECK-NEXT: add a1, sp, a1
+; CHECK-NEXT: addi a1, a1, 16
+; CHECK-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfmadd.vv v16, v24, v8, v0.t
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
@@ -954,25 +954,25 @@ define <32 x double> @vfma_vv_v32f64_unmasked(<32 x double> %va, <32 x double> %
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: li a0, 16
; CHECK-NEXT: .LBB51_2:
+; CHECK-NEXT: csrr a1, vlenb
+; CHECK-NEXT: slli a1, a1, 3
+; CHECK-NEXT: add a1, sp, a1
+; CHECK-NEXT: addi a1, a1, 16
+; CHECK-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vfmadd.vv v0, v8, v24
; CHECK-NEXT: addi a0, a4, -16
; CHECK-NEXT: sltu a1, a4, a0
; CHECK-NEXT: addi a1, a1, -1
; CHECK-NEXT: and a0, a1, a0
+; CHECK-NEXT: csrr a1, vlenb
+; CHECK-NEXT: slli a1, a1, 4
+; CHECK-NEXT: add a1, sp, a1
+; CHECK-NEXT: addi a1, a1, 16
+; CHECK-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
+; CHECK-NEXT: addi a1, sp, 16
+; CHECK-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 4
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vfmadd.vv v24, v16, v8
; CHECK-NEXT: vmv8r.v v8, v0
; CHECK-NEXT: vmv.v.v v16, v24
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfmax-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfmax-vp.ll
index ffa88e28d7dc8..c83a298cb501e 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfmax-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfmax-vp.ll
@@ -390,37 +390,46 @@ define <32 x double> @vfmax_vv_v32f64(<32 x double> %va, <32 x double> %vb, <32
; CHECK-NEXT: addi sp, sp, -16
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a1, a1, 3
+; CHECK-NEXT: slli a1, a1, 4
; CHECK-NEXT: sub sp, sp, a1
-; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
; CHECK-NEXT: addi a1, a0, 128
; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; CHECK-NEXT: vle64.v v24, (a1)
-; CHECK-NEXT: addi a1, sp, 16
+; CHECK-NEXT: csrr a1, vlenb
+; CHECK-NEXT: slli a1, a1, 3
+; CHECK-NEXT: add a1, sp, a1
+; CHECK-NEXT: addi a1, a1, 16
; CHECK-NEXT: vs8r.v v24, (a1) # Unknown-size Folded Spill
-; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
-; CHECK-NEXT: vslidedown.vi v7, v0, 2
-; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; CHECK-NEXT: vle64.v v24, (a0)
+; CHECK-NEXT: addi a0, sp, 16
+; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: li a1, 16
+; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
+; CHECK-NEXT: vslidedown.vi v7, v0, 2
; CHECK-NEXT: mv a0, a2
; CHECK-NEXT: bltu a2, a1, .LBB26_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: li a0, 16
; CHECK-NEXT: .LBB26_2:
+; CHECK-NEXT: addi a1, sp, 16
+; CHECK-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfmax.vv v8, v8, v24, v0.t
; CHECK-NEXT: addi a0, a2, -16
; CHECK-NEXT: sltu a1, a2, a0
; CHECK-NEXT: addi a1, a1, -1
; CHECK-NEXT: and a0, a1, a0
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v7
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: csrr a1, vlenb
+; CHECK-NEXT: slli a1, a1, 3
+; CHECK-NEXT: add a1, sp, a1
+; CHECK-NEXT: addi a1, a1, 16
+; CHECK-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfmax.vv v16, v16, v24, v0.t
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
+; CHECK-NEXT: slli a0, a0, 4
; CHECK-NEXT: add sp, sp, a0
; CHECK-NEXT: addi sp, sp, 16
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfmin-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfmin-vp.ll
index 17f851e172f81..60dbededb90a5 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfmin-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfmin-vp.ll
@@ -390,37 +390,46 @@ define <32 x double> @vfmin_vv_v32f64(<32 x double> %va, <32 x double> %vb, <32
; CHECK-NEXT: addi sp, sp, -16
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a1, a1, 3
+; CHECK-NEXT: slli a1, a1, 4
; CHECK-NEXT: sub sp, sp, a1
-; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
; CHECK-NEXT: addi a1, a0, 128
; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; CHECK-NEXT: vle64.v v24, (a1)
-; CHECK-NEXT: addi a1, sp, 16
+; CHECK-NEXT: csrr a1, vlenb
+; CHECK-NEXT: slli a1, a1, 3
+; CHECK-NEXT: add a1, sp, a1
+; CHECK-NEXT: addi a1, a1, 16
; CHECK-NEXT: vs8r.v v24, (a1) # Unknown-size Folded Spill
-; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
-; CHECK-NEXT: vslidedown.vi v7, v0, 2
-; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; CHECK-NEXT: vle64.v v24, (a0)
+; CHECK-NEXT: addi a0, sp, 16
+; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: li a1, 16
+; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
+; CHECK-NEXT: vslidedown.vi v7, v0, 2
; CHECK-NEXT: mv a0, a2
; CHECK-NEXT: bltu a2, a1, .LBB26_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: li a0, 16
; CHECK-NEXT: .LBB26_2:
+; CHECK-NEXT: addi a1, sp, 16
+; CHECK-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfmin.vv v8, v8, v24, v0.t
; CHECK-NEXT: addi a0, a2, -16
; CHECK-NEXT: sltu a1, a2, a0
; CHECK-NEXT: addi a1, a1, -1
; CHECK-NEXT: and a0, a1, a0
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v7
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: csrr a1, vlenb
+; CHECK-NEXT: slli a1, a1, 3
+; CHECK-NEXT: add a1, sp, a1
+; CHECK-NEXT: addi a1, a1, 16
+; CHECK-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfmin.vv v16, v16, v24, v0.t
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
+; CHECK-NEXT: slli a0, a0, 4
; CHECK-NEXT: add sp, sp, a0
; CHECK-NEXT: addi sp, sp, 16
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfmuladd-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfmuladd-vp.ll
index 288efb0f1fc27..6c695b43d2718 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfmuladd-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfmuladd-vp.ll
@@ -626,51 +626,51 @@ define <32 x double> @vfma_vv_v32f64(<32 x double> %va, <32 x double> %b, <32 x
; CHECK-NEXT: add a1, sp, a1
; CHECK-NEXT: addi a1, a1, 16
; CHECK-NEXT: vs8r.v v24, (a1) # Unknown-size Folded Spill
-; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
-; CHECK-NEXT: vslidedown.vi v7, v0, 2
-; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; CHECK-NEXT: vle64.v v24, (a2)
; CHECK-NEXT: addi a1, sp, 16
; CHECK-NEXT: vs8r.v v24, (a1) # Unknown-size Folded Spill
; CHECK-NEXT: vle64.v v24, (a0)
-; CHECK-NEXT: li a1, 16
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: li a1, 16
; CHECK-NEXT: mv a0, a4
+; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
+; CHECK-NEXT: vslidedown.vi v7, v0, 2
; CHECK-NEXT: bltu a4, a1, .LBB50_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: li a0, 16
; CHECK-NEXT: .LBB50_2:
+; CHECK-NEXT: addi a1, sp, 16
+; CHECK-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vfmadd.vv v24, v8, v16, v0.t
+; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: addi a0, a4, -16
; CHECK-NEXT: sltu a1, a4, a0
; CHECK-NEXT: addi a1, a1, -1
; CHECK-NEXT: and a0, a1, a0
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v7
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a1, 24
-; CHECK-NEXT: mul a0, a0, a1
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 4
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: csrr a1, vlenb
+; CHECK-NEXT: li a2, 24
+; CHECK-NEXT: mul a1, a1, a2
+; CHECK-NEXT: add a1, sp, a1
+; CHECK-NEXT: addi a1, a1, 16
+; CHECK-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
+; CHECK-NEXT: csrr a1, vlenb
+; CHECK-NEXT: slli a1, a1, 4
+; CHECK-NEXT: add a1, sp, a1
+; CHECK-NEXT: addi a1, a1, 16
+; CHECK-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
+; CHECK-NEXT: csrr a1, vlenb
+; CHECK-NEXT: slli a1, a1, 3
+; CHECK-NEXT: add a1, sp, a1
+; CHECK-NEXT: addi a1, a1, 16
+; CHECK-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfmadd.vv v16, v24, v8, v0.t
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
@@ -718,25 +718,25 @@ define <32 x double> @vfma_vv_v32f64_unmasked(<32 x double> %va, <32 x double> %
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: li a0, 16
; CHECK-NEXT: .LBB51_2:
+; CHECK-NEXT: csrr a1, vlenb
+; CHECK-NEXT: slli a1, a1, 3
+; CHECK-NEXT: add a1, sp, a1
+; CHECK-NEXT: addi a1, a1, 16
+; CHECK-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vfmadd.vv v0, v8, v24
; CHECK-NEXT: addi a0, a4, -16
; CHECK-NEXT: sltu a1, a4, a0
; CHECK-NEXT: addi a1, a1, -1
; CHECK-NEXT: and a0, a1, a0
+; CHECK-NEXT: csrr a1, vlenb
+; CHECK-NEXT: slli a1, a1, 4
+; CHECK-NEXT: add a1, sp, a1
+; CHECK-NEXT: addi a1, a1, 16
+; CHECK-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
+; CHECK-NEXT: addi a1, sp, 16
+; CHECK-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 4
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vfmadd.vv v24, v16, v8
; CHECK-NEXT: vmv8r.v v8, v0
; CHECK-NEXT: vmv.v.v v16, v24
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfneg-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfneg-vp.ll
index c36ec25c04f93..fbc4c56a91134 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfneg-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfneg-vp.ll
@@ -379,8 +379,8 @@ declare <32 x double> @llvm.vp.fneg.v32f64(<32 x double>, <32 x i1>, i32)
define <32 x double> @vfneg_vv_v32f64(<32 x double> %va, <32 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfneg_vv_v32f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; CHECK-NEXT: li a2, 16
+; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; CHECK-NEXT: vslidedown.vi v24, v0, 2
; CHECK-NEXT: mv a1, a0
; CHECK-NEXT: bltu a0, a2, .LBB26_2
@@ -393,8 +393,8 @@ define <32 x double> @vfneg_vv_v32f64(<32 x double> %va, <32 x i1> %m, i32 zeroe
; CHECK-NEXT: sltu a0, a0, a1
; CHECK-NEXT: addi a0, a0, -1
; CHECK-NEXT: and a0, a0, a1
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfneg.v v16, v16, v0.t
; CHECK-NEXT: ret
%v = call <32 x double> @llvm.vp.fneg.v32f64(<32 x double> %va, <32 x i1> %m, i32 %evl)
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfsqrt-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfsqrt-vp.ll
index 6004eb4fe217a..988b200ae5365 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfsqrt-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfsqrt-vp.ll
@@ -379,8 +379,8 @@ declare <32 x double> @llvm.vp.sqrt.v32f64(<32 x double>, <32 x i1>, i32)
define <32 x double> @vfsqrt_vv_v32f64(<32 x double> %va, <32 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfsqrt_vv_v32f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; CHECK-NEXT: li a2, 16
+; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; CHECK-NEXT: vslidedown.vi v24, v0, 2
; CHECK-NEXT: mv a1, a0
; CHECK-NEXT: bltu a0, a2, .LBB26_2
@@ -393,8 +393,8 @@ define <32 x double> @vfsqrt_vv_v32f64(<32 x double> %va, <32 x i1> %m, i32 zero
; CHECK-NEXT: sltu a0, a0, a1
; CHECK-NEXT: addi a0, a0, -1
; CHECK-NEXT: and a0, a0, a1
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfsqrt.v v16, v16, v0.t
; CHECK-NEXT: ret
%v = call <32 x double> @llvm.vp.sqrt.v32f64(<32 x double> %va, <32 x i1> %m, i32 %evl)
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfwadd.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfwadd.ll
index dd3a50cfd7737..05c7bd990642c 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfwadd.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfwadd.ll
@@ -105,13 +105,12 @@ define <64 x float> @vfwadd_v64f16(ptr %x, ptr %y) {
; CHECK-NEXT: vslidedown.vx v16, v8, a0
; CHECK-NEXT: vslidedown.vx v8, v0, a0
; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, ma
-; CHECK-NEXT: vmv4r.v v24, v8
-; CHECK-NEXT: vfwadd.vv v8, v16, v24
+; CHECK-NEXT: vfwadd.vv v24, v16, v8
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vfwadd.vv v8, v16, v0
@@ -216,13 +215,12 @@ define <32 x double> @vfwadd_v32f32(ptr %x, ptr %y) {
; CHECK-NEXT: vslidedown.vi v16, v8, 16
; CHECK-NEXT: vslidedown.vi v8, v0, 16
; CHECK-NEXT: vsetivli zero, 16, e32, m4, ta, ma
-; CHECK-NEXT: vmv4r.v v24, v8
-; CHECK-NEXT: vfwadd.vv v8, v16, v24
+; CHECK-NEXT: vfwadd.vv v24, v16, v8
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vfwadd.vv v8, v16, v0
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfwmul.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfwmul.ll
index 7eaa1856ce221..5a57801d33b40 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfwmul.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfwmul.ll
@@ -105,13 +105,12 @@ define <64 x float> @vfwmul_v64f16(ptr %x, ptr %y) {
; CHECK-NEXT: vslidedown.vx v16, v8, a0
; CHECK-NEXT: vslidedown.vx v8, v0, a0
; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, ma
-; CHECK-NEXT: vmv4r.v v24, v8
-; CHECK-NEXT: vfwmul.vv v8, v16, v24
+; CHECK-NEXT: vfwmul.vv v24, v16, v8
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vfwmul.vv v8, v16, v0
@@ -216,13 +215,12 @@ define <32 x double> @vfwmul_v32f32(ptr %x, ptr %y) {
; CHECK-NEXT: vslidedown.vi v16, v8, 16
; CHECK-NEXT: vslidedown.vi v8, v0, 16
; CHECK-NEXT: vsetivli zero, 16, e32, m4, ta, ma
-; CHECK-NEXT: vmv4r.v v24, v8
-; CHECK-NEXT: vfwmul.vv v8, v16, v24
+; CHECK-NEXT: vfwmul.vv v24, v16, v8
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vfwmul.vv v8, v16, v0
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfwsub.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfwsub.ll
index 8cf7c5f175865..2c706cad9742f 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfwsub.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfwsub.ll
@@ -105,13 +105,12 @@ define <64 x float> @vfwsub_v64f16(ptr %x, ptr %y) {
; CHECK-NEXT: vslidedown.vx v16, v8, a0
; CHECK-NEXT: vslidedown.vx v8, v0, a0
; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, ma
-; CHECK-NEXT: vmv4r.v v24, v8
-; CHECK-NEXT: vfwsub.vv v8, v16, v24
+; CHECK-NEXT: vfwsub.vv v24, v16, v8
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vfwsub.vv v8, v16, v0
@@ -216,13 +215,12 @@ define <32 x double> @vfwsub_v32f32(ptr %x, ptr %y) {
; CHECK-NEXT: vslidedown.vi v16, v8, 16
; CHECK-NEXT: vslidedown.vi v8, v0, 16
; CHECK-NEXT: vsetivli zero, 16, e32, m4, ta, ma
-; CHECK-NEXT: vmv4r.v v24, v8
-; CHECK-NEXT: vfwsub.vv v8, v16, v24
+; CHECK-NEXT: vfwsub.vv v24, v16, v8
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vfwsub.vv v8, v16, v0
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vmax-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vmax-vp.ll
index 3db44e87109bd..9789afda9344a 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vmax-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vmax-vp.ll
@@ -282,8 +282,8 @@ define <256 x i8> @vmax_vx_v258i8(<256 x i8> %va, i8 %b, <256 x i1> %m, i32 zero
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: li a2, 128
; CHECK-NEXT: .LBB22_2:
-; CHECK-NEXT: vsetvli zero, a2, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a2, e8, m8, ta, ma
; CHECK-NEXT: vmax.vx v8, v8, a0, v0.t
; CHECK-NEXT: ret
%elt.head = insertelement <256 x i8> poison, i8 %b, i32 0
@@ -325,8 +325,8 @@ define <256 x i8> @vmax_vx_v258i8_evl129(<256 x i8> %va, i8 %b, <256 x i1> %m) {
; CHECK-NEXT: vsetvli zero, a2, e8, m8, ta, ma
; CHECK-NEXT: vlm.v v24, (a1)
; CHECK-NEXT: vmax.vx v8, v8, a0, v0.t
-; CHECK-NEXT: vsetivli zero, 1, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetivli zero, 1, e8, m8, ta, ma
; CHECK-NEXT: vmax.vx v16, v16, a0, v0.t
; CHECK-NEXT: ret
%elt.head = insertelement <256 x i8> poison, i8 %b, i32 0
@@ -1021,8 +1021,8 @@ declare <32 x i64> @llvm.vp.smax.v32i64(<32 x i64>, <32 x i64>, <32 x i1>, i32)
define <32 x i64> @vmax_vx_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %evl) {
; RV32-LABEL: vmax_vx_v32i64:
; RV32: # %bb.0:
-; RV32-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV32-NEXT: li a2, 16
+; RV32-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV32-NEXT: vslidedown.vi v7, v0, 2
; RV32-NEXT: mv a1, a0
; RV32-NEXT: bltu a0, a2, .LBB74_2
@@ -1038,15 +1038,15 @@ define <32 x i64> @vmax_vx_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %evl
; RV32-NEXT: sltu a0, a0, a1
; RV32-NEXT: addi a0, a0, -1
; RV32-NEXT: and a0, a0, a1
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vmv1r.v v0, v7
+; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vmax.vv v16, v16, v24, v0.t
; RV32-NEXT: ret
;
; RV64-LABEL: vmax_vx_v32i64:
; RV64: # %bb.0:
-; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64-NEXT: li a2, 16
+; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64-NEXT: vslidedown.vi v24, v0, 2
; RV64-NEXT: mv a1, a0
; RV64-NEXT: bltu a0, a2, .LBB74_2
@@ -1060,8 +1060,8 @@ define <32 x i64> @vmax_vx_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %evl
; RV64-NEXT: sltu a0, a0, a1
; RV64-NEXT: addi a0, a0, -1
; RV64-NEXT: and a0, a0, a1
-; RV64-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV64-NEXT: vmv1r.v v0, v24
+; RV64-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV64-NEXT: vmax.vx v16, v16, a2, v0.t
; RV64-NEXT: ret
%v = call <32 x i64> @llvm.vp.smax.v32i64(<32 x i64> %va, <32 x i64> splat (i64 -1), <32 x i1> %m, i32 %evl)
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vmaxu-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vmaxu-vp.ll
index c97c2232715f5..36b0a4642b616 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vmaxu-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vmaxu-vp.ll
@@ -281,8 +281,8 @@ define <256 x i8> @vmaxu_vx_v258i8(<256 x i8> %va, i8 %b, <256 x i1> %m, i32 zer
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: li a2, 128
; CHECK-NEXT: .LBB22_2:
-; CHECK-NEXT: vsetvli zero, a2, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a2, e8, m8, ta, ma
; CHECK-NEXT: vmaxu.vx v8, v8, a0, v0.t
; CHECK-NEXT: ret
%elt.head = insertelement <256 x i8> poison, i8 %b, i32 0
@@ -324,8 +324,8 @@ define <256 x i8> @vmaxu_vx_v258i8_evl129(<256 x i8> %va, i8 %b, <256 x i1> %m)
; CHECK-NEXT: vsetvli zero, a2, e8, m8, ta, ma
; CHECK-NEXT: vlm.v v24, (a1)
; CHECK-NEXT: vmaxu.vx v8, v8, a0, v0.t
-; CHECK-NEXT: vsetivli zero, 1, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetivli zero, 1, e8, m8, ta, ma
; CHECK-NEXT: vmaxu.vx v16, v16, a0, v0.t
; CHECK-NEXT: ret
%elt.head = insertelement <256 x i8> poison, i8 %b, i32 0
@@ -1020,8 +1020,8 @@ declare <32 x i64> @llvm.vp.umax.v32i64(<32 x i64>, <32 x i64>, <32 x i1>, i32)
define <32 x i64> @vmaxu_vx_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %evl) {
; RV32-LABEL: vmaxu_vx_v32i64:
; RV32: # %bb.0:
-; RV32-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV32-NEXT: li a2, 16
+; RV32-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV32-NEXT: vslidedown.vi v7, v0, 2
; RV32-NEXT: mv a1, a0
; RV32-NEXT: bltu a0, a2, .LBB74_2
@@ -1037,15 +1037,15 @@ define <32 x i64> @vmaxu_vx_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %ev
; RV32-NEXT: sltu a0, a0, a1
; RV32-NEXT: addi a0, a0, -1
; RV32-NEXT: and a0, a0, a1
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vmv1r.v v0, v7
+; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vmaxu.vv v16, v16, v24, v0.t
; RV32-NEXT: ret
;
; RV64-LABEL: vmaxu_vx_v32i64:
; RV64: # %bb.0:
-; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64-NEXT: li a2, 16
+; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64-NEXT: vslidedown.vi v24, v0, 2
; RV64-NEXT: mv a1, a0
; RV64-NEXT: bltu a0, a2, .LBB74_2
@@ -1059,8 +1059,8 @@ define <32 x i64> @vmaxu_vx_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %ev
; RV64-NEXT: sltu a0, a0, a1
; RV64-NEXT: addi a0, a0, -1
; RV64-NEXT: and a0, a0, a1
-; RV64-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV64-NEXT: vmv1r.v v0, v24
+; RV64-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV64-NEXT: vmaxu.vx v16, v16, a2, v0.t
; RV64-NEXT: ret
%v = call <32 x i64> @llvm.vp.umax.v32i64(<32 x i64> %va, <32 x i64> splat (i64 -1), <32 x i1> %m, i32 %evl)
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vmin-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vmin-vp.ll
index eaa19110a2a28..adb0a30f34d35 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vmin-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vmin-vp.ll
@@ -282,8 +282,8 @@ define <256 x i8> @vmin_vx_v258i8(<256 x i8> %va, i8 %b, <256 x i1> %m, i32 zero
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: li a2, 128
; CHECK-NEXT: .LBB22_2:
-; CHECK-NEXT: vsetvli zero, a2, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a2, e8, m8, ta, ma
; CHECK-NEXT: vmin.vx v8, v8, a0, v0.t
; CHECK-NEXT: ret
%elt.head = insertelement <256 x i8> poison, i8 %b, i32 0
@@ -325,8 +325,8 @@ define <256 x i8> @vmin_vx_v258i8_evl129(<256 x i8> %va, i8 %b, <256 x i1> %m) {
; CHECK-NEXT: vsetvli zero, a2, e8, m8, ta, ma
; CHECK-NEXT: vlm.v v24, (a1)
; CHECK-NEXT: vmin.vx v8, v8, a0, v0.t
-; CHECK-NEXT: vsetivli zero, 1, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetivli zero, 1, e8, m8, ta, ma
; CHECK-NEXT: vmin.vx v16, v16, a0, v0.t
; CHECK-NEXT: ret
%elt.head = insertelement <256 x i8> poison, i8 %b, i32 0
@@ -1021,8 +1021,8 @@ declare <32 x i64> @llvm.vp.smin.v32i64(<32 x i64>, <32 x i64>, <32 x i1>, i32)
define <32 x i64> @vmin_vx_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %evl) {
; RV32-LABEL: vmin_vx_v32i64:
; RV32: # %bb.0:
-; RV32-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV32-NEXT: li a2, 16
+; RV32-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV32-NEXT: vslidedown.vi v7, v0, 2
; RV32-NEXT: mv a1, a0
; RV32-NEXT: bltu a0, a2, .LBB74_2
@@ -1038,15 +1038,15 @@ define <32 x i64> @vmin_vx_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %evl
; RV32-NEXT: sltu a0, a0, a1
; RV32-NEXT: addi a0, a0, -1
; RV32-NEXT: and a0, a0, a1
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vmv1r.v v0, v7
+; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vmin.vv v16, v16, v24, v0.t
; RV32-NEXT: ret
;
; RV64-LABEL: vmin_vx_v32i64:
; RV64: # %bb.0:
-; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64-NEXT: li a2, 16
+; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64-NEXT: vslidedown.vi v24, v0, 2
; RV64-NEXT: mv a1, a0
; RV64-NEXT: bltu a0, a2, .LBB74_2
@@ -1060,8 +1060,8 @@ define <32 x i64> @vmin_vx_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %evl
; RV64-NEXT: sltu a0, a0, a1
; RV64-NEXT: addi a0, a0, -1
; RV64-NEXT: and a0, a0, a1
-; RV64-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV64-NEXT: vmv1r.v v0, v24
+; RV64-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV64-NEXT: vmin.vx v16, v16, a2, v0.t
; RV64-NEXT: ret
%v = call <32 x i64> @llvm.vp.smin.v32i64(<32 x i64> %va, <32 x i64> splat (i64 -1), <32 x i1> %m, i32 %evl)
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vminu-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vminu-vp.ll
index 48175e5b905ba..671ce82d4ae79 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vminu-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vminu-vp.ll
@@ -281,8 +281,8 @@ define <256 x i8> @vminu_vx_v258i8(<256 x i8> %va, i8 %b, <256 x i1> %m, i32 zer
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: li a2, 128
; CHECK-NEXT: .LBB22_2:
-; CHECK-NEXT: vsetvli zero, a2, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a2, e8, m8, ta, ma
; CHECK-NEXT: vminu.vx v8, v8, a0, v0.t
; CHECK-NEXT: ret
%elt.head = insertelement <256 x i8> poison, i8 %b, i32 0
@@ -324,8 +324,8 @@ define <256 x i8> @vminu_vx_v258i8_evl129(<256 x i8> %va, i8 %b, <256 x i1> %m)
; CHECK-NEXT: vsetvli zero, a2, e8, m8, ta, ma
; CHECK-NEXT: vlm.v v24, (a1)
; CHECK-NEXT: vminu.vx v8, v8, a0, v0.t
-; CHECK-NEXT: vsetivli zero, 1, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetivli zero, 1, e8, m8, ta, ma
; CHECK-NEXT: vminu.vx v16, v16, a0, v0.t
; CHECK-NEXT: ret
%elt.head = insertelement <256 x i8> poison, i8 %b, i32 0
@@ -1020,8 +1020,8 @@ declare <32 x i64> @llvm.vp.umin.v32i64(<32 x i64>, <32 x i64>, <32 x i1>, i32)
define <32 x i64> @vminu_vx_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %evl) {
; RV32-LABEL: vminu_vx_v32i64:
; RV32: # %bb.0:
-; RV32-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV32-NEXT: li a2, 16
+; RV32-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV32-NEXT: vslidedown.vi v7, v0, 2
; RV32-NEXT: mv a1, a0
; RV32-NEXT: bltu a0, a2, .LBB74_2
@@ -1037,15 +1037,15 @@ define <32 x i64> @vminu_vx_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %ev
; RV32-NEXT: sltu a0, a0, a1
; RV32-NEXT: addi a0, a0, -1
; RV32-NEXT: and a0, a0, a1
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vmv1r.v v0, v7
+; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vminu.vv v16, v16, v24, v0.t
; RV32-NEXT: ret
;
; RV64-LABEL: vminu_vx_v32i64:
; RV64: # %bb.0:
-; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64-NEXT: li a2, 16
+; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64-NEXT: vslidedown.vi v24, v0, 2
; RV64-NEXT: mv a1, a0
; RV64-NEXT: bltu a0, a2, .LBB74_2
@@ -1059,8 +1059,8 @@ define <32 x i64> @vminu_vx_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %ev
; RV64-NEXT: sltu a0, a0, a1
; RV64-NEXT: addi a0, a0, -1
; RV64-NEXT: and a0, a0, a1
-; RV64-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV64-NEXT: vmv1r.v v0, v24
+; RV64-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV64-NEXT: vminu.vx v16, v16, a2, v0.t
; RV64-NEXT: ret
%v = call <32 x i64> @llvm.vp.umin.v32i64(<32 x i64> %va, <32 x i64> splat (i64 -1), <32 x i1> %m, i32 %evl)
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vpgather.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vpgather.ll
index a13f1eed8efb1..028fb9a626f02 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vpgather.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vpgather.ll
@@ -297,10 +297,10 @@ define <32 x i8> @vpgather_baseidx_v32i8(ptr %base, <32 x i8> %idxs, <32 x i1> %
; RV64-NEXT: and a1, a1, a2
; RV64-NEXT: vsetivli zero, 16, e8, m2, ta, ma
; RV64-NEXT: vslidedown.vi v8, v8, 16
-; RV64-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV64-NEXT: vsext.vf8 v16, v8
; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64-NEXT: vslidedown.vi v0, v0, 2
+; RV64-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; RV64-NEXT: vsext.vf8 v16, v8
; RV64-NEXT: vsetvli zero, a1, e8, m1, ta, ma
; RV64-NEXT: vluxei64.v v8, (a0), v16, v0.t
; RV64-NEXT: li a0, 32
@@ -1882,10 +1882,10 @@ define <32 x double> @vpgather_v32f64(<32 x ptr> %ptrs, <32 x i1> %m, i32 zeroex
; RV32-NEXT: sltu a0, a0, a1
; RV32-NEXT: addi a0, a0, -1
; RV32-NEXT: and a0, a0, a1
-; RV32-NEXT: vsetivli zero, 16, e32, m8, ta, ma
-; RV32-NEXT: vslidedown.vi v8, v8, 16
; RV32-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV32-NEXT: vslidedown.vi v0, v0, 2
+; RV32-NEXT: vsetivli zero, 16, e32, m8, ta, ma
+; RV32-NEXT: vslidedown.vi v8, v8, 16
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vluxei32.v v16, (zero), v8, v0.t
; RV32-NEXT: vmv8r.v v8, v24
@@ -1904,9 +1904,9 @@ define <32 x double> @vpgather_v32f64(<32 x ptr> %ptrs, <32 x i1> %m, i32 zeroex
; RV64-NEXT: addi a1, a0, -16
; RV64-NEXT: sltu a0, a0, a1
; RV64-NEXT: addi a0, a0, -1
-; RV64-NEXT: and a0, a0, a1
; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64-NEXT: vslidedown.vi v0, v0, 2
+; RV64-NEXT: and a0, a0, a1
; RV64-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV64-NEXT: vluxei64.v v16, (zero), v16, v0.t
; RV64-NEXT: ret
@@ -1933,10 +1933,10 @@ define <32 x double> @vpgather_baseidx_v32i8_v32f64(ptr %base, <32 x i8> %idxs,
; RV32-NEXT: sltu a1, a1, a2
; RV32-NEXT: addi a1, a1, -1
; RV32-NEXT: and a1, a1, a2
-; RV32-NEXT: vsetivli zero, 16, e32, m8, ta, ma
-; RV32-NEXT: vslidedown.vi v24, v16, 16
; RV32-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV32-NEXT: vslidedown.vi v0, v0, 2
+; RV32-NEXT: vsetivli zero, 16, e32, m8, ta, ma
+; RV32-NEXT: vslidedown.vi v24, v16, 16
; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; RV32-NEXT: vluxei32.v v16, (a0), v24, v0.t
; RV32-NEXT: ret
@@ -1961,9 +1961,9 @@ define <32 x double> @vpgather_baseidx_v32i8_v32f64(ptr %base, <32 x i8> %idxs,
; RV64-NEXT: addi a2, a1, -16
; RV64-NEXT: sltu a1, a1, a2
; RV64-NEXT: addi a1, a1, -1
-; RV64-NEXT: and a1, a1, a2
; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64-NEXT: vslidedown.vi v0, v0, 2
+; RV64-NEXT: and a1, a1, a2
; RV64-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; RV64-NEXT: vluxei64.v v16, (a0), v16, v0.t
; RV64-NEXT: ret
@@ -1991,10 +1991,10 @@ define <32 x double> @vpgather_baseidx_sext_v32i8_v32f64(ptr %base, <32 x i8> %i
; RV32-NEXT: sltu a1, a1, a2
; RV32-NEXT: addi a1, a1, -1
; RV32-NEXT: and a1, a1, a2
-; RV32-NEXT: vsetivli zero, 16, e32, m8, ta, ma
-; RV32-NEXT: vslidedown.vi v24, v16, 16
; RV32-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV32-NEXT: vslidedown.vi v0, v0, 2
+; RV32-NEXT: vsetivli zero, 16, e32, m8, ta, ma
+; RV32-NEXT: vslidedown.vi v24, v16, 16
; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; RV32-NEXT: vluxei32.v v16, (a0), v24, v0.t
; RV32-NEXT: ret
@@ -2020,9 +2020,9 @@ define <32 x double> @vpgather_baseidx_sext_v32i8_v32f64(ptr %base, <32 x i8> %i
; RV64-NEXT: addi a2, a1, -16
; RV64-NEXT: sltu a1, a1, a2
; RV64-NEXT: addi a1, a1, -1
-; RV64-NEXT: and a1, a1, a2
; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64-NEXT: vslidedown.vi v0, v0, 2
+; RV64-NEXT: and a1, a1, a2
; RV64-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; RV64-NEXT: vluxei64.v v16, (a0), v16, v0.t
; RV64-NEXT: ret
@@ -2051,10 +2051,10 @@ define <32 x double> @vpgather_baseidx_zext_v32i8_v32f64(ptr %base, <32 x i8> %i
; RV32-NEXT: sltu a1, a1, a2
; RV32-NEXT: addi a1, a1, -1
; RV32-NEXT: and a1, a1, a2
-; RV32-NEXT: vsetivli zero, 16, e16, m4, ta, ma
-; RV32-NEXT: vslidedown.vi v24, v16, 16
; RV32-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV32-NEXT: vslidedown.vi v0, v0, 2
+; RV32-NEXT: vsetivli zero, 16, e16, m4, ta, ma
+; RV32-NEXT: vslidedown.vi v24, v16, 16
; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; RV32-NEXT: vluxei16.v v16, (a0), v24, v0.t
; RV32-NEXT: ret
@@ -2077,10 +2077,10 @@ define <32 x double> @vpgather_baseidx_zext_v32i8_v32f64(ptr %base, <32 x i8> %i
; RV64-NEXT: sltu a1, a1, a2
; RV64-NEXT: addi a1, a1, -1
; RV64-NEXT: and a1, a1, a2
-; RV64-NEXT: vsetivli zero, 16, e16, m4, ta, ma
-; RV64-NEXT: vslidedown.vi v24, v16, 16
; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64-NEXT: vslidedown.vi v0, v0, 2
+; RV64-NEXT: vsetivli zero, 16, e16, m4, ta, ma
+; RV64-NEXT: vslidedown.vi v24, v16, 16
; RV64-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; RV64-NEXT: vluxei16.v v16, (a0), v24, v0.t
; RV64-NEXT: ret
@@ -2109,10 +2109,10 @@ define <32 x double> @vpgather_baseidx_v32i16_v32f64(ptr %base, <32 x i16> %idxs
; RV32-NEXT: sltu a1, a1, a2
; RV32-NEXT: addi a1, a1, -1
; RV32-NEXT: and a1, a1, a2
-; RV32-NEXT: vsetivli zero, 16, e32, m8, ta, ma
-; RV32-NEXT: vslidedown.vi v24, v16, 16
; RV32-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV32-NEXT: vslidedown.vi v0, v0, 2
+; RV32-NEXT: vsetivli zero, 16, e32, m8, ta, ma
+; RV32-NEXT: vslidedown.vi v24, v16, 16
; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; RV32-NEXT: vluxei32.v v16, (a0), v24, v0.t
; RV32-NEXT: ret
@@ -2137,9 +2137,9 @@ define <32 x double> @vpgather_baseidx_v32i16_v32f64(ptr %base, <32 x i16> %idxs
; RV64-NEXT: addi a2, a1, -16
; RV64-NEXT: sltu a1, a1, a2
; RV64-NEXT: addi a1, a1, -1
-; RV64-NEXT: and a1, a1, a2
; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64-NEXT: vslidedown.vi v0, v0, 2
+; RV64-NEXT: and a1, a1, a2
; RV64-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; RV64-NEXT: vluxei64.v v16, (a0), v16, v0.t
; RV64-NEXT: ret
@@ -2167,10 +2167,10 @@ define <32 x double> @vpgather_baseidx_sext_v32i16_v32f64(ptr %base, <32 x i16>
; RV32-NEXT: sltu a1, a1, a2
; RV32-NEXT: addi a1, a1, -1
; RV32-NEXT: and a1, a1, a2
-; RV32-NEXT: vsetivli zero, 16, e32, m8, ta, ma
-; RV32-NEXT: vslidedown.vi v24, v16, 16
; RV32-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV32-NEXT: vslidedown.vi v0, v0, 2
+; RV32-NEXT: vsetivli zero, 16, e32, m8, ta, ma
+; RV32-NEXT: vslidedown.vi v24, v16, 16
; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; RV32-NEXT: vluxei32.v v16, (a0), v24, v0.t
; RV32-NEXT: ret
@@ -2196,9 +2196,9 @@ define <32 x double> @vpgather_baseidx_sext_v32i16_v32f64(ptr %base, <32 x i16>
; RV64-NEXT: addi a2, a1, -16
; RV64-NEXT: sltu a1, a1, a2
; RV64-NEXT: addi a1, a1, -1
-; RV64-NEXT: and a1, a1, a2
; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64-NEXT: vslidedown.vi v0, v0, 2
+; RV64-NEXT: and a1, a1, a2
; RV64-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; RV64-NEXT: vluxei64.v v16, (a0), v16, v0.t
; RV64-NEXT: ret
@@ -2227,10 +2227,10 @@ define <32 x double> @vpgather_baseidx_zext_v32i16_v32f64(ptr %base, <32 x i16>
; RV32-NEXT: sltu a1, a1, a2
; RV32-NEXT: addi a1, a1, -1
; RV32-NEXT: and a1, a1, a2
-; RV32-NEXT: vsetivli zero, 16, e32, m8, ta, ma
-; RV32-NEXT: vslidedown.vi v24, v16, 16
; RV32-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV32-NEXT: vslidedown.vi v0, v0, 2
+; RV32-NEXT: vsetivli zero, 16, e32, m8, ta, ma
+; RV32-NEXT: vslidedown.vi v24, v16, 16
; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; RV32-NEXT: vluxei32.v v16, (a0), v24, v0.t
; RV32-NEXT: ret
@@ -2253,10 +2253,10 @@ define <32 x double> @vpgather_baseidx_zext_v32i16_v32f64(ptr %base, <32 x i16>
; RV64-NEXT: sltu a1, a1, a2
; RV64-NEXT: addi a1, a1, -1
; RV64-NEXT: and a1, a1, a2
-; RV64-NEXT: vsetivli zero, 16, e32, m8, ta, ma
-; RV64-NEXT: vslidedown.vi v24, v16, 16
; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64-NEXT: vslidedown.vi v0, v0, 2
+; RV64-NEXT: vsetivli zero, 16, e32, m8, ta, ma
+; RV64-NEXT: vslidedown.vi v24, v16, 16
; RV64-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; RV64-NEXT: vluxei32.v v16, (a0), v24, v0.t
; RV64-NEXT: ret
@@ -2270,8 +2270,8 @@ define <32 x double> @vpgather_baseidx_v32i32_v32f64(ptr %base, <32 x i32> %idxs
; RV32-LABEL: vpgather_baseidx_v32i32_v32f64:
; RV32: # %bb.0:
; RV32-NEXT: li a2, 32
-; RV32-NEXT: vsetvli zero, a2, e32, m8, ta, ma
; RV32-NEXT: li a3, 16
+; RV32-NEXT: vsetvli zero, a2, e32, m8, ta, ma
; RV32-NEXT: vsll.vi v16, v8, 3
; RV32-NEXT: mv a2, a1
; RV32-NEXT: bltu a1, a3, .LBB93_2
@@ -2284,10 +2284,10 @@ define <32 x double> @vpgather_baseidx_v32i32_v32f64(ptr %base, <32 x i32> %idxs
; RV32-NEXT: sltu a1, a1, a2
; RV32-NEXT: addi a1, a1, -1
; RV32-NEXT: and a1, a1, a2
-; RV32-NEXT: vsetivli zero, 16, e32, m8, ta, ma
-; RV32-NEXT: vslidedown.vi v24, v16, 16
; RV32-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV32-NEXT: vslidedown.vi v0, v0, 2
+; RV32-NEXT: vsetivli zero, 16, e32, m8, ta, ma
+; RV32-NEXT: vslidedown.vi v24, v16, 16
; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; RV32-NEXT: vluxei32.v v16, (a0), v24, v0.t
; RV32-NEXT: ret
@@ -2312,9 +2312,9 @@ define <32 x double> @vpgather_baseidx_v32i32_v32f64(ptr %base, <32 x i32> %idxs
; RV64-NEXT: addi a2, a1, -16
; RV64-NEXT: sltu a1, a1, a2
; RV64-NEXT: addi a1, a1, -1
-; RV64-NEXT: and a1, a1, a2
; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64-NEXT: vslidedown.vi v0, v0, 2
+; RV64-NEXT: and a1, a1, a2
; RV64-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; RV64-NEXT: vluxei64.v v16, (a0), v16, v0.t
; RV64-NEXT: ret
@@ -2327,8 +2327,8 @@ define <32 x double> @vpgather_baseidx_sext_v32i32_v32f64(ptr %base, <32 x i32>
; RV32-LABEL: vpgather_baseidx_sext_v32i32_v32f64:
; RV32: # %bb.0:
; RV32-NEXT: li a2, 32
-; RV32-NEXT: vsetvli zero, a2, e32, m8, ta, ma
; RV32-NEXT: li a3, 16
+; RV32-NEXT: vsetvli zero, a2, e32, m8, ta, ma
; RV32-NEXT: vsll.vi v16, v8, 3
; RV32-NEXT: mv a2, a1
; RV32-NEXT: bltu a1, a3, .LBB94_2
@@ -2341,10 +2341,10 @@ define <32 x double> @vpgather_baseidx_sext_v32i32_v32f64(ptr %base, <32 x i32>
; RV32-NEXT: sltu a1, a1, a2
; RV32-NEXT: addi a1, a1, -1
; RV32-NEXT: and a1, a1, a2
-; RV32-NEXT: vsetivli zero, 16, e32, m8, ta, ma
-; RV32-NEXT: vslidedown.vi v24, v16, 16
; RV32-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV32-NEXT: vslidedown.vi v0, v0, 2
+; RV32-NEXT: vsetivli zero, 16, e32, m8, ta, ma
+; RV32-NEXT: vslidedown.vi v24, v16, 16
; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; RV32-NEXT: vluxei32.v v16, (a0), v24, v0.t
; RV32-NEXT: ret
@@ -2370,9 +2370,9 @@ define <32 x double> @vpgather_baseidx_sext_v32i32_v32f64(ptr %base, <32 x i32>
; RV64-NEXT: addi a2, a1, -16
; RV64-NEXT: sltu a1, a1, a2
; RV64-NEXT: addi a1, a1, -1
-; RV64-NEXT: and a1, a1, a2
; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64-NEXT: vslidedown.vi v0, v0, 2
+; RV64-NEXT: and a1, a1, a2
; RV64-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; RV64-NEXT: vluxei64.v v16, (a0), v16, v0.t
; RV64-NEXT: ret
@@ -2386,8 +2386,8 @@ define <32 x double> @vpgather_baseidx_zext_v32i32_v32f64(ptr %base, <32 x i32>
; RV32-LABEL: vpgather_baseidx_zext_v32i32_v32f64:
; RV32: # %bb.0:
; RV32-NEXT: li a2, 32
-; RV32-NEXT: vsetvli zero, a2, e32, m8, ta, ma
; RV32-NEXT: li a3, 16
+; RV32-NEXT: vsetvli zero, a2, e32, m8, ta, ma
; RV32-NEXT: vsll.vi v16, v8, 3
; RV32-NEXT: mv a2, a1
; RV32-NEXT: bltu a1, a3, .LBB95_2
@@ -2400,10 +2400,10 @@ define <32 x double> @vpgather_baseidx_zext_v32i32_v32f64(ptr %base, <32 x i32>
; RV32-NEXT: sltu a1, a1, a2
; RV32-NEXT: addi a1, a1, -1
; RV32-NEXT: and a1, a1, a2
-; RV32-NEXT: vsetivli zero, 16, e32, m8, ta, ma
-; RV32-NEXT: vslidedown.vi v24, v16, 16
; RV32-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV32-NEXT: vslidedown.vi v0, v0, 2
+; RV32-NEXT: vsetivli zero, 16, e32, m8, ta, ma
+; RV32-NEXT: vslidedown.vi v24, v16, 16
; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; RV32-NEXT: vluxei32.v v16, (a0), v24, v0.t
; RV32-NEXT: ret
@@ -2429,9 +2429,9 @@ define <32 x double> @vpgather_baseidx_zext_v32i32_v32f64(ptr %base, <32 x i32>
; RV64-NEXT: addi a2, a1, -16
; RV64-NEXT: sltu a1, a1, a2
; RV64-NEXT: addi a1, a1, -1
-; RV64-NEXT: and a1, a1, a2
; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64-NEXT: vslidedown.vi v0, v0, 2
+; RV64-NEXT: and a1, a1, a2
; RV64-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; RV64-NEXT: vluxei64.v v16, (a0), v16, v0.t
; RV64-NEXT: ret
@@ -2457,9 +2457,9 @@ define <32 x double> @vpgather_baseidx_v32f64(ptr %base, <32 x i64> %idxs, <32 x
; RV32-NEXT: addi a2, a1, -16
; RV32-NEXT: sltu a3, a1, a2
; RV32-NEXT: addi a3, a3, -1
-; RV32-NEXT: and a2, a3, a2
; RV32-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV32-NEXT: vslidedown.vi v0, v0, 2
+; RV32-NEXT: and a2, a3, a2
; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
; RV32-NEXT: vluxei32.v v16, (a0), v8, v0.t
; RV32-NEXT: li a2, 16
@@ -2467,8 +2467,8 @@ define <32 x double> @vpgather_baseidx_v32f64(ptr %base, <32 x i64> %idxs, <32 x
; RV32-NEXT: # %bb.1:
; RV32-NEXT: li a1, 16
; RV32-NEXT: .LBB96_2:
-; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; RV32-NEXT: vmv1r.v v0, v7
+; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; RV32-NEXT: vluxei32.v v8, (a0), v24, v0.t
; RV32-NEXT: ret
;
@@ -2488,9 +2488,9 @@ define <32 x double> @vpgather_baseidx_v32f64(ptr %base, <32 x i64> %idxs, <32 x
; RV64-NEXT: addi a2, a1, -16
; RV64-NEXT: sltu a1, a1, a2
; RV64-NEXT: addi a1, a1, -1
-; RV64-NEXT: and a1, a1, a2
; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64-NEXT: vslidedown.vi v0, v0, 2
+; RV64-NEXT: and a1, a1, a2
; RV64-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; RV64-NEXT: vluxei64.v v16, (a0), v16, v0.t
; RV64-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vpload.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vpload.ll
index 9ef89352e65e5..f204d812c14f6 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vpload.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vpload.ll
@@ -377,9 +377,9 @@ define <32 x double> @vpload_v32f64(ptr %ptr, <32 x i1> %m, i32 zeroext %evl) {
; CHECK-NEXT: sltu a1, a1, a2
; CHECK-NEXT: addi a1, a1, -1
; CHECK-NEXT: and a1, a1, a2
-; CHECK-NEXT: addi a0, a0, 128
; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; CHECK-NEXT: vslidedown.vi v0, v0, 2
+; CHECK-NEXT: addi a0, a0, 128
; CHECK-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; CHECK-NEXT: vle64.v v16, (a0), v0.t
; CHECK-NEXT: ret
@@ -405,9 +405,9 @@ define <33 x double> @vpload_v33f64(ptr %ptr, <33 x i1> %m, i32 zeroext %evl) {
; CHECK-NEXT: sltu a3, a3, a4
; CHECK-NEXT: addi a3, a3, -1
; CHECK-NEXT: and a3, a3, a4
-; CHECK-NEXT: addi a4, a1, 128
; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; CHECK-NEXT: vslidedown.vi v0, v8, 2
+; CHECK-NEXT: addi a4, a1, 128
; CHECK-NEXT: vsetvli zero, a3, e64, m8, ta, ma
; CHECK-NEXT: vle64.v v16, (a4), v0.t
; CHECK-NEXT: addi a3, a2, -32
@@ -419,17 +419,17 @@ define <33 x double> @vpload_v33f64(ptr %ptr, <33 x i1> %m, i32 zeroext %evl) {
; CHECK-NEXT: # %bb.3:
; CHECK-NEXT: li a4, 16
; CHECK-NEXT: .LBB32_4:
-; CHECK-NEXT: addi a5, a1, 256
; CHECK-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; CHECK-NEXT: vslidedown.vi v0, v8, 4
+; CHECK-NEXT: addi a5, a1, 256
; CHECK-NEXT: vsetvli zero, a4, e64, m8, ta, ma
; CHECK-NEXT: vle64.v v24, (a5), v0.t
; CHECK-NEXT: bltu a2, a3, .LBB32_6
; CHECK-NEXT: # %bb.5:
; CHECK-NEXT: li a2, 16
; CHECK-NEXT: .LBB32_6:
-; CHECK-NEXT: vsetvli zero, a2, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a2, e64, m8, ta, ma
; CHECK-NEXT: vle64.v v8, (a1), v0.t
; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; CHECK-NEXT: vse64.v v8, (a0)
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vpmerge.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vpmerge.ll
index 466448a7a05a2..9f0561b394b81 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vpmerge.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vpmerge.ll
@@ -1193,17 +1193,17 @@ define <32 x double> @vpmerge_vv_v32f64(<32 x double> %va, <32 x double> %vb, <3
; CHECK-NEXT: addi a0, a2, -16
; CHECK-NEXT: sltu a1, a2, a0
; CHECK-NEXT: addi a1, a1, -1
-; CHECK-NEXT: and a0, a1, a0
; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; CHECK-NEXT: vslidedown.vi v0, v0, 2
+; CHECK-NEXT: and a0, a1, a0
+; CHECK-NEXT: csrr a1, vlenb
+; CHECK-NEXT: slli a1, a1, 3
+; CHECK-NEXT: add a1, sp, a1
+; CHECK-NEXT: addi a1, a1, 16
+; CHECK-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
+; CHECK-NEXT: addi a1, sp, 16
+; CHECK-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, a0, e64, m8, tu, ma
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vmerge.vvm v16, v16, v8, v0
; CHECK-NEXT: vmv8r.v v8, v24
; CHECK-NEXT: csrr a0, vlenb
@@ -1229,9 +1229,9 @@ define <32 x double> @vpmerge_vf_v32f64(double %a, <32 x double> %vb, <32 x i1>
; CHECK-NEXT: addi a1, a0, -16
; CHECK-NEXT: sltu a0, a0, a1
; CHECK-NEXT: addi a0, a0, -1
-; CHECK-NEXT: and a0, a0, a1
; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; CHECK-NEXT: vslidedown.vi v0, v0, 2
+; CHECK-NEXT: and a0, a0, a1
; CHECK-NEXT: vsetvli zero, a0, e64, m8, tu, ma
; CHECK-NEXT: vfmerge.vfm v16, v16, fa0, v0
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vpscatter.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vpscatter.ll
index cd9a38d5167d5..0c180cd148b81 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vpscatter.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vpscatter.ll
@@ -1685,10 +1685,10 @@ define void @vpscatter_v32f64(<32 x double> %val, <32 x ptr> %ptrs, <32 x i1> %m
; RV32-NEXT: sltu a1, a1, a0
; RV32-NEXT: addi a1, a1, -1
; RV32-NEXT: and a0, a1, a0
-; RV32-NEXT: vsetivli zero, 16, e32, m8, ta, ma
-; RV32-NEXT: vslidedown.vi v8, v24, 16
; RV32-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV32-NEXT: vslidedown.vi v0, v0, 2
+; RV32-NEXT: vsetivli zero, 16, e32, m8, ta, ma
+; RV32-NEXT: vslidedown.vi v8, v24, 16
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vsoxei32.v v16, (zero), v8, v0.t
; RV32-NEXT: ret
@@ -1718,12 +1718,12 @@ define void @vpscatter_v32f64(<32 x double> %val, <32 x ptr> %ptrs, <32 x i1> %m
; RV64-NEXT: addi a0, a2, -16
; RV64-NEXT: sltu a1, a2, a0
; RV64-NEXT: addi a1, a1, -1
-; RV64-NEXT: and a0, a1, a0
; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64-NEXT: vslidedown.vi v0, v0, 2
+; RV64-NEXT: and a0, a1, a0
+; RV64-NEXT: addi a1, sp, 16
+; RV64-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
; RV64-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV64-NEXT: addi a0, sp, 16
-; RV64-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; RV64-NEXT: vsoxei64.v v16, (zero), v8, v0.t
; RV64-NEXT: csrr a0, vlenb
; RV64-NEXT: slli a0, a0, 3
@@ -1753,10 +1753,10 @@ define void @vpscatter_baseidx_v32i32_v32f64(<32 x double> %val, ptr %base, <32
; RV32-NEXT: sltu a2, a2, a1
; RV32-NEXT: addi a2, a2, -1
; RV32-NEXT: and a1, a2, a1
-; RV32-NEXT: vsetivli zero, 16, e32, m8, ta, ma
-; RV32-NEXT: vslidedown.vi v8, v24, 16
; RV32-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV32-NEXT: vslidedown.vi v0, v0, 2
+; RV32-NEXT: vsetivli zero, 16, e32, m8, ta, ma
+; RV32-NEXT: vslidedown.vi v8, v24, 16
; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; RV32-NEXT: vsoxei32.v v16, (a0), v8, v0.t
; RV32-NEXT: ret
@@ -1766,51 +1766,44 @@ define void @vpscatter_baseidx_v32i32_v32f64(<32 x double> %val, ptr %base, <32
; RV64-NEXT: addi sp, sp, -16
; RV64-NEXT: .cfi_def_cfa_offset 16
; RV64-NEXT: csrr a3, vlenb
-; RV64-NEXT: li a4, 10
-; RV64-NEXT: mul a3, a3, a4
+; RV64-NEXT: slli a3, a3, 3
; RV64-NEXT: sub sp, sp, a3
-; RV64-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x0a, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 10 * vlenb
+; RV64-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
; RV64-NEXT: li a3, 32
; RV64-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; RV64-NEXT: vle32.v v24, (a1)
+; RV64-NEXT: vmv1r.v v7, v0
; RV64-NEXT: addi a1, sp, 16
-; RV64-NEXT: vs1r.v v0, (a1) # Unknown-size Folded Spill
-; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: add a1, sp, a1
-; RV64-NEXT: addi a1, a1, 16
; RV64-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
; RV64-NEXT: vsetivli zero, 16, e32, m8, ta, ma
-; RV64-NEXT: vslidedown.vi v0, v24, 16
+; RV64-NEXT: vslidedown.vi v16, v24, 16
+; RV64-NEXT: vmv4r.v v0, v24
; RV64-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV64-NEXT: vsext.vf2 v16, v0
-; RV64-NEXT: vsll.vi v16, v16, 3
-; RV64-NEXT: vsext.vf2 v0, v24
+; RV64-NEXT: vsext.vf2 v24, v16
+; RV64-NEXT: vsll.vi v16, v24, 3
+; RV64-NEXT: vsext.vf2 v24, v0
; RV64-NEXT: li a3, 16
-; RV64-NEXT: vsll.vi v24, v0, 3
+; RV64-NEXT: vsll.vi v24, v24, 3
; RV64-NEXT: mv a1, a2
; RV64-NEXT: bltu a2, a3, .LBB80_2
; RV64-NEXT: # %bb.1:
; RV64-NEXT: li a1, 16
; RV64-NEXT: .LBB80_2:
+; RV64-NEXT: vmv1r.v v0, v7
; RV64-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; RV64-NEXT: addi a1, sp, 16
-; RV64-NEXT: vl1r.v v0, (a1) # Unknown-size Folded Reload
; RV64-NEXT: vsoxei64.v v8, (a0), v24, v0.t
; RV64-NEXT: addi a1, a2, -16
; RV64-NEXT: sltu a2, a2, a1
; RV64-NEXT: addi a2, a2, -1
-; RV64-NEXT: and a1, a2, a1
; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
-; RV64-NEXT: vslidedown.vi v0, v0, 2
+; RV64-NEXT: vslidedown.vi v0, v7, 2
+; RV64-NEXT: and a1, a2, a1
+; RV64-NEXT: addi a2, sp, 16
+; RV64-NEXT: vl8r.v v8, (a2) # Unknown-size Folded Reload
; RV64-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: add a1, sp, a1
-; RV64-NEXT: addi a1, a1, 16
-; RV64-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
; RV64-NEXT: vsoxei64.v v8, (a0), v16, v0.t
; RV64-NEXT: csrr a0, vlenb
-; RV64-NEXT: li a1, 10
-; RV64-NEXT: mul a0, a0, a1
+; RV64-NEXT: slli a0, a0, 3
; RV64-NEXT: add sp, sp, a0
; RV64-NEXT: addi sp, sp, 16
; RV64-NEXT: ret
@@ -1838,10 +1831,10 @@ define void @vpscatter_baseidx_sext_v32i32_v32f64(<32 x double> %val, ptr %base,
; RV32-NEXT: sltu a2, a2, a1
; RV32-NEXT: addi a2, a2, -1
; RV32-NEXT: and a1, a2, a1
-; RV32-NEXT: vsetivli zero, 16, e32, m8, ta, ma
-; RV32-NEXT: vslidedown.vi v8, v24, 16
; RV32-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV32-NEXT: vslidedown.vi v0, v0, 2
+; RV32-NEXT: vsetivli zero, 16, e32, m8, ta, ma
+; RV32-NEXT: vslidedown.vi v8, v24, 16
; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; RV32-NEXT: vsoxei32.v v16, (a0), v8, v0.t
; RV32-NEXT: ret
@@ -1878,21 +1871,21 @@ define void @vpscatter_baseidx_sext_v32i32_v32f64(<32 x double> %val, ptr %base,
; RV64-NEXT: # %bb.1:
; RV64-NEXT: li a1, 16
; RV64-NEXT: .LBB81_2:
+; RV64-NEXT: addi a3, sp, 16
+; RV64-NEXT: vl1r.v v0, (a3) # Unknown-size Folded Reload
; RV64-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; RV64-NEXT: addi a1, sp, 16
-; RV64-NEXT: vl1r.v v0, (a1) # Unknown-size Folded Reload
; RV64-NEXT: vsoxei64.v v8, (a0), v24, v0.t
; RV64-NEXT: addi a1, a2, -16
; RV64-NEXT: sltu a2, a2, a1
; RV64-NEXT: addi a2, a2, -1
-; RV64-NEXT: and a1, a2, a1
; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64-NEXT: vslidedown.vi v0, v0, 2
+; RV64-NEXT: and a1, a2, a1
+; RV64-NEXT: csrr a2, vlenb
+; RV64-NEXT: add a2, sp, a2
+; RV64-NEXT: addi a2, a2, 16
+; RV64-NEXT: vl8r.v v8, (a2) # Unknown-size Folded Reload
; RV64-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: add a1, sp, a1
-; RV64-NEXT: addi a1, a1, 16
-; RV64-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
; RV64-NEXT: vsoxei64.v v8, (a0), v16, v0.t
; RV64-NEXT: csrr a0, vlenb
; RV64-NEXT: li a1, 10
@@ -1925,10 +1918,10 @@ define void @vpscatter_baseidx_zext_v32i32_v32f64(<32 x double> %val, ptr %base,
; RV32-NEXT: sltu a2, a2, a1
; RV32-NEXT: addi a2, a2, -1
; RV32-NEXT: and a1, a2, a1
-; RV32-NEXT: vsetivli zero, 16, e32, m8, ta, ma
-; RV32-NEXT: vslidedown.vi v8, v24, 16
; RV32-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV32-NEXT: vslidedown.vi v0, v0, 2
+; RV32-NEXT: vsetivli zero, 16, e32, m8, ta, ma
+; RV32-NEXT: vslidedown.vi v8, v24, 16
; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; RV32-NEXT: vsoxei32.v v16, (a0), v8, v0.t
; RV32-NEXT: ret
@@ -1965,21 +1958,21 @@ define void @vpscatter_baseidx_zext_v32i32_v32f64(<32 x double> %val, ptr %base,
; RV64-NEXT: # %bb.1:
; RV64-NEXT: li a1, 16
; RV64-NEXT: .LBB82_2:
+; RV64-NEXT: addi a3, sp, 16
+; RV64-NEXT: vl1r.v v0, (a3) # Unknown-size Folded Reload
; RV64-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; RV64-NEXT: addi a1, sp, 16
-; RV64-NEXT: vl1r.v v0, (a1) # Unknown-size Folded Reload
; RV64-NEXT: vsoxei64.v v8, (a0), v24, v0.t
; RV64-NEXT: addi a1, a2, -16
; RV64-NEXT: sltu a2, a2, a1
; RV64-NEXT: addi a2, a2, -1
-; RV64-NEXT: and a1, a2, a1
; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64-NEXT: vslidedown.vi v0, v0, 2
+; RV64-NEXT: and a1, a2, a1
+; RV64-NEXT: csrr a2, vlenb
+; RV64-NEXT: add a2, sp, a2
+; RV64-NEXT: addi a2, a2, 16
+; RV64-NEXT: vl8r.v v8, (a2) # Unknown-size Folded Reload
; RV64-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: add a1, sp, a1
-; RV64-NEXT: addi a1, a1, 16
-; RV64-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
; RV64-NEXT: vsoxei64.v v8, (a0), v16, v0.t
; RV64-NEXT: csrr a0, vlenb
; RV64-NEXT: li a1, 10
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vpstore.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vpstore.ll
index c0aa735614b21..f396790f4f178 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vpstore.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vpstore.ll
@@ -295,9 +295,9 @@ define void @vpstore_v32f64(<32 x double> %val, ptr %ptr, <32 x i1> %m, i32 zero
; CHECK-NEXT: sltu a1, a1, a2
; CHECK-NEXT: addi a1, a1, -1
; CHECK-NEXT: and a1, a1, a2
-; CHECK-NEXT: addi a0, a0, 128
; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; CHECK-NEXT: vslidedown.vi v0, v0, 2
+; CHECK-NEXT: addi a0, a0, 128
; CHECK-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; CHECK-NEXT: vse64.v v16, (a0), v0.t
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vsadd-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vsadd-vp.ll
index 291629de6dcfa..df2c83028e5df 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vsadd-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vsadd-vp.ll
@@ -386,8 +386,8 @@ define <256 x i8> @vsadd_vi_v258i8(<256 x i8> %va, <256 x i1> %m, i32 zeroext %e
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: li a1, 128
; CHECK-NEXT: .LBB32_2:
-; CHECK-NEXT: vsetvli zero, a1, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a1, e8, m8, ta, ma
; CHECK-NEXT: vsadd.vi v8, v8, -1, v0.t
; CHECK-NEXT: ret
%v = call <256 x i8> @llvm.vp.sadd.sat.v258i8(<256 x i8> %va, <256 x i8> splat (i8 -1), <256 x i1> %m, i32 %evl)
@@ -425,8 +425,8 @@ define <256 x i8> @vsadd_vi_v258i8_evl129(<256 x i8> %va, <256 x i1> %m) {
; CHECK-NEXT: vsetvli zero, a1, e8, m8, ta, ma
; CHECK-NEXT: vlm.v v24, (a0)
; CHECK-NEXT: vsadd.vi v8, v8, -1, v0.t
-; CHECK-NEXT: vsetivli zero, 1, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetivli zero, 1, e8, m8, ta, ma
; CHECK-NEXT: vsadd.vi v16, v16, -1, v0.t
; CHECK-NEXT: ret
%v = call <256 x i8> @llvm.vp.sadd.sat.v258i8(<256 x i8> %va, <256 x i8> splat (i8 -1), <256 x i1> %m, i32 129)
@@ -442,8 +442,8 @@ define <256 x i8> @vsadd_vi_v258i8_evl128(<256 x i8> %va, <256 x i1> %m) {
; CHECK-NEXT: vsetvli zero, a1, e8, m8, ta, ma
; CHECK-NEXT: vlm.v v24, (a0)
; CHECK-NEXT: vsadd.vi v8, v8, -1, v0.t
-; CHECK-NEXT: vsetivli zero, 0, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetivli zero, 0, e8, m8, ta, ma
; CHECK-NEXT: vsadd.vi v16, v16, -1, v0.t
; CHECK-NEXT: ret
%v = call <256 x i8> @llvm.vp.sadd.sat.v258i8(<256 x i8> %va, <256 x i8> splat (i8 -1), <256 x i1> %m, i32 128)
@@ -1361,8 +1361,8 @@ declare <32 x i64> @llvm.vp.sadd.sat.v32i64(<32 x i64>, <32 x i64>, <32 x i1>, i
define <32 x i64> @vsadd_vx_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %evl) {
; RV32-LABEL: vsadd_vx_v32i64:
; RV32: # %bb.0:
-; RV32-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV32-NEXT: li a2, 16
+; RV32-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV32-NEXT: vslidedown.vi v7, v0, 2
; RV32-NEXT: mv a1, a0
; RV32-NEXT: bltu a0, a2, .LBB108_2
@@ -1378,15 +1378,15 @@ define <32 x i64> @vsadd_vx_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %ev
; RV32-NEXT: sltu a0, a0, a1
; RV32-NEXT: addi a0, a0, -1
; RV32-NEXT: and a0, a0, a1
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vmv1r.v v0, v7
+; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vsadd.vv v16, v16, v24, v0.t
; RV32-NEXT: ret
;
; RV64-LABEL: vsadd_vx_v32i64:
; RV64: # %bb.0:
-; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64-NEXT: li a2, 16
+; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64-NEXT: vslidedown.vi v24, v0, 2
; RV64-NEXT: mv a1, a0
; RV64-NEXT: bltu a0, a2, .LBB108_2
@@ -1399,8 +1399,8 @@ define <32 x i64> @vsadd_vx_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %ev
; RV64-NEXT: sltu a0, a0, a1
; RV64-NEXT: addi a0, a0, -1
; RV64-NEXT: and a0, a0, a1
-; RV64-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV64-NEXT: vmv1r.v v0, v24
+; RV64-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV64-NEXT: vsadd.vi v16, v16, -1, v0.t
; RV64-NEXT: ret
%v = call <32 x i64> @llvm.vp.sadd.sat.v32i64(<32 x i64> %va, <32 x i64> splat (i64 -1), <32 x i1> %m, i32 %evl)
@@ -1462,8 +1462,8 @@ define <32 x i64> @vsadd_vx_v32i64_evl12(<32 x i64> %va, <32 x i1> %m) {
; RV32-NEXT: vmv.v.i v24, -1
; RV32-NEXT: vsetivli zero, 12, e64, m8, ta, ma
; RV32-NEXT: vsadd.vv v8, v8, v24, v0.t
-; RV32-NEXT: vsetivli zero, 0, e64, m8, ta, ma
; RV32-NEXT: vmv1r.v v0, v7
+; RV32-NEXT: vsetivli zero, 0, e64, m8, ta, ma
; RV32-NEXT: vsadd.vv v16, v16, v24, v0.t
; RV32-NEXT: ret
;
@@ -1473,8 +1473,8 @@ define <32 x i64> @vsadd_vx_v32i64_evl12(<32 x i64> %va, <32 x i1> %m) {
; RV64-NEXT: vslidedown.vi v24, v0, 2
; RV64-NEXT: vsetivli zero, 12, e64, m8, ta, ma
; RV64-NEXT: vsadd.vi v8, v8, -1, v0.t
-; RV64-NEXT: vsetivli zero, 0, e64, m8, ta, ma
; RV64-NEXT: vmv1r.v v0, v24
+; RV64-NEXT: vsetivli zero, 0, e64, m8, ta, ma
; RV64-NEXT: vsadd.vi v16, v16, -1, v0.t
; RV64-NEXT: ret
%v = call <32 x i64> @llvm.vp.sadd.sat.v32i64(<32 x i64> %va, <32 x i64> splat (i64 -1), <32 x i1> %m, i32 12)
@@ -1491,8 +1491,8 @@ define <32 x i64> @vsadd_vx_v32i64_evl27(<32 x i64> %va, <32 x i1> %m) {
; RV32-NEXT: vmv.v.i v24, -1
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV32-NEXT: vsadd.vv v8, v8, v24, v0.t
-; RV32-NEXT: vsetivli zero, 11, e64, m8, ta, ma
; RV32-NEXT: vmv1r.v v0, v7
+; RV32-NEXT: vsetivli zero, 11, e64, m8, ta, ma
; RV32-NEXT: vsadd.vv v16, v16, v24, v0.t
; RV32-NEXT: ret
;
@@ -1502,8 +1502,8 @@ define <32 x i64> @vsadd_vx_v32i64_evl27(<32 x i64> %va, <32 x i1> %m) {
; RV64-NEXT: vslidedown.vi v24, v0, 2
; RV64-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV64-NEXT: vsadd.vi v8, v8, -1, v0.t
-; RV64-NEXT: vsetivli zero, 11, e64, m8, ta, ma
; RV64-NEXT: vmv1r.v v0, v24
+; RV64-NEXT: vsetivli zero, 11, e64, m8, ta, ma
; RV64-NEXT: vsadd.vi v16, v16, -1, v0.t
; RV64-NEXT: ret
%v = call <32 x i64> @llvm.vp.sadd.sat.v32i64(<32 x i64> %va, <32 x i64> splat (i64 -1), <32 x i1> %m, i32 27)
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vsaddu-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vsaddu-vp.ll
index d38ee1148e894..f50dadf019910 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vsaddu-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vsaddu-vp.ll
@@ -382,8 +382,8 @@ define <256 x i8> @vsaddu_vi_v258i8(<256 x i8> %va, <256 x i1> %m, i32 zeroext %
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: li a1, 128
; CHECK-NEXT: .LBB32_2:
-; CHECK-NEXT: vsetvli zero, a1, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a1, e8, m8, ta, ma
; CHECK-NEXT: vsaddu.vi v8, v8, -1, v0.t
; CHECK-NEXT: ret
%v = call <256 x i8> @llvm.vp.uadd.sat.v258i8(<256 x i8> %va, <256 x i8> splat (i8 -1), <256 x i1> %m, i32 %evl)
@@ -421,8 +421,8 @@ define <256 x i8> @vsaddu_vi_v258i8_evl129(<256 x i8> %va, <256 x i1> %m) {
; CHECK-NEXT: vsetvli zero, a1, e8, m8, ta, ma
; CHECK-NEXT: vlm.v v24, (a0)
; CHECK-NEXT: vsaddu.vi v8, v8, -1, v0.t
-; CHECK-NEXT: vsetivli zero, 1, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetivli zero, 1, e8, m8, ta, ma
; CHECK-NEXT: vsaddu.vi v16, v16, -1, v0.t
; CHECK-NEXT: ret
%v = call <256 x i8> @llvm.vp.uadd.sat.v258i8(<256 x i8> %va, <256 x i8> splat (i8 -1), <256 x i1> %m, i32 129)
@@ -438,8 +438,8 @@ define <256 x i8> @vsaddu_vi_v258i8_evl128(<256 x i8> %va, <256 x i1> %m) {
; CHECK-NEXT: vsetvli zero, a1, e8, m8, ta, ma
; CHECK-NEXT: vlm.v v24, (a0)
; CHECK-NEXT: vsaddu.vi v8, v8, -1, v0.t
-; CHECK-NEXT: vsetivli zero, 0, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetivli zero, 0, e8, m8, ta, ma
; CHECK-NEXT: vsaddu.vi v16, v16, -1, v0.t
; CHECK-NEXT: ret
%v = call <256 x i8> @llvm.vp.uadd.sat.v258i8(<256 x i8> %va, <256 x i8> splat (i8 -1), <256 x i1> %m, i32 128)
@@ -1357,8 +1357,8 @@ declare <32 x i64> @llvm.vp.uadd.sat.v32i64(<32 x i64>, <32 x i64>, <32 x i1>, i
define <32 x i64> @vsaddu_vx_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %evl) {
; RV32-LABEL: vsaddu_vx_v32i64:
; RV32: # %bb.0:
-; RV32-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV32-NEXT: li a2, 16
+; RV32-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV32-NEXT: vslidedown.vi v7, v0, 2
; RV32-NEXT: mv a1, a0
; RV32-NEXT: bltu a0, a2, .LBB108_2
@@ -1374,15 +1374,15 @@ define <32 x i64> @vsaddu_vx_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %e
; RV32-NEXT: sltu a0, a0, a1
; RV32-NEXT: addi a0, a0, -1
; RV32-NEXT: and a0, a0, a1
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vmv1r.v v0, v7
+; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vsaddu.vv v16, v16, v24, v0.t
; RV32-NEXT: ret
;
; RV64-LABEL: vsaddu_vx_v32i64:
; RV64: # %bb.0:
-; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64-NEXT: li a2, 16
+; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64-NEXT: vslidedown.vi v24, v0, 2
; RV64-NEXT: mv a1, a0
; RV64-NEXT: bltu a0, a2, .LBB108_2
@@ -1395,8 +1395,8 @@ define <32 x i64> @vsaddu_vx_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %e
; RV64-NEXT: sltu a0, a0, a1
; RV64-NEXT: addi a0, a0, -1
; RV64-NEXT: and a0, a0, a1
-; RV64-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV64-NEXT: vmv1r.v v0, v24
+; RV64-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV64-NEXT: vsaddu.vi v16, v16, -1, v0.t
; RV64-NEXT: ret
%v = call <32 x i64> @llvm.vp.uadd.sat.v32i64(<32 x i64> %va, <32 x i64> splat (i64 -1), <32 x i1> %m, i32 %evl)
@@ -1458,8 +1458,8 @@ define <32 x i64> @vsaddu_vx_v32i64_evl12(<32 x i64> %va, <32 x i1> %m) {
; RV32-NEXT: vmv.v.i v24, -1
; RV32-NEXT: vsetivli zero, 12, e64, m8, ta, ma
; RV32-NEXT: vsaddu.vv v8, v8, v24, v0.t
-; RV32-NEXT: vsetivli zero, 0, e64, m8, ta, ma
; RV32-NEXT: vmv1r.v v0, v7
+; RV32-NEXT: vsetivli zero, 0, e64, m8, ta, ma
; RV32-NEXT: vsaddu.vv v16, v16, v24, v0.t
; RV32-NEXT: ret
;
@@ -1469,8 +1469,8 @@ define <32 x i64> @vsaddu_vx_v32i64_evl12(<32 x i64> %va, <32 x i1> %m) {
; RV64-NEXT: vslidedown.vi v24, v0, 2
; RV64-NEXT: vsetivli zero, 12, e64, m8, ta, ma
; RV64-NEXT: vsaddu.vi v8, v8, -1, v0.t
-; RV64-NEXT: vsetivli zero, 0, e64, m8, ta, ma
; RV64-NEXT: vmv1r.v v0, v24
+; RV64-NEXT: vsetivli zero, 0, e64, m8, ta, ma
; RV64-NEXT: vsaddu.vi v16, v16, -1, v0.t
; RV64-NEXT: ret
%v = call <32 x i64> @llvm.vp.uadd.sat.v32i64(<32 x i64> %va, <32 x i64> splat (i64 -1), <32 x i1> %m, i32 12)
@@ -1487,8 +1487,8 @@ define <32 x i64> @vsaddu_vx_v32i64_evl27(<32 x i64> %va, <32 x i1> %m) {
; RV32-NEXT: vmv.v.i v24, -1
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV32-NEXT: vsaddu.vv v8, v8, v24, v0.t
-; RV32-NEXT: vsetivli zero, 11, e64, m8, ta, ma
; RV32-NEXT: vmv1r.v v0, v7
+; RV32-NEXT: vsetivli zero, 11, e64, m8, ta, ma
; RV32-NEXT: vsaddu.vv v16, v16, v24, v0.t
; RV32-NEXT: ret
;
@@ -1498,8 +1498,8 @@ define <32 x i64> @vsaddu_vx_v32i64_evl27(<32 x i64> %va, <32 x i1> %m) {
; RV64-NEXT: vslidedown.vi v24, v0, 2
; RV64-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV64-NEXT: vsaddu.vi v8, v8, -1, v0.t
-; RV64-NEXT: vsetivli zero, 11, e64, m8, ta, ma
; RV64-NEXT: vmv1r.v v0, v24
+; RV64-NEXT: vsetivli zero, 11, e64, m8, ta, ma
; RV64-NEXT: vsaddu.vi v16, v16, -1, v0.t
; RV64-NEXT: ret
%v = call <32 x i64> @llvm.vp.uadd.sat.v32i64(<32 x i64> %va, <32 x i64> splat (i64 -1), <32 x i1> %m, i32 27)
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vscale-range.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vscale-range.ll
index 12d96fbfb88d6..4f533f2055bf3 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vscale-range.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vscale-range.ll
@@ -24,17 +24,17 @@ define <512 x i8> @vadd_v512i8_zvl128(<512 x i8> %a, <512 x i8> %b) #0 {
; CHECK-NEXT: addi a2, a2, 16
; CHECK-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
; CHECK-NEXT: li a2, 128
+; CHECK-NEXT: addi a4, a3, 128
+; CHECK-NEXT: addi a5, a3, 384
; CHECK-NEXT: vsetvli zero, a2, e8, m8, ta, ma
-; CHECK-NEXT: addi a2, a3, 128
-; CHECK-NEXT: addi a4, a3, 384
-; CHECK-NEXT: vle8.v v8, (a4)
-; CHECK-NEXT: csrr a4, vlenb
+; CHECK-NEXT: vle8.v v8, (a5)
+; CHECK-NEXT: csrr a2, vlenb
; CHECK-NEXT: li a5, 24
-; CHECK-NEXT: mul a4, a4, a5
-; CHECK-NEXT: add a4, sp, a4
-; CHECK-NEXT: addi a4, a4, 16
-; CHECK-NEXT: vs8r.v v8, (a4) # Unknown-size Folded Spill
-; CHECK-NEXT: addi a4, a1, 128
+; CHECK-NEXT: mul a2, a2, a5
+; CHECK-NEXT: add a2, sp, a2
+; CHECK-NEXT: addi a2, a2, 16
+; CHECK-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
+; CHECK-NEXT: addi a2, a1, 128
; CHECK-NEXT: vle8.v v8, (a1)
; CHECK-NEXT: csrr a1, vlenb
; CHECK-NEXT: slli a1, a1, 4
@@ -48,10 +48,10 @@ define <512 x i8> @vadd_v512i8_zvl128(<512 x i8> %a, <512 x i8> %b) #0 {
; CHECK-NEXT: add a1, sp, a1
; CHECK-NEXT: addi a1, a1, 16
; CHECK-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
-; CHECK-NEXT: vle8.v v8, (a4)
+; CHECK-NEXT: vle8.v v8, (a2)
; CHECK-NEXT: addi a1, sp, 16
; CHECK-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
-; CHECK-NEXT: vle8.v v24, (a2)
+; CHECK-NEXT: vle8.v v24, (a4)
; CHECK-NEXT: vle8.v v0, (a3)
; CHECK-NEXT: csrr a1, vlenb
; CHECK-NEXT: slli a1, a1, 4
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vselect-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vselect-vp.ll
index d05f580ea7d22..0a2ed3eb1ffbf 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vselect-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vselect-vp.ll
@@ -175,19 +175,18 @@ define <256 x i8> @select_v256i8(<256 x i1> %a, <256 x i8> %b, <256 x i8> %c, i3
; CHECK-NEXT: vle8.v v16, (a0)
; CHECK-NEXT: addi a0, a3, -128
; CHECK-NEXT: sltu a4, a3, a0
-; CHECK-NEXT: addi a4, a4, -1
; CHECK-NEXT: vle8.v v0, (a1)
; CHECK-NEXT: addi a1, sp, 16
; CHECK-NEXT: vs8r.v v0, (a1) # Unknown-size Folded Spill
+; CHECK-NEXT: addi a4, a4, -1
; CHECK-NEXT: and a0, a4, a0
-; CHECK-NEXT: vsetvli zero, a0, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, m8, ta, ma
; CHECK-NEXT: vmerge.vvm v24, v16, v24, v0
; CHECK-NEXT: bltu a3, a2, .LBB11_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: li a3, 128
; CHECK-NEXT: .LBB11_2:
-; CHECK-NEXT: vsetvli zero, a3, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
@@ -196,6 +195,7 @@ define <256 x i8> @select_v256i8(<256 x i1> %a, <256 x i8> %b, <256 x i8> %c, i3
; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, a3, e8, m8, ta, ma
; CHECK-NEXT: vmerge.vvm v8, v16, v8, v0
; CHECK-NEXT: vmv8r.v v16, v24
; CHECK-NEXT: csrr a0, vlenb
@@ -221,39 +221,39 @@ define <256 x i8> @select_evl_v256i8(<256 x i1> %a, <256 x i8> %b, <256 x i8> %c
; CHECK-NEXT: vsetvli zero, a2, e8, m8, ta, ma
; CHECK-NEXT: vle8.v v24, (a0)
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
+; CHECK-NEXT: slli a0, a0, 4
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: addi a0, a1, 128
; CHECK-NEXT: vle8.v v24, (a0)
+; CHECK-NEXT: addi a0, sp, 16
+; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vle8.v v24, (a1)
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 4
+; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vle8.v v16, (a1)
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: vsetivli zero, 1, e8, m8, ta, ma
+; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vmv1r.v v7, v0
; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
+; CHECK-NEXT: slli a0, a0, 4
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vmerge.vvm v24, v24, v16, v0
-; CHECK-NEXT: vsetvli zero, a2, e8, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: addi a0, sp, 16
+; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetivli zero, 1, e8, m8, ta, ma
+; CHECK-NEXT: vmerge.vvm v24, v8, v24, v0
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 4
+; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vmerge.vvm v8, v16, v8, v0
+; CHECK-NEXT: vsetvli zero, a2, e8, m8, ta, ma
+; CHECK-NEXT: vmerge.vvm v8, v8, v16, v0
; CHECK-NEXT: vmv8r.v v16, v24
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: li a1, 24
@@ -437,12 +437,12 @@ define <32 x i64> @select_v32i64(<32 x i1> %a, <32 x i64> %b, <32 x i64> %c, i32
; CHECK-NEXT: addi a0, a2, -16
; CHECK-NEXT: sltu a1, a2, a0
; CHECK-NEXT: addi a1, a1, -1
-; CHECK-NEXT: and a0, a1, a0
; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; CHECK-NEXT: vslidedown.vi v0, v0, 2
+; CHECK-NEXT: and a0, a1, a0
+; CHECK-NEXT: addi a1, sp, 16
+; CHECK-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vmerge.vvm v16, v24, v16, v0
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
@@ -456,15 +456,41 @@ define <32 x i64> @select_v32i64(<32 x i1> %a, <32 x i64> %b, <32 x i64> %c, i32
define <32 x i64> @select_evl_v32i64(<32 x i1> %a, <32 x i64> %b, <32 x i64> %c) {
; CHECK-LABEL: select_evl_v32i64:
; CHECK: # %bb.0:
+; CHECK-NEXT: addi sp, sp, -16
+; CHECK-NEXT: .cfi_def_cfa_offset 16
+; CHECK-NEXT: csrr a1, vlenb
+; CHECK-NEXT: slli a1, a1, 4
+; CHECK-NEXT: sub sp, sp, a1
+; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
+; CHECK-NEXT: addi a1, a0, 128
; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; CHECK-NEXT: vle64.v v24, (a0)
-; CHECK-NEXT: vmerge.vvm v8, v24, v8, v0
-; CHECK-NEXT: addi a0, a0, 128
-; CHECK-NEXT: vle64.v v24, (a0)
+; CHECK-NEXT: addi a0, sp, 16
+; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vle64.v v24, (a1)
+; CHECK-NEXT: csrr a0, vlenb
+; CHECK-NEXT: slli a0, a0, 3
+; CHECK-NEXT: add a0, sp, a0
+; CHECK-NEXT: addi a0, a0, 16
+; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
-; CHECK-NEXT: vslidedown.vi v0, v0, 2
+; CHECK-NEXT: vslidedown.vi v7, v0, 2
+; CHECK-NEXT: addi a0, sp, 16
+; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; CHECK-NEXT: vmerge.vvm v8, v24, v8, v0
+; CHECK-NEXT: vmv1r.v v0, v7
+; CHECK-NEXT: csrr a0, vlenb
+; CHECK-NEXT: slli a0, a0, 3
+; CHECK-NEXT: add a0, sp, a0
+; CHECK-NEXT: addi a0, a0, 16
+; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vsetivli zero, 1, e64, m8, ta, ma
; CHECK-NEXT: vmerge.vvm v16, v24, v16, v0
+; CHECK-NEXT: csrr a0, vlenb
+; CHECK-NEXT: slli a0, a0, 4
+; CHECK-NEXT: add sp, sp, a0
+; CHECK-NEXT: addi sp, sp, 16
; CHECK-NEXT: ret
%v = call <32 x i64> @llvm.vp.select.v32i64(<32 x i1> %a, <32 x i64> %b, <32 x i64> %c, i32 17)
ret <32 x i64> %v
@@ -594,12 +620,12 @@ define <64 x float> @select_v64f32(<64 x i1> %a, <64 x float> %b, <64 x float> %
; CHECK-NEXT: addi a0, a2, -32
; CHECK-NEXT: sltu a1, a2, a0
; CHECK-NEXT: addi a1, a1, -1
-; CHECK-NEXT: and a0, a1, a0
; CHECK-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
; CHECK-NEXT: vslidedown.vi v0, v0, 4
+; CHECK-NEXT: and a0, a1, a0
+; CHECK-NEXT: addi a1, sp, 16
+; CHECK-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vmerge.vvm v16, v24, v16, v0
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vselect.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vselect.ll
index 7dcd4c4199827..ed2ed2a2ebfaa 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vselect.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vselect.ll
@@ -5,8 +5,8 @@
define void @vselect_vv_v6i32(ptr %a, ptr %b, ptr %cc, ptr %z) {
; RV32-LABEL: vselect_vv_v6i32:
; RV32: # %bb.0:
-; RV32-NEXT: vsetivli zero, 6, e32, m2, ta, ma
; RV32-NEXT: lbu a2, 0(a2)
+; RV32-NEXT: vsetivli zero, 6, e32, m2, ta, ma
; RV32-NEXT: vle32.v v8, (a1)
; RV32-NEXT: slli a1, a2, 30
; RV32-NEXT: srli a1, a1, 31
@@ -35,8 +35,8 @@ define void @vselect_vv_v6i32(ptr %a, ptr %b, ptr %cc, ptr %z) {
;
; RV64-LABEL: vselect_vv_v6i32:
; RV64: # %bb.0:
-; RV64-NEXT: vsetivli zero, 6, e32, m2, ta, ma
; RV64-NEXT: lbu a2, 0(a2)
+; RV64-NEXT: vsetivli zero, 6, e32, m2, ta, ma
; RV64-NEXT: vle32.v v8, (a1)
; RV64-NEXT: slli a1, a2, 62
; RV64-NEXT: srli a1, a1, 63
@@ -73,8 +73,8 @@ define void @vselect_vv_v6i32(ptr %a, ptr %b, ptr %cc, ptr %z) {
define void @vselect_vx_v6i32(i32 %a, ptr %b, ptr %cc, ptr %z) {
; RV32-LABEL: vselect_vx_v6i32:
; RV32: # %bb.0:
-; RV32-NEXT: vsetivli zero, 6, e32, m2, ta, ma
; RV32-NEXT: lbu a2, 0(a2)
+; RV32-NEXT: vsetivli zero, 6, e32, m2, ta, ma
; RV32-NEXT: vle32.v v8, (a1)
; RV32-NEXT: slli a1, a2, 30
; RV32-NEXT: srli a1, a1, 31
@@ -104,8 +104,8 @@ define void @vselect_vx_v6i32(i32 %a, ptr %b, ptr %cc, ptr %z) {
;
; RV64-LABEL: vselect_vx_v6i32:
; RV64: # %bb.0:
-; RV64-NEXT: vsetivli zero, 6, e32, m2, ta, ma
; RV64-NEXT: lbu a2, 0(a2)
+; RV64-NEXT: vsetivli zero, 6, e32, m2, ta, ma
; RV64-NEXT: vle32.v v8, (a1)
; RV64-NEXT: slli a1, a2, 62
; RV64-NEXT: srli a1, a1, 63
@@ -144,8 +144,8 @@ define void @vselect_vx_v6i32(i32 %a, ptr %b, ptr %cc, ptr %z) {
define void @vselect_vi_v6i32(ptr %b, ptr %cc, ptr %z) {
; RV32-LABEL: vselect_vi_v6i32:
; RV32: # %bb.0:
-; RV32-NEXT: vsetivli zero, 6, e32, m2, ta, ma
; RV32-NEXT: lbu a1, 0(a1)
+; RV32-NEXT: vsetivli zero, 6, e32, m2, ta, ma
; RV32-NEXT: vle32.v v8, (a0)
; RV32-NEXT: slli a0, a1, 30
; RV32-NEXT: srli a0, a0, 31
@@ -175,8 +175,8 @@ define void @vselect_vi_v6i32(ptr %b, ptr %cc, ptr %z) {
;
; RV64-LABEL: vselect_vi_v6i32:
; RV64: # %bb.0:
-; RV64-NEXT: vsetivli zero, 6, e32, m2, ta, ma
; RV64-NEXT: lbu a1, 0(a1)
+; RV64-NEXT: vsetivli zero, 6, e32, m2, ta, ma
; RV64-NEXT: vle32.v v8, (a0)
; RV64-NEXT: slli a0, a1, 62
; RV64-NEXT: srli a0, a0, 63
@@ -214,8 +214,8 @@ define void @vselect_vi_v6i32(ptr %b, ptr %cc, ptr %z) {
define void @vselect_vv_v6f32(ptr %a, ptr %b, ptr %cc, ptr %z) {
; RV32-LABEL: vselect_vv_v6f32:
; RV32: # %bb.0:
-; RV32-NEXT: vsetivli zero, 6, e32, m2, ta, ma
; RV32-NEXT: lbu a2, 0(a2)
+; RV32-NEXT: vsetivli zero, 6, e32, m2, ta, ma
; RV32-NEXT: vle32.v v8, (a1)
; RV32-NEXT: slli a1, a2, 30
; RV32-NEXT: srli a1, a1, 31
@@ -244,8 +244,8 @@ define void @vselect_vv_v6f32(ptr %a, ptr %b, ptr %cc, ptr %z) {
;
; RV64-LABEL: vselect_vv_v6f32:
; RV64: # %bb.0:
-; RV64-NEXT: vsetivli zero, 6, e32, m2, ta, ma
; RV64-NEXT: lbu a2, 0(a2)
+; RV64-NEXT: vsetivli zero, 6, e32, m2, ta, ma
; RV64-NEXT: vle32.v v8, (a1)
; RV64-NEXT: slli a1, a2, 62
; RV64-NEXT: srli a1, a1, 63
@@ -282,8 +282,8 @@ define void @vselect_vv_v6f32(ptr %a, ptr %b, ptr %cc, ptr %z) {
define void @vselect_vx_v6f32(float %a, ptr %b, ptr %cc, ptr %z) {
; RV32-LABEL: vselect_vx_v6f32:
; RV32: # %bb.0:
-; RV32-NEXT: vsetivli zero, 6, e32, m2, ta, ma
; RV32-NEXT: lbu a1, 0(a1)
+; RV32-NEXT: vsetivli zero, 6, e32, m2, ta, ma
; RV32-NEXT: vle32.v v8, (a0)
; RV32-NEXT: slli a0, a1, 30
; RV32-NEXT: srli a0, a0, 31
@@ -313,8 +313,8 @@ define void @vselect_vx_v6f32(float %a, ptr %b, ptr %cc, ptr %z) {
;
; RV64-LABEL: vselect_vx_v6f32:
; RV64: # %bb.0:
-; RV64-NEXT: vsetivli zero, 6, e32, m2, ta, ma
; RV64-NEXT: lbu a1, 0(a1)
+; RV64-NEXT: vsetivli zero, 6, e32, m2, ta, ma
; RV64-NEXT: vle32.v v8, (a0)
; RV64-NEXT: slli a0, a1, 62
; RV64-NEXT: srli a0, a0, 63
@@ -353,8 +353,8 @@ define void @vselect_vx_v6f32(float %a, ptr %b, ptr %cc, ptr %z) {
define void @vselect_vfpzero_v6f32(ptr %b, ptr %cc, ptr %z) {
; RV32-LABEL: vselect_vfpzero_v6f32:
; RV32: # %bb.0:
-; RV32-NEXT: vsetivli zero, 6, e32, m2, ta, ma
; RV32-NEXT: lbu a1, 0(a1)
+; RV32-NEXT: vsetivli zero, 6, e32, m2, ta, ma
; RV32-NEXT: vle32.v v8, (a0)
; RV32-NEXT: slli a0, a1, 30
; RV32-NEXT: srli a0, a0, 31
@@ -384,8 +384,8 @@ define void @vselect_vfpzero_v6f32(ptr %b, ptr %cc, ptr %z) {
;
; RV64-LABEL: vselect_vfpzero_v6f32:
; RV64: # %bb.0:
-; RV64-NEXT: vsetivli zero, 6, e32, m2, ta, ma
; RV64-NEXT: lbu a1, 0(a1)
+; RV64-NEXT: vsetivli zero, 6, e32, m2, ta, ma
; RV64-NEXT: vle32.v v8, (a0)
; RV64-NEXT: slli a0, a1, 62
; RV64-NEXT: srli a0, a0, 63
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vssub-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vssub-vp.ll
index 2caa2ff41a7d9..b82ca70477ba3 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vssub-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vssub-vp.ll
@@ -399,8 +399,8 @@ define <256 x i8> @vssub_vi_v258i8(<256 x i8> %va, <256 x i1> %m, i32 zeroext %e
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: li a1, 128
; CHECK-NEXT: .LBB32_2:
-; CHECK-NEXT: vsetvli zero, a1, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a1, e8, m8, ta, ma
; CHECK-NEXT: vssub.vx v8, v8, a0, v0.t
; CHECK-NEXT: ret
%v = call <256 x i8> @llvm.vp.ssub.sat.v258i8(<256 x i8> %va, <256 x i8> splat (i8 -1), <256 x i1> %m, i32 %evl)
@@ -440,8 +440,8 @@ define <256 x i8> @vssub_vi_v258i8_evl129(<256 x i8> %va, <256 x i1> %m) {
; CHECK-NEXT: vlm.v v24, (a0)
; CHECK-NEXT: li a0, -1
; CHECK-NEXT: vssub.vx v8, v8, a0, v0.t
-; CHECK-NEXT: vsetivli zero, 1, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetivli zero, 1, e8, m8, ta, ma
; CHECK-NEXT: vssub.vx v16, v16, a0, v0.t
; CHECK-NEXT: ret
%v = call <256 x i8> @llvm.vp.ssub.sat.v258i8(<256 x i8> %va, <256 x i8> splat (i8 -1), <256 x i1> %m, i32 129)
@@ -458,8 +458,8 @@ define <256 x i8> @vssub_vi_v258i8_evl128(<256 x i8> %va, <256 x i1> %m) {
; CHECK-NEXT: vlm.v v24, (a0)
; CHECK-NEXT: li a0, -1
; CHECK-NEXT: vssub.vx v8, v8, a0, v0.t
-; CHECK-NEXT: vsetivli zero, 0, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetivli zero, 0, e8, m8, ta, ma
; CHECK-NEXT: vssub.vx v16, v16, a0, v0.t
; CHECK-NEXT: ret
%v = call <256 x i8> @llvm.vp.ssub.sat.v258i8(<256 x i8> %va, <256 x i8> splat (i8 -1), <256 x i1> %m, i32 128)
@@ -1401,8 +1401,8 @@ declare <32 x i64> @llvm.vp.ssub.sat.v32i64(<32 x i64>, <32 x i64>, <32 x i1>, i
define <32 x i64> @vssub_vx_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %evl) {
; RV32-LABEL: vssub_vx_v32i64:
; RV32: # %bb.0:
-; RV32-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV32-NEXT: li a2, 16
+; RV32-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV32-NEXT: vslidedown.vi v7, v0, 2
; RV32-NEXT: mv a1, a0
; RV32-NEXT: bltu a0, a2, .LBB108_2
@@ -1418,15 +1418,15 @@ define <32 x i64> @vssub_vx_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %ev
; RV32-NEXT: sltu a0, a0, a1
; RV32-NEXT: addi a0, a0, -1
; RV32-NEXT: and a0, a0, a1
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vmv1r.v v0, v7
+; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vssub.vv v16, v16, v24, v0.t
; RV32-NEXT: ret
;
; RV64-LABEL: vssub_vx_v32i64:
; RV64: # %bb.0:
-; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64-NEXT: li a2, 16
+; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64-NEXT: vslidedown.vi v24, v0, 2
; RV64-NEXT: mv a1, a0
; RV64-NEXT: bltu a0, a2, .LBB108_2
@@ -1440,8 +1440,8 @@ define <32 x i64> @vssub_vx_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %ev
; RV64-NEXT: sltu a0, a0, a1
; RV64-NEXT: addi a0, a0, -1
; RV64-NEXT: and a0, a0, a1
-; RV64-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV64-NEXT: vmv1r.v v0, v24
+; RV64-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV64-NEXT: vssub.vx v16, v16, a2, v0.t
; RV64-NEXT: ret
%v = call <32 x i64> @llvm.vp.ssub.sat.v32i64(<32 x i64> %va, <32 x i64> splat (i64 -1), <32 x i1> %m, i32 %evl)
@@ -1504,8 +1504,8 @@ define <32 x i64> @vssub_vx_v32i64_evl12(<32 x i64> %va, <32 x i1> %m) {
; RV32-NEXT: vmv.v.i v24, -1
; RV32-NEXT: vsetivli zero, 12, e64, m8, ta, ma
; RV32-NEXT: vssub.vv v8, v8, v24, v0.t
-; RV32-NEXT: vsetivli zero, 0, e64, m8, ta, ma
; RV32-NEXT: vmv1r.v v0, v7
+; RV32-NEXT: vsetivli zero, 0, e64, m8, ta, ma
; RV32-NEXT: vssub.vv v16, v16, v24, v0.t
; RV32-NEXT: ret
;
@@ -1516,8 +1516,8 @@ define <32 x i64> @vssub_vx_v32i64_evl12(<32 x i64> %va, <32 x i1> %m) {
; RV64-NEXT: li a0, -1
; RV64-NEXT: vsetivli zero, 12, e64, m8, ta, ma
; RV64-NEXT: vssub.vx v8, v8, a0, v0.t
-; RV64-NEXT: vsetivli zero, 0, e64, m8, ta, ma
; RV64-NEXT: vmv1r.v v0, v24
+; RV64-NEXT: vsetivli zero, 0, e64, m8, ta, ma
; RV64-NEXT: vssub.vx v16, v16, a0, v0.t
; RV64-NEXT: ret
%v = call <32 x i64> @llvm.vp.ssub.sat.v32i64(<32 x i64> %va, <32 x i64> splat (i64 -1), <32 x i1> %m, i32 12)
@@ -1534,8 +1534,8 @@ define <32 x i64> @vssub_vx_v32i64_evl27(<32 x i64> %va, <32 x i1> %m) {
; RV32-NEXT: vmv.v.i v24, -1
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV32-NEXT: vssub.vv v8, v8, v24, v0.t
-; RV32-NEXT: vsetivli zero, 11, e64, m8, ta, ma
; RV32-NEXT: vmv1r.v v0, v7
+; RV32-NEXT: vsetivli zero, 11, e64, m8, ta, ma
; RV32-NEXT: vssub.vv v16, v16, v24, v0.t
; RV32-NEXT: ret
;
@@ -1546,8 +1546,8 @@ define <32 x i64> @vssub_vx_v32i64_evl27(<32 x i64> %va, <32 x i1> %m) {
; RV64-NEXT: li a0, -1
; RV64-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV64-NEXT: vssub.vx v8, v8, a0, v0.t
-; RV64-NEXT: vsetivli zero, 11, e64, m8, ta, ma
; RV64-NEXT: vmv1r.v v0, v24
+; RV64-NEXT: vsetivli zero, 11, e64, m8, ta, ma
; RV64-NEXT: vssub.vx v16, v16, a0, v0.t
; RV64-NEXT: ret
%v = call <32 x i64> @llvm.vp.ssub.sat.v32i64(<32 x i64> %va, <32 x i64> splat (i64 -1), <32 x i1> %m, i32 27)
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vssubu-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vssubu-vp.ll
index 6313f31bc1a61..6d8ed563f02bd 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vssubu-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vssubu-vp.ll
@@ -394,8 +394,8 @@ define <256 x i8> @vssubu_vi_v258i8(<256 x i8> %va, <256 x i1> %m, i32 zeroext %
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: li a1, 128
; CHECK-NEXT: .LBB32_2:
-; CHECK-NEXT: vsetvli zero, a1, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a1, e8, m8, ta, ma
; CHECK-NEXT: vssubu.vx v8, v8, a0, v0.t
; CHECK-NEXT: ret
%v = call <256 x i8> @llvm.vp.usub.sat.v258i8(<256 x i8> %va, <256 x i8> splat (i8 -1), <256 x i1> %m, i32 %evl)
@@ -435,8 +435,8 @@ define <256 x i8> @vssubu_vi_v258i8_evl129(<256 x i8> %va, <256 x i1> %m) {
; CHECK-NEXT: vlm.v v24, (a0)
; CHECK-NEXT: li a0, -1
; CHECK-NEXT: vssubu.vx v8, v8, a0, v0.t
-; CHECK-NEXT: vsetivli zero, 1, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetivli zero, 1, e8, m8, ta, ma
; CHECK-NEXT: vssubu.vx v16, v16, a0, v0.t
; CHECK-NEXT: ret
%v = call <256 x i8> @llvm.vp.usub.sat.v258i8(<256 x i8> %va, <256 x i8> splat (i8 -1), <256 x i1> %m, i32 129)
@@ -453,8 +453,8 @@ define <256 x i8> @vssubu_vi_v258i8_evl128(<256 x i8> %va, <256 x i1> %m) {
; CHECK-NEXT: vlm.v v24, (a0)
; CHECK-NEXT: li a0, -1
; CHECK-NEXT: vssubu.vx v8, v8, a0, v0.t
-; CHECK-NEXT: vsetivli zero, 0, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetivli zero, 0, e8, m8, ta, ma
; CHECK-NEXT: vssubu.vx v16, v16, a0, v0.t
; CHECK-NEXT: ret
%v = call <256 x i8> @llvm.vp.usub.sat.v258i8(<256 x i8> %va, <256 x i8> splat (i8 -1), <256 x i1> %m, i32 128)
@@ -1396,8 +1396,8 @@ declare <32 x i64> @llvm.vp.usub.sat.v32i64(<32 x i64>, <32 x i64>, <32 x i1>, i
define <32 x i64> @vssubu_vx_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %evl) {
; RV32-LABEL: vssubu_vx_v32i64:
; RV32: # %bb.0:
-; RV32-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV32-NEXT: li a2, 16
+; RV32-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV32-NEXT: vslidedown.vi v7, v0, 2
; RV32-NEXT: mv a1, a0
; RV32-NEXT: bltu a0, a2, .LBB108_2
@@ -1413,15 +1413,15 @@ define <32 x i64> @vssubu_vx_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %e
; RV32-NEXT: sltu a0, a0, a1
; RV32-NEXT: addi a0, a0, -1
; RV32-NEXT: and a0, a0, a1
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vmv1r.v v0, v7
+; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vssubu.vv v16, v16, v24, v0.t
; RV32-NEXT: ret
;
; RV64-LABEL: vssubu_vx_v32i64:
; RV64: # %bb.0:
-; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64-NEXT: li a2, 16
+; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; RV64-NEXT: vslidedown.vi v24, v0, 2
; RV64-NEXT: mv a1, a0
; RV64-NEXT: bltu a0, a2, .LBB108_2
@@ -1435,8 +1435,8 @@ define <32 x i64> @vssubu_vx_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %e
; RV64-NEXT: sltu a0, a0, a1
; RV64-NEXT: addi a0, a0, -1
; RV64-NEXT: and a0, a0, a1
-; RV64-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV64-NEXT: vmv1r.v v0, v24
+; RV64-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV64-NEXT: vssubu.vx v16, v16, a2, v0.t
; RV64-NEXT: ret
%v = call <32 x i64> @llvm.vp.usub.sat.v32i64(<32 x i64> %va, <32 x i64> splat (i64 -1), <32 x i1> %m, i32 %evl)
@@ -1499,8 +1499,8 @@ define <32 x i64> @vssubu_vx_v32i64_evl12(<32 x i64> %va, <32 x i1> %m) {
; RV32-NEXT: vmv.v.i v24, -1
; RV32-NEXT: vsetivli zero, 12, e64, m8, ta, ma
; RV32-NEXT: vssubu.vv v8, v8, v24, v0.t
-; RV32-NEXT: vsetivli zero, 0, e64, m8, ta, ma
; RV32-NEXT: vmv1r.v v0, v7
+; RV32-NEXT: vsetivli zero, 0, e64, m8, ta, ma
; RV32-NEXT: vssubu.vv v16, v16, v24, v0.t
; RV32-NEXT: ret
;
@@ -1511,8 +1511,8 @@ define <32 x i64> @vssubu_vx_v32i64_evl12(<32 x i64> %va, <32 x i1> %m) {
; RV64-NEXT: li a0, -1
; RV64-NEXT: vsetivli zero, 12, e64, m8, ta, ma
; RV64-NEXT: vssubu.vx v8, v8, a0, v0.t
-; RV64-NEXT: vsetivli zero, 0, e64, m8, ta, ma
; RV64-NEXT: vmv1r.v v0, v24
+; RV64-NEXT: vsetivli zero, 0, e64, m8, ta, ma
; RV64-NEXT: vssubu.vx v16, v16, a0, v0.t
; RV64-NEXT: ret
%v = call <32 x i64> @llvm.vp.usub.sat.v32i64(<32 x i64> %va, <32 x i64> splat (i64 -1), <32 x i1> %m, i32 12)
@@ -1529,8 +1529,8 @@ define <32 x i64> @vssubu_vx_v32i64_evl27(<32 x i64> %va, <32 x i1> %m) {
; RV32-NEXT: vmv.v.i v24, -1
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV32-NEXT: vssubu.vv v8, v8, v24, v0.t
-; RV32-NEXT: vsetivli zero, 11, e64, m8, ta, ma
; RV32-NEXT: vmv1r.v v0, v7
+; RV32-NEXT: vsetivli zero, 11, e64, m8, ta, ma
; RV32-NEXT: vssubu.vv v16, v16, v24, v0.t
; RV32-NEXT: ret
;
@@ -1541,8 +1541,8 @@ define <32 x i64> @vssubu_vx_v32i64_evl27(<32 x i64> %va, <32 x i1> %m) {
; RV64-NEXT: li a0, -1
; RV64-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV64-NEXT: vssubu.vx v8, v8, a0, v0.t
-; RV64-NEXT: vsetivli zero, 11, e64, m8, ta, ma
; RV64-NEXT: vmv1r.v v0, v24
+; RV64-NEXT: vsetivli zero, 11, e64, m8, ta, ma
; RV64-NEXT: vssubu.vx v16, v16, a0, v0.t
; RV64-NEXT: ret
%v = call <32 x i64> @llvm.vp.usub.sat.v32i64(<32 x i64> %va, <32 x i64> splat (i64 -1), <32 x i1> %m, i32 27)
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vwadd.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vwadd.ll
index a4a5917fd4f9e..d6ca6c5a4b83d 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vwadd.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vwadd.ll
@@ -263,13 +263,12 @@ define <128 x i16> @vwadd_v128i16(ptr %x, ptr %y) nounwind {
; CHECK-NEXT: vslidedown.vx v16, v8, a0
; CHECK-NEXT: vslidedown.vx v8, v0, a0
; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, ma
-; CHECK-NEXT: vmv4r.v v24, v8
-; CHECK-NEXT: vwadd.vv v8, v16, v24
+; CHECK-NEXT: vwadd.vv v24, v16, v8
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vwadd.vv v8, v16, v0
@@ -309,13 +308,12 @@ define <64 x i32> @vwadd_v64i32(ptr %x, ptr %y) nounwind {
; CHECK-NEXT: vslidedown.vx v16, v8, a0
; CHECK-NEXT: vslidedown.vx v8, v0, a0
; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, ma
-; CHECK-NEXT: vmv4r.v v24, v8
-; CHECK-NEXT: vwadd.vv v8, v16, v24
+; CHECK-NEXT: vwadd.vv v24, v16, v8
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vwadd.vv v8, v16, v0
@@ -354,13 +352,12 @@ define <32 x i64> @vwadd_v32i64(ptr %x, ptr %y) nounwind {
; CHECK-NEXT: vslidedown.vi v16, v8, 16
; CHECK-NEXT: vslidedown.vi v8, v0, 16
; CHECK-NEXT: vsetivli zero, 16, e32, m4, ta, ma
-; CHECK-NEXT: vmv4r.v v24, v8
-; CHECK-NEXT: vwadd.vv v8, v16, v24
+; CHECK-NEXT: vwadd.vv v24, v16, v8
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vwadd.vv v8, v16, v0
@@ -768,8 +765,8 @@ define <4 x i32> @vwadd_vx_v4i32_i32(ptr %x, ptr %y) {
define <2 x i64> @vwadd_vx_v2i64_i8(ptr %x, ptr %y) nounwind {
; RV32-LABEL: vwadd_vx_v2i64_i8:
; RV32: # %bb.0:
-; RV32-NEXT: vsetivli zero, 2, e64, m1, ta, ma
; RV32-NEXT: lb a1, 0(a1)
+; RV32-NEXT: vsetivli zero, 2, e64, m1, ta, ma
; RV32-NEXT: vle32.v v9, (a0)
; RV32-NEXT: vmv.v.x v8, a1
; RV32-NEXT: vsetvli zero, zero, e32, mf2, ta, ma
@@ -796,8 +793,8 @@ define <2 x i64> @vwadd_vx_v2i64_i8(ptr %x, ptr %y) nounwind {
define <2 x i64> @vwadd_vx_v2i64_i16(ptr %x, ptr %y) nounwind {
; RV32-LABEL: vwadd_vx_v2i64_i16:
; RV32: # %bb.0:
-; RV32-NEXT: vsetivli zero, 2, e64, m1, ta, ma
; RV32-NEXT: lh a1, 0(a1)
+; RV32-NEXT: vsetivli zero, 2, e64, m1, ta, ma
; RV32-NEXT: vle32.v v9, (a0)
; RV32-NEXT: vmv.v.x v8, a1
; RV32-NEXT: vsetvli zero, zero, e32, mf2, ta, ma
@@ -824,8 +821,8 @@ define <2 x i64> @vwadd_vx_v2i64_i16(ptr %x, ptr %y) nounwind {
define <2 x i64> @vwadd_vx_v2i64_i32(ptr %x, ptr %y) nounwind {
; RV32-LABEL: vwadd_vx_v2i64_i32:
; RV32: # %bb.0:
-; RV32-NEXT: vsetivli zero, 2, e64, m1, ta, ma
; RV32-NEXT: lw a1, 0(a1)
+; RV32-NEXT: vsetivli zero, 2, e64, m1, ta, ma
; RV32-NEXT: vle32.v v9, (a0)
; RV32-NEXT: vmv.v.x v8, a1
; RV32-NEXT: vsetvli zero, zero, e32, mf2, ta, ma
@@ -853,9 +850,9 @@ define <2 x i64> @vwadd_vx_v2i64_i64(ptr %x, ptr %y) nounwind {
; RV32-LABEL: vwadd_vx_v2i64_i64:
; RV32: # %bb.0:
; RV32-NEXT: addi sp, sp, -16
-; RV32-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
; RV32-NEXT: lw a2, 4(a1)
; RV32-NEXT: lw a1, 0(a1)
+; RV32-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
; RV32-NEXT: vle32.v v9, (a0)
; RV32-NEXT: sw a2, 12(sp)
; RV32-NEXT: sw a1, 8(sp)
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vwaddu.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vwaddu.ll
index bc0bf5dd76ad4..61378a424ecba 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vwaddu.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vwaddu.ll
@@ -263,13 +263,12 @@ define <128 x i16> @vwaddu_v128i16(ptr %x, ptr %y) nounwind {
; CHECK-NEXT: vslidedown.vx v16, v8, a0
; CHECK-NEXT: vslidedown.vx v8, v0, a0
; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, ma
-; CHECK-NEXT: vmv4r.v v24, v8
-; CHECK-NEXT: vwaddu.vv v8, v16, v24
+; CHECK-NEXT: vwaddu.vv v24, v16, v8
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vwaddu.vv v8, v16, v0
@@ -309,13 +308,12 @@ define <64 x i32> @vwaddu_v64i32(ptr %x, ptr %y) nounwind {
; CHECK-NEXT: vslidedown.vx v16, v8, a0
; CHECK-NEXT: vslidedown.vx v8, v0, a0
; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, ma
-; CHECK-NEXT: vmv4r.v v24, v8
-; CHECK-NEXT: vwaddu.vv v8, v16, v24
+; CHECK-NEXT: vwaddu.vv v24, v16, v8
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vwaddu.vv v8, v16, v0
@@ -354,13 +352,12 @@ define <32 x i64> @vwaddu_v32i64(ptr %x, ptr %y) nounwind {
; CHECK-NEXT: vslidedown.vi v16, v8, 16
; CHECK-NEXT: vslidedown.vi v8, v0, 16
; CHECK-NEXT: vsetivli zero, 16, e32, m4, ta, ma
-; CHECK-NEXT: vmv4r.v v24, v8
-; CHECK-NEXT: vwaddu.vv v8, v16, v24
+; CHECK-NEXT: vwaddu.vv v24, v16, v8
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vwaddu.vv v8, v16, v0
@@ -769,8 +766,8 @@ define <2 x i64> @vwaddu_vx_v2i64_i8(ptr %x, ptr %y) nounwind {
; RV32-LABEL: vwaddu_vx_v2i64_i8:
; RV32: # %bb.0:
; RV32-NEXT: addi sp, sp, -16
-; RV32-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
; RV32-NEXT: lbu a1, 0(a1)
+; RV32-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
; RV32-NEXT: vle32.v v9, (a0)
; RV32-NEXT: sw zero, 12(sp)
; RV32-NEXT: sw a1, 8(sp)
@@ -801,8 +798,8 @@ define <2 x i64> @vwaddu_vx_v2i64_i16(ptr %x, ptr %y) nounwind {
; RV32-LABEL: vwaddu_vx_v2i64_i16:
; RV32: # %bb.0:
; RV32-NEXT: addi sp, sp, -16
-; RV32-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
; RV32-NEXT: lhu a1, 0(a1)
+; RV32-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
; RV32-NEXT: vle32.v v9, (a0)
; RV32-NEXT: sw zero, 12(sp)
; RV32-NEXT: sw a1, 8(sp)
@@ -833,8 +830,8 @@ define <2 x i64> @vwaddu_vx_v2i64_i32(ptr %x, ptr %y) nounwind {
; RV32-LABEL: vwaddu_vx_v2i64_i32:
; RV32: # %bb.0:
; RV32-NEXT: addi sp, sp, -16
-; RV32-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
; RV32-NEXT: lw a1, 0(a1)
+; RV32-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
; RV32-NEXT: vle32.v v9, (a0)
; RV32-NEXT: sw zero, 12(sp)
; RV32-NEXT: sw a1, 8(sp)
@@ -865,9 +862,9 @@ define <2 x i64> @vwaddu_vx_v2i64_i64(ptr %x, ptr %y) nounwind {
; RV32-LABEL: vwaddu_vx_v2i64_i64:
; RV32: # %bb.0:
; RV32-NEXT: addi sp, sp, -16
-; RV32-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
; RV32-NEXT: lw a2, 4(a1)
; RV32-NEXT: lw a1, 0(a1)
+; RV32-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
; RV32-NEXT: vle32.v v9, (a0)
; RV32-NEXT: sw a2, 12(sp)
; RV32-NEXT: sw a1, 8(sp)
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vwmul.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vwmul.ll
index 2abd34f01c14c..93927e10e607e 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vwmul.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vwmul.ll
@@ -289,13 +289,12 @@ define <128 x i16> @vwmul_v128i16(ptr %x, ptr %y) {
; CHECK-NEXT: vslidedown.vx v16, v8, a0
; CHECK-NEXT: vslidedown.vx v8, v0, a0
; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, ma
-; CHECK-NEXT: vmv4r.v v24, v8
-; CHECK-NEXT: vwmul.vv v8, v16, v24
+; CHECK-NEXT: vwmul.vv v24, v16, v8
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vwmul.vv v8, v16, v0
@@ -337,13 +336,12 @@ define <64 x i32> @vwmul_v64i32(ptr %x, ptr %y) {
; CHECK-NEXT: vslidedown.vx v16, v8, a0
; CHECK-NEXT: vslidedown.vx v8, v0, a0
; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, ma
-; CHECK-NEXT: vmv4r.v v24, v8
-; CHECK-NEXT: vwmul.vv v8, v16, v24
+; CHECK-NEXT: vwmul.vv v24, v16, v8
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vwmul.vv v8, v16, v0
@@ -384,13 +382,12 @@ define <32 x i64> @vwmul_v32i64(ptr %x, ptr %y) {
; CHECK-NEXT: vslidedown.vi v16, v8, 16
; CHECK-NEXT: vslidedown.vi v8, v0, 16
; CHECK-NEXT: vsetivli zero, 16, e32, m4, ta, ma
-; CHECK-NEXT: vmv4r.v v24, v8
-; CHECK-NEXT: vwmul.vv v8, v16, v24
+; CHECK-NEXT: vwmul.vv v24, v16, v8
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vwmul.vv v8, v16, v0
@@ -883,9 +880,9 @@ define <2 x i64> @vwmul_vx_v2i64_i64(ptr %x, ptr %y) {
; RV32: # %bb.0:
; RV32-NEXT: addi sp, sp, -16
; RV32-NEXT: .cfi_def_cfa_offset 16
-; RV32-NEXT: vsetivli zero, 2, e64, m1, ta, ma
; RV32-NEXT: lw a2, 4(a1)
; RV32-NEXT: lw a1, 0(a1)
+; RV32-NEXT: vsetivli zero, 2, e64, m1, ta, ma
; RV32-NEXT: vle32.v v8, (a0)
; RV32-NEXT: sw a2, 12(sp)
; RV32-NEXT: sw a1, 8(sp)
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vwmulsu.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vwmulsu.ll
index 921037db2ea99..ee114350a4323 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vwmulsu.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vwmulsu.ll
@@ -281,13 +281,12 @@ define <128 x i16> @vwmulsu_v128i16(ptr %x, ptr %y) {
; CHECK-NEXT: vslidedown.vx v16, v8, a0
; CHECK-NEXT: vslidedown.vx v8, v0, a0
; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, ma
-; CHECK-NEXT: vmv4r.v v24, v8
-; CHECK-NEXT: vwmulsu.vv v8, v24, v16
+; CHECK-NEXT: vwmulsu.vv v24, v8, v16
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vwmulsu.vv v8, v0, v16
@@ -329,13 +328,12 @@ define <64 x i32> @vwmulsu_v64i32(ptr %x, ptr %y) {
; CHECK-NEXT: vslidedown.vx v16, v8, a0
; CHECK-NEXT: vslidedown.vx v8, v0, a0
; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, ma
-; CHECK-NEXT: vmv4r.v v24, v8
-; CHECK-NEXT: vwmulsu.vv v8, v24, v16
+; CHECK-NEXT: vwmulsu.vv v24, v8, v16
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vwmulsu.vv v8, v0, v16
@@ -376,13 +374,12 @@ define <32 x i64> @vwmulsu_v32i64(ptr %x, ptr %y) {
; CHECK-NEXT: vslidedown.vi v16, v8, 16
; CHECK-NEXT: vslidedown.vi v8, v0, 16
; CHECK-NEXT: vsetivli zero, 16, e32, m4, ta, ma
-; CHECK-NEXT: vmv4r.v v24, v8
-; CHECK-NEXT: vwmulsu.vv v8, v24, v16
+; CHECK-NEXT: vwmulsu.vv v24, v8, v16
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vwmulsu.vv v8, v0, v16
@@ -793,8 +790,8 @@ define <2 x i64> @vwmulsu_vx_v2i64_i8(ptr %x, ptr %y) {
; RV32: # %bb.0:
; RV32-NEXT: addi sp, sp, -16
; RV32-NEXT: .cfi_def_cfa_offset 16
-; RV32-NEXT: vsetivli zero, 2, e64, m1, ta, ma
; RV32-NEXT: lbu a1, 0(a1)
+; RV32-NEXT: vsetivli zero, 2, e64, m1, ta, ma
; RV32-NEXT: vle32.v v8, (a0)
; RV32-NEXT: sw zero, 12(sp)
; RV32-NEXT: sw a1, 8(sp)
@@ -827,8 +824,8 @@ define <2 x i64> @vwmulsu_vx_v2i64_i16(ptr %x, ptr %y) {
; RV32: # %bb.0:
; RV32-NEXT: addi sp, sp, -16
; RV32-NEXT: .cfi_def_cfa_offset 16
-; RV32-NEXT: vsetivli zero, 2, e64, m1, ta, ma
; RV32-NEXT: lhu a1, 0(a1)
+; RV32-NEXT: vsetivli zero, 2, e64, m1, ta, ma
; RV32-NEXT: vle32.v v8, (a0)
; RV32-NEXT: sw zero, 12(sp)
; RV32-NEXT: sw a1, 8(sp)
@@ -861,8 +858,8 @@ define <2 x i64> @vwmulsu_vx_v2i64_i32(ptr %x, ptr %y) {
; RV32: # %bb.0:
; RV32-NEXT: addi sp, sp, -16
; RV32-NEXT: .cfi_def_cfa_offset 16
-; RV32-NEXT: vsetivli zero, 2, e64, m1, ta, ma
; RV32-NEXT: lw a1, 0(a1)
+; RV32-NEXT: vsetivli zero, 2, e64, m1, ta, ma
; RV32-NEXT: vle32.v v8, (a0)
; RV32-NEXT: sw zero, 12(sp)
; RV32-NEXT: sw a1, 8(sp)
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vwmulu.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vwmulu.ll
index b97c9654ad3cb..17a76ae5e7f75 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vwmulu.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vwmulu.ll
@@ -265,13 +265,12 @@ define <128 x i16> @vwmulu_v128i16(ptr %x, ptr %y) {
; CHECK-NEXT: vslidedown.vx v16, v8, a0
; CHECK-NEXT: vslidedown.vx v8, v0, a0
; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, ma
-; CHECK-NEXT: vmv4r.v v24, v8
-; CHECK-NEXT: vwmulu.vv v8, v16, v24
+; CHECK-NEXT: vwmulu.vv v24, v16, v8
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vwmulu.vv v8, v16, v0
@@ -313,13 +312,12 @@ define <64 x i32> @vwmulu_v64i32(ptr %x, ptr %y) {
; CHECK-NEXT: vslidedown.vx v16, v8, a0
; CHECK-NEXT: vslidedown.vx v8, v0, a0
; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, ma
-; CHECK-NEXT: vmv4r.v v24, v8
-; CHECK-NEXT: vwmulu.vv v8, v16, v24
+; CHECK-NEXT: vwmulu.vv v24, v16, v8
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vwmulu.vv v8, v16, v0
@@ -360,13 +358,12 @@ define <32 x i64> @vwmulu_v32i64(ptr %x, ptr %y) {
; CHECK-NEXT: vslidedown.vi v16, v8, 16
; CHECK-NEXT: vslidedown.vi v8, v0, 16
; CHECK-NEXT: vsetivli zero, 16, e32, m4, ta, ma
-; CHECK-NEXT: vmv4r.v v24, v8
-; CHECK-NEXT: vwmulu.vv v8, v16, v24
+; CHECK-NEXT: vwmulu.vv v24, v16, v8
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vwmulu.vv v8, v16, v0
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vwsub.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vwsub.ll
index 154093d759d6d..a2675d59ade93 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vwsub.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vwsub.ll
@@ -263,13 +263,12 @@ define <128 x i16> @vwsub_v128i16(ptr %x, ptr %y) nounwind {
; CHECK-NEXT: vslidedown.vx v16, v8, a0
; CHECK-NEXT: vslidedown.vx v8, v0, a0
; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, ma
-; CHECK-NEXT: vmv4r.v v24, v8
-; CHECK-NEXT: vwsub.vv v8, v16, v24
+; CHECK-NEXT: vwsub.vv v24, v16, v8
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vwsub.vv v8, v16, v0
@@ -309,13 +308,12 @@ define <64 x i32> @vwsub_v64i32(ptr %x, ptr %y) nounwind {
; CHECK-NEXT: vslidedown.vx v16, v8, a0
; CHECK-NEXT: vslidedown.vx v8, v0, a0
; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, ma
-; CHECK-NEXT: vmv4r.v v24, v8
-; CHECK-NEXT: vwsub.vv v8, v16, v24
+; CHECK-NEXT: vwsub.vv v24, v16, v8
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vwsub.vv v8, v16, v0
@@ -354,13 +352,12 @@ define <32 x i64> @vwsub_v32i64(ptr %x, ptr %y) nounwind {
; CHECK-NEXT: vslidedown.vi v16, v8, 16
; CHECK-NEXT: vslidedown.vi v8, v0, 16
; CHECK-NEXT: vsetivli zero, 16, e32, m4, ta, ma
-; CHECK-NEXT: vmv4r.v v24, v8
-; CHECK-NEXT: vwsub.vv v8, v16, v24
+; CHECK-NEXT: vwsub.vv v24, v16, v8
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vwsub.vv v8, v16, v0
@@ -715,8 +712,8 @@ define <8 x i16> @vwsub_vx_v8i16_i16(ptr %x, ptr %y) {
define <4 x i32> @vwsub_vx_v4i32_i8(ptr %x, ptr %y) {
; CHECK-LABEL: vwsub_vx_v4i32_i8:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetivli zero, 4, e16, mf2, ta, ma
; CHECK-NEXT: lb a1, 0(a1)
+; CHECK-NEXT: vsetivli zero, 4, e16, mf2, ta, ma
; CHECK-NEXT: vle16.v v9, (a0)
; CHECK-NEXT: vmv.v.x v10, a1
; CHECK-NEXT: vwsub.vv v8, v10, v9
@@ -769,8 +766,8 @@ define <4 x i32> @vwsub_vx_v4i32_i32(ptr %x, ptr %y) {
define <2 x i64> @vwsub_vx_v2i64_i8(ptr %x, ptr %y) nounwind {
; RV32-LABEL: vwsub_vx_v2i64_i8:
; RV32: # %bb.0:
-; RV32-NEXT: vsetivli zero, 2, e64, m1, ta, ma
; RV32-NEXT: lb a1, 0(a1)
+; RV32-NEXT: vsetivli zero, 2, e64, m1, ta, ma
; RV32-NEXT: vle32.v v9, (a0)
; RV32-NEXT: vmv.v.x v8, a1
; RV32-NEXT: vsetvli zero, zero, e32, mf2, ta, ma
@@ -779,8 +776,8 @@ define <2 x i64> @vwsub_vx_v2i64_i8(ptr %x, ptr %y) nounwind {
;
; RV64-LABEL: vwsub_vx_v2i64_i8:
; RV64: # %bb.0:
-; RV64-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
; RV64-NEXT: lb a1, 0(a1)
+; RV64-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
; RV64-NEXT: vle32.v v9, (a0)
; RV64-NEXT: vmv.v.x v10, a1
; RV64-NEXT: vwsub.vv v8, v10, v9
@@ -798,8 +795,8 @@ define <2 x i64> @vwsub_vx_v2i64_i8(ptr %x, ptr %y) nounwind {
define <2 x i64> @vwsub_vx_v2i64_i16(ptr %x, ptr %y) nounwind {
; RV32-LABEL: vwsub_vx_v2i64_i16:
; RV32: # %bb.0:
-; RV32-NEXT: vsetivli zero, 2, e64, m1, ta, ma
; RV32-NEXT: lh a1, 0(a1)
+; RV32-NEXT: vsetivli zero, 2, e64, m1, ta, ma
; RV32-NEXT: vle32.v v9, (a0)
; RV32-NEXT: vmv.v.x v8, a1
; RV32-NEXT: vsetvli zero, zero, e32, mf2, ta, ma
@@ -808,8 +805,8 @@ define <2 x i64> @vwsub_vx_v2i64_i16(ptr %x, ptr %y) nounwind {
;
; RV64-LABEL: vwsub_vx_v2i64_i16:
; RV64: # %bb.0:
-; RV64-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
; RV64-NEXT: lh a1, 0(a1)
+; RV64-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
; RV64-NEXT: vle32.v v9, (a0)
; RV64-NEXT: vmv.v.x v10, a1
; RV64-NEXT: vwsub.vv v8, v10, v9
@@ -827,8 +824,8 @@ define <2 x i64> @vwsub_vx_v2i64_i16(ptr %x, ptr %y) nounwind {
define <2 x i64> @vwsub_vx_v2i64_i32(ptr %x, ptr %y) nounwind {
; RV32-LABEL: vwsub_vx_v2i64_i32:
; RV32: # %bb.0:
-; RV32-NEXT: vsetivli zero, 2, e64, m1, ta, ma
; RV32-NEXT: lw a1, 0(a1)
+; RV32-NEXT: vsetivli zero, 2, e64, m1, ta, ma
; RV32-NEXT: vle32.v v9, (a0)
; RV32-NEXT: vmv.v.x v8, a1
; RV32-NEXT: vsetvli zero, zero, e32, mf2, ta, ma
@@ -856,9 +853,9 @@ define <2 x i64> @vwsub_vx_v2i64_i64(ptr %x, ptr %y) nounwind {
; RV32-LABEL: vwsub_vx_v2i64_i64:
; RV32: # %bb.0:
; RV32-NEXT: addi sp, sp, -16
-; RV32-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
; RV32-NEXT: lw a2, 4(a1)
; RV32-NEXT: lw a1, 0(a1)
+; RV32-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
; RV32-NEXT: vle32.v v9, (a0)
; RV32-NEXT: sw a2, 12(sp)
; RV32-NEXT: sw a1, 8(sp)
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vwsubu.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vwsubu.ll
index a084b5383b403..1a9e3aac00341 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vwsubu.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vwsubu.ll
@@ -263,13 +263,12 @@ define <128 x i16> @vwsubu_v128i16(ptr %x, ptr %y) nounwind {
; CHECK-NEXT: vslidedown.vx v16, v8, a0
; CHECK-NEXT: vslidedown.vx v8, v0, a0
; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, ma
-; CHECK-NEXT: vmv4r.v v24, v8
-; CHECK-NEXT: vwsubu.vv v8, v16, v24
+; CHECK-NEXT: vwsubu.vv v24, v16, v8
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vwsubu.vv v8, v16, v0
@@ -309,13 +308,12 @@ define <64 x i32> @vwsubu_v64i32(ptr %x, ptr %y) nounwind {
; CHECK-NEXT: vslidedown.vx v16, v8, a0
; CHECK-NEXT: vslidedown.vx v8, v0, a0
; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, ma
-; CHECK-NEXT: vmv4r.v v24, v8
-; CHECK-NEXT: vwsubu.vv v8, v16, v24
+; CHECK-NEXT: vwsubu.vv v24, v16, v8
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vwsubu.vv v8, v16, v0
@@ -354,13 +352,12 @@ define <32 x i64> @vwsubu_v32i64(ptr %x, ptr %y) nounwind {
; CHECK-NEXT: vslidedown.vi v16, v8, 16
; CHECK-NEXT: vslidedown.vi v8, v0, 16
; CHECK-NEXT: vsetivli zero, 16, e32, m4, ta, ma
-; CHECK-NEXT: vmv4r.v v24, v8
-; CHECK-NEXT: vwsubu.vv v8, v16, v24
+; CHECK-NEXT: vwsubu.vv v24, v16, v8
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vwsubu.vv v8, v16, v0
@@ -715,8 +712,8 @@ define <8 x i16> @vwsubu_vx_v8i16_i16(ptr %x, ptr %y) {
define <4 x i32> @vwsubu_vx_v4i32_i8(ptr %x, ptr %y) {
; CHECK-LABEL: vwsubu_vx_v4i32_i8:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetivli zero, 4, e16, mf2, ta, ma
; CHECK-NEXT: lbu a1, 0(a1)
+; CHECK-NEXT: vsetivli zero, 4, e16, mf2, ta, ma
; CHECK-NEXT: vle16.v v9, (a0)
; CHECK-NEXT: vmv.v.x v10, a1
; CHECK-NEXT: vwsubu.vv v8, v10, v9
@@ -770,8 +767,8 @@ define <2 x i64> @vwsubu_vx_v2i64_i8(ptr %x, ptr %y) nounwind {
; RV32-LABEL: vwsubu_vx_v2i64_i8:
; RV32: # %bb.0:
; RV32-NEXT: addi sp, sp, -16
-; RV32-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
; RV32-NEXT: lbu a1, 0(a1)
+; RV32-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
; RV32-NEXT: vle32.v v9, (a0)
; RV32-NEXT: sw zero, 12(sp)
; RV32-NEXT: sw a1, 8(sp)
@@ -783,8 +780,8 @@ define <2 x i64> @vwsubu_vx_v2i64_i8(ptr %x, ptr %y) nounwind {
;
; RV64-LABEL: vwsubu_vx_v2i64_i8:
; RV64: # %bb.0:
-; RV64-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
; RV64-NEXT: lbu a1, 0(a1)
+; RV64-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
; RV64-NEXT: vle32.v v9, (a0)
; RV64-NEXT: vmv.v.x v10, a1
; RV64-NEXT: vwsubu.vv v8, v10, v9
@@ -803,8 +800,8 @@ define <2 x i64> @vwsubu_vx_v2i64_i16(ptr %x, ptr %y) nounwind {
; RV32-LABEL: vwsubu_vx_v2i64_i16:
; RV32: # %bb.0:
; RV32-NEXT: addi sp, sp, -16
-; RV32-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
; RV32-NEXT: lhu a1, 0(a1)
+; RV32-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
; RV32-NEXT: vle32.v v9, (a0)
; RV32-NEXT: sw zero, 12(sp)
; RV32-NEXT: sw a1, 8(sp)
@@ -816,8 +813,8 @@ define <2 x i64> @vwsubu_vx_v2i64_i16(ptr %x, ptr %y) nounwind {
;
; RV64-LABEL: vwsubu_vx_v2i64_i16:
; RV64: # %bb.0:
-; RV64-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
; RV64-NEXT: lhu a1, 0(a1)
+; RV64-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
; RV64-NEXT: vle32.v v9, (a0)
; RV64-NEXT: vmv.v.x v10, a1
; RV64-NEXT: vwsubu.vv v8, v10, v9
@@ -836,8 +833,8 @@ define <2 x i64> @vwsubu_vx_v2i64_i32(ptr %x, ptr %y) nounwind {
; RV32-LABEL: vwsubu_vx_v2i64_i32:
; RV32: # %bb.0:
; RV32-NEXT: addi sp, sp, -16
-; RV32-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
; RV32-NEXT: lw a1, 0(a1)
+; RV32-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
; RV32-NEXT: vle32.v v9, (a0)
; RV32-NEXT: sw zero, 12(sp)
; RV32-NEXT: sw a1, 8(sp)
@@ -868,9 +865,9 @@ define <2 x i64> @vwsubu_vx_v2i64_i64(ptr %x, ptr %y) nounwind {
; RV32-LABEL: vwsubu_vx_v2i64_i64:
; RV32: # %bb.0:
; RV32-NEXT: addi sp, sp, -16
-; RV32-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
; RV32-NEXT: lw a2, 4(a1)
; RV32-NEXT: lw a1, 0(a1)
+; RV32-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
; RV32-NEXT: vle32.v v9, (a0)
; RV32-NEXT: sw a2, 12(sp)
; RV32-NEXT: sw a1, 8(sp)
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-zext-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-zext-vp.ll
index f4d679cd57cac..df90dae379c06 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-zext-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-zext-vp.ll
@@ -151,8 +151,8 @@ declare <32 x i64> @llvm.vp.zext.v32i64.v32i32(<32 x i32>, <32 x i1>, i32)
define <32 x i64> @vzext_v32i64_v32i32(<32 x i32> %va, <32 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vzext_v32i64_v32i32:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; CHECK-NEXT: li a2, 16
+; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; CHECK-NEXT: vslidedown.vi v16, v0, 2
; CHECK-NEXT: mv a1, a0
; CHECK-NEXT: bltu a0, a2, .LBB12_2
@@ -167,8 +167,8 @@ define <32 x i64> @vzext_v32i64_v32i32(<32 x i32> %va, <32 x i1> %m, i32 zeroext
; CHECK-NEXT: and a0, a0, a1
; CHECK-NEXT: vsetivli zero, 16, e32, m8, ta, ma
; CHECK-NEXT: vslidedown.vi v8, v8, 16
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vzext.vf2 v16, v8, v0.t
; CHECK-NEXT: vmv8r.v v8, v24
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/floor-vp.ll b/llvm/test/CodeGen/RISCV/rvv/floor-vp.ll
index d464b491bbbe2..26a3e053bf7aa 100644
--- a/llvm/test/CodeGen/RISCV/rvv/floor-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/floor-vp.ll
@@ -135,16 +135,16 @@ declare <vscale x 8 x half> @llvm.vp.floor.nxv8f16(<vscale x 8 x half>, <vscale
define <vscale x 8 x half> @vp_floor_nxv8f16(<vscale x 8 x half> %va, <vscale x 8 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_floor_nxv8f16:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v10, v0
; CHECK-NEXT: lui a1, %hi(.LCPI6_0)
; CHECK-NEXT: flh fa5, %lo(.LCPI6_0)(a1)
+; CHECK-NEXT: vmv1r.v v10, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, ma
; CHECK-NEXT: vfabs.v v12, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, mu
; CHECK-NEXT: vmflt.vf v10, v12, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 2
-; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
@@ -179,16 +179,16 @@ declare <vscale x 16 x half> @llvm.vp.floor.nxv16f16(<vscale x 16 x half>, <vsca
define <vscale x 16 x half> @vp_floor_nxv16f16(<vscale x 16 x half> %va, <vscale x 16 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_floor_nxv16f16:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v12, v0
; CHECK-NEXT: lui a1, %hi(.LCPI8_0)
; CHECK-NEXT: flh fa5, %lo(.LCPI8_0)(a1)
+; CHECK-NEXT: vmv1r.v v12, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, ma
; CHECK-NEXT: vfabs.v v16, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e16, m4, ta, mu
; CHECK-NEXT: vmflt.vf v12, v16, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 2
-; CHECK-NEXT: vsetvli zero, zero, e16, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, zero, e16, m4, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
@@ -223,16 +223,16 @@ declare <vscale x 32 x half> @llvm.vp.floor.nxv32f16(<vscale x 32 x half>, <vsca
define <vscale x 32 x half> @vp_floor_nxv32f16(<vscale x 32 x half> %va, <vscale x 32 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_floor_nxv32f16:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: lui a1, %hi(.LCPI10_0)
; CHECK-NEXT: flh fa5, %lo(.LCPI10_0)(a1)
+; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m8, ta, ma
; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e16, m8, ta, mu
; CHECK-NEXT: vmflt.vf v16, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 2
-; CHECK-NEXT: vsetvli zero, zero, e16, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vsetvli zero, zero, e16, m8, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
@@ -359,8 +359,8 @@ define <vscale x 4 x float> @vp_floor_nxv4f32(<vscale x 4 x float> %va, <vscale
; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, mu
; CHECK-NEXT: vmflt.vf v10, v12, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 2
-; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
@@ -403,8 +403,8 @@ define <vscale x 8 x float> @vp_floor_nxv8f32(<vscale x 8 x float> %va, <vscale
; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, mu
; CHECK-NEXT: vmflt.vf v12, v16, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 2
-; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
@@ -447,8 +447,8 @@ define <vscale x 16 x float> @vp_floor_nxv16f32(<vscale x 16 x float> %va, <vsca
; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, mu
; CHECK-NEXT: vmflt.vf v16, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 2
-; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
@@ -525,16 +525,16 @@ declare <vscale x 2 x double> @llvm.vp.floor.nxv2f64(<vscale x 2 x double>, <vsc
define <vscale x 2 x double> @vp_floor_nxv2f64(<vscale x 2 x double> %va, <vscale x 2 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_floor_nxv2f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v10, v0
; CHECK-NEXT: lui a1, %hi(.LCPI24_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI24_0)(a1)
+; CHECK-NEXT: vmv1r.v v10, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, ma
; CHECK-NEXT: vfabs.v v12, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, mu
; CHECK-NEXT: vmflt.vf v10, v12, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 2
-; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
@@ -569,16 +569,16 @@ declare <vscale x 4 x double> @llvm.vp.floor.nxv4f64(<vscale x 4 x double>, <vsc
define <vscale x 4 x double> @vp_floor_nxv4f64(<vscale x 4 x double> %va, <vscale x 4 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_floor_nxv4f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v12, v0
; CHECK-NEXT: lui a1, %hi(.LCPI26_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI26_0)(a1)
+; CHECK-NEXT: vmv1r.v v12, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, ma
; CHECK-NEXT: vfabs.v v16, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, mu
; CHECK-NEXT: vmflt.vf v12, v16, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 2
-; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
@@ -613,16 +613,16 @@ declare <vscale x 7 x double> @llvm.vp.floor.nxv7f64(<vscale x 7 x double>, <vsc
define <vscale x 7 x double> @vp_floor_nxv7f64(<vscale x 7 x double> %va, <vscale x 7 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_floor_nxv7f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: lui a1, %hi(.LCPI28_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI28_0)(a1)
+; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
; CHECK-NEXT: vmflt.vf v16, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 2
-; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
@@ -657,16 +657,16 @@ declare <vscale x 8 x double> @llvm.vp.floor.nxv8f64(<vscale x 8 x double>, <vsc
define <vscale x 8 x double> @vp_floor_nxv8f64(<vscale x 8 x double> %va, <vscale x 8 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_floor_nxv8f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: lui a1, %hi(.LCPI30_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI30_0)(a1)
+; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
; CHECK-NEXT: vmflt.vf v16, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 2
-; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
@@ -705,66 +705,56 @@ define <vscale x 16 x double> @vp_floor_nxv16f64(<vscale x 16 x double> %va, <vs
; CHECK-NEXT: addi sp, sp, -16
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a1, a1, 4
+; CHECK-NEXT: slli a1, a1, 3
; CHECK-NEXT: sub sp, sp, a1
-; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
-; CHECK-NEXT: vmv1r.v v24, v0
-; CHECK-NEXT: addi a1, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
+; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; CHECK-NEXT: vmv1r.v v7, v0
; CHECK-NEXT: csrr a1, vlenb
; CHECK-NEXT: srli a2, a1, 3
; CHECK-NEXT: vsetvli a3, zero, e8, mf4, ta, ma
-; CHECK-NEXT: vslidedown.vx v25, v0, a2
+; CHECK-NEXT: vslidedown.vx v6, v0, a2
; CHECK-NEXT: sub a2, a0, a1
+; CHECK-NEXT: lui a3, %hi(.LCPI32_0)
+; CHECK-NEXT: fld fa5, %lo(.LCPI32_0)(a3)
; CHECK-NEXT: sltu a3, a0, a2
; CHECK-NEXT: addi a3, a3, -1
; CHECK-NEXT: and a2, a3, a2
-; CHECK-NEXT: lui a3, %hi(.LCPI32_0)
-; CHECK-NEXT: fld fa5, %lo(.LCPI32_0)(a3)
+; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: vsetvli zero, a2, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v25
-; CHECK-NEXT: vfabs.v v8, v16, v0.t
+; CHECK-NEXT: vfabs.v v24, v16, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vmflt.vf v25, v8, fa5, v0.t
+; CHECK-NEXT: vmflt.vf v6, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a2, 2
+; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v25
-; CHECK-NEXT: vfcvt.x.f.v v8, v16, v0.t
+; CHECK-NEXT: vfcvt.x.f.v v24, v16, v0.t
+; CHECK-NEXT: addi a3, sp, 16
+; CHECK-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
; CHECK-NEXT: fsrm a2
-; CHECK-NEXT: vfcvt.f.x.v v8, v8, v0.t
+; CHECK-NEXT: addi a2, sp, 16
+; CHECK-NEXT: vl8r.v v24, (a2) # Unknown-size Folded Reload
+; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vfsgnj.vv v16, v8, v16, v0.t
-; CHECK-NEXT: csrr a2, vlenb
-; CHECK-NEXT: slli a2, a2, 3
-; CHECK-NEXT: add a2, sp, a2
-; CHECK-NEXT: addi a2, a2, 16
-; CHECK-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
+; CHECK-NEXT: vfsgnj.vv v16, v24, v16, v0.t
; CHECK-NEXT: bltu a0, a1, .LBB32_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB32_2:
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v24
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vfabs.v v16, v8, v0.t
+; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vmflt.vf v24, v16, fa5, v0.t
+; CHECK-NEXT: vmflt.vf v7, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 2
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v24
-; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
+; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: fsrm a0
-; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
+; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vfsgnj.vv v8, v16, v8, v0.t
+; CHECK-NEXT: vfsgnj.vv v8, v24, v8, v0.t
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 4
; CHECK-NEXT: add sp, sp, a0
; CHECK-NEXT: addi sp, sp, 16
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/fmaximum-sdnode.ll b/llvm/test/CodeGen/RISCV/rvv/fmaximum-sdnode.ll
index 386f23f68c357..05896d8ef6ffd 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fmaximum-sdnode.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fmaximum-sdnode.ll
@@ -214,10 +214,7 @@ define <vscale x 32 x half> @vfmax_nxv32f16_vv(<vscale x 32 x half> %a, <vscale
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 4
; ZVFHMIN-NEXT: sub sp, sp, a0
-; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 3
-; ZVFHMIN-NEXT: add a0, sp, a0
-; ZVFHMIN-NEXT: addi a0, a0, 16
+; ZVFHMIN-NEXT: addi a0, sp, 16
; ZVFHMIN-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vmv8r.v v0, v8
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
@@ -229,38 +226,31 @@ define <vscale x 32 x half> @vfmax_nxv32f16_vv(<vscale x 32 x half> %a, <vscale
; ZVFHMIN-NEXT: vmerge.vvm v16, v8, v24, v0
; ZVFHMIN-NEXT: vmv1r.v v0, v3
; ZVFHMIN-NEXT: vmerge.vvm v8, v24, v8, v0
-; ZVFHMIN-NEXT: vfmax.vv v24, v8, v16
-; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v24
-; ZVFHMIN-NEXT: addi a0, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vfmax.vv v8, v8, v16
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
+; ZVFHMIN-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: addi a0, sp, 16
; ZVFHMIN-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v20
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v4
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmfeq.vv v0, v16, v16
; ZVFHMIN-NEXT: vmfeq.vv v7, v8, v8
; ZVFHMIN-NEXT: vmerge.vvm v24, v16, v8, v0
-; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 3
-; ZVFHMIN-NEXT: add a0, sp, a0
-; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vmv1r.v v0, v7
-; ZVFHMIN-NEXT: vmerge.vvm v16, v8, v16, v0
+; ZVFHMIN-NEXT: vmerge.vvm v8, v8, v16, v0
+; ZVFHMIN-NEXT: vfmax.vv v16, v8, v24
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfmax.vv v16, v16, v24
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: addi a0, sp, 16
-; ZVFHMIN-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v24
; ZVFHMIN-NEXT: vfncvt.f.f.w v12, v16
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 4
diff --git a/llvm/test/CodeGen/RISCV/rvv/fmaximum-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fmaximum-vp.ll
index b78b8663eac90..ab07fff59b218 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fmaximum-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fmaximum-vp.ll
@@ -177,8 +177,8 @@ define <vscale x 4 x half> @vfmax_vv_nxv4f16(<vscale x 4 x half> %va, <vscale x
; ZVFHMIN-NEXT: vmfeq.vv v8, v12, v12, v0.t
; ZVFHMIN-NEXT: vsetvli a1, zero, e16, m1, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v14, v9
-; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m2, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v8
+; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m2, ta, ma
; ZVFHMIN-NEXT: vmerge.vvm v16, v12, v14, v0
; ZVFHMIN-NEXT: vmv1r.v v0, v10
; ZVFHMIN-NEXT: vmfeq.vv v8, v14, v14, v0.t
@@ -253,8 +253,8 @@ define <vscale x 8 x half> @vfmax_vv_nxv8f16(<vscale x 8 x half> %va, <vscale x
; ZVFHMIN-NEXT: vmfeq.vv v8, v16, v16, v0.t
; ZVFHMIN-NEXT: vsetvli a1, zero, e16, m2, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v20, v10
-; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m4, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v8
+; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m4, ta, ma
; ZVFHMIN-NEXT: vmerge.vvm v24, v16, v20, v0
; ZVFHMIN-NEXT: vmv1r.v v0, v12
; ZVFHMIN-NEXT: vmfeq.vv v8, v20, v20, v0.t
@@ -335,8 +335,8 @@ define <vscale x 16 x half> @vfmax_vv_nxv16f16(<vscale x 16 x half> %va, <vscale
; ZVFHMIN-NEXT: vmfeq.vv v8, v24, v24, v0.t
; ZVFHMIN-NEXT: vsetvli a1, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
-; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v8
+; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmerge.vvm v8, v24, v16, v0
; ZVFHMIN-NEXT: addi a0, sp, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
@@ -444,13 +444,8 @@ define <vscale x 32 x half> @vfmax_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
; ZVFHMIN-NEXT: mul a1, a1, a2
; ZVFHMIN-NEXT: sub sp, sp, a1
; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x22, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 34 * vlenb
-; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: li a2, 25
-; ZVFHMIN-NEXT: mul a1, a1, a2
-; ZVFHMIN-NEXT: add a1, sp, a1
-; ZVFHMIN-NEXT: addi a1, a1, 16
-; ZVFHMIN-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vmv8r.v v16, v8
+; ZVFHMIN-NEXT: vmv1r.v v24, v0
+; ZVFHMIN-NEXT: vmv8r.v v0, v8
; ZVFHMIN-NEXT: csrr a2, vlenb
; ZVFHMIN-NEXT: slli a1, a2, 1
; ZVFHMIN-NEXT: sub a3, a0, a1
@@ -458,35 +453,36 @@ define <vscale x 32 x half> @vfmax_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
; ZVFHMIN-NEXT: addi a4, a4, -1
; ZVFHMIN-NEXT: and a3, a4, a3
; ZVFHMIN-NEXT: srli a2, a2, 2
-; ZVFHMIN-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
; ZVFHMIN-NEXT: csrr a4, vlenb
-; ZVFHMIN-NEXT: li a5, 24
-; ZVFHMIN-NEXT: mul a4, a4, a5
+; ZVFHMIN-NEXT: slli a4, a4, 5
; ZVFHMIN-NEXT: add a4, sp, a4
; ZVFHMIN-NEXT: addi a4, a4, 16
-; ZVFHMIN-NEXT: vs1r.v v0, (a4) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vslidedown.vx v0, v0, a2
+; ZVFHMIN-NEXT: vs1r.v v24, (a4) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
+; ZVFHMIN-NEXT: vslidedown.vx v8, v24, a2
; ZVFHMIN-NEXT: addi a2, sp, 16
-; ZVFHMIN-NEXT: vs1r.v v0, (a2) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
+; ZVFHMIN-NEXT: vs1r.v v8, (a2) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a2, vlenb
; ZVFHMIN-NEXT: slli a2, a2, 4
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
-; ZVFHMIN-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20
+; ZVFHMIN-NEXT: vs8r.v v0, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v4
+; ZVFHMIN-NEXT: vmv1r.v v0, v8
; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmfeq.vv v12, v24, v24, v0.t
-; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
+; ZVFHMIN-NEXT: vmv8r.v v0, v16
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: li a4, 25
+; ZVFHMIN-NEXT: li a4, 24
; ZVFHMIN-NEXT: mul a2, a2, a4
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
-; ZVFHMIN-NEXT: vl8r.v v0, (a2) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v4
-; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v12
+; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmerge.vvm v8, v24, v16, v0
; ZVFHMIN-NEXT: csrr a2, vlenb
; ZVFHMIN-NEXT: slli a2, a2, 3
@@ -523,45 +519,34 @@ define <vscale x 32 x half> @vfmax_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
; ZVFHMIN-NEXT: addi a1, a1, 16
; ZVFHMIN-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v16
-; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: li a2, 24
-; ZVFHMIN-NEXT: mul a1, a1, a2
+; ZVFHMIN-NEXT: slli a1, a1, 5
; ZVFHMIN-NEXT: add a1, sp, a1
; ZVFHMIN-NEXT: addi a1, a1, 16
; ZVFHMIN-NEXT: vl1r.v v0, (a1) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmfeq.vv v8, v24, v24, v0.t
-; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a1, a1, 4
-; ZVFHMIN-NEXT: add a1, sp, a1
-; ZVFHMIN-NEXT: addi a1, a1, 16
-; ZVFHMIN-NEXT: vs1r.v v8, (a1) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: addi a1, sp, 16
; ZVFHMIN-NEXT: vs8r.v v24, (a1) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vsetvli a1, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: li a2, 25
+; ZVFHMIN-NEXT: li a2, 24
; ZVFHMIN-NEXT: mul a1, a1, a2
; ZVFHMIN-NEXT: add a1, sp, a1
; ZVFHMIN-NEXT: addi a1, a1, 16
; ZVFHMIN-NEXT: vl8r.v v0, (a1) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vsetvli a1, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v0
+; ZVFHMIN-NEXT: vmv1r.v v0, v8
; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
-; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 4
-; ZVFHMIN-NEXT: add a0, sp, a0
-; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vl1r.v v0, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vmerge.vvm v24, v24, v16, v0
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: li a1, 25
+; ZVFHMIN-NEXT: li a1, 24
; ZVFHMIN-NEXT: mul a0, a0, a1
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: li a1, 24
-; ZVFHMIN-NEXT: mul a0, a0, a1
+; ZVFHMIN-NEXT: slli a0, a0, 5
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vl1r.v v9, (a0) # Unknown-size Folded Reload
@@ -573,18 +558,18 @@ define <vscale x 32 x half> @vfmax_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
; ZVFHMIN-NEXT: vmerge.vvm v16, v16, v24, v0
; ZVFHMIN-NEXT: vmv1r.v v0, v9
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: li a1, 25
+; ZVFHMIN-NEXT: li a1, 24
; ZVFHMIN-NEXT: mul a0, a0, a1
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfmax.vv v16, v16, v24, v0.t
-; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v16
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: li a1, 34
@@ -627,19 +612,18 @@ define <vscale x 32 x half> @vfmax_vv_nxv32f16_unmasked(<vscale x 32 x half> %va
; ZVFHMIN-NEXT: vmset.m v24
; ZVFHMIN-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
; ZVFHMIN-NEXT: vslidedown.vx v7, v24, a2
-; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: csrr a2, vlenb
; ZVFHMIN-NEXT: slli a2, a2, 4
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vs8r.v v24, (a2) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v7
+; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmfeq.vv v12, v24, v24, v0.t
-; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vmv4r.v v8, v16
; ZVFHMIN-NEXT: csrr a2, vlenb
; ZVFHMIN-NEXT: li a4, 24
@@ -647,11 +631,12 @@ define <vscale x 32 x half> @vfmax_vv_nxv32f16_unmasked(<vscale x 32 x half> %va
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20
-; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v12
; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vl8r.v v16, (a2) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmerge.vvm v8, v16, v24, v0
; ZVFHMIN-NEXT: csrr a2, vlenb
; ZVFHMIN-NEXT: slli a2, a2, 3
@@ -683,13 +668,13 @@ define <vscale x 32 x half> @vfmax_vv_nxv32f16_unmasked(<vscale x 32 x half> %va
; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v16
; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmfeq.vv v0, v8, v8
-; ZVFHMIN-NEXT: vsetvli a1, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: csrr a1, vlenb
; ZVFHMIN-NEXT: li a2, 24
; ZVFHMIN-NEXT: mul a1, a1, a2
; ZVFHMIN-NEXT: add a1, sp, a1
; ZVFHMIN-NEXT: addi a1, a1, 16
; ZVFHMIN-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vsetvli a1, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v24
; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmfeq.vv v3, v16, v16
@@ -1012,77 +997,99 @@ define <vscale x 16 x double> @vfmax_vv_nxv16f64(<vscale x 16 x double> %va, <vs
; CHECK-NEXT: addi sp, sp, -16
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: li a3, 42
+; CHECK-NEXT: li a3, 36
; CHECK-NEXT: mul a1, a1, a3
; CHECK-NEXT: sub sp, sp, a1
-; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x2a, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 42 * vlenb
+; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x24, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 36 * vlenb
; CHECK-NEXT: csrr a1, vlenb
; CHECK-NEXT: slli a3, a1, 3
; CHECK-NEXT: add a3, a0, a3
; CHECK-NEXT: vl8re64.v v24, (a3)
; CHECK-NEXT: csrr a3, vlenb
-; CHECK-NEXT: slli a3, a3, 4
+; CHECK-NEXT: li a4, 27
+; CHECK-NEXT: mul a3, a3, a4
; CHECK-NEXT: add a3, sp, a3
; CHECK-NEXT: addi a3, a3, 16
; CHECK-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
; CHECK-NEXT: srli a3, a1, 3
-; CHECK-NEXT: vsetvli a4, zero, e8, mf4, ta, ma
; CHECK-NEXT: csrr a4, vlenb
-; CHECK-NEXT: slli a4, a4, 5
+; CHECK-NEXT: li a5, 18
+; CHECK-NEXT: mul a4, a4, a5
; CHECK-NEXT: add a4, sp, a4
; CHECK-NEXT: addi a4, a4, 16
; CHECK-NEXT: vs1r.v v0, (a4) # Unknown-size Folded Spill
-; CHECK-NEXT: vslidedown.vx v24, v0, a3
+; CHECK-NEXT: vsetvli a4, zero, e8, mf4, ta, ma
+; CHECK-NEXT: vslidedown.vx v7, v0, a3
+; CHECK-NEXT: csrr a3, vlenb
+; CHECK-NEXT: add a3, sp, a3
+; CHECK-NEXT: addi a3, a3, 16
+; CHECK-NEXT: vs1r.v v7, (a3) # Unknown-size Folded Spill
; CHECK-NEXT: sub a3, a2, a1
; CHECK-NEXT: sltu a4, a2, a3
; CHECK-NEXT: addi a4, a4, -1
; CHECK-NEXT: and a3, a4, a3
-; CHECK-NEXT: vl8re64.v v0, (a0)
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a4, a0, 5
-; CHECK-NEXT: add a0, a4, a0
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v0, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vmv1r.v v0, v7
+; CHECK-NEXT: csrr a4, vlenb
+; CHECK-NEXT: slli a4, a4, 1
+; CHECK-NEXT: add a4, sp, a4
+; CHECK-NEXT: addi a4, a4, 16
+; CHECK-NEXT: vs8r.v v16, (a4) # Unknown-size Folded Spill
; CHECK-NEXT: vsetvli zero, a3, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v7, v24
-; CHECK-NEXT: vmv1r.v v0, v24
; CHECK-NEXT: vmfeq.vv v26, v16, v16, v0.t
; CHECK-NEXT: vmv1r.v v0, v26
+; CHECK-NEXT: csrr a3, vlenb
+; CHECK-NEXT: li a4, 27
+; CHECK-NEXT: mul a3, a3, a4
+; CHECK-NEXT: add a3, sp, a3
+; CHECK-NEXT: addi a3, a3, 16
+; CHECK-NEXT: vl8r.v v24, (a3) # Unknown-size Folded Reload
+; CHECK-NEXT: vmerge.vvm v24, v16, v24, v0
+; CHECK-NEXT: csrr a3, vlenb
+; CHECK-NEXT: li a4, 10
+; CHECK-NEXT: mul a3, a3, a4
+; CHECK-NEXT: add a3, sp, a3
+; CHECK-NEXT: addi a3, a3, 16
+; CHECK-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
+; CHECK-NEXT: vmv1r.v v0, v7
+; CHECK-NEXT: csrr a3, vlenb
+; CHECK-NEXT: li a4, 27
+; CHECK-NEXT: mul a3, a3, a4
+; CHECK-NEXT: add a3, sp, a3
+; CHECK-NEXT: addi a3, a3, 16
+; CHECK-NEXT: vl8r.v v24, (a3) # Unknown-size Folded Reload
+; CHECK-NEXT: vmfeq.vv v16, v24, v24, v0.t
+; CHECK-NEXT: addi a3, sp, 16
+; CHECK-NEXT: vs1r.v v16, (a3) # Unknown-size Folded Spill
+; CHECK-NEXT: vl8re64.v v0, (a0)
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 4
+; CHECK-NEXT: li a3, 19
+; CHECK-NEXT: mul a0, a0, a3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vs8r.v v0, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vmv8r.v v16, v8
; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: vmerge.vvm v16, v16, v24, v0
+; CHECK-NEXT: vl1r.v v0, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
+; CHECK-NEXT: slli a0, a0, 1
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: vmv1r.v v0, v7
-; CHECK-NEXT: vmfeq.vv v17, v24, v24, v0.t
+; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vmerge.vvm v8, v24, v8, v0
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a3, 24
-; CHECK-NEXT: mul a0, a0, a3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: vmv1r.v v0, v17
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vmerge.vvm v8, v24, v8, v0
-; CHECK-NEXT: vmv1r.v v0, v7
+; CHECK-NEXT: vl1r.v v0, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
+; CHECK-NEXT: li a3, 10
+; CHECK-NEXT: mul a0, a0, a3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vfmax.vv v8, v8, v16, v0.t
+; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vfmax.vv v8, v8, v24, v0.t
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 4
+; CHECK-NEXT: li a3, 10
+; CHECK-NEXT: mul a0, a0, a3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
@@ -1090,51 +1097,65 @@ define <vscale x 16 x double> @vfmax_vv_nxv16f64(<vscale x 16 x double> %va, <vs
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a2, a1
; CHECK-NEXT: .LBB28_2:
+; CHECK-NEXT: csrr a0, vlenb
+; CHECK-NEXT: li a1, 18
+; CHECK-NEXT: mul a0, a0, a1
+; CHECK-NEXT: add a0, sp, a0
+; CHECK-NEXT: addi a0, a0, 16
+; CHECK-NEXT: vl1r.v v24, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vmv1r.v v0, v24
; CHECK-NEXT: vsetvli zero, a2, e64, m8, ta, ma
+; CHECK-NEXT: vmfeq.vv v25, v16, v16, v0.t
+; CHECK-NEXT: vmv1r.v v0, v25
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 5
+; CHECK-NEXT: li a1, 19
+; CHECK-NEXT: mul a0, a0, a1
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl1r.v v7, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vmv1r.v v0, v7
+; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a1, 24
+; CHECK-NEXT: li a1, 27
; CHECK-NEXT: mul a0, a0, a1
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vmfeq.vv v25, v16, v16, v0.t
-; CHECK-NEXT: vmv1r.v v0, v25
+; CHECK-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a1, a0, 5
-; CHECK-NEXT: add a0, a1, a0
+; CHECK-NEXT: li a1, 27
+; CHECK-NEXT: mul a0, a0, a1
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vmerge.vvm v24, v16, v8, v0
+; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vmerge.vvm v16, v16, v8, v0
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
+; CHECK-NEXT: slli a0, a0, 1
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: vmv1r.v v0, v7
+; CHECK-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vmv1r.v v0, v24
; CHECK-NEXT: vmfeq.vv v25, v8, v8, v0.t
; CHECK-NEXT: vmv1r.v v0, v25
+; CHECK-NEXT: csrr a0, vlenb
+; CHECK-NEXT: li a1, 27
+; CHECK-NEXT: mul a0, a0, a1
+; CHECK-NEXT: add a0, sp, a0
+; CHECK-NEXT: addi a0, a0, 16
+; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vmerge.vvm v8, v8, v16, v0
-; CHECK-NEXT: vmv1r.v v0, v7
+; CHECK-NEXT: vmv1r.v v0, v24
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
+; CHECK-NEXT: slli a0, a0, 1
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vfmax.vv v8, v8, v16, v0.t
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 4
+; CHECK-NEXT: li a1, 10
+; CHECK-NEXT: mul a0, a0, a1
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a1, 42
+; CHECK-NEXT: li a1, 36
; CHECK-NEXT: mul a0, a0, a1
; CHECK-NEXT: add sp, sp, a0
; CHECK-NEXT: addi sp, sp, 16
@@ -1162,19 +1183,19 @@ define <vscale x 16 x double> @vfmax_vv_nxv16f64_unmasked(<vscale x 16 x double>
; CHECK-NEXT: slli a3, a1, 3
; CHECK-NEXT: add a3, a0, a3
; CHECK-NEXT: vl8re64.v v24, (a3)
+; CHECK-NEXT: sub a3, a2, a1
+; CHECK-NEXT: sltu a4, a2, a3
+; CHECK-NEXT: addi a4, a4, -1
+; CHECK-NEXT: and a3, a4, a3
+; CHECK-NEXT: vsetvli zero, a3, e64, m8, ta, ma
+; CHECK-NEXT: vmfeq.vv v0, v16, v16
+; CHECK-NEXT: vmfeq.vv v7, v24, v24
; CHECK-NEXT: vl8re64.v v8, (a0)
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: sub a0, a2, a1
-; CHECK-NEXT: sltu a3, a2, a0
-; CHECK-NEXT: addi a3, a3, -1
-; CHECK-NEXT: and a0, a3, a0
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: vmfeq.vv v0, v16, v16
-; CHECK-NEXT: vmfeq.vv v7, v24, v24
; CHECK-NEXT: vmerge.vvm v8, v16, v24, v0
; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vmerge.vvm v16, v24, v16, v0
@@ -1185,12 +1206,12 @@ define <vscale x 16 x double> @vfmax_vv_nxv16f64_unmasked(<vscale x 16 x double>
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a2, a1
; CHECK-NEXT: .LBB29_2:
-; CHECK-NEXT: vsetvli zero, a2, e64, m8, ta, ma
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 4
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, a2, e64, m8, ta, ma
; CHECK-NEXT: vmfeq.vv v0, v16, v16
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
diff --git a/llvm/test/CodeGen/RISCV/rvv/fminimum-sdnode.ll b/llvm/test/CodeGen/RISCV/rvv/fminimum-sdnode.ll
index 48baa12aa2e59..e942593924987 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fminimum-sdnode.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fminimum-sdnode.ll
@@ -214,10 +214,7 @@ define <vscale x 32 x half> @vfmin_nxv32f16_vv(<vscale x 32 x half> %a, <vscale
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 4
; ZVFHMIN-NEXT: sub sp, sp, a0
-; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 3
-; ZVFHMIN-NEXT: add a0, sp, a0
-; ZVFHMIN-NEXT: addi a0, a0, 16
+; ZVFHMIN-NEXT: addi a0, sp, 16
; ZVFHMIN-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vmv8r.v v0, v8
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
@@ -229,38 +226,31 @@ define <vscale x 32 x half> @vfmin_nxv32f16_vv(<vscale x 32 x half> %a, <vscale
; ZVFHMIN-NEXT: vmerge.vvm v16, v8, v24, v0
; ZVFHMIN-NEXT: vmv1r.v v0, v3
; ZVFHMIN-NEXT: vmerge.vvm v8, v24, v8, v0
-; ZVFHMIN-NEXT: vfmin.vv v24, v8, v16
-; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v24
-; ZVFHMIN-NEXT: addi a0, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vfmin.vv v8, v8, v16
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
+; ZVFHMIN-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: addi a0, sp, 16
; ZVFHMIN-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v20
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v4
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmfeq.vv v0, v16, v16
; ZVFHMIN-NEXT: vmfeq.vv v7, v8, v8
; ZVFHMIN-NEXT: vmerge.vvm v24, v16, v8, v0
-; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 3
-; ZVFHMIN-NEXT: add a0, sp, a0
-; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vmv1r.v v0, v7
-; ZVFHMIN-NEXT: vmerge.vvm v16, v8, v16, v0
+; ZVFHMIN-NEXT: vmerge.vvm v8, v8, v16, v0
+; ZVFHMIN-NEXT: vfmin.vv v16, v8, v24
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfmin.vv v16, v16, v24
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: addi a0, sp, 16
-; ZVFHMIN-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v24
; ZVFHMIN-NEXT: vfncvt.f.f.w v12, v16
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 4
diff --git a/llvm/test/CodeGen/RISCV/rvv/fminimum-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fminimum-vp.ll
index 69c76152910e8..fc5b11284dab0 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fminimum-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fminimum-vp.ll
@@ -177,8 +177,8 @@ define <vscale x 4 x half> @vfmin_vv_nxv4f16(<vscale x 4 x half> %va, <vscale x
; ZVFHMIN-NEXT: vmfeq.vv v8, v12, v12, v0.t
; ZVFHMIN-NEXT: vsetvli a1, zero, e16, m1, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v14, v9
-; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m2, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v8
+; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m2, ta, ma
; ZVFHMIN-NEXT: vmerge.vvm v16, v12, v14, v0
; ZVFHMIN-NEXT: vmv1r.v v0, v10
; ZVFHMIN-NEXT: vmfeq.vv v8, v14, v14, v0.t
@@ -253,8 +253,8 @@ define <vscale x 8 x half> @vfmin_vv_nxv8f16(<vscale x 8 x half> %va, <vscale x
; ZVFHMIN-NEXT: vmfeq.vv v8, v16, v16, v0.t
; ZVFHMIN-NEXT: vsetvli a1, zero, e16, m2, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v20, v10
-; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m4, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v8
+; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m4, ta, ma
; ZVFHMIN-NEXT: vmerge.vvm v24, v16, v20, v0
; ZVFHMIN-NEXT: vmv1r.v v0, v12
; ZVFHMIN-NEXT: vmfeq.vv v8, v20, v20, v0.t
@@ -335,8 +335,8 @@ define <vscale x 16 x half> @vfmin_vv_nxv16f16(<vscale x 16 x half> %va, <vscale
; ZVFHMIN-NEXT: vmfeq.vv v8, v24, v24, v0.t
; ZVFHMIN-NEXT: vsetvli a1, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
-; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v8
+; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmerge.vvm v8, v24, v16, v0
; ZVFHMIN-NEXT: addi a0, sp, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
@@ -444,13 +444,8 @@ define <vscale x 32 x half> @vfmin_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
; ZVFHMIN-NEXT: mul a1, a1, a2
; ZVFHMIN-NEXT: sub sp, sp, a1
; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x22, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 34 * vlenb
-; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: li a2, 25
-; ZVFHMIN-NEXT: mul a1, a1, a2
-; ZVFHMIN-NEXT: add a1, sp, a1
-; ZVFHMIN-NEXT: addi a1, a1, 16
-; ZVFHMIN-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vmv8r.v v16, v8
+; ZVFHMIN-NEXT: vmv1r.v v24, v0
+; ZVFHMIN-NEXT: vmv8r.v v0, v8
; ZVFHMIN-NEXT: csrr a2, vlenb
; ZVFHMIN-NEXT: slli a1, a2, 1
; ZVFHMIN-NEXT: sub a3, a0, a1
@@ -458,35 +453,36 @@ define <vscale x 32 x half> @vfmin_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
; ZVFHMIN-NEXT: addi a4, a4, -1
; ZVFHMIN-NEXT: and a3, a4, a3
; ZVFHMIN-NEXT: srli a2, a2, 2
-; ZVFHMIN-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
; ZVFHMIN-NEXT: csrr a4, vlenb
-; ZVFHMIN-NEXT: li a5, 24
-; ZVFHMIN-NEXT: mul a4, a4, a5
+; ZVFHMIN-NEXT: slli a4, a4, 5
; ZVFHMIN-NEXT: add a4, sp, a4
; ZVFHMIN-NEXT: addi a4, a4, 16
-; ZVFHMIN-NEXT: vs1r.v v0, (a4) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vslidedown.vx v0, v0, a2
+; ZVFHMIN-NEXT: vs1r.v v24, (a4) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
+; ZVFHMIN-NEXT: vslidedown.vx v8, v24, a2
; ZVFHMIN-NEXT: addi a2, sp, 16
-; ZVFHMIN-NEXT: vs1r.v v0, (a2) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
+; ZVFHMIN-NEXT: vs1r.v v8, (a2) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a2, vlenb
; ZVFHMIN-NEXT: slli a2, a2, 4
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
-; ZVFHMIN-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20
+; ZVFHMIN-NEXT: vs8r.v v0, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v4
+; ZVFHMIN-NEXT: vmv1r.v v0, v8
; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmfeq.vv v12, v24, v24, v0.t
-; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
+; ZVFHMIN-NEXT: vmv8r.v v0, v16
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: li a4, 25
+; ZVFHMIN-NEXT: li a4, 24
; ZVFHMIN-NEXT: mul a2, a2, a4
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
-; ZVFHMIN-NEXT: vl8r.v v0, (a2) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v4
-; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v12
+; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmerge.vvm v8, v24, v16, v0
; ZVFHMIN-NEXT: csrr a2, vlenb
; ZVFHMIN-NEXT: slli a2, a2, 3
@@ -523,45 +519,34 @@ define <vscale x 32 x half> @vfmin_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
; ZVFHMIN-NEXT: addi a1, a1, 16
; ZVFHMIN-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v16
-; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: li a2, 24
-; ZVFHMIN-NEXT: mul a1, a1, a2
+; ZVFHMIN-NEXT: slli a1, a1, 5
; ZVFHMIN-NEXT: add a1, sp, a1
; ZVFHMIN-NEXT: addi a1, a1, 16
; ZVFHMIN-NEXT: vl1r.v v0, (a1) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmfeq.vv v8, v24, v24, v0.t
-; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a1, a1, 4
-; ZVFHMIN-NEXT: add a1, sp, a1
-; ZVFHMIN-NEXT: addi a1, a1, 16
-; ZVFHMIN-NEXT: vs1r.v v8, (a1) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: addi a1, sp, 16
; ZVFHMIN-NEXT: vs8r.v v24, (a1) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vsetvli a1, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: li a2, 25
+; ZVFHMIN-NEXT: li a2, 24
; ZVFHMIN-NEXT: mul a1, a1, a2
; ZVFHMIN-NEXT: add a1, sp, a1
; ZVFHMIN-NEXT: addi a1, a1, 16
; ZVFHMIN-NEXT: vl8r.v v0, (a1) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vsetvli a1, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v0
+; ZVFHMIN-NEXT: vmv1r.v v0, v8
; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
-; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 4
-; ZVFHMIN-NEXT: add a0, sp, a0
-; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vl1r.v v0, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vmerge.vvm v24, v24, v16, v0
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: li a1, 25
+; ZVFHMIN-NEXT: li a1, 24
; ZVFHMIN-NEXT: mul a0, a0, a1
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: li a1, 24
-; ZVFHMIN-NEXT: mul a0, a0, a1
+; ZVFHMIN-NEXT: slli a0, a0, 5
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vl1r.v v9, (a0) # Unknown-size Folded Reload
@@ -573,18 +558,18 @@ define <vscale x 32 x half> @vfmin_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
; ZVFHMIN-NEXT: vmerge.vvm v16, v16, v24, v0
; ZVFHMIN-NEXT: vmv1r.v v0, v9
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: li a1, 25
+; ZVFHMIN-NEXT: li a1, 24
; ZVFHMIN-NEXT: mul a0, a0, a1
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfmin.vv v16, v16, v24, v0.t
-; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v16
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: li a1, 34
@@ -627,19 +612,18 @@ define <vscale x 32 x half> @vfmin_vv_nxv32f16_unmasked(<vscale x 32 x half> %va
; ZVFHMIN-NEXT: vmset.m v24
; ZVFHMIN-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
; ZVFHMIN-NEXT: vslidedown.vx v7, v24, a2
-; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: csrr a2, vlenb
; ZVFHMIN-NEXT: slli a2, a2, 4
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vs8r.v v24, (a2) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v7
+; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmfeq.vv v12, v24, v24, v0.t
-; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vmv4r.v v8, v16
; ZVFHMIN-NEXT: csrr a2, vlenb
; ZVFHMIN-NEXT: li a4, 24
@@ -647,11 +631,12 @@ define <vscale x 32 x half> @vfmin_vv_nxv32f16_unmasked(<vscale x 32 x half> %va
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20
-; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v12
; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vl8r.v v16, (a2) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmerge.vvm v8, v16, v24, v0
; ZVFHMIN-NEXT: csrr a2, vlenb
; ZVFHMIN-NEXT: slli a2, a2, 3
@@ -683,13 +668,13 @@ define <vscale x 32 x half> @vfmin_vv_nxv32f16_unmasked(<vscale x 32 x half> %va
; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v16
; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmfeq.vv v0, v8, v8
-; ZVFHMIN-NEXT: vsetvli a1, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: csrr a1, vlenb
; ZVFHMIN-NEXT: li a2, 24
; ZVFHMIN-NEXT: mul a1, a1, a2
; ZVFHMIN-NEXT: add a1, sp, a1
; ZVFHMIN-NEXT: addi a1, a1, 16
; ZVFHMIN-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vsetvli a1, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v24
; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmfeq.vv v3, v16, v16
@@ -1012,77 +997,99 @@ define <vscale x 16 x double> @vfmin_vv_nxv16f64(<vscale x 16 x double> %va, <vs
; CHECK-NEXT: addi sp, sp, -16
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: li a3, 42
+; CHECK-NEXT: li a3, 36
; CHECK-NEXT: mul a1, a1, a3
; CHECK-NEXT: sub sp, sp, a1
-; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x2a, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 42 * vlenb
+; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x24, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 36 * vlenb
; CHECK-NEXT: csrr a1, vlenb
; CHECK-NEXT: slli a3, a1, 3
; CHECK-NEXT: add a3, a0, a3
; CHECK-NEXT: vl8re64.v v24, (a3)
; CHECK-NEXT: csrr a3, vlenb
-; CHECK-NEXT: slli a3, a3, 4
+; CHECK-NEXT: li a4, 27
+; CHECK-NEXT: mul a3, a3, a4
; CHECK-NEXT: add a3, sp, a3
; CHECK-NEXT: addi a3, a3, 16
; CHECK-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
; CHECK-NEXT: srli a3, a1, 3
-; CHECK-NEXT: vsetvli a4, zero, e8, mf4, ta, ma
; CHECK-NEXT: csrr a4, vlenb
-; CHECK-NEXT: slli a4, a4, 5
+; CHECK-NEXT: li a5, 18
+; CHECK-NEXT: mul a4, a4, a5
; CHECK-NEXT: add a4, sp, a4
; CHECK-NEXT: addi a4, a4, 16
; CHECK-NEXT: vs1r.v v0, (a4) # Unknown-size Folded Spill
-; CHECK-NEXT: vslidedown.vx v24, v0, a3
+; CHECK-NEXT: vsetvli a4, zero, e8, mf4, ta, ma
+; CHECK-NEXT: vslidedown.vx v7, v0, a3
+; CHECK-NEXT: csrr a3, vlenb
+; CHECK-NEXT: add a3, sp, a3
+; CHECK-NEXT: addi a3, a3, 16
+; CHECK-NEXT: vs1r.v v7, (a3) # Unknown-size Folded Spill
; CHECK-NEXT: sub a3, a2, a1
; CHECK-NEXT: sltu a4, a2, a3
; CHECK-NEXT: addi a4, a4, -1
; CHECK-NEXT: and a3, a4, a3
-; CHECK-NEXT: vl8re64.v v0, (a0)
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a4, a0, 5
-; CHECK-NEXT: add a0, a4, a0
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v0, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vmv1r.v v0, v7
+; CHECK-NEXT: csrr a4, vlenb
+; CHECK-NEXT: slli a4, a4, 1
+; CHECK-NEXT: add a4, sp, a4
+; CHECK-NEXT: addi a4, a4, 16
+; CHECK-NEXT: vs8r.v v16, (a4) # Unknown-size Folded Spill
; CHECK-NEXT: vsetvli zero, a3, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v7, v24
-; CHECK-NEXT: vmv1r.v v0, v24
; CHECK-NEXT: vmfeq.vv v26, v16, v16, v0.t
; CHECK-NEXT: vmv1r.v v0, v26
+; CHECK-NEXT: csrr a3, vlenb
+; CHECK-NEXT: li a4, 27
+; CHECK-NEXT: mul a3, a3, a4
+; CHECK-NEXT: add a3, sp, a3
+; CHECK-NEXT: addi a3, a3, 16
+; CHECK-NEXT: vl8r.v v24, (a3) # Unknown-size Folded Reload
+; CHECK-NEXT: vmerge.vvm v24, v16, v24, v0
+; CHECK-NEXT: csrr a3, vlenb
+; CHECK-NEXT: li a4, 10
+; CHECK-NEXT: mul a3, a3, a4
+; CHECK-NEXT: add a3, sp, a3
+; CHECK-NEXT: addi a3, a3, 16
+; CHECK-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
+; CHECK-NEXT: vmv1r.v v0, v7
+; CHECK-NEXT: csrr a3, vlenb
+; CHECK-NEXT: li a4, 27
+; CHECK-NEXT: mul a3, a3, a4
+; CHECK-NEXT: add a3, sp, a3
+; CHECK-NEXT: addi a3, a3, 16
+; CHECK-NEXT: vl8r.v v24, (a3) # Unknown-size Folded Reload
+; CHECK-NEXT: vmfeq.vv v16, v24, v24, v0.t
+; CHECK-NEXT: addi a3, sp, 16
+; CHECK-NEXT: vs1r.v v16, (a3) # Unknown-size Folded Spill
+; CHECK-NEXT: vl8re64.v v0, (a0)
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 4
+; CHECK-NEXT: li a3, 19
+; CHECK-NEXT: mul a0, a0, a3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vs8r.v v0, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vmv8r.v v16, v8
; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: vmerge.vvm v16, v16, v24, v0
+; CHECK-NEXT: vl1r.v v0, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
+; CHECK-NEXT: slli a0, a0, 1
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: vmv1r.v v0, v7
-; CHECK-NEXT: vmfeq.vv v17, v24, v24, v0.t
+; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vmerge.vvm v8, v24, v8, v0
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a3, 24
-; CHECK-NEXT: mul a0, a0, a3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: vmv1r.v v0, v17
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vmerge.vvm v8, v24, v8, v0
-; CHECK-NEXT: vmv1r.v v0, v7
+; CHECK-NEXT: vl1r.v v0, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
+; CHECK-NEXT: li a3, 10
+; CHECK-NEXT: mul a0, a0, a3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vfmin.vv v8, v8, v16, v0.t
+; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vfmin.vv v8, v8, v24, v0.t
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 4
+; CHECK-NEXT: li a3, 10
+; CHECK-NEXT: mul a0, a0, a3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
@@ -1090,51 +1097,65 @@ define <vscale x 16 x double> @vfmin_vv_nxv16f64(<vscale x 16 x double> %va, <vs
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a2, a1
; CHECK-NEXT: .LBB28_2:
+; CHECK-NEXT: csrr a0, vlenb
+; CHECK-NEXT: li a1, 18
+; CHECK-NEXT: mul a0, a0, a1
+; CHECK-NEXT: add a0, sp, a0
+; CHECK-NEXT: addi a0, a0, 16
+; CHECK-NEXT: vl1r.v v24, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vmv1r.v v0, v24
; CHECK-NEXT: vsetvli zero, a2, e64, m8, ta, ma
+; CHECK-NEXT: vmfeq.vv v25, v16, v16, v0.t
+; CHECK-NEXT: vmv1r.v v0, v25
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 5
+; CHECK-NEXT: li a1, 19
+; CHECK-NEXT: mul a0, a0, a1
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl1r.v v7, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vmv1r.v v0, v7
+; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a1, 24
+; CHECK-NEXT: li a1, 27
; CHECK-NEXT: mul a0, a0, a1
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vmfeq.vv v25, v16, v16, v0.t
-; CHECK-NEXT: vmv1r.v v0, v25
+; CHECK-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a1, a0, 5
-; CHECK-NEXT: add a0, a1, a0
+; CHECK-NEXT: li a1, 27
+; CHECK-NEXT: mul a0, a0, a1
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vmerge.vvm v24, v16, v8, v0
+; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vmerge.vvm v16, v16, v8, v0
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
+; CHECK-NEXT: slli a0, a0, 1
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: vmv1r.v v0, v7
+; CHECK-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vmv1r.v v0, v24
; CHECK-NEXT: vmfeq.vv v25, v8, v8, v0.t
; CHECK-NEXT: vmv1r.v v0, v25
+; CHECK-NEXT: csrr a0, vlenb
+; CHECK-NEXT: li a1, 27
+; CHECK-NEXT: mul a0, a0, a1
+; CHECK-NEXT: add a0, sp, a0
+; CHECK-NEXT: addi a0, a0, 16
+; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vmerge.vvm v8, v8, v16, v0
-; CHECK-NEXT: vmv1r.v v0, v7
+; CHECK-NEXT: vmv1r.v v0, v24
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
+; CHECK-NEXT: slli a0, a0, 1
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vfmin.vv v8, v8, v16, v0.t
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 4
+; CHECK-NEXT: li a1, 10
+; CHECK-NEXT: mul a0, a0, a1
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a1, 42
+; CHECK-NEXT: li a1, 36
; CHECK-NEXT: mul a0, a0, a1
; CHECK-NEXT: add sp, sp, a0
; CHECK-NEXT: addi sp, sp, 16
@@ -1162,19 +1183,19 @@ define <vscale x 16 x double> @vfmin_vv_nxv16f64_unmasked(<vscale x 16 x double>
; CHECK-NEXT: slli a3, a1, 3
; CHECK-NEXT: add a3, a0, a3
; CHECK-NEXT: vl8re64.v v24, (a3)
+; CHECK-NEXT: sub a3, a2, a1
+; CHECK-NEXT: sltu a4, a2, a3
+; CHECK-NEXT: addi a4, a4, -1
+; CHECK-NEXT: and a3, a4, a3
+; CHECK-NEXT: vsetvli zero, a3, e64, m8, ta, ma
+; CHECK-NEXT: vmfeq.vv v0, v16, v16
+; CHECK-NEXT: vmfeq.vv v7, v24, v24
; CHECK-NEXT: vl8re64.v v8, (a0)
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: sub a0, a2, a1
-; CHECK-NEXT: sltu a3, a2, a0
-; CHECK-NEXT: addi a3, a3, -1
-; CHECK-NEXT: and a0, a3, a0
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: vmfeq.vv v0, v16, v16
-; CHECK-NEXT: vmfeq.vv v7, v24, v24
; CHECK-NEXT: vmerge.vvm v8, v16, v24, v0
; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vmerge.vvm v16, v24, v16, v0
@@ -1185,12 +1206,12 @@ define <vscale x 16 x double> @vfmin_vv_nxv16f64_unmasked(<vscale x 16 x double>
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a2, a1
; CHECK-NEXT: .LBB29_2:
-; CHECK-NEXT: vsetvli zero, a2, e64, m8, ta, ma
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 4
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, a2, e64, m8, ta, ma
; CHECK-NEXT: vmfeq.vv v0, v16, v16
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
diff --git a/llvm/test/CodeGen/RISCV/rvv/fnearbyint-constrained-sdnode.ll b/llvm/test/CodeGen/RISCV/rvv/fnearbyint-constrained-sdnode.ll
index f90237b8d7e95..f88a9b3081a1a 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fnearbyint-constrained-sdnode.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fnearbyint-constrained-sdnode.ll
@@ -19,9 +19,9 @@ define <vscale x 1 x half> @nearbyint_nxv1f16(<vscale x 1 x half> %v) strictfp {
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v9, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v9, v9, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e16, mf4, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%r = call <vscale x 1 x half> @llvm.experimental.constrained.nearbyint.nxv1f16(<vscale x 1 x half> %v, metadata !"round.dynamic", metadata !"fpexcept.strict")
ret <vscale x 1 x half> %r
@@ -42,9 +42,9 @@ define <vscale x 2 x half> @nearbyint_nxv2f16(<vscale x 2 x half> %v) strictfp {
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v9, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v9, v9, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e16, mf2, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%r = call <vscale x 2 x half> @llvm.experimental.constrained.nearbyint.nxv2f16(<vscale x 2 x half> %v, metadata !"round.dynamic", metadata !"fpexcept.strict")
ret <vscale x 2 x half> %r
@@ -65,9 +65,9 @@ define <vscale x 4 x half> @nearbyint_nxv4f16(<vscale x 4 x half> %v) strictfp {
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v9, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v9, v9, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e16, m1, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%r = call <vscale x 4 x half> @llvm.experimental.constrained.nearbyint.nxv4f16(<vscale x 4 x half> %v, metadata !"round.dynamic", metadata !"fpexcept.strict")
ret <vscale x 4 x half> %r
@@ -88,9 +88,9 @@ define <vscale x 8 x half> @nearbyint_nxv8f16(<vscale x 8 x half> %v) strictfp {
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v10, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v10, v10, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v10, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%r = call <vscale x 8 x half> @llvm.experimental.constrained.nearbyint.nxv8f16(<vscale x 8 x half> %v, metadata !"round.dynamic", metadata !"fpexcept.strict")
ret <vscale x 8 x half> %r
@@ -111,9 +111,9 @@ define <vscale x 16 x half> @nearbyint_nxv16f16(<vscale x 16 x half> %v) strictf
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e16, m4, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v12, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%r = call <vscale x 16 x half> @llvm.experimental.constrained.nearbyint.nxv16f16(<vscale x 16 x half> %v, metadata !"round.dynamic", metadata !"fpexcept.strict")
ret <vscale x 16 x half> %r
@@ -134,9 +134,9 @@ define <vscale x 32 x half> @nearbyint_nxv32f16(<vscale x 32 x half> %v) strictf
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e16, m8, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v16, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%r = call <vscale x 32 x half> @llvm.experimental.constrained.nearbyint.nxv32f16(<vscale x 32 x half> %v, metadata !"round.dynamic", metadata !"fpexcept.strict")
ret <vscale x 32 x half> %r
@@ -157,9 +157,9 @@ define <vscale x 1 x float> @nearbyint_nxv1f32(<vscale x 1 x float> %v) strictfp
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v9, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v9, v9, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e32, mf2, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%r = call <vscale x 1 x float> @llvm.experimental.constrained.nearbyint.nxv1f32(<vscale x 1 x float> %v, metadata !"round.dynamic", metadata !"fpexcept.strict")
ret <vscale x 1 x float> %r
@@ -180,9 +180,9 @@ define <vscale x 2 x float> @nearbyint_nxv2f32(<vscale x 2 x float> %v) strictfp
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v9, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v9, v9, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e32, m1, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%r = call <vscale x 2 x float> @llvm.experimental.constrained.nearbyint.nxv2f32(<vscale x 2 x float> %v, metadata !"round.dynamic", metadata !"fpexcept.strict")
ret <vscale x 2 x float> %r
@@ -203,9 +203,9 @@ define <vscale x 4 x float> @nearbyint_nxv4f32(<vscale x 4 x float> %v) strictfp
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v10, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v10, v10, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v10, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%r = call <vscale x 4 x float> @llvm.experimental.constrained.nearbyint.nxv4f32(<vscale x 4 x float> %v, metadata !"round.dynamic", metadata !"fpexcept.strict")
ret <vscale x 4 x float> %r
@@ -226,9 +226,9 @@ define <vscale x 8 x float> @nearbyint_nxv8f32(<vscale x 8 x float> %v) strictfp
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v12, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%r = call <vscale x 8 x float> @llvm.experimental.constrained.nearbyint.nxv8f32(<vscale x 8 x float> %v, metadata !"round.dynamic", metadata !"fpexcept.strict")
ret <vscale x 8 x float> %r
@@ -249,9 +249,9 @@ define <vscale x 16 x float> @nearbyint_nxv16f32(<vscale x 16 x float> %v) stric
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v16, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%r = call <vscale x 16 x float> @llvm.experimental.constrained.nearbyint.nxv16f32(<vscale x 16 x float> %v, metadata !"round.dynamic", metadata !"fpexcept.strict")
ret <vscale x 16 x float> %r
@@ -272,9 +272,9 @@ define <vscale x 1 x double> @nearbyint_nxv1f64(<vscale x 1 x double> %v) strict
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v9, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v9, v9, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e64, m1, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%r = call <vscale x 1 x double> @llvm.experimental.constrained.nearbyint.nxv1f64(<vscale x 1 x double> %v, metadata !"round.dynamic", metadata !"fpexcept.strict")
ret <vscale x 1 x double> %r
@@ -295,9 +295,9 @@ define <vscale x 2 x double> @nearbyint_nxv2f64(<vscale x 2 x double> %v) strict
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v10, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v10, v10, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v10, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%r = call <vscale x 2 x double> @llvm.experimental.constrained.nearbyint.nxv2f64(<vscale x 2 x double> %v, metadata !"round.dynamic", metadata !"fpexcept.strict")
ret <vscale x 2 x double> %r
@@ -318,9 +318,9 @@ define <vscale x 4 x double> @nearbyint_nxv4f64(<vscale x 4 x double> %v) strict
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v12, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%r = call <vscale x 4 x double> @llvm.experimental.constrained.nearbyint.nxv4f64(<vscale x 4 x double> %v, metadata !"round.dynamic", metadata !"fpexcept.strict")
ret <vscale x 4 x double> %r
@@ -341,9 +341,9 @@ define <vscale x 8 x double> @nearbyint_nxv8f64(<vscale x 8 x double> %v) strict
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v16, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%r = call <vscale x 8 x double> @llvm.experimental.constrained.nearbyint.nxv8f64(<vscale x 8 x double> %v, metadata !"round.dynamic", metadata !"fpexcept.strict")
ret <vscale x 8 x double> %r
diff --git a/llvm/test/CodeGen/RISCV/rvv/fnearbyint-sdnode.ll b/llvm/test/CodeGen/RISCV/rvv/fnearbyint-sdnode.ll
index 9aa356b9b65e0..9e14852305caa 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fnearbyint-sdnode.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fnearbyint-sdnode.ll
@@ -15,9 +15,9 @@ define <vscale x 1 x half> @nearbyint_nxv1f16(<vscale x 1 x half> %x) {
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v9, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v9, v9, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e16, mf4, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%a = call <vscale x 1 x half> @llvm.nearbyint.nxv1f16(<vscale x 1 x half> %x)
ret <vscale x 1 x half> %a
@@ -35,9 +35,9 @@ define <vscale x 2 x half> @nearbyint_nxv2f16(<vscale x 2 x half> %x) {
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v9, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v9, v9, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e16, mf2, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%a = call <vscale x 2 x half> @llvm.nearbyint.nxv2f16(<vscale x 2 x half> %x)
ret <vscale x 2 x half> %a
@@ -55,9 +55,9 @@ define <vscale x 4 x half> @nearbyint_nxv4f16(<vscale x 4 x half> %x) {
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v9, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v9, v9, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e16, m1, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%a = call <vscale x 4 x half> @llvm.nearbyint.nxv4f16(<vscale x 4 x half> %x)
ret <vscale x 4 x half> %a
@@ -75,9 +75,9 @@ define <vscale x 8 x half> @nearbyint_nxv8f16(<vscale x 8 x half> %x) {
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v10, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v10, v10, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v10, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%a = call <vscale x 8 x half> @llvm.nearbyint.nxv8f16(<vscale x 8 x half> %x)
ret <vscale x 8 x half> %a
@@ -95,9 +95,9 @@ define <vscale x 16 x half> @nearbyint_nxv16f16(<vscale x 16 x half> %x) {
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e16, m4, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v12, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%a = call <vscale x 16 x half> @llvm.nearbyint.nxv16f16(<vscale x 16 x half> %x)
ret <vscale x 16 x half> %a
@@ -115,9 +115,9 @@ define <vscale x 32 x half> @nearbyint_nxv32f16(<vscale x 32 x half> %x) {
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e16, m8, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v16, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%a = call <vscale x 32 x half> @llvm.nearbyint.nxv32f16(<vscale x 32 x half> %x)
ret <vscale x 32 x half> %a
@@ -135,9 +135,9 @@ define <vscale x 1 x float> @nearbyint_nxv1f32(<vscale x 1 x float> %x) {
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v9, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v9, v9, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e32, mf2, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%a = call <vscale x 1 x float> @llvm.nearbyint.nxv1f32(<vscale x 1 x float> %x)
ret <vscale x 1 x float> %a
@@ -155,9 +155,9 @@ define <vscale x 2 x float> @nearbyint_nxv2f32(<vscale x 2 x float> %x) {
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v9, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v9, v9, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e32, m1, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%a = call <vscale x 2 x float> @llvm.nearbyint.nxv2f32(<vscale x 2 x float> %x)
ret <vscale x 2 x float> %a
@@ -175,9 +175,9 @@ define <vscale x 4 x float> @nearbyint_nxv4f32(<vscale x 4 x float> %x) {
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v10, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v10, v10, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v10, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%a = call <vscale x 4 x float> @llvm.nearbyint.nxv4f32(<vscale x 4 x float> %x)
ret <vscale x 4 x float> %a
@@ -195,9 +195,9 @@ define <vscale x 8 x float> @nearbyint_nxv8f32(<vscale x 8 x float> %x) {
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v12, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%a = call <vscale x 8 x float> @llvm.nearbyint.nxv8f32(<vscale x 8 x float> %x)
ret <vscale x 8 x float> %a
@@ -215,9 +215,9 @@ define <vscale x 16 x float> @nearbyint_nxv16f32(<vscale x 16 x float> %x) {
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v16, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%a = call <vscale x 16 x float> @llvm.nearbyint.nxv16f32(<vscale x 16 x float> %x)
ret <vscale x 16 x float> %a
@@ -235,9 +235,9 @@ define <vscale x 1 x double> @nearbyint_nxv1f64(<vscale x 1 x double> %x) {
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v9, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v9, v9, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e64, m1, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%a = call <vscale x 1 x double> @llvm.nearbyint.nxv1f64(<vscale x 1 x double> %x)
ret <vscale x 1 x double> %a
@@ -255,9 +255,9 @@ define <vscale x 2 x double> @nearbyint_nxv2f64(<vscale x 2 x double> %x) {
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v10, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v10, v10, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v10, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%a = call <vscale x 2 x double> @llvm.nearbyint.nxv2f64(<vscale x 2 x double> %x)
ret <vscale x 2 x double> %a
@@ -275,9 +275,9 @@ define <vscale x 4 x double> @nearbyint_nxv4f64(<vscale x 4 x double> %x) {
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v12, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%a = call <vscale x 4 x double> @llvm.nearbyint.nxv4f64(<vscale x 4 x double> %x)
ret <vscale x 4 x double> %a
@@ -295,9 +295,9 @@ define <vscale x 8 x double> @nearbyint_nxv8f64(<vscale x 8 x double> %x) {
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v16, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%a = call <vscale x 8 x double> @llvm.nearbyint.nxv8f64(<vscale x 8 x double> %x)
ret <vscale x 8 x double> %a
diff --git a/llvm/test/CodeGen/RISCV/rvv/fpclamptosat_vec.ll b/llvm/test/CodeGen/RISCV/rvv/fpclamptosat_vec.ll
index bb28ff5c6dc4f..aa845bd8bb0b7 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fpclamptosat_vec.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fpclamptosat_vec.ll
@@ -448,8 +448,8 @@ define <4 x i32> @stest_f16i32(<4 x half> %x) {
; CHECK-V-NEXT: fmv.w.x fa0, a0
; CHECK-V-NEXT: call __extendhfsf2
; CHECK-V-NEXT: fcvt.l.s a0, fa0, rtz
-; CHECK-V-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; CHECK-V-NEXT: fmv.w.x fa0, s2
+; CHECK-V-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; CHECK-V-NEXT: vmv.s.x v8, a0
; CHECK-V-NEXT: addi a0, sp, 16
; CHECK-V-NEXT: vs1r.v v8, (a0) # Unknown-size Folded Spill
@@ -467,8 +467,8 @@ define <4 x i32> @stest_f16i32(<4 x half> %x) {
; CHECK-V-NEXT: fmv.w.x fa0, s1
; CHECK-V-NEXT: call __extendhfsf2
; CHECK-V-NEXT: fcvt.l.s a0, fa0, rtz
-; CHECK-V-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; CHECK-V-NEXT: fmv.w.x fa0, s0
+; CHECK-V-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; CHECK-V-NEXT: vmv.s.x v8, a0
; CHECK-V-NEXT: addi a0, sp, 16
; CHECK-V-NEXT: vs1r.v v8, (a0) # Unknown-size Folded Spill
@@ -609,8 +609,8 @@ define <4 x i32> @utesth_f16i32(<4 x half> %x) {
; CHECK-V-NEXT: fmv.w.x fa0, a0
; CHECK-V-NEXT: call __extendhfsf2
; CHECK-V-NEXT: fcvt.lu.s a0, fa0, rtz
-; CHECK-V-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; CHECK-V-NEXT: fmv.w.x fa0, s2
+; CHECK-V-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; CHECK-V-NEXT: vmv.s.x v8, a0
; CHECK-V-NEXT: addi a0, sp, 16
; CHECK-V-NEXT: vs1r.v v8, (a0) # Unknown-size Folded Spill
@@ -628,8 +628,8 @@ define <4 x i32> @utesth_f16i32(<4 x half> %x) {
; CHECK-V-NEXT: fmv.w.x fa0, s1
; CHECK-V-NEXT: call __extendhfsf2
; CHECK-V-NEXT: fcvt.lu.s a0, fa0, rtz
-; CHECK-V-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; CHECK-V-NEXT: fmv.w.x fa0, s0
+; CHECK-V-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; CHECK-V-NEXT: vmv.s.x v8, a0
; CHECK-V-NEXT: addi a0, sp, 16
; CHECK-V-NEXT: vs1r.v v8, (a0) # Unknown-size Folded Spill
@@ -780,8 +780,8 @@ define <4 x i32> @ustest_f16i32(<4 x half> %x) {
; CHECK-V-NEXT: fmv.w.x fa0, a0
; CHECK-V-NEXT: call __extendhfsf2
; CHECK-V-NEXT: fcvt.l.s a0, fa0, rtz
-; CHECK-V-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; CHECK-V-NEXT: fmv.w.x fa0, s2
+; CHECK-V-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; CHECK-V-NEXT: vmv.s.x v8, a0
; CHECK-V-NEXT: addi a0, sp, 16
; CHECK-V-NEXT: vs1r.v v8, (a0) # Unknown-size Folded Spill
@@ -799,8 +799,8 @@ define <4 x i32> @ustest_f16i32(<4 x half> %x) {
; CHECK-V-NEXT: fmv.w.x fa0, s1
; CHECK-V-NEXT: call __extendhfsf2
; CHECK-V-NEXT: fcvt.l.s a0, fa0, rtz
-; CHECK-V-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; CHECK-V-NEXT: fmv.w.x fa0, s0
+; CHECK-V-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; CHECK-V-NEXT: vmv.s.x v8, a0
; CHECK-V-NEXT: addi a0, sp, 16
; CHECK-V-NEXT: vs1r.v v8, (a0) # Unknown-size Folded Spill
@@ -1397,8 +1397,8 @@ define <8 x i16> @stest_f16i16(<8 x half> %x) {
; CHECK-V-NEXT: fmv.w.x fa0, a0
; CHECK-V-NEXT: call __extendhfsf2
; CHECK-V-NEXT: fcvt.l.s a0, fa0, rtz
-; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: fmv.w.x fa0, s6
+; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: vmv.s.x v8, a0
; CHECK-V-NEXT: csrr a0, vlenb
; CHECK-V-NEXT: slli a0, a0, 1
@@ -1422,8 +1422,8 @@ define <8 x i16> @stest_f16i16(<8 x half> %x) {
; CHECK-V-NEXT: fmv.w.x fa0, s5
; CHECK-V-NEXT: call __extendhfsf2
; CHECK-V-NEXT: fcvt.l.s a0, fa0, rtz
-; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: fmv.w.x fa0, s4
+; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: vmv.s.x v8, a0
; CHECK-V-NEXT: csrr a0, vlenb
; CHECK-V-NEXT: slli a0, a0, 1
@@ -1440,11 +1440,11 @@ define <8 x i16> @stest_f16i16(<8 x half> %x) {
; CHECK-V-NEXT: addi a0, a0, 16
; CHECK-V-NEXT: vl1r.v v9, (a0) # Unknown-size Folded Reload
; CHECK-V-NEXT: vslideup.vi v8, v9, 1
-; CHECK-V-NEXT: vsetivli zero, 4, e32, m1, ta, ma
; CHECK-V-NEXT: csrr a0, vlenb
; CHECK-V-NEXT: add a0, sp, a0
; CHECK-V-NEXT: addi a0, a0, 16
; CHECK-V-NEXT: vl1r.v v9, (a0) # Unknown-size Folded Reload
+; CHECK-V-NEXT: vsetivli zero, 4, e32, m1, ta, ma
; CHECK-V-NEXT: vslideup.vi v8, v9, 2
; CHECK-V-NEXT: csrr a0, vlenb
; CHECK-V-NEXT: slli a0, a0, 1
@@ -1454,8 +1454,8 @@ define <8 x i16> @stest_f16i16(<8 x half> %x) {
; CHECK-V-NEXT: fmv.w.x fa0, s3
; CHECK-V-NEXT: call __extendhfsf2
; CHECK-V-NEXT: fcvt.l.s a0, fa0, rtz
-; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: fmv.w.x fa0, s2
+; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: vmv.s.x v8, a0
; CHECK-V-NEXT: addi a0, sp, 16
; CHECK-V-NEXT: vs1r.v v8, (a0) # Unknown-size Folded Spill
@@ -1473,8 +1473,8 @@ define <8 x i16> @stest_f16i16(<8 x half> %x) {
; CHECK-V-NEXT: fmv.w.x fa0, s1
; CHECK-V-NEXT: call __extendhfsf2
; CHECK-V-NEXT: fcvt.l.s a0, fa0, rtz
-; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: fmv.w.x fa0, s0
+; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: vmv.s.x v8, a0
; CHECK-V-NEXT: addi a0, sp, 16
; CHECK-V-NEXT: vs1r.v v8, (a0) # Unknown-size Folded Spill
@@ -1485,18 +1485,18 @@ define <8 x i16> @stest_f16i16(<8 x half> %x) {
; CHECK-V-NEXT: addi a0, sp, 16
; CHECK-V-NEXT: vl1r.v v8, (a0) # Unknown-size Folded Reload
; CHECK-V-NEXT: vslideup.vi v10, v8, 1
-; CHECK-V-NEXT: vsetivli zero, 4, e32, m1, ta, ma
; CHECK-V-NEXT: csrr a0, vlenb
; CHECK-V-NEXT: add a0, sp, a0
; CHECK-V-NEXT: addi a0, a0, 16
; CHECK-V-NEXT: vl1r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-V-NEXT: vsetivli zero, 4, e32, m1, ta, ma
; CHECK-V-NEXT: vslideup.vi v10, v8, 2
-; CHECK-V-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; CHECK-V-NEXT: csrr a0, vlenb
; CHECK-V-NEXT: slli a0, a0, 1
; CHECK-V-NEXT: add a0, sp, a0
; CHECK-V-NEXT: addi a0, a0, 16
; CHECK-V-NEXT: vl2r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-V-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; CHECK-V-NEXT: vslideup.vi v10, v8, 4
; CHECK-V-NEXT: vsetvli zero, zero, e16, m1, ta, ma
; CHECK-V-NEXT: vnclip.wi v8, v10, 0
@@ -1710,8 +1710,8 @@ define <8 x i16> @utesth_f16i16(<8 x half> %x) {
; CHECK-V-NEXT: fmv.w.x fa0, a0
; CHECK-V-NEXT: call __extendhfsf2
; CHECK-V-NEXT: fcvt.lu.s a0, fa0, rtz
-; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: fmv.w.x fa0, s6
+; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: vmv.s.x v8, a0
; CHECK-V-NEXT: csrr a0, vlenb
; CHECK-V-NEXT: slli a0, a0, 1
@@ -1735,8 +1735,8 @@ define <8 x i16> @utesth_f16i16(<8 x half> %x) {
; CHECK-V-NEXT: fmv.w.x fa0, s5
; CHECK-V-NEXT: call __extendhfsf2
; CHECK-V-NEXT: fcvt.lu.s a0, fa0, rtz
-; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: fmv.w.x fa0, s4
+; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: vmv.s.x v8, a0
; CHECK-V-NEXT: csrr a0, vlenb
; CHECK-V-NEXT: slli a0, a0, 1
@@ -1753,11 +1753,11 @@ define <8 x i16> @utesth_f16i16(<8 x half> %x) {
; CHECK-V-NEXT: addi a0, a0, 16
; CHECK-V-NEXT: vl1r.v v9, (a0) # Unknown-size Folded Reload
; CHECK-V-NEXT: vslideup.vi v8, v9, 1
-; CHECK-V-NEXT: vsetivli zero, 4, e32, m1, ta, ma
; CHECK-V-NEXT: csrr a0, vlenb
; CHECK-V-NEXT: add a0, sp, a0
; CHECK-V-NEXT: addi a0, a0, 16
; CHECK-V-NEXT: vl1r.v v9, (a0) # Unknown-size Folded Reload
+; CHECK-V-NEXT: vsetivli zero, 4, e32, m1, ta, ma
; CHECK-V-NEXT: vslideup.vi v8, v9, 2
; CHECK-V-NEXT: csrr a0, vlenb
; CHECK-V-NEXT: slli a0, a0, 1
@@ -1767,8 +1767,8 @@ define <8 x i16> @utesth_f16i16(<8 x half> %x) {
; CHECK-V-NEXT: fmv.w.x fa0, s3
; CHECK-V-NEXT: call __extendhfsf2
; CHECK-V-NEXT: fcvt.lu.s a0, fa0, rtz
-; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: fmv.w.x fa0, s2
+; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: vmv.s.x v8, a0
; CHECK-V-NEXT: addi a0, sp, 16
; CHECK-V-NEXT: vs1r.v v8, (a0) # Unknown-size Folded Spill
@@ -1786,8 +1786,8 @@ define <8 x i16> @utesth_f16i16(<8 x half> %x) {
; CHECK-V-NEXT: fmv.w.x fa0, s1
; CHECK-V-NEXT: call __extendhfsf2
; CHECK-V-NEXT: fcvt.lu.s a0, fa0, rtz
-; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: fmv.w.x fa0, s0
+; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: vmv.s.x v8, a0
; CHECK-V-NEXT: addi a0, sp, 16
; CHECK-V-NEXT: vs1r.v v8, (a0) # Unknown-size Folded Spill
@@ -1798,18 +1798,18 @@ define <8 x i16> @utesth_f16i16(<8 x half> %x) {
; CHECK-V-NEXT: addi a0, sp, 16
; CHECK-V-NEXT: vl1r.v v8, (a0) # Unknown-size Folded Reload
; CHECK-V-NEXT: vslideup.vi v10, v8, 1
-; CHECK-V-NEXT: vsetivli zero, 4, e32, m1, ta, ma
; CHECK-V-NEXT: csrr a0, vlenb
; CHECK-V-NEXT: add a0, sp, a0
; CHECK-V-NEXT: addi a0, a0, 16
; CHECK-V-NEXT: vl1r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-V-NEXT: vsetivli zero, 4, e32, m1, ta, ma
; CHECK-V-NEXT: vslideup.vi v10, v8, 2
-; CHECK-V-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; CHECK-V-NEXT: csrr a0, vlenb
; CHECK-V-NEXT: slli a0, a0, 1
; CHECK-V-NEXT: add a0, sp, a0
; CHECK-V-NEXT: addi a0, a0, 16
; CHECK-V-NEXT: vl2r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-V-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; CHECK-V-NEXT: vslideup.vi v10, v8, 4
; CHECK-V-NEXT: vsetvli zero, zero, e16, m1, ta, ma
; CHECK-V-NEXT: vnclipu.wi v8, v10, 0
@@ -2045,8 +2045,8 @@ define <8 x i16> @ustest_f16i16(<8 x half> %x) {
; CHECK-V-NEXT: fmv.w.x fa0, a0
; CHECK-V-NEXT: call __extendhfsf2
; CHECK-V-NEXT: fcvt.l.s a0, fa0, rtz
-; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: fmv.w.x fa0, s6
+; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: vmv.s.x v8, a0
; CHECK-V-NEXT: csrr a0, vlenb
; CHECK-V-NEXT: slli a0, a0, 1
@@ -2070,8 +2070,8 @@ define <8 x i16> @ustest_f16i16(<8 x half> %x) {
; CHECK-V-NEXT: fmv.w.x fa0, s5
; CHECK-V-NEXT: call __extendhfsf2
; CHECK-V-NEXT: fcvt.l.s a0, fa0, rtz
-; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: fmv.w.x fa0, s4
+; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: vmv.s.x v8, a0
; CHECK-V-NEXT: csrr a0, vlenb
; CHECK-V-NEXT: slli a0, a0, 1
@@ -2088,11 +2088,11 @@ define <8 x i16> @ustest_f16i16(<8 x half> %x) {
; CHECK-V-NEXT: addi a0, a0, 16
; CHECK-V-NEXT: vl1r.v v9, (a0) # Unknown-size Folded Reload
; CHECK-V-NEXT: vslideup.vi v8, v9, 1
-; CHECK-V-NEXT: vsetivli zero, 4, e32, m1, ta, ma
; CHECK-V-NEXT: csrr a0, vlenb
; CHECK-V-NEXT: add a0, sp, a0
; CHECK-V-NEXT: addi a0, a0, 16
; CHECK-V-NEXT: vl1r.v v9, (a0) # Unknown-size Folded Reload
+; CHECK-V-NEXT: vsetivli zero, 4, e32, m1, ta, ma
; CHECK-V-NEXT: vslideup.vi v8, v9, 2
; CHECK-V-NEXT: csrr a0, vlenb
; CHECK-V-NEXT: slli a0, a0, 1
@@ -2102,8 +2102,8 @@ define <8 x i16> @ustest_f16i16(<8 x half> %x) {
; CHECK-V-NEXT: fmv.w.x fa0, s3
; CHECK-V-NEXT: call __extendhfsf2
; CHECK-V-NEXT: fcvt.l.s a0, fa0, rtz
-; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: fmv.w.x fa0, s2
+; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: vmv.s.x v8, a0
; CHECK-V-NEXT: addi a0, sp, 16
; CHECK-V-NEXT: vs1r.v v8, (a0) # Unknown-size Folded Spill
@@ -2121,8 +2121,8 @@ define <8 x i16> @ustest_f16i16(<8 x half> %x) {
; CHECK-V-NEXT: fmv.w.x fa0, s1
; CHECK-V-NEXT: call __extendhfsf2
; CHECK-V-NEXT: fcvt.l.s a0, fa0, rtz
-; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: fmv.w.x fa0, s0
+; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: vmv.s.x v8, a0
; CHECK-V-NEXT: addi a0, sp, 16
; CHECK-V-NEXT: vs1r.v v8, (a0) # Unknown-size Folded Spill
@@ -2133,18 +2133,18 @@ define <8 x i16> @ustest_f16i16(<8 x half> %x) {
; CHECK-V-NEXT: addi a0, sp, 16
; CHECK-V-NEXT: vl1r.v v9, (a0) # Unknown-size Folded Reload
; CHECK-V-NEXT: vslideup.vi v8, v9, 1
-; CHECK-V-NEXT: vsetivli zero, 4, e32, m1, ta, ma
; CHECK-V-NEXT: csrr a0, vlenb
; CHECK-V-NEXT: add a0, sp, a0
; CHECK-V-NEXT: addi a0, a0, 16
; CHECK-V-NEXT: vl1r.v v9, (a0) # Unknown-size Folded Reload
+; CHECK-V-NEXT: vsetivli zero, 4, e32, m1, ta, ma
; CHECK-V-NEXT: vslideup.vi v8, v9, 2
-; CHECK-V-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; CHECK-V-NEXT: csrr a0, vlenb
; CHECK-V-NEXT: slli a0, a0, 1
; CHECK-V-NEXT: add a0, sp, a0
; CHECK-V-NEXT: addi a0, a0, 16
; CHECK-V-NEXT: vl2r.v v10, (a0) # Unknown-size Folded Reload
+; CHECK-V-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; CHECK-V-NEXT: vslideup.vi v8, v10, 4
; CHECK-V-NEXT: lui a0, 16
; CHECK-V-NEXT: addi a0, a0, -1
@@ -2279,9 +2279,9 @@ define <2 x i64> @stest_f64i64(<2 x double> %x) {
; CHECK-V-NEXT: call __fixdfti
; CHECK-V-NEXT: mv s0, a0
; CHECK-V-NEXT: mv s1, a1
-; CHECK-V-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; CHECK-V-NEXT: addi a0, sp, 32
; CHECK-V-NEXT: vl1r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-V-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; CHECK-V-NEXT: vfmv.f.s fa0, v8
; CHECK-V-NEXT: call __fixdfti
; CHECK-V-NEXT: li a2, -1
@@ -2412,9 +2412,9 @@ define <2 x i64> @utest_f64i64(<2 x double> %x) {
; CHECK-V-NEXT: call __fixunsdfti
; CHECK-V-NEXT: mv s0, a0
; CHECK-V-NEXT: mv s1, a1
-; CHECK-V-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; CHECK-V-NEXT: addi a0, sp, 32
; CHECK-V-NEXT: vl1r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-V-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; CHECK-V-NEXT: vfmv.f.s fa0, v8
; CHECK-V-NEXT: call __fixunsdfti
; CHECK-V-NEXT: snez a1, a1
@@ -2524,9 +2524,9 @@ define <2 x i64> @ustest_f64i64(<2 x double> %x) {
; CHECK-V-NEXT: call __fixdfti
; CHECK-V-NEXT: mv s0, a0
; CHECK-V-NEXT: mv s1, a1
-; CHECK-V-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; CHECK-V-NEXT: addi a0, sp, 32
; CHECK-V-NEXT: vl1r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-V-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; CHECK-V-NEXT: vfmv.f.s fa0, v8
; CHECK-V-NEXT: call __fixdfti
; CHECK-V-NEXT: mv a2, s1
@@ -2686,9 +2686,9 @@ define <2 x i64> @stest_f32i64(<2 x float> %x) {
; CHECK-V-NEXT: call __fixsfti
; CHECK-V-NEXT: mv s0, a0
; CHECK-V-NEXT: mv s1, a1
-; CHECK-V-NEXT: vsetivli zero, 1, e32, mf2, ta, ma
; CHECK-V-NEXT: addi a0, sp, 32
; CHECK-V-NEXT: vl1r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-V-NEXT: vsetivli zero, 1, e32, mf2, ta, ma
; CHECK-V-NEXT: vfmv.f.s fa0, v8
; CHECK-V-NEXT: call __fixsfti
; CHECK-V-NEXT: li a2, -1
@@ -2819,9 +2819,9 @@ define <2 x i64> @utest_f32i64(<2 x float> %x) {
; CHECK-V-NEXT: call __fixunssfti
; CHECK-V-NEXT: mv s0, a0
; CHECK-V-NEXT: mv s1, a1
-; CHECK-V-NEXT: vsetivli zero, 1, e32, mf2, ta, ma
; CHECK-V-NEXT: addi a0, sp, 32
; CHECK-V-NEXT: vl1r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-V-NEXT: vsetivli zero, 1, e32, mf2, ta, ma
; CHECK-V-NEXT: vfmv.f.s fa0, v8
; CHECK-V-NEXT: call __fixunssfti
; CHECK-V-NEXT: snez a1, a1
@@ -2931,9 +2931,9 @@ define <2 x i64> @ustest_f32i64(<2 x float> %x) {
; CHECK-V-NEXT: call __fixsfti
; CHECK-V-NEXT: mv s0, a0
; CHECK-V-NEXT: mv s1, a1
-; CHECK-V-NEXT: vsetivli zero, 1, e32, mf2, ta, ma
; CHECK-V-NEXT: addi a0, sp, 32
; CHECK-V-NEXT: vl1r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-V-NEXT: vsetivli zero, 1, e32, mf2, ta, ma
; CHECK-V-NEXT: vfmv.f.s fa0, v8
; CHECK-V-NEXT: call __fixsfti
; CHECK-V-NEXT: mv a2, s1
@@ -3819,8 +3819,8 @@ define <4 x i32> @stest_f16i32_mm(<4 x half> %x) {
; CHECK-V-NEXT: fmv.w.x fa0, a0
; CHECK-V-NEXT: call __extendhfsf2
; CHECK-V-NEXT: fcvt.l.s a0, fa0, rtz
-; CHECK-V-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; CHECK-V-NEXT: fmv.w.x fa0, s2
+; CHECK-V-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; CHECK-V-NEXT: vmv.s.x v8, a0
; CHECK-V-NEXT: addi a0, sp, 16
; CHECK-V-NEXT: vs1r.v v8, (a0) # Unknown-size Folded Spill
@@ -3838,8 +3838,8 @@ define <4 x i32> @stest_f16i32_mm(<4 x half> %x) {
; CHECK-V-NEXT: fmv.w.x fa0, s1
; CHECK-V-NEXT: call __extendhfsf2
; CHECK-V-NEXT: fcvt.l.s a0, fa0, rtz
-; CHECK-V-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; CHECK-V-NEXT: fmv.w.x fa0, s0
+; CHECK-V-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; CHECK-V-NEXT: vmv.s.x v8, a0
; CHECK-V-NEXT: addi a0, sp, 16
; CHECK-V-NEXT: vs1r.v v8, (a0) # Unknown-size Folded Spill
@@ -3978,8 +3978,8 @@ define <4 x i32> @utesth_f16i32_mm(<4 x half> %x) {
; CHECK-V-NEXT: fmv.w.x fa0, a0
; CHECK-V-NEXT: call __extendhfsf2
; CHECK-V-NEXT: fcvt.lu.s a0, fa0, rtz
-; CHECK-V-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; CHECK-V-NEXT: fmv.w.x fa0, s2
+; CHECK-V-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; CHECK-V-NEXT: vmv.s.x v8, a0
; CHECK-V-NEXT: addi a0, sp, 16
; CHECK-V-NEXT: vs1r.v v8, (a0) # Unknown-size Folded Spill
@@ -3997,8 +3997,8 @@ define <4 x i32> @utesth_f16i32_mm(<4 x half> %x) {
; CHECK-V-NEXT: fmv.w.x fa0, s1
; CHECK-V-NEXT: call __extendhfsf2
; CHECK-V-NEXT: fcvt.lu.s a0, fa0, rtz
-; CHECK-V-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; CHECK-V-NEXT: fmv.w.x fa0, s0
+; CHECK-V-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; CHECK-V-NEXT: vmv.s.x v8, a0
; CHECK-V-NEXT: addi a0, sp, 16
; CHECK-V-NEXT: vs1r.v v8, (a0) # Unknown-size Folded Spill
@@ -4148,8 +4148,8 @@ define <4 x i32> @ustest_f16i32_mm(<4 x half> %x) {
; CHECK-V-NEXT: fmv.w.x fa0, a0
; CHECK-V-NEXT: call __extendhfsf2
; CHECK-V-NEXT: fcvt.l.s a0, fa0, rtz
-; CHECK-V-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; CHECK-V-NEXT: fmv.w.x fa0, s2
+; CHECK-V-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; CHECK-V-NEXT: vmv.s.x v8, a0
; CHECK-V-NEXT: addi a0, sp, 16
; CHECK-V-NEXT: vs1r.v v8, (a0) # Unknown-size Folded Spill
@@ -4167,8 +4167,8 @@ define <4 x i32> @ustest_f16i32_mm(<4 x half> %x) {
; CHECK-V-NEXT: fmv.w.x fa0, s1
; CHECK-V-NEXT: call __extendhfsf2
; CHECK-V-NEXT: fcvt.l.s a0, fa0, rtz
-; CHECK-V-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; CHECK-V-NEXT: fmv.w.x fa0, s0
+; CHECK-V-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; CHECK-V-NEXT: vmv.s.x v8, a0
; CHECK-V-NEXT: addi a0, sp, 16
; CHECK-V-NEXT: vs1r.v v8, (a0) # Unknown-size Folded Spill
@@ -4753,8 +4753,8 @@ define <8 x i16> @stest_f16i16_mm(<8 x half> %x) {
; CHECK-V-NEXT: fmv.w.x fa0, a0
; CHECK-V-NEXT: call __extendhfsf2
; CHECK-V-NEXT: fcvt.l.s a0, fa0, rtz
-; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: fmv.w.x fa0, s6
+; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: vmv.s.x v8, a0
; CHECK-V-NEXT: csrr a0, vlenb
; CHECK-V-NEXT: slli a0, a0, 1
@@ -4778,8 +4778,8 @@ define <8 x i16> @stest_f16i16_mm(<8 x half> %x) {
; CHECK-V-NEXT: fmv.w.x fa0, s5
; CHECK-V-NEXT: call __extendhfsf2
; CHECK-V-NEXT: fcvt.l.s a0, fa0, rtz
-; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: fmv.w.x fa0, s4
+; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: vmv.s.x v8, a0
; CHECK-V-NEXT: csrr a0, vlenb
; CHECK-V-NEXT: slli a0, a0, 1
@@ -4796,11 +4796,11 @@ define <8 x i16> @stest_f16i16_mm(<8 x half> %x) {
; CHECK-V-NEXT: addi a0, a0, 16
; CHECK-V-NEXT: vl1r.v v9, (a0) # Unknown-size Folded Reload
; CHECK-V-NEXT: vslideup.vi v8, v9, 1
-; CHECK-V-NEXT: vsetivli zero, 4, e32, m1, ta, ma
; CHECK-V-NEXT: csrr a0, vlenb
; CHECK-V-NEXT: add a0, sp, a0
; CHECK-V-NEXT: addi a0, a0, 16
; CHECK-V-NEXT: vl1r.v v9, (a0) # Unknown-size Folded Reload
+; CHECK-V-NEXT: vsetivli zero, 4, e32, m1, ta, ma
; CHECK-V-NEXT: vslideup.vi v8, v9, 2
; CHECK-V-NEXT: csrr a0, vlenb
; CHECK-V-NEXT: slli a0, a0, 1
@@ -4810,8 +4810,8 @@ define <8 x i16> @stest_f16i16_mm(<8 x half> %x) {
; CHECK-V-NEXT: fmv.w.x fa0, s3
; CHECK-V-NEXT: call __extendhfsf2
; CHECK-V-NEXT: fcvt.l.s a0, fa0, rtz
-; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: fmv.w.x fa0, s2
+; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: vmv.s.x v8, a0
; CHECK-V-NEXT: addi a0, sp, 16
; CHECK-V-NEXT: vs1r.v v8, (a0) # Unknown-size Folded Spill
@@ -4829,8 +4829,8 @@ define <8 x i16> @stest_f16i16_mm(<8 x half> %x) {
; CHECK-V-NEXT: fmv.w.x fa0, s1
; CHECK-V-NEXT: call __extendhfsf2
; CHECK-V-NEXT: fcvt.l.s a0, fa0, rtz
-; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: fmv.w.x fa0, s0
+; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: vmv.s.x v8, a0
; CHECK-V-NEXT: addi a0, sp, 16
; CHECK-V-NEXT: vs1r.v v8, (a0) # Unknown-size Folded Spill
@@ -4841,18 +4841,18 @@ define <8 x i16> @stest_f16i16_mm(<8 x half> %x) {
; CHECK-V-NEXT: addi a0, sp, 16
; CHECK-V-NEXT: vl1r.v v8, (a0) # Unknown-size Folded Reload
; CHECK-V-NEXT: vslideup.vi v10, v8, 1
-; CHECK-V-NEXT: vsetivli zero, 4, e32, m1, ta, ma
; CHECK-V-NEXT: csrr a0, vlenb
; CHECK-V-NEXT: add a0, sp, a0
; CHECK-V-NEXT: addi a0, a0, 16
; CHECK-V-NEXT: vl1r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-V-NEXT: vsetivli zero, 4, e32, m1, ta, ma
; CHECK-V-NEXT: vslideup.vi v10, v8, 2
-; CHECK-V-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; CHECK-V-NEXT: csrr a0, vlenb
; CHECK-V-NEXT: slli a0, a0, 1
; CHECK-V-NEXT: add a0, sp, a0
; CHECK-V-NEXT: addi a0, a0, 16
; CHECK-V-NEXT: vl2r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-V-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; CHECK-V-NEXT: vslideup.vi v10, v8, 4
; CHECK-V-NEXT: vsetvli zero, zero, e16, m1, ta, ma
; CHECK-V-NEXT: vnclip.wi v8, v10, 0
@@ -5064,8 +5064,8 @@ define <8 x i16> @utesth_f16i16_mm(<8 x half> %x) {
; CHECK-V-NEXT: fmv.w.x fa0, a0
; CHECK-V-NEXT: call __extendhfsf2
; CHECK-V-NEXT: fcvt.lu.s a0, fa0, rtz
-; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: fmv.w.x fa0, s6
+; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: vmv.s.x v8, a0
; CHECK-V-NEXT: csrr a0, vlenb
; CHECK-V-NEXT: slli a0, a0, 1
@@ -5089,8 +5089,8 @@ define <8 x i16> @utesth_f16i16_mm(<8 x half> %x) {
; CHECK-V-NEXT: fmv.w.x fa0, s5
; CHECK-V-NEXT: call __extendhfsf2
; CHECK-V-NEXT: fcvt.lu.s a0, fa0, rtz
-; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: fmv.w.x fa0, s4
+; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: vmv.s.x v8, a0
; CHECK-V-NEXT: csrr a0, vlenb
; CHECK-V-NEXT: slli a0, a0, 1
@@ -5107,11 +5107,11 @@ define <8 x i16> @utesth_f16i16_mm(<8 x half> %x) {
; CHECK-V-NEXT: addi a0, a0, 16
; CHECK-V-NEXT: vl1r.v v9, (a0) # Unknown-size Folded Reload
; CHECK-V-NEXT: vslideup.vi v8, v9, 1
-; CHECK-V-NEXT: vsetivli zero, 4, e32, m1, ta, ma
; CHECK-V-NEXT: csrr a0, vlenb
; CHECK-V-NEXT: add a0, sp, a0
; CHECK-V-NEXT: addi a0, a0, 16
; CHECK-V-NEXT: vl1r.v v9, (a0) # Unknown-size Folded Reload
+; CHECK-V-NEXT: vsetivli zero, 4, e32, m1, ta, ma
; CHECK-V-NEXT: vslideup.vi v8, v9, 2
; CHECK-V-NEXT: csrr a0, vlenb
; CHECK-V-NEXT: slli a0, a0, 1
@@ -5121,8 +5121,8 @@ define <8 x i16> @utesth_f16i16_mm(<8 x half> %x) {
; CHECK-V-NEXT: fmv.w.x fa0, s3
; CHECK-V-NEXT: call __extendhfsf2
; CHECK-V-NEXT: fcvt.lu.s a0, fa0, rtz
-; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: fmv.w.x fa0, s2
+; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: vmv.s.x v8, a0
; CHECK-V-NEXT: addi a0, sp, 16
; CHECK-V-NEXT: vs1r.v v8, (a0) # Unknown-size Folded Spill
@@ -5140,8 +5140,8 @@ define <8 x i16> @utesth_f16i16_mm(<8 x half> %x) {
; CHECK-V-NEXT: fmv.w.x fa0, s1
; CHECK-V-NEXT: call __extendhfsf2
; CHECK-V-NEXT: fcvt.lu.s a0, fa0, rtz
-; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: fmv.w.x fa0, s0
+; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: vmv.s.x v8, a0
; CHECK-V-NEXT: addi a0, sp, 16
; CHECK-V-NEXT: vs1r.v v8, (a0) # Unknown-size Folded Spill
@@ -5152,18 +5152,18 @@ define <8 x i16> @utesth_f16i16_mm(<8 x half> %x) {
; CHECK-V-NEXT: addi a0, sp, 16
; CHECK-V-NEXT: vl1r.v v8, (a0) # Unknown-size Folded Reload
; CHECK-V-NEXT: vslideup.vi v10, v8, 1
-; CHECK-V-NEXT: vsetivli zero, 4, e32, m1, ta, ma
; CHECK-V-NEXT: csrr a0, vlenb
; CHECK-V-NEXT: add a0, sp, a0
; CHECK-V-NEXT: addi a0, a0, 16
; CHECK-V-NEXT: vl1r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-V-NEXT: vsetivli zero, 4, e32, m1, ta, ma
; CHECK-V-NEXT: vslideup.vi v10, v8, 2
-; CHECK-V-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; CHECK-V-NEXT: csrr a0, vlenb
; CHECK-V-NEXT: slli a0, a0, 1
; CHECK-V-NEXT: add a0, sp, a0
; CHECK-V-NEXT: addi a0, a0, 16
; CHECK-V-NEXT: vl2r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-V-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; CHECK-V-NEXT: vslideup.vi v10, v8, 4
; CHECK-V-NEXT: vsetvli zero, zero, e16, m1, ta, ma
; CHECK-V-NEXT: vnclipu.wi v8, v10, 0
@@ -5398,8 +5398,8 @@ define <8 x i16> @ustest_f16i16_mm(<8 x half> %x) {
; CHECK-V-NEXT: fmv.w.x fa0, a0
; CHECK-V-NEXT: call __extendhfsf2
; CHECK-V-NEXT: fcvt.l.s a0, fa0, rtz
-; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: fmv.w.x fa0, s6
+; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: vmv.s.x v8, a0
; CHECK-V-NEXT: csrr a0, vlenb
; CHECK-V-NEXT: slli a0, a0, 1
@@ -5423,8 +5423,8 @@ define <8 x i16> @ustest_f16i16_mm(<8 x half> %x) {
; CHECK-V-NEXT: fmv.w.x fa0, s5
; CHECK-V-NEXT: call __extendhfsf2
; CHECK-V-NEXT: fcvt.l.s a0, fa0, rtz
-; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: fmv.w.x fa0, s4
+; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: vmv.s.x v8, a0
; CHECK-V-NEXT: csrr a0, vlenb
; CHECK-V-NEXT: slli a0, a0, 1
@@ -5441,11 +5441,11 @@ define <8 x i16> @ustest_f16i16_mm(<8 x half> %x) {
; CHECK-V-NEXT: addi a0, a0, 16
; CHECK-V-NEXT: vl1r.v v9, (a0) # Unknown-size Folded Reload
; CHECK-V-NEXT: vslideup.vi v8, v9, 1
-; CHECK-V-NEXT: vsetivli zero, 4, e32, m1, ta, ma
; CHECK-V-NEXT: csrr a0, vlenb
; CHECK-V-NEXT: add a0, sp, a0
; CHECK-V-NEXT: addi a0, a0, 16
; CHECK-V-NEXT: vl1r.v v9, (a0) # Unknown-size Folded Reload
+; CHECK-V-NEXT: vsetivli zero, 4, e32, m1, ta, ma
; CHECK-V-NEXT: vslideup.vi v8, v9, 2
; CHECK-V-NEXT: csrr a0, vlenb
; CHECK-V-NEXT: slli a0, a0, 1
@@ -5455,8 +5455,8 @@ define <8 x i16> @ustest_f16i16_mm(<8 x half> %x) {
; CHECK-V-NEXT: fmv.w.x fa0, s3
; CHECK-V-NEXT: call __extendhfsf2
; CHECK-V-NEXT: fcvt.l.s a0, fa0, rtz
-; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: fmv.w.x fa0, s2
+; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: vmv.s.x v8, a0
; CHECK-V-NEXT: addi a0, sp, 16
; CHECK-V-NEXT: vs1r.v v8, (a0) # Unknown-size Folded Spill
@@ -5474,8 +5474,8 @@ define <8 x i16> @ustest_f16i16_mm(<8 x half> %x) {
; CHECK-V-NEXT: fmv.w.x fa0, s1
; CHECK-V-NEXT: call __extendhfsf2
; CHECK-V-NEXT: fcvt.l.s a0, fa0, rtz
-; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: fmv.w.x fa0, s0
+; CHECK-V-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; CHECK-V-NEXT: vmv.s.x v8, a0
; CHECK-V-NEXT: addi a0, sp, 16
; CHECK-V-NEXT: vs1r.v v8, (a0) # Unknown-size Folded Spill
@@ -5486,18 +5486,18 @@ define <8 x i16> @ustest_f16i16_mm(<8 x half> %x) {
; CHECK-V-NEXT: addi a0, sp, 16
; CHECK-V-NEXT: vl1r.v v9, (a0) # Unknown-size Folded Reload
; CHECK-V-NEXT: vslideup.vi v8, v9, 1
-; CHECK-V-NEXT: vsetivli zero, 4, e32, m1, ta, ma
; CHECK-V-NEXT: csrr a0, vlenb
; CHECK-V-NEXT: add a0, sp, a0
; CHECK-V-NEXT: addi a0, a0, 16
; CHECK-V-NEXT: vl1r.v v9, (a0) # Unknown-size Folded Reload
+; CHECK-V-NEXT: vsetivli zero, 4, e32, m1, ta, ma
; CHECK-V-NEXT: vslideup.vi v8, v9, 2
-; CHECK-V-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; CHECK-V-NEXT: csrr a0, vlenb
; CHECK-V-NEXT: slli a0, a0, 1
; CHECK-V-NEXT: add a0, sp, a0
; CHECK-V-NEXT: addi a0, a0, 16
; CHECK-V-NEXT: vl2r.v v10, (a0) # Unknown-size Folded Reload
+; CHECK-V-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; CHECK-V-NEXT: vslideup.vi v8, v10, 4
; CHECK-V-NEXT: lui a0, 16
; CHECK-V-NEXT: addi a0, a0, -1
@@ -5633,9 +5633,9 @@ define <2 x i64> @stest_f64i64_mm(<2 x double> %x) {
; CHECK-V-NEXT: call __fixdfti
; CHECK-V-NEXT: mv s0, a0
; CHECK-V-NEXT: mv s1, a1
-; CHECK-V-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; CHECK-V-NEXT: addi a0, sp, 32
; CHECK-V-NEXT: vl1r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-V-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; CHECK-V-NEXT: vfmv.f.s fa0, v8
; CHECK-V-NEXT: call __fixdfti
; CHECK-V-NEXT: li a2, -1
@@ -5766,9 +5766,9 @@ define <2 x i64> @utest_f64i64_mm(<2 x double> %x) {
; CHECK-V-NEXT: call __fixunsdfti
; CHECK-V-NEXT: mv s0, a0
; CHECK-V-NEXT: mv s1, a1
-; CHECK-V-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; CHECK-V-NEXT: addi a0, sp, 32
; CHECK-V-NEXT: vl1r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-V-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; CHECK-V-NEXT: vslidedown.vi v8, v8, 1
; CHECK-V-NEXT: vfmv.f.s fa0, v8
; CHECK-V-NEXT: call __fixunsdfti
@@ -5867,9 +5867,9 @@ define <2 x i64> @ustest_f64i64_mm(<2 x double> %x) {
; CHECK-V-NEXT: call __fixdfti
; CHECK-V-NEXT: mv s0, a0
; CHECK-V-NEXT: mv s1, a1
-; CHECK-V-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; CHECK-V-NEXT: addi a0, sp, 32
; CHECK-V-NEXT: vl1r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-V-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; CHECK-V-NEXT: vfmv.f.s fa0, v8
; CHECK-V-NEXT: call __fixdfti
; CHECK-V-NEXT: mv a2, a1
@@ -6019,9 +6019,9 @@ define <2 x i64> @stest_f32i64_mm(<2 x float> %x) {
; CHECK-V-NEXT: call __fixsfti
; CHECK-V-NEXT: mv s0, a0
; CHECK-V-NEXT: mv s1, a1
-; CHECK-V-NEXT: vsetivli zero, 1, e32, mf2, ta, ma
; CHECK-V-NEXT: addi a0, sp, 32
; CHECK-V-NEXT: vl1r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-V-NEXT: vsetivli zero, 1, e32, mf2, ta, ma
; CHECK-V-NEXT: vfmv.f.s fa0, v8
; CHECK-V-NEXT: call __fixsfti
; CHECK-V-NEXT: li a2, -1
@@ -6152,9 +6152,9 @@ define <2 x i64> @utest_f32i64_mm(<2 x float> %x) {
; CHECK-V-NEXT: call __fixunssfti
; CHECK-V-NEXT: mv s0, a0
; CHECK-V-NEXT: mv s1, a1
-; CHECK-V-NEXT: vsetivli zero, 1, e32, mf2, ta, ma
; CHECK-V-NEXT: addi a0, sp, 32
; CHECK-V-NEXT: vl1r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-V-NEXT: vsetivli zero, 1, e32, mf2, ta, ma
; CHECK-V-NEXT: vslidedown.vi v8, v8, 1
; CHECK-V-NEXT: vfmv.f.s fa0, v8
; CHECK-V-NEXT: call __fixunssfti
@@ -6253,9 +6253,9 @@ define <2 x i64> @ustest_f32i64_mm(<2 x float> %x) {
; CHECK-V-NEXT: call __fixsfti
; CHECK-V-NEXT: mv s0, a0
; CHECK-V-NEXT: mv s1, a1
-; CHECK-V-NEXT: vsetivli zero, 1, e32, mf2, ta, ma
; CHECK-V-NEXT: addi a0, sp, 32
; CHECK-V-NEXT: vl1r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-V-NEXT: vsetivli zero, 1, e32, mf2, ta, ma
; CHECK-V-NEXT: vfmv.f.s fa0, v8
; CHECK-V-NEXT: call __fixsfti
; CHECK-V-NEXT: mv a2, a1
diff --git a/llvm/test/CodeGen/RISCV/rvv/fptosi-sat.ll b/llvm/test/CodeGen/RISCV/rvv/fptosi-sat.ll
index 8f36aad817274..c45af61ced94f 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fptosi-sat.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fptosi-sat.ll
@@ -163,12 +163,11 @@ define <vscale x 4 x i16> @test_signed_v4f64_v4i16(<vscale x 4 x double> %f) {
; CHECK-NEXT: vfmin.vf v12, v12, fa4
; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; CHECK-NEXT: vfncvt.rtz.x.f.w v16, v12
-; CHECK-NEXT: vsetvli zero, zero, e16, m1, ta, ma
-; CHECK-NEXT: vnsrl.wi v12, v16, 0
; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, ma
; CHECK-NEXT: vmfne.vv v0, v8, v8
; CHECK-NEXT: vsetvli zero, zero, e16, m1, ta, ma
-; CHECK-NEXT: vmerge.vim v8, v12, 0, v0
+; CHECK-NEXT: vnsrl.wi v8, v16, 0
+; CHECK-NEXT: vmerge.vim v8, v8, 0, v0
; CHECK-NEXT: ret
%x = call <vscale x 4 x i16> @llvm.fptosi.sat.nxv4f64.nxv4i16(<vscale x 4 x double> %f)
ret <vscale x 4 x i16> %x
@@ -186,12 +185,11 @@ define <vscale x 8 x i16> @test_signed_v8f64_v8i16(<vscale x 8 x double> %f) {
; CHECK-NEXT: vfmin.vf v16, v16, fa4
; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; CHECK-NEXT: vfncvt.rtz.x.f.w v24, v16
-; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, ma
-; CHECK-NEXT: vnsrl.wi v16, v24, 0
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vmfne.vv v0, v8, v8
; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, ma
-; CHECK-NEXT: vmerge.vim v8, v16, 0, v0
+; CHECK-NEXT: vnsrl.wi v8, v24, 0
+; CHECK-NEXT: vmerge.vim v8, v8, 0, v0
; CHECK-NEXT: ret
%x = call <vscale x 8 x i16> @llvm.fptosi.sat.nxv8f64.nxv8i16(<vscale x 8 x double> %f)
ret <vscale x 8 x i16> %x
diff --git a/llvm/test/CodeGen/RISCV/rvv/fshr-fshl-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fshr-fshl-vp.ll
index 277cd7dcdabce..9c8991b52c251 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fshr-fshl-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fshr-fshl-vp.ll
@@ -960,10 +960,10 @@ define <vscale x 16 x i64> @fshr_v16i64(<vscale x 16 x i64> %a, <vscale x 16 x i
; CHECK-NEXT: addi sp, sp, -16
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: li a3, 56
+; CHECK-NEXT: li a3, 48
; CHECK-NEXT: mul a1, a1, a3
; CHECK-NEXT: sub sp, sp, a1
-; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x38, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 56 * vlenb
+; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x30, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 48 * vlenb
; CHECK-NEXT: vmv1r.v v24, v0
; CHECK-NEXT: csrr a1, vlenb
; CHECK-NEXT: li a3, 24
@@ -976,140 +976,121 @@ define <vscale x 16 x i64> @fshr_v16i64(<vscale x 16 x i64> %a, <vscale x 16 x i
; CHECK-NEXT: add a1, sp, a1
; CHECK-NEXT: addi a1, a1, 16
; CHECK-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
+; CHECK-NEXT: csrr a1, vlenb
+; CHECK-NEXT: slli a3, a1, 3
+; CHECK-NEXT: add a5, a0, a3
+; CHECK-NEXT: add a3, a2, a3
+; CHECK-NEXT: vl8re64.v v16, (a3)
; CHECK-NEXT: csrr a3, vlenb
-; CHECK-NEXT: slli a1, a3, 3
-; CHECK-NEXT: add a5, a0, a1
+; CHECK-NEXT: slli a3, a3, 4
+; CHECK-NEXT: add a3, sp, a3
+; CHECK-NEXT: addi a3, a3, 16
+; CHECK-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
+; CHECK-NEXT: sub a3, a4, a1
+; CHECK-NEXT: sltu a6, a4, a3
+; CHECK-NEXT: addi a6, a6, -1
+; CHECK-NEXT: and a6, a6, a3
+; CHECK-NEXT: srli a3, a1, 3
; CHECK-NEXT: vl8re64.v v8, (a5)
; CHECK-NEXT: csrr a5, vlenb
+; CHECK-NEXT: li a7, 40
+; CHECK-NEXT: mul a5, a5, a7
+; CHECK-NEXT: add a5, sp, a5
+; CHECK-NEXT: addi a5, a5, 16
+; CHECK-NEXT: vs8r.v v8, (a5) # Unknown-size Folded Spill
+; CHECK-NEXT: vsetvli a5, zero, e8, mf4, ta, ma
+; CHECK-NEXT: vslidedown.vx v0, v0, a3
+; CHECK-NEXT: li a3, 63
+; CHECK-NEXT: vsetvli zero, a6, e64, m8, ta, ma
+; CHECK-NEXT: vand.vx v8, v16, a3, v0.t
+; CHECK-NEXT: csrr a5, vlenb
+; CHECK-NEXT: li a6, 40
+; CHECK-NEXT: mul a5, a5, a6
+; CHECK-NEXT: add a5, sp, a5
+; CHECK-NEXT: addi a5, a5, 16
+; CHECK-NEXT: vl8r.v v16, (a5) # Unknown-size Folded Reload
+; CHECK-NEXT: vsrl.vv v8, v16, v8, v0.t
+; CHECK-NEXT: csrr a5, vlenb
; CHECK-NEXT: slli a5, a5, 3
; CHECK-NEXT: add a5, sp, a5
; CHECK-NEXT: addi a5, a5, 16
; CHECK-NEXT: vs8r.v v8, (a5) # Unknown-size Folded Spill
-; CHECK-NEXT: srli a5, a3, 3
-; CHECK-NEXT: vsetvli a6, zero, e8, mf4, ta, ma
-; CHECK-NEXT: vslidedown.vx v0, v0, a5
-; CHECK-NEXT: add a5, a2, a1
-; CHECK-NEXT: sub a1, a4, a3
-; CHECK-NEXT: sltu a6, a4, a1
-; CHECK-NEXT: addi a6, a6, -1
-; CHECK-NEXT: and a6, a6, a1
-; CHECK-NEXT: li a1, 63
-; CHECK-NEXT: vl8re64.v v8, (a5)
; CHECK-NEXT: csrr a5, vlenb
-; CHECK-NEXT: li a7, 40
-; CHECK-NEXT: mul a5, a5, a7
+; CHECK-NEXT: slli a5, a5, 4
; CHECK-NEXT: add a5, sp, a5
; CHECK-NEXT: addi a5, a5, 16
+; CHECK-NEXT: vl8r.v v8, (a5) # Unknown-size Folded Reload
+; CHECK-NEXT: vnot.v v8, v8, v0.t
+; CHECK-NEXT: vand.vx v8, v8, a3, v0.t
+; CHECK-NEXT: addi a5, sp, 16
; CHECK-NEXT: vs8r.v v8, (a5) # Unknown-size Folded Spill
-; CHECK-NEXT: vl8re64.v v16, (a0)
+; CHECK-NEXT: vl8re64.v v8, (a0)
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 4
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: vl8re64.v v16, (a2)
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a2, 48
-; CHECK-NEXT: mul a0, a0, a2
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: vsetvli zero, a6, e64, m8, ta, ma
+; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vl8re64.v v8, (a2)
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: li a2, 40
; CHECK-NEXT: mul a0, a0, a2
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vand.vx v8, v8, a1, v0.t
-; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
+; CHECK-NEXT: li a2, 24
+; CHECK-NEXT: mul a0, a0, a2
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vsll.vi v16, v8, 1, v0.t
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vsrl.vv v16, v16, v8, v0.t
+; CHECK-NEXT: vsll.vv v16, v16, v8, v0.t
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a2, 40
-; CHECK-NEXT: mul a0, a0, a2
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vnot.v v8, v8, v0.t
-; CHECK-NEXT: vand.vx v16, v8, a1, v0.t
+; CHECK-NEXT: vor.vv v8, v16, v8, v0.t
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: li a2, 24
; CHECK-NEXT: mul a0, a0, a2
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vsll.vi v8, v8, 1, v0.t
-; CHECK-NEXT: vsll.vv v8, v8, v16, v0.t
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vor.vv v8, v8, v16, v0.t
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a2, 40
-; CHECK-NEXT: mul a0, a0, a2
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: bltu a4, a3, .LBB46_2
+; CHECK-NEXT: bltu a4, a1, .LBB46_2
; CHECK-NEXT: # %bb.1:
-; CHECK-NEXT: mv a4, a3
+; CHECK-NEXT: mv a4, a1
; CHECK-NEXT: .LBB46_2:
-; CHECK-NEXT: vsetvli zero, a4, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a2, 48
-; CHECK-NEXT: mul a0, a0, a2
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vand.vx v8, v8, a1, v0.t
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a2, 24
-; CHECK-NEXT: mul a0, a0, a2
+; CHECK-NEXT: li a1, 40
+; CHECK-NEXT: mul a0, a0, a1
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, a4, e64, m8, ta, ma
+; CHECK-NEXT: vand.vx v8, v16, a3, v0.t
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 4
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a2, 24
-; CHECK-NEXT: mul a0, a0, a2
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vsrl.vv v8, v8, v16, v0.t
+; CHECK-NEXT: vsrl.vv v8, v16, v8, v0.t
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a2, 24
-; CHECK-NEXT: mul a0, a0, a2
+; CHECK-NEXT: slli a0, a0, 4
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a2, 48
-; CHECK-NEXT: mul a0, a0, a2
+; CHECK-NEXT: li a1, 40
+; CHECK-NEXT: mul a0, a0, a1
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vnot.v v16, v8, v0.t
-; CHECK-NEXT: vand.vx v16, v16, a1, v0.t
+; CHECK-NEXT: vand.vx v16, v16, a3, v0.t
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 5
; CHECK-NEXT: add a0, sp, a0
@@ -1118,20 +1099,19 @@ define <vscale x 16 x i64> @fshr_v16i64(<vscale x 16 x i64> %a, <vscale x 16 x i
; CHECK-NEXT: vsll.vi v8, v8, 1, v0.t
; CHECK-NEXT: vsll.vv v8, v8, v16, v0.t
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a1, 24
-; CHECK-NEXT: mul a0, a0, a1
+; CHECK-NEXT: slli a0, a0, 4
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vor.vv v8, v8, v16, v0.t
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a1, 40
+; CHECK-NEXT: li a1, 24
; CHECK-NEXT: mul a0, a0, a1
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a1, 56
+; CHECK-NEXT: li a1, 48
; CHECK-NEXT: mul a0, a0, a1
; CHECK-NEXT: add sp, sp, a0
; CHECK-NEXT: addi sp, sp, 16
@@ -1147,174 +1127,147 @@ define <vscale x 16 x i64> @fshl_v16i64(<vscale x 16 x i64> %a, <vscale x 16 x i
; CHECK-NEXT: addi sp, sp, -16
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: li a3, 56
+; CHECK-NEXT: li a3, 40
; CHECK-NEXT: mul a1, a1, a3
; CHECK-NEXT: sub sp, sp, a1
-; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x38, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 56 * vlenb
+; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x28, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 40 * vlenb
; CHECK-NEXT: vmv1r.v v24, v0
; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a1, a1, 3
+; CHECK-NEXT: slli a1, a1, 5
; CHECK-NEXT: add a1, sp, a1
; CHECK-NEXT: addi a1, a1, 16
; CHECK-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a1, a1, 5
+; CHECK-NEXT: li a3, 24
+; CHECK-NEXT: mul a1, a1, a3
; CHECK-NEXT: add a1, sp, a1
; CHECK-NEXT: addi a1, a1, 16
; CHECK-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
; CHECK-NEXT: csrr a3, vlenb
-; CHECK-NEXT: slli a1, a3, 3
-; CHECK-NEXT: add a5, a0, a1
-; CHECK-NEXT: vl8re64.v v8, (a5)
-; CHECK-NEXT: csrr a5, vlenb
-; CHECK-NEXT: slli a5, a5, 4
-; CHECK-NEXT: add a5, sp, a5
-; CHECK-NEXT: addi a5, a5, 16
-; CHECK-NEXT: vs8r.v v8, (a5) # Unknown-size Folded Spill
-; CHECK-NEXT: srli a5, a3, 3
+; CHECK-NEXT: slli a5, a3, 3
+; CHECK-NEXT: srli a1, a3, 3
; CHECK-NEXT: vsetvli a6, zero, e8, mf4, ta, ma
-; CHECK-NEXT: vslidedown.vx v0, v0, a5
-; CHECK-NEXT: add a5, a2, a1
+; CHECK-NEXT: vslidedown.vx v0, v0, a1
+; CHECK-NEXT: add a1, a2, a5
+; CHECK-NEXT: vl8re64.v v8, (a1)
+; CHECK-NEXT: csrr a1, vlenb
+; CHECK-NEXT: slli a1, a1, 4
+; CHECK-NEXT: add a1, sp, a1
+; CHECK-NEXT: addi a1, a1, 16
+; CHECK-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
; CHECK-NEXT: sub a1, a4, a3
; CHECK-NEXT: sltu a6, a4, a1
; CHECK-NEXT: addi a6, a6, -1
; CHECK-NEXT: and a6, a6, a1
; CHECK-NEXT: li a1, 63
-; CHECK-NEXT: vl8re64.v v8, (a5)
-; CHECK-NEXT: csrr a5, vlenb
-; CHECK-NEXT: li a7, 40
-; CHECK-NEXT: mul a5, a5, a7
-; CHECK-NEXT: add a5, sp, a5
-; CHECK-NEXT: addi a5, a5, 16
-; CHECK-NEXT: vs8r.v v8, (a5) # Unknown-size Folded Spill
-; CHECK-NEXT: vl8re64.v v16, (a0)
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a5, 24
-; CHECK-NEXT: mul a0, a0, a5
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: vl8re64.v v16, (a2)
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a2, 48
-; CHECK-NEXT: mul a0, a0, a2
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: vsetvli zero, a6, e64, m8, ta, ma
+; CHECK-NEXT: vand.vx v8, v8, a1, v0.t
+; CHECK-NEXT: csrr a6, vlenb
+; CHECK-NEXT: slli a6, a6, 5
+; CHECK-NEXT: add a6, sp, a6
+; CHECK-NEXT: addi a6, a6, 16
+; CHECK-NEXT: vl8r.v v16, (a6) # Unknown-size Folded Reload
+; CHECK-NEXT: vsll.vv v16, v16, v8, v0.t
+; CHECK-NEXT: csrr a6, vlenb
+; CHECK-NEXT: slli a6, a6, 3
+; CHECK-NEXT: add a6, sp, a6
+; CHECK-NEXT: addi a6, a6, 16
+; CHECK-NEXT: vs8r.v v16, (a6) # Unknown-size Folded Spill
+; CHECK-NEXT: add a5, a0, a5
+; CHECK-NEXT: csrr a6, vlenb
+; CHECK-NEXT: slli a6, a6, 4
+; CHECK-NEXT: add a6, sp, a6
+; CHECK-NEXT: addi a6, a6, 16
+; CHECK-NEXT: vl8r.v v8, (a6) # Unknown-size Folded Reload
+; CHECK-NEXT: vnot.v v8, v8, v0.t
+; CHECK-NEXT: vl8re64.v v16, (a5)
+; CHECK-NEXT: vand.vx v8, v8, a1, v0.t
+; CHECK-NEXT: addi a5, sp, 16
+; CHECK-NEXT: vs8r.v v8, (a5) # Unknown-size Folded Spill
+; CHECK-NEXT: vl8re64.v v8, (a0)
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a2, 40
-; CHECK-NEXT: mul a0, a0, a2
+; CHECK-NEXT: slli a0, a0, 4
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vand.vx v8, v8, a1, v0.t
-; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vl8re64.v v8, (a2)
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
+; CHECK-NEXT: slli a0, a0, 5
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vsrl.vi v16, v16, 1, v0.t
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vsll.vv v16, v16, v8, v0.t
+; CHECK-NEXT: vsrl.vv v16, v16, v8, v0.t
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a2, 40
-; CHECK-NEXT: mul a0, a0, a2
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vnot.v v8, v8, v0.t
-; CHECK-NEXT: vand.vx v16, v8, a1, v0.t
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 4
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vsrl.vi v8, v8, 1, v0.t
-; CHECK-NEXT: vsrl.vv v8, v8, v16, v0.t
+; CHECK-NEXT: vor.vv v8, v8, v16, v0.t
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vor.vv v8, v16, v8, v0.t
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a2, 40
-; CHECK-NEXT: mul a0, a0, a2
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: bltu a4, a3, .LBB47_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a4, a3
; CHECK-NEXT: .LBB47_2:
-; CHECK-NEXT: vsetvli zero, a4, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a2, 48
-; CHECK-NEXT: mul a0, a0, a2
+; CHECK-NEXT: slli a0, a0, 5
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, a4, e64, m8, ta, ma
; CHECK-NEXT: vand.vx v8, v8, a1, v0.t
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 4
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
+; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 5
+; CHECK-NEXT: li a2, 24
+; CHECK-NEXT: mul a0, a0, a2
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
+; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vsll.vv v8, v8, v16, v0.t
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 4
+; CHECK-NEXT: li a2, 24
+; CHECK-NEXT: mul a0, a0, a2
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vsll.vv v16, v16, v8, v0.t
+; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 5
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a2, 48
-; CHECK-NEXT: mul a0, a0, a2
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vnot.v v16, v8, v0.t
-; CHECK-NEXT: vand.vx v16, v16, a1, v0.t
+; CHECK-NEXT: vnot.v v8, v8, v0.t
+; CHECK-NEXT: vand.vx v16, v8, a1, v0.t
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a1, 24
-; CHECK-NEXT: mul a0, a0, a1
+; CHECK-NEXT: slli a0, a0, 4
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vsrl.vi v8, v8, 1, v0.t
; CHECK-NEXT: vsrl.vv v8, v8, v16, v0.t
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 5
+; CHECK-NEXT: li a1, 24
+; CHECK-NEXT: mul a0, a0, a1
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vor.vv v8, v16, v8, v0.t
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a1, 40
-; CHECK-NEXT: mul a0, a0, a1
+; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a1, 56
+; CHECK-NEXT: li a1, 40
; CHECK-NEXT: mul a0, a0, a1
; CHECK-NEXT: add sp, sp, a0
; CHECK-NEXT: addi sp, sp, 16
diff --git a/llvm/test/CodeGen/RISCV/rvv/llrint-vp.ll b/llvm/test/CodeGen/RISCV/rvv/llrint-vp.ll
index 6d8763d34ec1b..e34b4a81b631b 100644
--- a/llvm/test/CodeGen/RISCV/rvv/llrint-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/llrint-vp.ll
@@ -70,8 +70,8 @@ define <vscale x 16 x i64> @llrint_nxv16i64_nxv16f32(<vscale x 16 x float> %x, <
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB4_2:
-; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, ma
; CHECK-NEXT: vfwcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: vmv8r.v v8, v24
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/lrint-vp.ll b/llvm/test/CodeGen/RISCV/rvv/lrint-vp.ll
index 9fa8807ed4add..c9f91bf9def2c 100644
--- a/llvm/test/CodeGen/RISCV/rvv/lrint-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/lrint-vp.ll
@@ -132,8 +132,8 @@ define <vscale x 16 x iXLen> @lrint_nxv16f32(<vscale x 16 x float> %x, <vscale x
; RV64-i64-NEXT: # %bb.1:
; RV64-i64-NEXT: mv a0, a1
; RV64-i64-NEXT: .LBB4_2:
-; RV64-i64-NEXT: vsetvli zero, a0, e32, m4, ta, ma
; RV64-i64-NEXT: vmv1r.v v0, v24
+; RV64-i64-NEXT: vsetvli zero, a0, e32, m4, ta, ma
; RV64-i64-NEXT: vfwcvt.x.f.v v24, v8, v0.t
; RV64-i64-NEXT: vmv8r.v v8, v24
; RV64-i64-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/masked-tama.ll b/llvm/test/CodeGen/RISCV/rvv/masked-tama.ll
index f87fa3ec6f160..3aca3130cc545 100644
--- a/llvm/test/CodeGen/RISCV/rvv/masked-tama.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/masked-tama.ll
@@ -1289,8 +1289,8 @@ define <vscale x 1 x i8> @intrinsic_viota_mask_m_nxv1i8_nxv1i1(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_viota_mask_m_nxv1i8_nxv1i1:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, ma
; CHECK-NEXT: viota.m v8, v9, v0.t
; CHECK-NEXT: ret
entry:
@@ -1313,8 +1313,8 @@ define <vscale x 1 x i1> @intrinsic_vmsbf_mask_m_nxv1i1_nxv1i1(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmsbf_mask_m_nxv1i1_nxv1i1:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, ma
; CHECK-NEXT: vmsbf.m v8, v9, v0.t
; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
@@ -1444,8 +1444,8 @@ define <vscale x 64 x i1> @intrinsic_vmsbf_mask_m_nxv64i1_nxv64i1(<vscale x 64 x
; CHECK-LABEL: intrinsic_vmsbf_mask_m_nxv64i1_nxv64i1:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a0, e8, m8, ta, ma
; CHECK-NEXT: vmsbf.m v8, v9, v0.t
; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/mgather-sdnode.ll b/llvm/test/CodeGen/RISCV/rvv/mgather-sdnode.ll
index e12f1cf7603b8..be37be06f0e77 100644
--- a/llvm/test/CodeGen/RISCV/rvv/mgather-sdnode.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/mgather-sdnode.ll
@@ -1200,13 +1200,13 @@ define void @mgather_nxv16i64(<vscale x 8 x ptr> %ptrs0, <vscale x 8 x ptr> %ptr
; RV32-LABEL: mgather_nxv16i64:
; RV32: # %bb.0:
; RV32-NEXT: vl8re64.v v24, (a0)
-; RV32-NEXT: vsetvli a0, zero, e64, m8, ta, mu
-; RV32-NEXT: vluxei32.v v16, (zero), v8, v0.t
; RV32-NEXT: csrr a0, vlenb
; RV32-NEXT: srli a2, a0, 3
; RV32-NEXT: vsetvli a3, zero, e8, mf4, ta, ma
-; RV32-NEXT: vslidedown.vx v0, v0, a2
+; RV32-NEXT: vslidedown.vx v7, v0, a2
; RV32-NEXT: vsetvli a2, zero, e64, m8, ta, mu
+; RV32-NEXT: vluxei32.v v16, (zero), v8, v0.t
+; RV32-NEXT: vmv1r.v v0, v7
; RV32-NEXT: vluxei32.v v24, (zero), v12, v0.t
; RV32-NEXT: slli a0, a0, 3
; RV32-NEXT: add a0, a1, a0
@@ -1222,18 +1222,18 @@ define void @mgather_nxv16i64(<vscale x 8 x ptr> %ptrs0, <vscale x 8 x ptr> %ptr
; RV64-NEXT: slli a3, a3, 3
; RV64-NEXT: sub sp, sp, a3
; RV64-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
-; RV64-NEXT: vl8re64.v v24, (a0)
-; RV64-NEXT: addi a0, sp, 16
-; RV64-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
+; RV64-NEXT: addi a3, sp, 16
+; RV64-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
; RV64-NEXT: vmv8r.v v16, v8
-; RV64-NEXT: vl8re64.v v8, (a1)
-; RV64-NEXT: vsetvli a0, zero, e64, m8, ta, mu
-; RV64-NEXT: vluxei64.v v24, (zero), v16, v0.t
+; RV64-NEXT: vl8re64.v v24, (a0)
; RV64-NEXT: csrr a0, vlenb
+; RV64-NEXT: vl8re64.v v8, (a1)
; RV64-NEXT: srli a1, a0, 3
; RV64-NEXT: vsetvli a3, zero, e8, mf4, ta, ma
-; RV64-NEXT: vslidedown.vx v0, v0, a1
+; RV64-NEXT: vslidedown.vx v7, v0, a1
; RV64-NEXT: vsetvli a1, zero, e64, m8, ta, mu
+; RV64-NEXT: vluxei64.v v24, (zero), v16, v0.t
+; RV64-NEXT: vmv1r.v v0, v7
; RV64-NEXT: addi a1, sp, 16
; RV64-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
; RV64-NEXT: vluxei64.v v8, (zero), v16, v0.t
@@ -2131,8 +2131,8 @@ define <vscale x 32 x i8> @mgather_baseidx_nxv32i8(ptr %base, <vscale x 32 x i8>
; RV64-NEXT: vluxei64.v v15, (a0), v16, v0.t
; RV64-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; RV64-NEXT: vsext.vf8 v16, v10
-; RV64-NEXT: vsetvli zero, zero, e8, m1, ta, mu
; RV64-NEXT: vmv1r.v v0, v8
+; RV64-NEXT: vsetvli zero, zero, e8, m1, ta, mu
; RV64-NEXT: vluxei64.v v14, (a0), v16, v0.t
; RV64-NEXT: vmv4r.v v8, v12
; RV64-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/mscatter-sdnode.ll b/llvm/test/CodeGen/RISCV/rvv/mscatter-sdnode.ll
index 0e09f59b6a20f..9bfa0f31dc3a6 100644
--- a/llvm/test/CodeGen/RISCV/rvv/mscatter-sdnode.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/mscatter-sdnode.ll
@@ -1691,15 +1691,15 @@ declare <vscale x 16 x ptr> @llvm.vector.insert.nxv8p0.nxv16p0(<vscale x 16 x pt
define void @mscatter_nxv16f64(<vscale x 8 x double> %val0, <vscale x 8 x double> %val1, <vscale x 8 x ptr> %ptrs0, <vscale x 8 x ptr> %ptrs1, <vscale x 16 x i1> %m) {
; RV32-LABEL: mscatter_nxv16f64:
; RV32: # %bb.0:
-; RV32-NEXT: vl4re32.v v24, (a0)
; RV32-NEXT: vl4re32.v v28, (a1)
-; RV32-NEXT: vsetvli a0, zero, e64, m8, ta, ma
-; RV32-NEXT: vsoxei32.v v8, (zero), v24, v0.t
+; RV32-NEXT: vl4re32.v v4, (a0)
; RV32-NEXT: csrr a0, vlenb
; RV32-NEXT: srli a0, a0, 3
; RV32-NEXT: vsetvli a1, zero, e8, mf4, ta, ma
-; RV32-NEXT: vslidedown.vx v0, v0, a0
+; RV32-NEXT: vslidedown.vx v24, v0, a0
; RV32-NEXT: vsetvli a0, zero, e64, m8, ta, ma
+; RV32-NEXT: vsoxei32.v v8, (zero), v4, v0.t
+; RV32-NEXT: vmv1r.v v0, v24
; RV32-NEXT: vsoxei32.v v16, (zero), v28, v0.t
; RV32-NEXT: ret
;
@@ -1708,25 +1708,36 @@ define void @mscatter_nxv16f64(<vscale x 8 x double> %val0, <vscale x 8 x double
; RV64-NEXT: addi sp, sp, -16
; RV64-NEXT: .cfi_def_cfa_offset 16
; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: slli a2, a2, 3
+; RV64-NEXT: slli a2, a2, 4
; RV64-NEXT: sub sp, sp, a2
-; RV64-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
-; RV64-NEXT: vl8re64.v v24, (a0)
-; RV64-NEXT: addi a0, sp, 16
-; RV64-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
-; RV64-NEXT: vl8re64.v v16, (a1)
-; RV64-NEXT: vsetvli a0, zero, e64, m8, ta, ma
-; RV64-NEXT: vsoxei64.v v8, (zero), v24, v0.t
+; RV64-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
+; RV64-NEXT: csrr a2, vlenb
+; RV64-NEXT: slli a2, a2, 3
+; RV64-NEXT: add a2, sp, a2
+; RV64-NEXT: addi a2, a2, 16
+; RV64-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
+; RV64-NEXT: vmv8r.v v16, v8
+; RV64-NEXT: vl8re64.v v8, (a1)
+; RV64-NEXT: addi a1, sp, 16
+; RV64-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
+; RV64-NEXT: vl8re64.v v8, (a0)
; RV64-NEXT: csrr a0, vlenb
; RV64-NEXT: srli a0, a0, 3
; RV64-NEXT: vsetvli a1, zero, e8, mf4, ta, ma
-; RV64-NEXT: vslidedown.vx v0, v0, a0
+; RV64-NEXT: vslidedown.vx v24, v0, a0
; RV64-NEXT: vsetvli a0, zero, e64, m8, ta, ma
-; RV64-NEXT: addi a0, sp, 16
+; RV64-NEXT: vsoxei64.v v16, (zero), v8, v0.t
+; RV64-NEXT: vmv1r.v v0, v24
+; RV64-NEXT: csrr a0, vlenb
+; RV64-NEXT: slli a0, a0, 3
+; RV64-NEXT: add a0, sp, a0
+; RV64-NEXT: addi a0, a0, 16
; RV64-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; RV64-NEXT: addi a0, sp, 16
+; RV64-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; RV64-NEXT: vsoxei64.v v8, (zero), v16, v0.t
; RV64-NEXT: csrr a0, vlenb
-; RV64-NEXT: slli a0, a0, 3
+; RV64-NEXT: slli a0, a0, 4
; RV64-NEXT: add sp, sp, a0
; RV64-NEXT: addi sp, sp, 16
; RV64-NEXT: ret
@@ -1745,13 +1756,13 @@ define void @mscatter_baseidx_nxv16i8_nxv16f64(<vscale x 8 x double> %val0, <vsc
; RV32-NEXT: vsetvli a1, zero, e32, m8, ta, ma
; RV32-NEXT: vsext.vf4 v24, v6
; RV32-NEXT: vsll.vi v24, v24, 3
-; RV32-NEXT: vsetvli a1, zero, e64, m8, ta, ma
-; RV32-NEXT: vsoxei32.v v8, (a0), v24, v0.t
; RV32-NEXT: csrr a1, vlenb
; RV32-NEXT: srli a1, a1, 3
; RV32-NEXT: vsetvli a2, zero, e8, mf4, ta, ma
-; RV32-NEXT: vslidedown.vx v0, v0, a1
+; RV32-NEXT: vslidedown.vx v7, v0, a1
; RV32-NEXT: vsetvli a1, zero, e64, m8, ta, ma
+; RV32-NEXT: vsoxei32.v v8, (a0), v24, v0.t
+; RV32-NEXT: vmv1r.v v0, v7
; RV32-NEXT: vsoxei32.v v16, (a0), v28, v0.t
; RV32-NEXT: ret
;
@@ -1763,12 +1774,12 @@ define void @mscatter_baseidx_nxv16i8_nxv16f64(<vscale x 8 x double> %val0, <vsc
; RV64-NEXT: vsll.vi v24, v24, 3
; RV64-NEXT: vsoxei64.v v8, (a0), v24, v0.t
; RV64-NEXT: vsext.vf8 v8, v7
-; RV64-NEXT: vsll.vi v8, v8, 3
; RV64-NEXT: csrr a1, vlenb
; RV64-NEXT: srli a1, a1, 3
; RV64-NEXT: vsetvli a2, zero, e8, mf4, ta, ma
; RV64-NEXT: vslidedown.vx v0, v0, a1
; RV64-NEXT: vsetvli a1, zero, e64, m8, ta, ma
+; RV64-NEXT: vsll.vi v8, v8, 3
; RV64-NEXT: vsoxei64.v v16, (a0), v8, v0.t
; RV64-NEXT: ret
%ptrs = getelementptr inbounds double, ptr %base, <vscale x 16 x i8> %idxs
@@ -1785,13 +1796,13 @@ define void @mscatter_baseidx_nxv16i16_nxv16f64(<vscale x 8 x double> %val0, <vs
; RV32-NEXT: vsetvli a1, zero, e32, m8, ta, ma
; RV32-NEXT: vsext.vf2 v24, v4
; RV32-NEXT: vsll.vi v24, v24, 3
-; RV32-NEXT: vsetvli a1, zero, e64, m8, ta, ma
-; RV32-NEXT: vsoxei32.v v8, (a0), v24, v0.t
; RV32-NEXT: csrr a1, vlenb
; RV32-NEXT: srli a1, a1, 3
; RV32-NEXT: vsetvli a2, zero, e8, mf4, ta, ma
-; RV32-NEXT: vslidedown.vx v0, v0, a1
+; RV32-NEXT: vslidedown.vx v7, v0, a1
; RV32-NEXT: vsetvli a1, zero, e64, m8, ta, ma
+; RV32-NEXT: vsoxei32.v v8, (a0), v24, v0.t
+; RV32-NEXT: vmv1r.v v0, v7
; RV32-NEXT: vsoxei32.v v16, (a0), v28, v0.t
; RV32-NEXT: ret
;
@@ -1803,12 +1814,12 @@ define void @mscatter_baseidx_nxv16i16_nxv16f64(<vscale x 8 x double> %val0, <vs
; RV64-NEXT: vsll.vi v24, v24, 3
; RV64-NEXT: vsoxei64.v v8, (a0), v24, v0.t
; RV64-NEXT: vsext.vf4 v8, v6
-; RV64-NEXT: vsll.vi v8, v8, 3
; RV64-NEXT: csrr a1, vlenb
; RV64-NEXT: srli a1, a1, 3
; RV64-NEXT: vsetvli a2, zero, e8, mf4, ta, ma
; RV64-NEXT: vslidedown.vx v0, v0, a1
; RV64-NEXT: vsetvli a1, zero, e64, m8, ta, ma
+; RV64-NEXT: vsll.vi v8, v8, 3
; RV64-NEXT: vsoxei64.v v16, (a0), v8, v0.t
; RV64-NEXT: ret
%ptrs = getelementptr inbounds double, ptr %base, <vscale x 16 x i16> %idxs
diff --git a/llvm/test/CodeGen/RISCV/rvv/nearbyint-vp.ll b/llvm/test/CodeGen/RISCV/rvv/nearbyint-vp.ll
index 8bc2334282653..a3ea462b6a737 100644
--- a/llvm/test/CodeGen/RISCV/rvv/nearbyint-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/nearbyint-vp.ll
@@ -23,9 +23,9 @@ define <vscale x 1 x half> @vp_nearbyint_nxv1f16(<vscale x 1 x half> %va, <vscal
; ZVFH-NEXT: vsetvli zero, zero, e16, mf4, ta, ma
; ZVFH-NEXT: vfcvt.x.f.v v9, v8, v0.t
; ZVFH-NEXT: vfcvt.f.x.v v9, v9, v0.t
-; ZVFH-NEXT: fsflags a0
; ZVFH-NEXT: vsetvli zero, zero, e16, mf4, ta, mu
; ZVFH-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; ZVFH-NEXT: fsflags a0
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vp_nearbyint_nxv1f16:
@@ -42,11 +42,11 @@ define <vscale x 1 x half> @vp_nearbyint_nxv1f16(<vscale x 1 x half> %va, <vscal
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, mf2, ta, ma
; ZVFHMIN-NEXT: vfcvt.x.f.v v8, v9, v0.t
; ZVFHMIN-NEXT: vfcvt.f.x.v v8, v8, v0.t
-; ZVFHMIN-NEXT: fsflags a0
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, mf2, ta, mu
; ZVFHMIN-NEXT: vfsgnj.vv v9, v8, v9, v0.t
-; ZVFHMIN-NEXT: vsetvli a0, zero, e16, mf4, ta, ma
+; ZVFHMIN-NEXT: vsetvli a1, zero, e16, mf4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v9
+; ZVFHMIN-NEXT: fsflags a0
; ZVFHMIN-NEXT: ret
%v = call <vscale x 1 x half> @llvm.vp.nearbyint.nxv1f16(<vscale x 1 x half> %va, <vscale x 1 x i1> %m, i32 %evl)
ret <vscale x 1 x half> %v
@@ -63,9 +63,9 @@ define <vscale x 1 x half> @vp_nearbyint_nxv1f16_unmasked(<vscale x 1 x half> %v
; ZVFH-NEXT: frflags a0
; ZVFH-NEXT: vfcvt.x.f.v v9, v8, v0.t
; ZVFH-NEXT: vfcvt.f.x.v v9, v9, v0.t
-; ZVFH-NEXT: fsflags a0
; ZVFH-NEXT: vsetvli zero, zero, e16, mf4, ta, mu
; ZVFH-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; ZVFH-NEXT: fsflags a0
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vp_nearbyint_nxv1f16_unmasked:
@@ -80,11 +80,11 @@ define <vscale x 1 x half> @vp_nearbyint_nxv1f16_unmasked(<vscale x 1 x half> %v
; ZVFHMIN-NEXT: frflags a0
; ZVFHMIN-NEXT: vfcvt.x.f.v v8, v9, v0.t
; ZVFHMIN-NEXT: vfcvt.f.x.v v8, v8, v0.t
-; ZVFHMIN-NEXT: fsflags a0
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, mf2, ta, mu
; ZVFHMIN-NEXT: vfsgnj.vv v9, v8, v9, v0.t
-; ZVFHMIN-NEXT: vsetvli a0, zero, e16, mf4, ta, ma
+; ZVFHMIN-NEXT: vsetvli a1, zero, e16, mf4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v9
+; ZVFHMIN-NEXT: fsflags a0
; ZVFHMIN-NEXT: ret
%v = call <vscale x 1 x half> @llvm.vp.nearbyint.nxv1f16(<vscale x 1 x half> %va, <vscale x 1 x i1> splat (i1 true), i32 %evl)
ret <vscale x 1 x half> %v
@@ -105,9 +105,9 @@ define <vscale x 2 x half> @vp_nearbyint_nxv2f16(<vscale x 2 x half> %va, <vscal
; ZVFH-NEXT: vsetvli zero, zero, e16, mf2, ta, ma
; ZVFH-NEXT: vfcvt.x.f.v v9, v8, v0.t
; ZVFH-NEXT: vfcvt.f.x.v v9, v9, v0.t
-; ZVFH-NEXT: fsflags a0
; ZVFH-NEXT: vsetvli zero, zero, e16, mf2, ta, mu
; ZVFH-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; ZVFH-NEXT: fsflags a0
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vp_nearbyint_nxv2f16:
@@ -124,11 +124,11 @@ define <vscale x 2 x half> @vp_nearbyint_nxv2f16(<vscale x 2 x half> %va, <vscal
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m1, ta, ma
; ZVFHMIN-NEXT: vfcvt.x.f.v v8, v9, v0.t
; ZVFHMIN-NEXT: vfcvt.f.x.v v8, v8, v0.t
-; ZVFHMIN-NEXT: fsflags a0
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m1, ta, mu
; ZVFHMIN-NEXT: vfsgnj.vv v9, v8, v9, v0.t
-; ZVFHMIN-NEXT: vsetvli a0, zero, e16, mf2, ta, ma
+; ZVFHMIN-NEXT: vsetvli a1, zero, e16, mf2, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v9
+; ZVFHMIN-NEXT: fsflags a0
; ZVFHMIN-NEXT: ret
%v = call <vscale x 2 x half> @llvm.vp.nearbyint.nxv2f16(<vscale x 2 x half> %va, <vscale x 2 x i1> %m, i32 %evl)
ret <vscale x 2 x half> %v
@@ -145,9 +145,9 @@ define <vscale x 2 x half> @vp_nearbyint_nxv2f16_unmasked(<vscale x 2 x half> %v
; ZVFH-NEXT: frflags a0
; ZVFH-NEXT: vfcvt.x.f.v v9, v8, v0.t
; ZVFH-NEXT: vfcvt.f.x.v v9, v9, v0.t
-; ZVFH-NEXT: fsflags a0
; ZVFH-NEXT: vsetvli zero, zero, e16, mf2, ta, mu
; ZVFH-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; ZVFH-NEXT: fsflags a0
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vp_nearbyint_nxv2f16_unmasked:
@@ -162,11 +162,11 @@ define <vscale x 2 x half> @vp_nearbyint_nxv2f16_unmasked(<vscale x 2 x half> %v
; ZVFHMIN-NEXT: frflags a0
; ZVFHMIN-NEXT: vfcvt.x.f.v v8, v9, v0.t
; ZVFHMIN-NEXT: vfcvt.f.x.v v8, v8, v0.t
-; ZVFHMIN-NEXT: fsflags a0
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m1, ta, mu
; ZVFHMIN-NEXT: vfsgnj.vv v9, v8, v9, v0.t
-; ZVFHMIN-NEXT: vsetvli a0, zero, e16, mf2, ta, ma
+; ZVFHMIN-NEXT: vsetvli a1, zero, e16, mf2, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v9
+; ZVFHMIN-NEXT: fsflags a0
; ZVFHMIN-NEXT: ret
%v = call <vscale x 2 x half> @llvm.vp.nearbyint.nxv2f16(<vscale x 2 x half> %va, <vscale x 2 x i1> splat (i1 true), i32 %evl)
ret <vscale x 2 x half> %v
@@ -187,9 +187,9 @@ define <vscale x 4 x half> @vp_nearbyint_nxv4f16(<vscale x 4 x half> %va, <vscal
; ZVFH-NEXT: vsetvli zero, zero, e16, m1, ta, ma
; ZVFH-NEXT: vfcvt.x.f.v v9, v8, v0.t
; ZVFH-NEXT: vfcvt.f.x.v v9, v9, v0.t
-; ZVFH-NEXT: fsflags a0
; ZVFH-NEXT: vsetvli zero, zero, e16, m1, ta, mu
; ZVFH-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; ZVFH-NEXT: fsflags a0
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vp_nearbyint_nxv4f16:
@@ -204,15 +204,15 @@ define <vscale x 4 x half> @vp_nearbyint_nxv4f16(<vscale x 4 x half> %va, <vscal
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m2, ta, mu
; ZVFHMIN-NEXT: vmflt.vf v9, v12, fa5, v0.t
; ZVFHMIN-NEXT: frflags a0
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v9
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; ZVFHMIN-NEXT: vfcvt.x.f.v v12, v10, v0.t
; ZVFHMIN-NEXT: vfcvt.f.x.v v12, v12, v0.t
-; ZVFHMIN-NEXT: fsflags a0
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m2, ta, mu
; ZVFHMIN-NEXT: vfsgnj.vv v10, v12, v10, v0.t
-; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m1, ta, ma
+; ZVFHMIN-NEXT: vsetvli a1, zero, e16, m1, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v10
+; ZVFHMIN-NEXT: fsflags a0
; ZVFHMIN-NEXT: ret
%v = call <vscale x 4 x half> @llvm.vp.nearbyint.nxv4f16(<vscale x 4 x half> %va, <vscale x 4 x i1> %m, i32 %evl)
ret <vscale x 4 x half> %v
@@ -229,9 +229,9 @@ define <vscale x 4 x half> @vp_nearbyint_nxv4f16_unmasked(<vscale x 4 x half> %v
; ZVFH-NEXT: frflags a0
; ZVFH-NEXT: vfcvt.x.f.v v9, v8, v0.t
; ZVFH-NEXT: vfcvt.f.x.v v9, v9, v0.t
-; ZVFH-NEXT: fsflags a0
; ZVFH-NEXT: vsetvli zero, zero, e16, m1, ta, mu
; ZVFH-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; ZVFH-NEXT: fsflags a0
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vp_nearbyint_nxv4f16_unmasked:
@@ -246,11 +246,11 @@ define <vscale x 4 x half> @vp_nearbyint_nxv4f16_unmasked(<vscale x 4 x half> %v
; ZVFHMIN-NEXT: frflags a0
; ZVFHMIN-NEXT: vfcvt.x.f.v v8, v10, v0.t
; ZVFHMIN-NEXT: vfcvt.f.x.v v8, v8, v0.t
-; ZVFHMIN-NEXT: fsflags a0
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m2, ta, mu
; ZVFHMIN-NEXT: vfsgnj.vv v10, v8, v10, v0.t
-; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m1, ta, ma
+; ZVFHMIN-NEXT: vsetvli a1, zero, e16, m1, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v10
+; ZVFHMIN-NEXT: fsflags a0
; ZVFHMIN-NEXT: ret
%v = call <vscale x 4 x half> @llvm.vp.nearbyint.nxv4f16(<vscale x 4 x half> %va, <vscale x 4 x i1> splat (i1 true), i32 %evl)
ret <vscale x 4 x half> %v
@@ -261,21 +261,21 @@ declare <vscale x 8 x half> @llvm.vp.nearbyint.nxv8f16(<vscale x 8 x half>, <vsc
define <vscale x 8 x half> @vp_nearbyint_nxv8f16(<vscale x 8 x half> %va, <vscale x 8 x i1> %m, i32 zeroext %evl) {
; ZVFH-LABEL: vp_nearbyint_nxv8f16:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vmv1r.v v10, v0
; ZVFH-NEXT: lui a1, %hi(.LCPI6_0)
; ZVFH-NEXT: flh fa5, %lo(.LCPI6_0)(a1)
+; ZVFH-NEXT: vmv1r.v v10, v0
; ZVFH-NEXT: vsetvli zero, a0, e16, m2, ta, ma
; ZVFH-NEXT: vfabs.v v12, v8, v0.t
; ZVFH-NEXT: vsetvli zero, zero, e16, m2, ta, mu
; ZVFH-NEXT: vmflt.vf v10, v12, fa5, v0.t
; ZVFH-NEXT: frflags a0
-; ZVFH-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; ZVFH-NEXT: vmv1r.v v0, v10
+; ZVFH-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; ZVFH-NEXT: vfcvt.x.f.v v12, v8, v0.t
; ZVFH-NEXT: vfcvt.f.x.v v12, v12, v0.t
-; ZVFH-NEXT: fsflags a0
; ZVFH-NEXT: vsetvli zero, zero, e16, m2, ta, mu
; ZVFH-NEXT: vfsgnj.vv v8, v12, v8, v0.t
+; ZVFH-NEXT: fsflags a0
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vp_nearbyint_nxv8f16:
@@ -290,15 +290,15 @@ define <vscale x 8 x half> @vp_nearbyint_nxv8f16(<vscale x 8 x half> %va, <vscal
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m4, ta, mu
; ZVFHMIN-NEXT: vmflt.vf v10, v16, fa5, v0.t
; ZVFHMIN-NEXT: frflags a0
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v10
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; ZVFHMIN-NEXT: vfcvt.x.f.v v16, v12, v0.t
; ZVFHMIN-NEXT: vfcvt.f.x.v v16, v16, v0.t
-; ZVFHMIN-NEXT: fsflags a0
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m4, ta, mu
; ZVFHMIN-NEXT: vfsgnj.vv v12, v16, v12, v0.t
-; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m2, ta, ma
+; ZVFHMIN-NEXT: vsetvli a1, zero, e16, m2, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v12
+; ZVFHMIN-NEXT: fsflags a0
; ZVFHMIN-NEXT: ret
%v = call <vscale x 8 x half> @llvm.vp.nearbyint.nxv8f16(<vscale x 8 x half> %va, <vscale x 8 x i1> %m, i32 %evl)
ret <vscale x 8 x half> %v
@@ -315,9 +315,9 @@ define <vscale x 8 x half> @vp_nearbyint_nxv8f16_unmasked(<vscale x 8 x half> %v
; ZVFH-NEXT: frflags a0
; ZVFH-NEXT: vfcvt.x.f.v v10, v8, v0.t
; ZVFH-NEXT: vfcvt.f.x.v v10, v10, v0.t
-; ZVFH-NEXT: fsflags a0
; ZVFH-NEXT: vsetvli zero, zero, e16, m2, ta, mu
; ZVFH-NEXT: vfsgnj.vv v8, v10, v8, v0.t
+; ZVFH-NEXT: fsflags a0
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vp_nearbyint_nxv8f16_unmasked:
@@ -332,11 +332,11 @@ define <vscale x 8 x half> @vp_nearbyint_nxv8f16_unmasked(<vscale x 8 x half> %v
; ZVFHMIN-NEXT: frflags a0
; ZVFHMIN-NEXT: vfcvt.x.f.v v8, v12, v0.t
; ZVFHMIN-NEXT: vfcvt.f.x.v v8, v8, v0.t
-; ZVFHMIN-NEXT: fsflags a0
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m4, ta, mu
; ZVFHMIN-NEXT: vfsgnj.vv v12, v8, v12, v0.t
-; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m2, ta, ma
+; ZVFHMIN-NEXT: vsetvli a1, zero, e16, m2, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v12
+; ZVFHMIN-NEXT: fsflags a0
; ZVFHMIN-NEXT: ret
%v = call <vscale x 8 x half> @llvm.vp.nearbyint.nxv8f16(<vscale x 8 x half> %va, <vscale x 8 x i1> splat (i1 true), i32 %evl)
ret <vscale x 8 x half> %v
@@ -347,21 +347,21 @@ declare <vscale x 16 x half> @llvm.vp.nearbyint.nxv16f16(<vscale x 16 x half>, <
define <vscale x 16 x half> @vp_nearbyint_nxv16f16(<vscale x 16 x half> %va, <vscale x 16 x i1> %m, i32 zeroext %evl) {
; ZVFH-LABEL: vp_nearbyint_nxv16f16:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vmv1r.v v12, v0
; ZVFH-NEXT: lui a1, %hi(.LCPI8_0)
; ZVFH-NEXT: flh fa5, %lo(.LCPI8_0)(a1)
+; ZVFH-NEXT: vmv1r.v v12, v0
; ZVFH-NEXT: vsetvli zero, a0, e16, m4, ta, ma
; ZVFH-NEXT: vfabs.v v16, v8, v0.t
; ZVFH-NEXT: vsetvli zero, zero, e16, m4, ta, mu
; ZVFH-NEXT: vmflt.vf v12, v16, fa5, v0.t
; ZVFH-NEXT: frflags a0
-; ZVFH-NEXT: vsetvli zero, zero, e16, m4, ta, ma
; ZVFH-NEXT: vmv1r.v v0, v12
+; ZVFH-NEXT: vsetvli zero, zero, e16, m4, ta, ma
; ZVFH-NEXT: vfcvt.x.f.v v16, v8, v0.t
; ZVFH-NEXT: vfcvt.f.x.v v16, v16, v0.t
-; ZVFH-NEXT: fsflags a0
; ZVFH-NEXT: vsetvli zero, zero, e16, m4, ta, mu
; ZVFH-NEXT: vfsgnj.vv v8, v16, v8, v0.t
+; ZVFH-NEXT: fsflags a0
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vp_nearbyint_nxv16f16:
@@ -376,15 +376,15 @@ define <vscale x 16 x half> @vp_nearbyint_nxv16f16(<vscale x 16 x half> %va, <vs
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, mu
; ZVFHMIN-NEXT: vmflt.vf v12, v24, fa5, v0.t
; ZVFHMIN-NEXT: frflags a0
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v12
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfcvt.x.f.v v24, v16, v0.t
; ZVFHMIN-NEXT: vfcvt.f.x.v v24, v24, v0.t
-; ZVFHMIN-NEXT: fsflags a0
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, mu
; ZVFHMIN-NEXT: vfsgnj.vv v16, v24, v16, v0.t
-; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
+; ZVFHMIN-NEXT: vsetvli a1, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v16
+; ZVFHMIN-NEXT: fsflags a0
; ZVFHMIN-NEXT: ret
%v = call <vscale x 16 x half> @llvm.vp.nearbyint.nxv16f16(<vscale x 16 x half> %va, <vscale x 16 x i1> %m, i32 %evl)
ret <vscale x 16 x half> %v
@@ -401,9 +401,9 @@ define <vscale x 16 x half> @vp_nearbyint_nxv16f16_unmasked(<vscale x 16 x half>
; ZVFH-NEXT: frflags a0
; ZVFH-NEXT: vfcvt.x.f.v v12, v8, v0.t
; ZVFH-NEXT: vfcvt.f.x.v v12, v12, v0.t
-; ZVFH-NEXT: fsflags a0
; ZVFH-NEXT: vsetvli zero, zero, e16, m4, ta, mu
; ZVFH-NEXT: vfsgnj.vv v8, v12, v8, v0.t
+; ZVFH-NEXT: fsflags a0
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vp_nearbyint_nxv16f16_unmasked:
@@ -418,11 +418,11 @@ define <vscale x 16 x half> @vp_nearbyint_nxv16f16_unmasked(<vscale x 16 x half>
; ZVFHMIN-NEXT: frflags a0
; ZVFHMIN-NEXT: vfcvt.x.f.v v8, v16, v0.t
; ZVFHMIN-NEXT: vfcvt.f.x.v v8, v8, v0.t
-; ZVFHMIN-NEXT: fsflags a0
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, mu
; ZVFHMIN-NEXT: vfsgnj.vv v16, v8, v16, v0.t
-; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
+; ZVFHMIN-NEXT: vsetvli a1, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v16
+; ZVFHMIN-NEXT: fsflags a0
; ZVFHMIN-NEXT: ret
%v = call <vscale x 16 x half> @llvm.vp.nearbyint.nxv16f16(<vscale x 16 x half> %va, <vscale x 16 x i1> splat (i1 true), i32 %evl)
ret <vscale x 16 x half> %v
@@ -433,21 +433,21 @@ declare <vscale x 32 x half> @llvm.vp.nearbyint.nxv32f16(<vscale x 32 x half>, <
define <vscale x 32 x half> @vp_nearbyint_nxv32f16(<vscale x 32 x half> %va, <vscale x 32 x i1> %m, i32 zeroext %evl) {
; ZVFH-LABEL: vp_nearbyint_nxv32f16:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vmv1r.v v16, v0
; ZVFH-NEXT: lui a1, %hi(.LCPI10_0)
; ZVFH-NEXT: flh fa5, %lo(.LCPI10_0)(a1)
+; ZVFH-NEXT: vmv1r.v v16, v0
; ZVFH-NEXT: vsetvli zero, a0, e16, m8, ta, ma
; ZVFH-NEXT: vfabs.v v24, v8, v0.t
; ZVFH-NEXT: vsetvli zero, zero, e16, m8, ta, mu
; ZVFH-NEXT: vmflt.vf v16, v24, fa5, v0.t
; ZVFH-NEXT: frflags a0
-; ZVFH-NEXT: vsetvli zero, zero, e16, m8, ta, ma
; ZVFH-NEXT: vmv1r.v v0, v16
+; ZVFH-NEXT: vsetvli zero, zero, e16, m8, ta, ma
; ZVFH-NEXT: vfcvt.x.f.v v24, v8, v0.t
; ZVFH-NEXT: vfcvt.f.x.v v24, v24, v0.t
-; ZVFH-NEXT: fsflags a0
; ZVFH-NEXT: vsetvli zero, zero, e16, m8, ta, mu
; ZVFH-NEXT: vfsgnj.vv v8, v24, v8, v0.t
+; ZVFH-NEXT: fsflags a0
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vp_nearbyint_nxv32f16:
@@ -468,20 +468,20 @@ define <vscale x 32 x half> @vp_nearbyint_nxv32f16(<vscale x 32 x half> %va, <vs
; ZVFHMIN-NEXT: srli a2, a2, 2
; ZVFHMIN-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
; ZVFHMIN-NEXT: vslidedown.vx v17, v0, a2
-; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
-; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v17
+; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfabs.v v8, v24, v0.t
; ZVFHMIN-NEXT: lui a2, 307200
; ZVFHMIN-NEXT: fmv.w.x fa5, a2
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, mu
; ZVFHMIN-NEXT: vmflt.vf v17, v8, fa5, v0.t
; ZVFHMIN-NEXT: frflags a2
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v17
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfcvt.x.f.v v8, v24, v0.t
; ZVFHMIN-NEXT: vfcvt.f.x.v v8, v8, v0.t
; ZVFHMIN-NEXT: fsflags a2
@@ -496,21 +496,21 @@ define <vscale x 32 x half> @vp_nearbyint_nxv32f16(<vscale x 32 x half> %va, <vs
; ZVFHMIN-NEXT: addi a1, sp, 16
; ZVFHMIN-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v16
-; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v7
+; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfabs.v v16, v24, v0.t
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, mu
; ZVFHMIN-NEXT: vmflt.vf v7, v16, fa5, v0.t
; ZVFHMIN-NEXT: frflags a0
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v7
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfcvt.x.f.v v16, v24, v0.t
; ZVFHMIN-NEXT: vfcvt.f.x.v v16, v16, v0.t
-; ZVFHMIN-NEXT: fsflags a0
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, mu
; ZVFHMIN-NEXT: vfsgnj.vv v24, v16, v24, v0.t
-; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
+; ZVFHMIN-NEXT: vsetvli a1, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v24
+; ZVFHMIN-NEXT: fsflags a0
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: add sp, sp, a0
@@ -531,9 +531,9 @@ define <vscale x 32 x half> @vp_nearbyint_nxv32f16_unmasked(<vscale x 32 x half>
; ZVFH-NEXT: frflags a0
; ZVFH-NEXT: vfcvt.x.f.v v16, v8, v0.t
; ZVFH-NEXT: vfcvt.f.x.v v16, v16, v0.t
-; ZVFH-NEXT: fsflags a0
; ZVFH-NEXT: vsetvli zero, zero, e16, m8, ta, mu
; ZVFH-NEXT: vfsgnj.vv v8, v16, v8, v0.t
+; ZVFH-NEXT: fsflags a0
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vp_nearbyint_nxv32f16_unmasked:
@@ -555,20 +555,20 @@ define <vscale x 32 x half> @vp_nearbyint_nxv32f16_unmasked(<vscale x 32 x half>
; ZVFHMIN-NEXT: vmset.m v16
; ZVFHMIN-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
; ZVFHMIN-NEXT: vslidedown.vx v16, v16, a2
-; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
-; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v16
+; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfabs.v v8, v24, v0.t
; ZVFHMIN-NEXT: lui a2, 307200
; ZVFHMIN-NEXT: fmv.w.x fa5, a2
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, mu
; ZVFHMIN-NEXT: vmflt.vf v16, v8, fa5, v0.t
; ZVFHMIN-NEXT: frflags a2
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v16
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfcvt.x.f.v v8, v24, v0.t
; ZVFHMIN-NEXT: vfcvt.f.x.v v8, v8, v0.t
; ZVFHMIN-NEXT: fsflags a2
@@ -589,11 +589,11 @@ define <vscale x 32 x half> @vp_nearbyint_nxv32f16_unmasked(<vscale x 32 x half>
; ZVFHMIN-NEXT: frflags a0
; ZVFHMIN-NEXT: vfcvt.x.f.v v24, v16, v0.t
; ZVFHMIN-NEXT: vfcvt.f.x.v v24, v24, v0.t
-; ZVFHMIN-NEXT: fsflags a0
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, mu
; ZVFHMIN-NEXT: vfsgnj.vv v16, v24, v16, v0.t
-; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
+; ZVFHMIN-NEXT: vsetvli a1, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v16
+; ZVFHMIN-NEXT: fsflags a0
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: add sp, sp, a0
@@ -618,9 +618,9 @@ define <vscale x 1 x float> @vp_nearbyint_nxv1f32(<vscale x 1 x float> %va, <vsc
; CHECK-NEXT: vsetvli zero, zero, e32, mf2, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v9, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v9, v9, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e32, mf2, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <vscale x 1 x float> @llvm.vp.nearbyint.nxv1f32(<vscale x 1 x float> %va, <vscale x 1 x i1> %m, i32 %evl)
ret <vscale x 1 x float> %v
@@ -637,9 +637,9 @@ define <vscale x 1 x float> @vp_nearbyint_nxv1f32_unmasked(<vscale x 1 x float>
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v9, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v9, v9, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e32, mf2, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <vscale x 1 x float> @llvm.vp.nearbyint.nxv1f32(<vscale x 1 x float> %va, <vscale x 1 x i1> splat (i1 true), i32 %evl)
ret <vscale x 1 x float> %v
@@ -660,9 +660,9 @@ define <vscale x 2 x float> @vp_nearbyint_nxv2f32(<vscale x 2 x float> %va, <vsc
; CHECK-NEXT: vsetvli zero, zero, e32, m1, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v9, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v9, v9, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e32, m1, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <vscale x 2 x float> @llvm.vp.nearbyint.nxv2f32(<vscale x 2 x float> %va, <vscale x 2 x i1> %m, i32 %evl)
ret <vscale x 2 x float> %v
@@ -679,9 +679,9 @@ define <vscale x 2 x float> @vp_nearbyint_nxv2f32_unmasked(<vscale x 2 x float>
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v9, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v9, v9, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e32, m1, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <vscale x 2 x float> @llvm.vp.nearbyint.nxv2f32(<vscale x 2 x float> %va, <vscale x 2 x i1> splat (i1 true), i32 %evl)
ret <vscale x 2 x float> %v
@@ -700,13 +700,13 @@ define <vscale x 4 x float> @vp_nearbyint_nxv4f32(<vscale x 4 x float> %va, <vsc
; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, mu
; CHECK-NEXT: vmflt.vf v10, v12, fa5, v0.t
; CHECK-NEXT: frflags a0
-; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v12, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <vscale x 4 x float> @llvm.vp.nearbyint.nxv4f32(<vscale x 4 x float> %va, <vscale x 4 x i1> %m, i32 %evl)
ret <vscale x 4 x float> %v
@@ -723,9 +723,9 @@ define <vscale x 4 x float> @vp_nearbyint_nxv4f32_unmasked(<vscale x 4 x float>
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v10, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v10, v10, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v10, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <vscale x 4 x float> @llvm.vp.nearbyint.nxv4f32(<vscale x 4 x float> %va, <vscale x 4 x i1> splat (i1 true), i32 %evl)
ret <vscale x 4 x float> %v
@@ -744,13 +744,13 @@ define <vscale x 8 x float> @vp_nearbyint_nxv8f32(<vscale x 8 x float> %va, <vsc
; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, mu
; CHECK-NEXT: vmflt.vf v12, v16, fa5, v0.t
; CHECK-NEXT: frflags a0
-; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v16, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <vscale x 8 x float> @llvm.vp.nearbyint.nxv8f32(<vscale x 8 x float> %va, <vscale x 8 x i1> %m, i32 %evl)
ret <vscale x 8 x float> %v
@@ -767,9 +767,9 @@ define <vscale x 8 x float> @vp_nearbyint_nxv8f32_unmasked(<vscale x 8 x float>
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v12, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <vscale x 8 x float> @llvm.vp.nearbyint.nxv8f32(<vscale x 8 x float> %va, <vscale x 8 x i1> splat (i1 true), i32 %evl)
ret <vscale x 8 x float> %v
@@ -788,13 +788,13 @@ define <vscale x 16 x float> @vp_nearbyint_nxv16f32(<vscale x 16 x float> %va, <
; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, mu
; CHECK-NEXT: vmflt.vf v16, v24, fa5, v0.t
; CHECK-NEXT: frflags a0
-; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v24, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <vscale x 16 x float> @llvm.vp.nearbyint.nxv16f32(<vscale x 16 x float> %va, <vscale x 16 x i1> %m, i32 %evl)
ret <vscale x 16 x float> %v
@@ -811,9 +811,9 @@ define <vscale x 16 x float> @vp_nearbyint_nxv16f32_unmasked(<vscale x 16 x floa
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v16, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <vscale x 16 x float> @llvm.vp.nearbyint.nxv16f32(<vscale x 16 x float> %va, <vscale x 16 x i1> splat (i1 true), i32 %evl)
ret <vscale x 16 x float> %v
@@ -834,9 +834,9 @@ define <vscale x 1 x double> @vp_nearbyint_nxv1f64(<vscale x 1 x double> %va, <v
; CHECK-NEXT: vsetvli zero, zero, e64, m1, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v9, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v9, v9, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e64, m1, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <vscale x 1 x double> @llvm.vp.nearbyint.nxv1f64(<vscale x 1 x double> %va, <vscale x 1 x i1> %m, i32 %evl)
ret <vscale x 1 x double> %v
@@ -853,9 +853,9 @@ define <vscale x 1 x double> @vp_nearbyint_nxv1f64_unmasked(<vscale x 1 x double
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v9, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v9, v9, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e64, m1, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <vscale x 1 x double> @llvm.vp.nearbyint.nxv1f64(<vscale x 1 x double> %va, <vscale x 1 x i1> splat (i1 true), i32 %evl)
ret <vscale x 1 x double> %v
@@ -866,21 +866,21 @@ declare <vscale x 2 x double> @llvm.vp.nearbyint.nxv2f64(<vscale x 2 x double>,
define <vscale x 2 x double> @vp_nearbyint_nxv2f64(<vscale x 2 x double> %va, <vscale x 2 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_nearbyint_nxv2f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v10, v0
; CHECK-NEXT: lui a1, %hi(.LCPI24_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI24_0)(a1)
+; CHECK-NEXT: vmv1r.v v10, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, ma
; CHECK-NEXT: vfabs.v v12, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, mu
; CHECK-NEXT: vmflt.vf v10, v12, fa5, v0.t
; CHECK-NEXT: frflags a0
-; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v12, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <vscale x 2 x double> @llvm.vp.nearbyint.nxv2f64(<vscale x 2 x double> %va, <vscale x 2 x i1> %m, i32 %evl)
ret <vscale x 2 x double> %v
@@ -897,9 +897,9 @@ define <vscale x 2 x double> @vp_nearbyint_nxv2f64_unmasked(<vscale x 2 x double
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v10, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v10, v10, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v10, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <vscale x 2 x double> @llvm.vp.nearbyint.nxv2f64(<vscale x 2 x double> %va, <vscale x 2 x i1> splat (i1 true), i32 %evl)
ret <vscale x 2 x double> %v
@@ -910,21 +910,21 @@ declare <vscale x 4 x double> @llvm.vp.nearbyint.nxv4f64(<vscale x 4 x double>,
define <vscale x 4 x double> @vp_nearbyint_nxv4f64(<vscale x 4 x double> %va, <vscale x 4 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_nearbyint_nxv4f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v12, v0
; CHECK-NEXT: lui a1, %hi(.LCPI26_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI26_0)(a1)
+; CHECK-NEXT: vmv1r.v v12, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, ma
; CHECK-NEXT: vfabs.v v16, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, mu
; CHECK-NEXT: vmflt.vf v12, v16, fa5, v0.t
; CHECK-NEXT: frflags a0
-; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v16, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <vscale x 4 x double> @llvm.vp.nearbyint.nxv4f64(<vscale x 4 x double> %va, <vscale x 4 x i1> %m, i32 %evl)
ret <vscale x 4 x double> %v
@@ -941,9 +941,9 @@ define <vscale x 4 x double> @vp_nearbyint_nxv4f64_unmasked(<vscale x 4 x double
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v12, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <vscale x 4 x double> @llvm.vp.nearbyint.nxv4f64(<vscale x 4 x double> %va, <vscale x 4 x i1> splat (i1 true), i32 %evl)
ret <vscale x 4 x double> %v
@@ -954,21 +954,21 @@ declare <vscale x 7 x double> @llvm.vp.nearbyint.nxv7f64(<vscale x 7 x double>,
define <vscale x 7 x double> @vp_nearbyint_nxv7f64(<vscale x 7 x double> %va, <vscale x 7 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_nearbyint_nxv7f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: lui a1, %hi(.LCPI28_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI28_0)(a1)
+; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
; CHECK-NEXT: vmflt.vf v16, v24, fa5, v0.t
; CHECK-NEXT: frflags a0
-; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v24, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <vscale x 7 x double> @llvm.vp.nearbyint.nxv7f64(<vscale x 7 x double> %va, <vscale x 7 x i1> %m, i32 %evl)
ret <vscale x 7 x double> %v
@@ -985,9 +985,9 @@ define <vscale x 7 x double> @vp_nearbyint_nxv7f64_unmasked(<vscale x 7 x double
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v16, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <vscale x 7 x double> @llvm.vp.nearbyint.nxv7f64(<vscale x 7 x double> %va, <vscale x 7 x i1> splat (i1 true), i32 %evl)
ret <vscale x 7 x double> %v
@@ -998,21 +998,21 @@ declare <vscale x 8 x double> @llvm.vp.nearbyint.nxv8f64(<vscale x 8 x double>,
define <vscale x 8 x double> @vp_nearbyint_nxv8f64(<vscale x 8 x double> %va, <vscale x 8 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_nearbyint_nxv8f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: lui a1, %hi(.LCPI30_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI30_0)(a1)
+; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
; CHECK-NEXT: vmflt.vf v16, v24, fa5, v0.t
; CHECK-NEXT: frflags a0
-; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v24, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <vscale x 8 x double> @llvm.vp.nearbyint.nxv8f64(<vscale x 8 x double> %va, <vscale x 8 x i1> %m, i32 %evl)
ret <vscale x 8 x double> %v
@@ -1029,9 +1029,9 @@ define <vscale x 8 x double> @vp_nearbyint_nxv8f64_unmasked(<vscale x 8 x double
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v16, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <vscale x 8 x double> @llvm.vp.nearbyint.nxv8f64(<vscale x 8 x double> %va, <vscale x 8 x i1> splat (i1 true), i32 %evl)
ret <vscale x 8 x double> %v
@@ -1043,82 +1043,47 @@ declare <vscale x 16 x double> @llvm.vp.nearbyint.nxv16f64(<vscale x 16 x double
define <vscale x 16 x double> @vp_nearbyint_nxv16f64(<vscale x 16 x double> %va, <vscale x 16 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_nearbyint_nxv16f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: addi sp, sp, -16
-; CHECK-NEXT: .cfi_def_cfa_offset 16
-; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a1, a1, 4
-; CHECK-NEXT: sub sp, sp, a1
-; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
-; CHECK-NEXT: vmv1r.v v24, v0
-; CHECK-NEXT: addi a1, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
+; CHECK-NEXT: vmv1r.v v7, v0
; CHECK-NEXT: csrr a1, vlenb
; CHECK-NEXT: srli a2, a1, 3
; CHECK-NEXT: vsetvli a3, zero, e8, mf4, ta, ma
-; CHECK-NEXT: vslidedown.vx v25, v0, a2
+; CHECK-NEXT: vslidedown.vx v6, v0, a2
; CHECK-NEXT: sub a2, a0, a1
+; CHECK-NEXT: lui a3, %hi(.LCPI32_0)
+; CHECK-NEXT: fld fa5, %lo(.LCPI32_0)(a3)
; CHECK-NEXT: sltu a3, a0, a2
; CHECK-NEXT: addi a3, a3, -1
; CHECK-NEXT: and a2, a3, a2
-; CHECK-NEXT: lui a3, %hi(.LCPI32_0)
-; CHECK-NEXT: fld fa5, %lo(.LCPI32_0)(a3)
+; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: vsetvli zero, a2, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v25
-; CHECK-NEXT: vmv8r.v v8, v16
-; CHECK-NEXT: csrr a2, vlenb
-; CHECK-NEXT: slli a2, a2, 3
-; CHECK-NEXT: add a2, sp, a2
-; CHECK-NEXT: addi a2, a2, 16
-; CHECK-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
-; CHECK-NEXT: vfabs.v v16, v16, v0.t
+; CHECK-NEXT: vfabs.v v24, v16, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vmflt.vf v25, v16, fa5, v0.t
+; CHECK-NEXT: vmflt.vf v6, v24, fa5, v0.t
; CHECK-NEXT: frflags a2
+; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v25
-; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
-; CHECK-NEXT: vfcvt.f.x.v v8, v16, v0.t
+; CHECK-NEXT: vfcvt.x.f.v v24, v16, v0.t
+; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
; CHECK-NEXT: fsflags a2
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: csrr a2, vlenb
-; CHECK-NEXT: slli a2, a2, 3
-; CHECK-NEXT: add a2, sp, a2
-; CHECK-NEXT: addi a2, a2, 16
-; CHECK-NEXT: vl8r.v v16, (a2) # Unknown-size Folded Reload
-; CHECK-NEXT: vfsgnj.vv v16, v8, v16, v0.t
-; CHECK-NEXT: csrr a2, vlenb
-; CHECK-NEXT: slli a2, a2, 3
-; CHECK-NEXT: add a2, sp, a2
-; CHECK-NEXT: addi a2, a2, 16
-; CHECK-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
+; CHECK-NEXT: vfsgnj.vv v16, v24, v16, v0.t
; CHECK-NEXT: bltu a0, a1, .LBB32_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB32_2:
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v24
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vfabs.v v16, v8, v0.t
+; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vmflt.vf v24, v16, fa5, v0.t
+; CHECK-NEXT: vmflt.vf v7, v24, fa5, v0.t
; CHECK-NEXT: frflags a0
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v24
-; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
-; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
-; CHECK-NEXT: fsflags a0
+; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
+; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vfsgnj.vv v8, v16, v8, v0.t
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 4
-; CHECK-NEXT: add sp, sp, a0
-; CHECK-NEXT: addi sp, sp, 16
+; CHECK-NEXT: vfsgnj.vv v8, v24, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <vscale x 16 x double> @llvm.vp.nearbyint.nxv16f64(<vscale x 16 x double> %va, <vscale x 16 x i1> %m, i32 %evl)
ret <vscale x 16 x double> %v
@@ -1153,9 +1118,9 @@ define <vscale x 16 x double> @vp_nearbyint_nxv16f64_unmasked(<vscale x 16 x dou
; CHECK-NEXT: frflags a0
; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
-; CHECK-NEXT: fsflags a0
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
; CHECK-NEXT: vfsgnj.vv v8, v24, v8, v0.t
+; CHECK-NEXT: fsflags a0
; CHECK-NEXT: ret
%v = call <vscale x 16 x double> @llvm.vp.nearbyint.nxv16f64(<vscale x 16 x double> %va, <vscale x 16 x i1> splat (i1 true), i32 %evl)
ret <vscale x 16 x double> %v
diff --git a/llvm/test/CodeGen/RISCV/rvv/pr63596.ll b/llvm/test/CodeGen/RISCV/rvv/pr63596.ll
index d13d67fd0a882..8bb62eaa8e9e9 100644
--- a/llvm/test/CodeGen/RISCV/rvv/pr63596.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/pr63596.ll
@@ -27,20 +27,18 @@ define <4 x float> @foo(ptr %0) nounwind {
; CHECK-NEXT: fsw fa0, 0(sp)
; CHECK-NEXT: addi a0, sp, 4
; CHECK-NEXT: vsetivli zero, 1, e32, mf2, ta, ma
-; CHECK-NEXT: vle32.v v8, (a0)
-; CHECK-NEXT: addi a0, sp, 12
; CHECK-NEXT: vle32.v v9, (a0)
-; CHECK-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
-; CHECK-NEXT: vslideup.vi v9, v8, 1
-; CHECK-NEXT: addi a0, sp, 8
-; CHECK-NEXT: vsetivli zero, 1, e32, mf2, ta, ma
+; CHECK-NEXT: addi a0, sp, 12
; CHECK-NEXT: vle32.v v10, (a0)
+; CHECK-NEXT: addi a0, sp, 8
+; CHECK-NEXT: vle32.v v11, (a0)
; CHECK-NEXT: mv a0, sp
; CHECK-NEXT: vle32.v v8, (a0)
; CHECK-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
-; CHECK-NEXT: vslideup.vi v8, v10, 1
+; CHECK-NEXT: vslideup.vi v10, v9, 1
+; CHECK-NEXT: vslideup.vi v8, v11, 1
; CHECK-NEXT: vsetivli zero, 4, e32, m1, ta, ma
-; CHECK-NEXT: vslideup.vi v8, v9, 2
+; CHECK-NEXT: vslideup.vi v8, v10, 2
; CHECK-NEXT: ld ra, 40(sp) # 8-byte Folded Reload
; CHECK-NEXT: ld s0, 32(sp) # 8-byte Folded Reload
; CHECK-NEXT: ld s1, 24(sp) # 8-byte Folded Reload
diff --git a/llvm/test/CodeGen/RISCV/rvv/rint-vp.ll b/llvm/test/CodeGen/RISCV/rvv/rint-vp.ll
index f934127f978dc..88bd92c6ec161 100644
--- a/llvm/test/CodeGen/RISCV/rvv/rint-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/rint-vp.ll
@@ -185,8 +185,8 @@ define <vscale x 4 x half> @vp_rint_nxv4f16(<vscale x 4 x half> %va, <vscale x 4
; ZVFHMIN-NEXT: fmv.w.x fa5, a0
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m2, ta, mu
; ZVFHMIN-NEXT: vmflt.vf v9, v12, fa5, v0.t
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v9
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; ZVFHMIN-NEXT: vfcvt.x.f.v v12, v10, v0.t
; ZVFHMIN-NEXT: vfcvt.f.x.v v12, v12, v0.t
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m2, ta, mu
@@ -237,15 +237,15 @@ declare <vscale x 8 x half> @llvm.vp.rint.nxv8f16(<vscale x 8 x half>, <vscale x
define <vscale x 8 x half> @vp_rint_nxv8f16(<vscale x 8 x half> %va, <vscale x 8 x i1> %m, i32 zeroext %evl) {
; ZVFH-LABEL: vp_rint_nxv8f16:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vmv1r.v v10, v0
; ZVFH-NEXT: lui a1, %hi(.LCPI6_0)
; ZVFH-NEXT: flh fa5, %lo(.LCPI6_0)(a1)
+; ZVFH-NEXT: vmv1r.v v10, v0
; ZVFH-NEXT: vsetvli zero, a0, e16, m2, ta, ma
; ZVFH-NEXT: vfabs.v v12, v8, v0.t
; ZVFH-NEXT: vsetvli zero, zero, e16, m2, ta, mu
; ZVFH-NEXT: vmflt.vf v10, v12, fa5, v0.t
-; ZVFH-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; ZVFH-NEXT: vmv1r.v v0, v10
+; ZVFH-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; ZVFH-NEXT: vfcvt.x.f.v v12, v8, v0.t
; ZVFH-NEXT: vfcvt.f.x.v v12, v12, v0.t
; ZVFH-NEXT: vsetvli zero, zero, e16, m2, ta, mu
@@ -263,8 +263,8 @@ define <vscale x 8 x half> @vp_rint_nxv8f16(<vscale x 8 x half> %va, <vscale x 8
; ZVFHMIN-NEXT: fmv.w.x fa5, a0
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m4, ta, mu
; ZVFHMIN-NEXT: vmflt.vf v10, v16, fa5, v0.t
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v10
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; ZVFHMIN-NEXT: vfcvt.x.f.v v16, v12, v0.t
; ZVFHMIN-NEXT: vfcvt.f.x.v v16, v16, v0.t
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m4, ta, mu
@@ -315,15 +315,15 @@ declare <vscale x 16 x half> @llvm.vp.rint.nxv16f16(<vscale x 16 x half>, <vscal
define <vscale x 16 x half> @vp_rint_nxv16f16(<vscale x 16 x half> %va, <vscale x 16 x i1> %m, i32 zeroext %evl) {
; ZVFH-LABEL: vp_rint_nxv16f16:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vmv1r.v v12, v0
; ZVFH-NEXT: lui a1, %hi(.LCPI8_0)
; ZVFH-NEXT: flh fa5, %lo(.LCPI8_0)(a1)
+; ZVFH-NEXT: vmv1r.v v12, v0
; ZVFH-NEXT: vsetvli zero, a0, e16, m4, ta, ma
; ZVFH-NEXT: vfabs.v v16, v8, v0.t
; ZVFH-NEXT: vsetvli zero, zero, e16, m4, ta, mu
; ZVFH-NEXT: vmflt.vf v12, v16, fa5, v0.t
-; ZVFH-NEXT: vsetvli zero, zero, e16, m4, ta, ma
; ZVFH-NEXT: vmv1r.v v0, v12
+; ZVFH-NEXT: vsetvli zero, zero, e16, m4, ta, ma
; ZVFH-NEXT: vfcvt.x.f.v v16, v8, v0.t
; ZVFH-NEXT: vfcvt.f.x.v v16, v16, v0.t
; ZVFH-NEXT: vsetvli zero, zero, e16, m4, ta, mu
@@ -341,8 +341,8 @@ define <vscale x 16 x half> @vp_rint_nxv16f16(<vscale x 16 x half> %va, <vscale
; ZVFHMIN-NEXT: fmv.w.x fa5, a0
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, mu
; ZVFHMIN-NEXT: vmflt.vf v12, v24, fa5, v0.t
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v12
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfcvt.x.f.v v24, v16, v0.t
; ZVFHMIN-NEXT: vfcvt.f.x.v v24, v24, v0.t
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, mu
@@ -393,15 +393,15 @@ declare <vscale x 32 x half> @llvm.vp.rint.nxv32f16(<vscale x 32 x half>, <vscal
define <vscale x 32 x half> @vp_rint_nxv32f16(<vscale x 32 x half> %va, <vscale x 32 x i1> %m, i32 zeroext %evl) {
; ZVFH-LABEL: vp_rint_nxv32f16:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vmv1r.v v16, v0
; ZVFH-NEXT: lui a1, %hi(.LCPI10_0)
; ZVFH-NEXT: flh fa5, %lo(.LCPI10_0)(a1)
+; ZVFH-NEXT: vmv1r.v v16, v0
; ZVFH-NEXT: vsetvli zero, a0, e16, m8, ta, ma
; ZVFH-NEXT: vfabs.v v24, v8, v0.t
; ZVFH-NEXT: vsetvli zero, zero, e16, m8, ta, mu
; ZVFH-NEXT: vmflt.vf v16, v24, fa5, v0.t
-; ZVFH-NEXT: vsetvli zero, zero, e16, m8, ta, ma
; ZVFH-NEXT: vmv1r.v v0, v16
+; ZVFH-NEXT: vsetvli zero, zero, e16, m8, ta, ma
; ZVFH-NEXT: vfcvt.x.f.v v24, v8, v0.t
; ZVFH-NEXT: vfcvt.f.x.v v24, v24, v0.t
; ZVFH-NEXT: vsetvli zero, zero, e16, m8, ta, mu
@@ -426,46 +426,50 @@ define <vscale x 32 x half> @vp_rint_nxv32f16(<vscale x 32 x half> %va, <vscale
; ZVFHMIN-NEXT: srli a2, a2, 2
; ZVFHMIN-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
; ZVFHMIN-NEXT: vslidedown.vx v17, v0, a2
-; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
-; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v17
+; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfabs.v v8, v24, v0.t
; ZVFHMIN-NEXT: lui a2, 307200
; ZVFHMIN-NEXT: fmv.w.x fa5, a2
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, mu
; ZVFHMIN-NEXT: vmflt.vf v17, v8, fa5, v0.t
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v17
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfcvt.x.f.v v8, v24, v0.t
; ZVFHMIN-NEXT: vfcvt.f.x.v v8, v8, v0.t
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, mu
; ZVFHMIN-NEXT: vfsgnj.vv v24, v8, v24, v0.t
; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v20, v24
+; ZVFHMIN-NEXT: vfncvt.f.f.w v12, v24
; ZVFHMIN-NEXT: bltu a0, a1, .LBB10_2
; ZVFHMIN-NEXT: # %bb.1:
; ZVFHMIN-NEXT: mv a0, a1
; ZVFHMIN-NEXT: .LBB10_2:
; ZVFHMIN-NEXT: addi a1, sp, 16
-; ZVFHMIN-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v8
-; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
+; ZVFHMIN-NEXT: vl8r.v v0, (a1) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v0
; ZVFHMIN-NEXT: vmv1r.v v0, v16
-; ZVFHMIN-NEXT: vfabs.v v8, v24, v0.t
+; ZVFHMIN-NEXT: vmv1r.v v8, v16
+; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
+; ZVFHMIN-NEXT: vfabs.v v16, v24, v0.t
+; ZVFHMIN-NEXT: addi a0, sp, 16
+; ZVFHMIN-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vmv1r.v v0, v8
+; ZVFHMIN-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, mu
-; ZVFHMIN-NEXT: vmflt.vf v16, v8, fa5, v0.t
+; ZVFHMIN-NEXT: vmflt.vf v8, v16, fa5, v0.t
+; ZVFHMIN-NEXT: vmv1r.v v0, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vmv1r.v v0, v16
-; ZVFHMIN-NEXT: vfcvt.x.f.v v8, v24, v0.t
-; ZVFHMIN-NEXT: vfcvt.f.x.v v8, v8, v0.t
+; ZVFHMIN-NEXT: vfcvt.x.f.v v16, v24, v0.t
+; ZVFHMIN-NEXT: vfcvt.f.x.v v16, v16, v0.t
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, mu
-; ZVFHMIN-NEXT: vfsgnj.vv v24, v8, v24, v0.t
+; ZVFHMIN-NEXT: vfsgnj.vv v24, v16, v24, v0.t
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v16, v24
-; ZVFHMIN-NEXT: vmv8r.v v8, v16
+; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v24
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: add sp, sp, a0
@@ -508,19 +512,19 @@ define <vscale x 32 x half> @vp_rint_nxv32f16_unmasked(<vscale x 32 x half> %va,
; ZVFHMIN-NEXT: vmset.m v16
; ZVFHMIN-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
; ZVFHMIN-NEXT: vslidedown.vx v16, v16, a2
-; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
-; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v16
+; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfabs.v v8, v24, v0.t
; ZVFHMIN-NEXT: lui a2, 307200
; ZVFHMIN-NEXT: fmv.w.x fa5, a2
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, mu
; ZVFHMIN-NEXT: vmflt.vf v16, v8, fa5, v0.t
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v16
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfcvt.x.f.v v8, v24, v0.t
; ZVFHMIN-NEXT: vfcvt.f.x.v v8, v8, v0.t
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, mu
@@ -640,8 +644,8 @@ define <vscale x 4 x float> @vp_rint_nxv4f32(<vscale x 4 x float> %va, <vscale x
; CHECK-NEXT: fmv.w.x fa5, a0
; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, mu
; CHECK-NEXT: vmflt.vf v10, v12, fa5, v0.t
-; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, mu
@@ -680,8 +684,8 @@ define <vscale x 8 x float> @vp_rint_nxv8f32(<vscale x 8 x float> %va, <vscale x
; CHECK-NEXT: fmv.w.x fa5, a0
; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, mu
; CHECK-NEXT: vmflt.vf v12, v16, fa5, v0.t
-; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, mu
@@ -720,8 +724,8 @@ define <vscale x 16 x float> @vp_rint_nxv16f32(<vscale x 16 x float> %va, <vscal
; CHECK-NEXT: fmv.w.x fa5, a0
; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, mu
; CHECK-NEXT: vmflt.vf v16, v24, fa5, v0.t
-; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, mu
@@ -791,15 +795,15 @@ declare <vscale x 2 x double> @llvm.vp.rint.nxv2f64(<vscale x 2 x double>, <vsca
define <vscale x 2 x double> @vp_rint_nxv2f64(<vscale x 2 x double> %va, <vscale x 2 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_rint_nxv2f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v10, v0
; CHECK-NEXT: lui a1, %hi(.LCPI24_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI24_0)(a1)
+; CHECK-NEXT: vmv1r.v v10, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, ma
; CHECK-NEXT: vfabs.v v12, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, mu
; CHECK-NEXT: vmflt.vf v10, v12, fa5, v0.t
-; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, mu
@@ -831,15 +835,15 @@ declare <vscale x 4 x double> @llvm.vp.rint.nxv4f64(<vscale x 4 x double>, <vsca
define <vscale x 4 x double> @vp_rint_nxv4f64(<vscale x 4 x double> %va, <vscale x 4 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_rint_nxv4f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v12, v0
; CHECK-NEXT: lui a1, %hi(.LCPI26_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI26_0)(a1)
+; CHECK-NEXT: vmv1r.v v12, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, ma
; CHECK-NEXT: vfabs.v v16, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, mu
; CHECK-NEXT: vmflt.vf v12, v16, fa5, v0.t
-; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, mu
@@ -871,15 +875,15 @@ declare <vscale x 7 x double> @llvm.vp.rint.nxv7f64(<vscale x 7 x double>, <vsca
define <vscale x 7 x double> @vp_rint_nxv7f64(<vscale x 7 x double> %va, <vscale x 7 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_rint_nxv7f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: lui a1, %hi(.LCPI28_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI28_0)(a1)
+; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
; CHECK-NEXT: vmflt.vf v16, v24, fa5, v0.t
-; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
@@ -911,15 +915,15 @@ declare <vscale x 8 x double> @llvm.vp.rint.nxv8f64(<vscale x 8 x double>, <vsca
define <vscale x 8 x double> @vp_rint_nxv8f64(<vscale x 8 x double> %va, <vscale x 8 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_rint_nxv8f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: lui a1, %hi(.LCPI30_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI30_0)(a1)
+; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
; CHECK-NEXT: vmflt.vf v16, v24, fa5, v0.t
-; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
@@ -955,62 +959,51 @@ define <vscale x 16 x double> @vp_rint_nxv16f64(<vscale x 16 x double> %va, <vsc
; CHECK-NEXT: addi sp, sp, -16
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a1, a1, 4
+; CHECK-NEXT: slli a1, a1, 3
; CHECK-NEXT: sub sp, sp, a1
-; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
-; CHECK-NEXT: vmv1r.v v24, v0
-; CHECK-NEXT: addi a1, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
+; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; CHECK-NEXT: vmv1r.v v7, v0
; CHECK-NEXT: csrr a1, vlenb
; CHECK-NEXT: srli a2, a1, 3
; CHECK-NEXT: vsetvli a3, zero, e8, mf4, ta, ma
-; CHECK-NEXT: vslidedown.vx v25, v0, a2
+; CHECK-NEXT: vslidedown.vx v6, v0, a2
; CHECK-NEXT: sub a2, a0, a1
+; CHECK-NEXT: lui a3, %hi(.LCPI32_0)
+; CHECK-NEXT: fld fa5, %lo(.LCPI32_0)(a3)
; CHECK-NEXT: sltu a3, a0, a2
; CHECK-NEXT: addi a3, a3, -1
; CHECK-NEXT: and a2, a3, a2
-; CHECK-NEXT: lui a3, %hi(.LCPI32_0)
-; CHECK-NEXT: fld fa5, %lo(.LCPI32_0)(a3)
+; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: vsetvli zero, a2, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v25
-; CHECK-NEXT: vfabs.v v8, v16, v0.t
+; CHECK-NEXT: vfabs.v v24, v16, v0.t
+; CHECK-NEXT: addi a2, sp, 16
+; CHECK-NEXT: vs8r.v v24, (a2) # Unknown-size Folded Spill
+; CHECK-NEXT: vl8r.v v24, (a2) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vmflt.vf v25, v8, fa5, v0.t
+; CHECK-NEXT: vmflt.vf v6, v24, fa5, v0.t
+; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v25
-; CHECK-NEXT: vfcvt.x.f.v v8, v16, v0.t
-; CHECK-NEXT: vfcvt.f.x.v v8, v8, v0.t
+; CHECK-NEXT: vfcvt.x.f.v v24, v16, v0.t
+; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vfsgnj.vv v16, v8, v16, v0.t
-; CHECK-NEXT: csrr a2, vlenb
-; CHECK-NEXT: slli a2, a2, 3
-; CHECK-NEXT: add a2, sp, a2
-; CHECK-NEXT: addi a2, a2, 16
-; CHECK-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
+; CHECK-NEXT: vfsgnj.vv v16, v24, v16, v0.t
; CHECK-NEXT: bltu a0, a1, .LBB32_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB32_2:
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v24
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vfabs.v v16, v8, v0.t
+; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vmflt.vf v24, v16, fa5, v0.t
+; CHECK-NEXT: vmflt.vf v7, v24, fa5, v0.t
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v24
-; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
-; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
+; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
+; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vfsgnj.vv v8, v16, v8, v0.t
+; CHECK-NEXT: vfsgnj.vv v8, v24, v8, v0.t
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 4
; CHECK-NEXT: add sp, sp, a0
; CHECK-NEXT: addi sp, sp, 16
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/round-vp.ll b/llvm/test/CodeGen/RISCV/rvv/round-vp.ll
index edeac1acf3b01..1ddadcc493736 100644
--- a/llvm/test/CodeGen/RISCV/rvv/round-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/round-vp.ll
@@ -204,8 +204,8 @@ define <vscale x 4 x half> @vp_round_nxv4f16(<vscale x 4 x half> %va, <vscale x
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m2, ta, mu
; ZVFHMIN-NEXT: vmflt.vf v9, v12, fa5, v0.t
; ZVFHMIN-NEXT: fsrmi a0, 4
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v9
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; ZVFHMIN-NEXT: vfcvt.x.f.v v12, v10, v0.t
; ZVFHMIN-NEXT: fsrm a0
; ZVFHMIN-NEXT: vfcvt.f.x.v v12, v12, v0.t
@@ -261,16 +261,16 @@ declare <vscale x 8 x half> @llvm.vp.round.nxv8f16(<vscale x 8 x half>, <vscale
define <vscale x 8 x half> @vp_round_nxv8f16(<vscale x 8 x half> %va, <vscale x 8 x i1> %m, i32 zeroext %evl) {
; ZVFH-LABEL: vp_round_nxv8f16:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vmv1r.v v10, v0
; ZVFH-NEXT: lui a1, %hi(.LCPI6_0)
; ZVFH-NEXT: flh fa5, %lo(.LCPI6_0)(a1)
+; ZVFH-NEXT: vmv1r.v v10, v0
; ZVFH-NEXT: vsetvli zero, a0, e16, m2, ta, ma
; ZVFH-NEXT: vfabs.v v12, v8, v0.t
; ZVFH-NEXT: vsetvli zero, zero, e16, m2, ta, mu
; ZVFH-NEXT: vmflt.vf v10, v12, fa5, v0.t
; ZVFH-NEXT: fsrmi a0, 4
-; ZVFH-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; ZVFH-NEXT: vmv1r.v v0, v10
+; ZVFH-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; ZVFH-NEXT: vfcvt.x.f.v v12, v8, v0.t
; ZVFH-NEXT: fsrm a0
; ZVFH-NEXT: vfcvt.f.x.v v12, v12, v0.t
@@ -290,8 +290,8 @@ define <vscale x 8 x half> @vp_round_nxv8f16(<vscale x 8 x half> %va, <vscale x
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m4, ta, mu
; ZVFHMIN-NEXT: vmflt.vf v10, v16, fa5, v0.t
; ZVFHMIN-NEXT: fsrmi a0, 4
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v10
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; ZVFHMIN-NEXT: vfcvt.x.f.v v16, v12, v0.t
; ZVFHMIN-NEXT: fsrm a0
; ZVFHMIN-NEXT: vfcvt.f.x.v v16, v16, v0.t
@@ -347,16 +347,16 @@ declare <vscale x 16 x half> @llvm.vp.round.nxv16f16(<vscale x 16 x half>, <vsca
define <vscale x 16 x half> @vp_round_nxv16f16(<vscale x 16 x half> %va, <vscale x 16 x i1> %m, i32 zeroext %evl) {
; ZVFH-LABEL: vp_round_nxv16f16:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vmv1r.v v12, v0
; ZVFH-NEXT: lui a1, %hi(.LCPI8_0)
; ZVFH-NEXT: flh fa5, %lo(.LCPI8_0)(a1)
+; ZVFH-NEXT: vmv1r.v v12, v0
; ZVFH-NEXT: vsetvli zero, a0, e16, m4, ta, ma
; ZVFH-NEXT: vfabs.v v16, v8, v0.t
; ZVFH-NEXT: vsetvli zero, zero, e16, m4, ta, mu
; ZVFH-NEXT: vmflt.vf v12, v16, fa5, v0.t
; ZVFH-NEXT: fsrmi a0, 4
-; ZVFH-NEXT: vsetvli zero, zero, e16, m4, ta, ma
; ZVFH-NEXT: vmv1r.v v0, v12
+; ZVFH-NEXT: vsetvli zero, zero, e16, m4, ta, ma
; ZVFH-NEXT: vfcvt.x.f.v v16, v8, v0.t
; ZVFH-NEXT: fsrm a0
; ZVFH-NEXT: vfcvt.f.x.v v16, v16, v0.t
@@ -376,8 +376,8 @@ define <vscale x 16 x half> @vp_round_nxv16f16(<vscale x 16 x half> %va, <vscale
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, mu
; ZVFHMIN-NEXT: vmflt.vf v12, v24, fa5, v0.t
; ZVFHMIN-NEXT: fsrmi a0, 4
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v12
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfcvt.x.f.v v24, v16, v0.t
; ZVFHMIN-NEXT: fsrm a0
; ZVFHMIN-NEXT: vfcvt.f.x.v v24, v24, v0.t
@@ -433,16 +433,16 @@ declare <vscale x 32 x half> @llvm.vp.round.nxv32f16(<vscale x 32 x half>, <vsca
define <vscale x 32 x half> @vp_round_nxv32f16(<vscale x 32 x half> %va, <vscale x 32 x i1> %m, i32 zeroext %evl) {
; ZVFH-LABEL: vp_round_nxv32f16:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vmv1r.v v16, v0
; ZVFH-NEXT: lui a1, %hi(.LCPI10_0)
; ZVFH-NEXT: flh fa5, %lo(.LCPI10_0)(a1)
+; ZVFH-NEXT: vmv1r.v v16, v0
; ZVFH-NEXT: vsetvli zero, a0, e16, m8, ta, ma
; ZVFH-NEXT: vfabs.v v24, v8, v0.t
; ZVFH-NEXT: vsetvli zero, zero, e16, m8, ta, mu
; ZVFH-NEXT: vmflt.vf v16, v24, fa5, v0.t
; ZVFH-NEXT: fsrmi a0, 4
-; ZVFH-NEXT: vsetvli zero, zero, e16, m8, ta, ma
; ZVFH-NEXT: vmv1r.v v0, v16
+; ZVFH-NEXT: vsetvli zero, zero, e16, m8, ta, ma
; ZVFH-NEXT: vfcvt.x.f.v v24, v8, v0.t
; ZVFH-NEXT: fsrm a0
; ZVFH-NEXT: vfcvt.f.x.v v24, v24, v0.t
@@ -458,7 +458,6 @@ define <vscale x 32 x half> @vp_round_nxv32f16(<vscale x 32 x half> %va, <vscale
; ZVFHMIN-NEXT: slli a1, a1, 3
; ZVFHMIN-NEXT: sub sp, sp, a1
; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
-; ZVFHMIN-NEXT: vmv1r.v v16, v0
; ZVFHMIN-NEXT: csrr a2, vlenb
; ZVFHMIN-NEXT: slli a1, a2, 1
; ZVFHMIN-NEXT: sub a3, a0, a1
@@ -466,52 +465,60 @@ define <vscale x 32 x half> @vp_round_nxv32f16(<vscale x 32 x half> %va, <vscale
; ZVFHMIN-NEXT: addi a4, a4, -1
; ZVFHMIN-NEXT: and a3, a4, a3
; ZVFHMIN-NEXT: srli a2, a2, 2
+; ZVFHMIN-NEXT: vmv1r.v v16, v0
; ZVFHMIN-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
; ZVFHMIN-NEXT: vslidedown.vx v17, v0, a2
-; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
-; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v17
+; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfabs.v v8, v24, v0.t
; ZVFHMIN-NEXT: lui a2, 307200
; ZVFHMIN-NEXT: fmv.w.x fa5, a2
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, mu
; ZVFHMIN-NEXT: vmflt.vf v17, v8, fa5, v0.t
; ZVFHMIN-NEXT: fsrmi a2, 4
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v17
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfcvt.x.f.v v8, v24, v0.t
; ZVFHMIN-NEXT: fsrm a2
; ZVFHMIN-NEXT: vfcvt.f.x.v v8, v8, v0.t
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, mu
; ZVFHMIN-NEXT: vfsgnj.vv v24, v8, v24, v0.t
; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v20, v24
+; ZVFHMIN-NEXT: vfncvt.f.f.w v12, v24
; ZVFHMIN-NEXT: bltu a0, a1, .LBB10_2
; ZVFHMIN-NEXT: # %bb.1:
; ZVFHMIN-NEXT: mv a0, a1
; ZVFHMIN-NEXT: .LBB10_2:
; ZVFHMIN-NEXT: addi a1, sp, 16
-; ZVFHMIN-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v8
-; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
+; ZVFHMIN-NEXT: vl8r.v v0, (a1) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v0
+; ZVFHMIN-NEXT: vmv1r.v v8, v16
; ZVFHMIN-NEXT: vmv1r.v v0, v16
-; ZVFHMIN-NEXT: vfabs.v v8, v24, v0.t
+; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
+; ZVFHMIN-NEXT: vfabs.v v16, v24, v0.t
+; ZVFHMIN-NEXT: addi a0, sp, 16
+; ZVFHMIN-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vmv1r.v v0, v8
+; ZVFHMIN-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, mu
-; ZVFHMIN-NEXT: vmflt.vf v16, v8, fa5, v0.t
+; ZVFHMIN-NEXT: vmflt.vf v8, v16, fa5, v0.t
; ZVFHMIN-NEXT: fsrmi a0, 4
+; ZVFHMIN-NEXT: vmv1r.v v0, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vmv1r.v v0, v16
-; ZVFHMIN-NEXT: vfcvt.x.f.v v8, v24, v0.t
+; ZVFHMIN-NEXT: vfcvt.x.f.v v16, v24, v0.t
+; ZVFHMIN-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: fsrm a0
-; ZVFHMIN-NEXT: vfcvt.f.x.v v8, v8, v0.t
+; ZVFHMIN-NEXT: addi a0, sp, 16
+; ZVFHMIN-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vfcvt.f.x.v v16, v16, v0.t
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, mu
-; ZVFHMIN-NEXT: vfsgnj.vv v24, v8, v24, v0.t
+; ZVFHMIN-NEXT: vfsgnj.vv v24, v16, v24, v0.t
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v16, v24
-; ZVFHMIN-NEXT: vmv8r.v v8, v16
+; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v24
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: add sp, sp, a0
@@ -556,20 +563,20 @@ define <vscale x 32 x half> @vp_round_nxv32f16_unmasked(<vscale x 32 x half> %va
; ZVFHMIN-NEXT: vmset.m v16
; ZVFHMIN-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
; ZVFHMIN-NEXT: vslidedown.vx v16, v16, a2
-; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
-; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v16
+; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfabs.v v8, v24, v0.t
; ZVFHMIN-NEXT: lui a2, 307200
; ZVFHMIN-NEXT: fmv.w.x fa5, a2
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, mu
; ZVFHMIN-NEXT: vmflt.vf v16, v8, fa5, v0.t
; ZVFHMIN-NEXT: fsrmi a2, 4
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v16
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfcvt.x.f.v v8, v24, v0.t
; ZVFHMIN-NEXT: fsrm a2
; ZVFHMIN-NEXT: vfcvt.f.x.v v8, v8, v0.t
@@ -701,8 +708,8 @@ define <vscale x 4 x float> @vp_round_nxv4f32(<vscale x 4 x float> %va, <vscale
; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, mu
; CHECK-NEXT: vmflt.vf v10, v12, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 4
-; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
@@ -745,8 +752,8 @@ define <vscale x 8 x float> @vp_round_nxv8f32(<vscale x 8 x float> %va, <vscale
; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, mu
; CHECK-NEXT: vmflt.vf v12, v16, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 4
-; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
@@ -789,8 +796,8 @@ define <vscale x 16 x float> @vp_round_nxv16f32(<vscale x 16 x float> %va, <vsca
; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, mu
; CHECK-NEXT: vmflt.vf v16, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 4
-; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
@@ -867,16 +874,16 @@ declare <vscale x 2 x double> @llvm.vp.round.nxv2f64(<vscale x 2 x double>, <vsc
define <vscale x 2 x double> @vp_round_nxv2f64(<vscale x 2 x double> %va, <vscale x 2 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_round_nxv2f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v10, v0
; CHECK-NEXT: lui a1, %hi(.LCPI24_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI24_0)(a1)
+; CHECK-NEXT: vmv1r.v v10, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, ma
; CHECK-NEXT: vfabs.v v12, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, mu
; CHECK-NEXT: vmflt.vf v10, v12, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 4
-; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
@@ -911,16 +918,16 @@ declare <vscale x 4 x double> @llvm.vp.round.nxv4f64(<vscale x 4 x double>, <vsc
define <vscale x 4 x double> @vp_round_nxv4f64(<vscale x 4 x double> %va, <vscale x 4 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_round_nxv4f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v12, v0
; CHECK-NEXT: lui a1, %hi(.LCPI26_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI26_0)(a1)
+; CHECK-NEXT: vmv1r.v v12, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, ma
; CHECK-NEXT: vfabs.v v16, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, mu
; CHECK-NEXT: vmflt.vf v12, v16, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 4
-; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
@@ -955,16 +962,16 @@ declare <vscale x 7 x double> @llvm.vp.round.nxv7f64(<vscale x 7 x double>, <vsc
define <vscale x 7 x double> @vp_round_nxv7f64(<vscale x 7 x double> %va, <vscale x 7 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_round_nxv7f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: lui a1, %hi(.LCPI28_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI28_0)(a1)
+; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
; CHECK-NEXT: vmflt.vf v16, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 4
-; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
@@ -999,16 +1006,16 @@ declare <vscale x 8 x double> @llvm.vp.round.nxv8f64(<vscale x 8 x double>, <vsc
define <vscale x 8 x double> @vp_round_nxv8f64(<vscale x 8 x double> %va, <vscale x 8 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_round_nxv8f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: lui a1, %hi(.LCPI30_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI30_0)(a1)
+; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
; CHECK-NEXT: vmflt.vf v16, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 4
-; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
@@ -1047,66 +1054,56 @@ define <vscale x 16 x double> @vp_round_nxv16f64(<vscale x 16 x double> %va, <vs
; CHECK-NEXT: addi sp, sp, -16
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a1, a1, 4
+; CHECK-NEXT: slli a1, a1, 3
; CHECK-NEXT: sub sp, sp, a1
-; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
-; CHECK-NEXT: vmv1r.v v24, v0
-; CHECK-NEXT: addi a1, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
+; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; CHECK-NEXT: vmv1r.v v7, v0
; CHECK-NEXT: csrr a1, vlenb
; CHECK-NEXT: srli a2, a1, 3
; CHECK-NEXT: vsetvli a3, zero, e8, mf4, ta, ma
-; CHECK-NEXT: vslidedown.vx v25, v0, a2
+; CHECK-NEXT: vslidedown.vx v6, v0, a2
; CHECK-NEXT: sub a2, a0, a1
+; CHECK-NEXT: lui a3, %hi(.LCPI32_0)
+; CHECK-NEXT: fld fa5, %lo(.LCPI32_0)(a3)
; CHECK-NEXT: sltu a3, a0, a2
; CHECK-NEXT: addi a3, a3, -1
; CHECK-NEXT: and a2, a3, a2
-; CHECK-NEXT: lui a3, %hi(.LCPI32_0)
-; CHECK-NEXT: fld fa5, %lo(.LCPI32_0)(a3)
+; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: vsetvli zero, a2, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v25
-; CHECK-NEXT: vfabs.v v8, v16, v0.t
+; CHECK-NEXT: vfabs.v v24, v16, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vmflt.vf v25, v8, fa5, v0.t
+; CHECK-NEXT: vmflt.vf v6, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a2, 4
+; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v25
-; CHECK-NEXT: vfcvt.x.f.v v8, v16, v0.t
+; CHECK-NEXT: vfcvt.x.f.v v24, v16, v0.t
+; CHECK-NEXT: addi a3, sp, 16
+; CHECK-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
; CHECK-NEXT: fsrm a2
-; CHECK-NEXT: vfcvt.f.x.v v8, v8, v0.t
+; CHECK-NEXT: addi a2, sp, 16
+; CHECK-NEXT: vl8r.v v24, (a2) # Unknown-size Folded Reload
+; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vfsgnj.vv v16, v8, v16, v0.t
-; CHECK-NEXT: csrr a2, vlenb
-; CHECK-NEXT: slli a2, a2, 3
-; CHECK-NEXT: add a2, sp, a2
-; CHECK-NEXT: addi a2, a2, 16
-; CHECK-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
+; CHECK-NEXT: vfsgnj.vv v16, v24, v16, v0.t
; CHECK-NEXT: bltu a0, a1, .LBB32_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB32_2:
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v24
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vfabs.v v16, v8, v0.t
+; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vmflt.vf v24, v16, fa5, v0.t
+; CHECK-NEXT: vmflt.vf v7, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 4
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v24
-; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
+; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: fsrm a0
-; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
+; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vfsgnj.vv v8, v16, v8, v0.t
+; CHECK-NEXT: vfsgnj.vv v8, v24, v8, v0.t
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 4
; CHECK-NEXT: add sp, sp, a0
; CHECK-NEXT: addi sp, sp, 16
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/roundeven-vp.ll b/llvm/test/CodeGen/RISCV/rvv/roundeven-vp.ll
index a77c58ba9ec57..8c5a7bb2dea6a 100644
--- a/llvm/test/CodeGen/RISCV/rvv/roundeven-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/roundeven-vp.ll
@@ -204,8 +204,8 @@ define <vscale x 4 x half> @vp_roundeven_nxv4f16(<vscale x 4 x half> %va, <vscal
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m2, ta, mu
; ZVFHMIN-NEXT: vmflt.vf v9, v12, fa5, v0.t
; ZVFHMIN-NEXT: fsrmi a0, 0
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v9
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; ZVFHMIN-NEXT: vfcvt.x.f.v v12, v10, v0.t
; ZVFHMIN-NEXT: fsrm a0
; ZVFHMIN-NEXT: vfcvt.f.x.v v12, v12, v0.t
@@ -261,16 +261,16 @@ declare <vscale x 8 x half> @llvm.vp.roundeven.nxv8f16(<vscale x 8 x half>, <vsc
define <vscale x 8 x half> @vp_roundeven_nxv8f16(<vscale x 8 x half> %va, <vscale x 8 x i1> %m, i32 zeroext %evl) {
; ZVFH-LABEL: vp_roundeven_nxv8f16:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vmv1r.v v10, v0
; ZVFH-NEXT: lui a1, %hi(.LCPI6_0)
; ZVFH-NEXT: flh fa5, %lo(.LCPI6_0)(a1)
+; ZVFH-NEXT: vmv1r.v v10, v0
; ZVFH-NEXT: vsetvli zero, a0, e16, m2, ta, ma
; ZVFH-NEXT: vfabs.v v12, v8, v0.t
; ZVFH-NEXT: vsetvli zero, zero, e16, m2, ta, mu
; ZVFH-NEXT: vmflt.vf v10, v12, fa5, v0.t
; ZVFH-NEXT: fsrmi a0, 0
-; ZVFH-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; ZVFH-NEXT: vmv1r.v v0, v10
+; ZVFH-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; ZVFH-NEXT: vfcvt.x.f.v v12, v8, v0.t
; ZVFH-NEXT: fsrm a0
; ZVFH-NEXT: vfcvt.f.x.v v12, v12, v0.t
@@ -290,8 +290,8 @@ define <vscale x 8 x half> @vp_roundeven_nxv8f16(<vscale x 8 x half> %va, <vscal
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m4, ta, mu
; ZVFHMIN-NEXT: vmflt.vf v10, v16, fa5, v0.t
; ZVFHMIN-NEXT: fsrmi a0, 0
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v10
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; ZVFHMIN-NEXT: vfcvt.x.f.v v16, v12, v0.t
; ZVFHMIN-NEXT: fsrm a0
; ZVFHMIN-NEXT: vfcvt.f.x.v v16, v16, v0.t
@@ -347,16 +347,16 @@ declare <vscale x 16 x half> @llvm.vp.roundeven.nxv16f16(<vscale x 16 x half>, <
define <vscale x 16 x half> @vp_roundeven_nxv16f16(<vscale x 16 x half> %va, <vscale x 16 x i1> %m, i32 zeroext %evl) {
; ZVFH-LABEL: vp_roundeven_nxv16f16:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vmv1r.v v12, v0
; ZVFH-NEXT: lui a1, %hi(.LCPI8_0)
; ZVFH-NEXT: flh fa5, %lo(.LCPI8_0)(a1)
+; ZVFH-NEXT: vmv1r.v v12, v0
; ZVFH-NEXT: vsetvli zero, a0, e16, m4, ta, ma
; ZVFH-NEXT: vfabs.v v16, v8, v0.t
; ZVFH-NEXT: vsetvli zero, zero, e16, m4, ta, mu
; ZVFH-NEXT: vmflt.vf v12, v16, fa5, v0.t
; ZVFH-NEXT: fsrmi a0, 0
-; ZVFH-NEXT: vsetvli zero, zero, e16, m4, ta, ma
; ZVFH-NEXT: vmv1r.v v0, v12
+; ZVFH-NEXT: vsetvli zero, zero, e16, m4, ta, ma
; ZVFH-NEXT: vfcvt.x.f.v v16, v8, v0.t
; ZVFH-NEXT: fsrm a0
; ZVFH-NEXT: vfcvt.f.x.v v16, v16, v0.t
@@ -376,8 +376,8 @@ define <vscale x 16 x half> @vp_roundeven_nxv16f16(<vscale x 16 x half> %va, <vs
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, mu
; ZVFHMIN-NEXT: vmflt.vf v12, v24, fa5, v0.t
; ZVFHMIN-NEXT: fsrmi a0, 0
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v12
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfcvt.x.f.v v24, v16, v0.t
; ZVFHMIN-NEXT: fsrm a0
; ZVFHMIN-NEXT: vfcvt.f.x.v v24, v24, v0.t
@@ -433,16 +433,16 @@ declare <vscale x 32 x half> @llvm.vp.roundeven.nxv32f16(<vscale x 32 x half>, <
define <vscale x 32 x half> @vp_roundeven_nxv32f16(<vscale x 32 x half> %va, <vscale x 32 x i1> %m, i32 zeroext %evl) {
; ZVFH-LABEL: vp_roundeven_nxv32f16:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vmv1r.v v16, v0
; ZVFH-NEXT: lui a1, %hi(.LCPI10_0)
; ZVFH-NEXT: flh fa5, %lo(.LCPI10_0)(a1)
+; ZVFH-NEXT: vmv1r.v v16, v0
; ZVFH-NEXT: vsetvli zero, a0, e16, m8, ta, ma
; ZVFH-NEXT: vfabs.v v24, v8, v0.t
; ZVFH-NEXT: vsetvli zero, zero, e16, m8, ta, mu
; ZVFH-NEXT: vmflt.vf v16, v24, fa5, v0.t
; ZVFH-NEXT: fsrmi a0, 0
-; ZVFH-NEXT: vsetvli zero, zero, e16, m8, ta, ma
; ZVFH-NEXT: vmv1r.v v0, v16
+; ZVFH-NEXT: vsetvli zero, zero, e16, m8, ta, ma
; ZVFH-NEXT: vfcvt.x.f.v v24, v8, v0.t
; ZVFH-NEXT: fsrm a0
; ZVFH-NEXT: vfcvt.f.x.v v24, v24, v0.t
@@ -458,7 +458,6 @@ define <vscale x 32 x half> @vp_roundeven_nxv32f16(<vscale x 32 x half> %va, <vs
; ZVFHMIN-NEXT: slli a1, a1, 3
; ZVFHMIN-NEXT: sub sp, sp, a1
; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
-; ZVFHMIN-NEXT: vmv1r.v v16, v0
; ZVFHMIN-NEXT: csrr a2, vlenb
; ZVFHMIN-NEXT: slli a1, a2, 1
; ZVFHMIN-NEXT: sub a3, a0, a1
@@ -466,52 +465,60 @@ define <vscale x 32 x half> @vp_roundeven_nxv32f16(<vscale x 32 x half> %va, <vs
; ZVFHMIN-NEXT: addi a4, a4, -1
; ZVFHMIN-NEXT: and a3, a4, a3
; ZVFHMIN-NEXT: srli a2, a2, 2
+; ZVFHMIN-NEXT: vmv1r.v v16, v0
; ZVFHMIN-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
; ZVFHMIN-NEXT: vslidedown.vx v17, v0, a2
-; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
-; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v17
+; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfabs.v v8, v24, v0.t
; ZVFHMIN-NEXT: lui a2, 307200
; ZVFHMIN-NEXT: fmv.w.x fa5, a2
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, mu
; ZVFHMIN-NEXT: vmflt.vf v17, v8, fa5, v0.t
; ZVFHMIN-NEXT: fsrmi a2, 0
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v17
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfcvt.x.f.v v8, v24, v0.t
; ZVFHMIN-NEXT: fsrm a2
; ZVFHMIN-NEXT: vfcvt.f.x.v v8, v8, v0.t
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, mu
; ZVFHMIN-NEXT: vfsgnj.vv v24, v8, v24, v0.t
; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v20, v24
+; ZVFHMIN-NEXT: vfncvt.f.f.w v12, v24
; ZVFHMIN-NEXT: bltu a0, a1, .LBB10_2
; ZVFHMIN-NEXT: # %bb.1:
; ZVFHMIN-NEXT: mv a0, a1
; ZVFHMIN-NEXT: .LBB10_2:
; ZVFHMIN-NEXT: addi a1, sp, 16
-; ZVFHMIN-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v8
-; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
+; ZVFHMIN-NEXT: vl8r.v v0, (a1) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v0
+; ZVFHMIN-NEXT: vmv1r.v v8, v16
; ZVFHMIN-NEXT: vmv1r.v v0, v16
-; ZVFHMIN-NEXT: vfabs.v v8, v24, v0.t
+; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
+; ZVFHMIN-NEXT: vfabs.v v16, v24, v0.t
+; ZVFHMIN-NEXT: addi a0, sp, 16
+; ZVFHMIN-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vmv1r.v v0, v8
+; ZVFHMIN-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, mu
-; ZVFHMIN-NEXT: vmflt.vf v16, v8, fa5, v0.t
+; ZVFHMIN-NEXT: vmflt.vf v8, v16, fa5, v0.t
; ZVFHMIN-NEXT: fsrmi a0, 0
+; ZVFHMIN-NEXT: vmv1r.v v0, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vmv1r.v v0, v16
-; ZVFHMIN-NEXT: vfcvt.x.f.v v8, v24, v0.t
+; ZVFHMIN-NEXT: vfcvt.x.f.v v16, v24, v0.t
+; ZVFHMIN-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: fsrm a0
-; ZVFHMIN-NEXT: vfcvt.f.x.v v8, v8, v0.t
+; ZVFHMIN-NEXT: addi a0, sp, 16
+; ZVFHMIN-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vfcvt.f.x.v v16, v16, v0.t
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, mu
-; ZVFHMIN-NEXT: vfsgnj.vv v24, v8, v24, v0.t
+; ZVFHMIN-NEXT: vfsgnj.vv v24, v16, v24, v0.t
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v16, v24
-; ZVFHMIN-NEXT: vmv8r.v v8, v16
+; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v24
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: add sp, sp, a0
@@ -556,20 +563,20 @@ define <vscale x 32 x half> @vp_roundeven_nxv32f16_unmasked(<vscale x 32 x half>
; ZVFHMIN-NEXT: vmset.m v16
; ZVFHMIN-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
; ZVFHMIN-NEXT: vslidedown.vx v16, v16, a2
-; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
-; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v16
+; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfabs.v v8, v24, v0.t
; ZVFHMIN-NEXT: lui a2, 307200
; ZVFHMIN-NEXT: fmv.w.x fa5, a2
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, mu
; ZVFHMIN-NEXT: vmflt.vf v16, v8, fa5, v0.t
; ZVFHMIN-NEXT: fsrmi a2, 0
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v16
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfcvt.x.f.v v8, v24, v0.t
; ZVFHMIN-NEXT: fsrm a2
; ZVFHMIN-NEXT: vfcvt.f.x.v v8, v8, v0.t
@@ -701,8 +708,8 @@ define <vscale x 4 x float> @vp_roundeven_nxv4f32(<vscale x 4 x float> %va, <vsc
; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, mu
; CHECK-NEXT: vmflt.vf v10, v12, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 0
-; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
@@ -745,8 +752,8 @@ define <vscale x 8 x float> @vp_roundeven_nxv8f32(<vscale x 8 x float> %va, <vsc
; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, mu
; CHECK-NEXT: vmflt.vf v12, v16, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 0
-; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
@@ -789,8 +796,8 @@ define <vscale x 16 x float> @vp_roundeven_nxv16f32(<vscale x 16 x float> %va, <
; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, mu
; CHECK-NEXT: vmflt.vf v16, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 0
-; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
@@ -867,16 +874,16 @@ declare <vscale x 2 x double> @llvm.vp.roundeven.nxv2f64(<vscale x 2 x double>,
define <vscale x 2 x double> @vp_roundeven_nxv2f64(<vscale x 2 x double> %va, <vscale x 2 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_roundeven_nxv2f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v10, v0
; CHECK-NEXT: lui a1, %hi(.LCPI24_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI24_0)(a1)
+; CHECK-NEXT: vmv1r.v v10, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, ma
; CHECK-NEXT: vfabs.v v12, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, mu
; CHECK-NEXT: vmflt.vf v10, v12, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 0
-; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
@@ -911,16 +918,16 @@ declare <vscale x 4 x double> @llvm.vp.roundeven.nxv4f64(<vscale x 4 x double>,
define <vscale x 4 x double> @vp_roundeven_nxv4f64(<vscale x 4 x double> %va, <vscale x 4 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_roundeven_nxv4f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v12, v0
; CHECK-NEXT: lui a1, %hi(.LCPI26_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI26_0)(a1)
+; CHECK-NEXT: vmv1r.v v12, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, ma
; CHECK-NEXT: vfabs.v v16, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, mu
; CHECK-NEXT: vmflt.vf v12, v16, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 0
-; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
@@ -955,16 +962,16 @@ declare <vscale x 7 x double> @llvm.vp.roundeven.nxv7f64(<vscale x 7 x double>,
define <vscale x 7 x double> @vp_roundeven_nxv7f64(<vscale x 7 x double> %va, <vscale x 7 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_roundeven_nxv7f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: lui a1, %hi(.LCPI28_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI28_0)(a1)
+; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
; CHECK-NEXT: vmflt.vf v16, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 0
-; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
@@ -999,16 +1006,16 @@ declare <vscale x 8 x double> @llvm.vp.roundeven.nxv8f64(<vscale x 8 x double>,
define <vscale x 8 x double> @vp_roundeven_nxv8f64(<vscale x 8 x double> %va, <vscale x 8 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_roundeven_nxv8f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: lui a1, %hi(.LCPI30_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI30_0)(a1)
+; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
; CHECK-NEXT: vmflt.vf v16, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 0
-; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
@@ -1047,66 +1054,56 @@ define <vscale x 16 x double> @vp_roundeven_nxv16f64(<vscale x 16 x double> %va,
; CHECK-NEXT: addi sp, sp, -16
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a1, a1, 4
+; CHECK-NEXT: slli a1, a1, 3
; CHECK-NEXT: sub sp, sp, a1
-; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
-; CHECK-NEXT: vmv1r.v v24, v0
-; CHECK-NEXT: addi a1, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
+; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; CHECK-NEXT: vmv1r.v v7, v0
; CHECK-NEXT: csrr a1, vlenb
; CHECK-NEXT: srli a2, a1, 3
; CHECK-NEXT: vsetvli a3, zero, e8, mf4, ta, ma
-; CHECK-NEXT: vslidedown.vx v25, v0, a2
+; CHECK-NEXT: vslidedown.vx v6, v0, a2
; CHECK-NEXT: sub a2, a0, a1
+; CHECK-NEXT: lui a3, %hi(.LCPI32_0)
+; CHECK-NEXT: fld fa5, %lo(.LCPI32_0)(a3)
; CHECK-NEXT: sltu a3, a0, a2
; CHECK-NEXT: addi a3, a3, -1
; CHECK-NEXT: and a2, a3, a2
-; CHECK-NEXT: lui a3, %hi(.LCPI32_0)
-; CHECK-NEXT: fld fa5, %lo(.LCPI32_0)(a3)
+; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: vsetvli zero, a2, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v25
-; CHECK-NEXT: vfabs.v v8, v16, v0.t
+; CHECK-NEXT: vfabs.v v24, v16, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vmflt.vf v25, v8, fa5, v0.t
+; CHECK-NEXT: vmflt.vf v6, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a2, 0
+; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v25
-; CHECK-NEXT: vfcvt.x.f.v v8, v16, v0.t
+; CHECK-NEXT: vfcvt.x.f.v v24, v16, v0.t
+; CHECK-NEXT: addi a3, sp, 16
+; CHECK-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
; CHECK-NEXT: fsrm a2
-; CHECK-NEXT: vfcvt.f.x.v v8, v8, v0.t
+; CHECK-NEXT: addi a2, sp, 16
+; CHECK-NEXT: vl8r.v v24, (a2) # Unknown-size Folded Reload
+; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vfsgnj.vv v16, v8, v16, v0.t
-; CHECK-NEXT: csrr a2, vlenb
-; CHECK-NEXT: slli a2, a2, 3
-; CHECK-NEXT: add a2, sp, a2
-; CHECK-NEXT: addi a2, a2, 16
-; CHECK-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
+; CHECK-NEXT: vfsgnj.vv v16, v24, v16, v0.t
; CHECK-NEXT: bltu a0, a1, .LBB32_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB32_2:
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v24
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vfabs.v v16, v8, v0.t
+; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vmflt.vf v24, v16, fa5, v0.t
+; CHECK-NEXT: vmflt.vf v7, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 0
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v24
-; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
+; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: fsrm a0
-; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
+; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vfsgnj.vv v8, v16, v8, v0.t
+; CHECK-NEXT: vfsgnj.vv v8, v24, v8, v0.t
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 4
; CHECK-NEXT: add sp, sp, a0
; CHECK-NEXT: addi sp, sp, 16
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/roundtozero-vp.ll b/llvm/test/CodeGen/RISCV/rvv/roundtozero-vp.ll
index 71a53c525551c..1227e73a02432 100644
--- a/llvm/test/CodeGen/RISCV/rvv/roundtozero-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/roundtozero-vp.ll
@@ -204,8 +204,8 @@ define <vscale x 4 x half> @vp_roundtozero_nxv4f16(<vscale x 4 x half> %va, <vsc
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m2, ta, mu
; ZVFHMIN-NEXT: vmflt.vf v9, v12, fa5, v0.t
; ZVFHMIN-NEXT: fsrmi a0, 1
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v9
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; ZVFHMIN-NEXT: vfcvt.x.f.v v12, v10, v0.t
; ZVFHMIN-NEXT: fsrm a0
; ZVFHMIN-NEXT: vfcvt.f.x.v v12, v12, v0.t
@@ -261,16 +261,16 @@ declare <vscale x 8 x half> @llvm.vp.roundtozero.nxv8f16(<vscale x 8 x half>, <v
define <vscale x 8 x half> @vp_roundtozero_nxv8f16(<vscale x 8 x half> %va, <vscale x 8 x i1> %m, i32 zeroext %evl) {
; ZVFH-LABEL: vp_roundtozero_nxv8f16:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vmv1r.v v10, v0
; ZVFH-NEXT: lui a1, %hi(.LCPI6_0)
; ZVFH-NEXT: flh fa5, %lo(.LCPI6_0)(a1)
+; ZVFH-NEXT: vmv1r.v v10, v0
; ZVFH-NEXT: vsetvli zero, a0, e16, m2, ta, ma
; ZVFH-NEXT: vfabs.v v12, v8, v0.t
; ZVFH-NEXT: vsetvli zero, zero, e16, m2, ta, mu
; ZVFH-NEXT: vmflt.vf v10, v12, fa5, v0.t
; ZVFH-NEXT: fsrmi a0, 1
-; ZVFH-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; ZVFH-NEXT: vmv1r.v v0, v10
+; ZVFH-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; ZVFH-NEXT: vfcvt.x.f.v v12, v8, v0.t
; ZVFH-NEXT: fsrm a0
; ZVFH-NEXT: vfcvt.f.x.v v12, v12, v0.t
@@ -290,8 +290,8 @@ define <vscale x 8 x half> @vp_roundtozero_nxv8f16(<vscale x 8 x half> %va, <vsc
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m4, ta, mu
; ZVFHMIN-NEXT: vmflt.vf v10, v16, fa5, v0.t
; ZVFHMIN-NEXT: fsrmi a0, 1
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v10
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; ZVFHMIN-NEXT: vfcvt.x.f.v v16, v12, v0.t
; ZVFHMIN-NEXT: fsrm a0
; ZVFHMIN-NEXT: vfcvt.f.x.v v16, v16, v0.t
@@ -347,16 +347,16 @@ declare <vscale x 16 x half> @llvm.vp.roundtozero.nxv16f16(<vscale x 16 x half>,
define <vscale x 16 x half> @vp_roundtozero_nxv16f16(<vscale x 16 x half> %va, <vscale x 16 x i1> %m, i32 zeroext %evl) {
; ZVFH-LABEL: vp_roundtozero_nxv16f16:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vmv1r.v v12, v0
; ZVFH-NEXT: lui a1, %hi(.LCPI8_0)
; ZVFH-NEXT: flh fa5, %lo(.LCPI8_0)(a1)
+; ZVFH-NEXT: vmv1r.v v12, v0
; ZVFH-NEXT: vsetvli zero, a0, e16, m4, ta, ma
; ZVFH-NEXT: vfabs.v v16, v8, v0.t
; ZVFH-NEXT: vsetvli zero, zero, e16, m4, ta, mu
; ZVFH-NEXT: vmflt.vf v12, v16, fa5, v0.t
; ZVFH-NEXT: fsrmi a0, 1
-; ZVFH-NEXT: vsetvli zero, zero, e16, m4, ta, ma
; ZVFH-NEXT: vmv1r.v v0, v12
+; ZVFH-NEXT: vsetvli zero, zero, e16, m4, ta, ma
; ZVFH-NEXT: vfcvt.x.f.v v16, v8, v0.t
; ZVFH-NEXT: fsrm a0
; ZVFH-NEXT: vfcvt.f.x.v v16, v16, v0.t
@@ -376,8 +376,8 @@ define <vscale x 16 x half> @vp_roundtozero_nxv16f16(<vscale x 16 x half> %va, <
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, mu
; ZVFHMIN-NEXT: vmflt.vf v12, v24, fa5, v0.t
; ZVFHMIN-NEXT: fsrmi a0, 1
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v12
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfcvt.x.f.v v24, v16, v0.t
; ZVFHMIN-NEXT: fsrm a0
; ZVFHMIN-NEXT: vfcvt.f.x.v v24, v24, v0.t
@@ -433,16 +433,16 @@ declare <vscale x 32 x half> @llvm.vp.roundtozero.nxv32f16(<vscale x 32 x half>,
define <vscale x 32 x half> @vp_roundtozero_nxv32f16(<vscale x 32 x half> %va, <vscale x 32 x i1> %m, i32 zeroext %evl) {
; ZVFH-LABEL: vp_roundtozero_nxv32f16:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vmv1r.v v16, v0
; ZVFH-NEXT: lui a1, %hi(.LCPI10_0)
; ZVFH-NEXT: flh fa5, %lo(.LCPI10_0)(a1)
+; ZVFH-NEXT: vmv1r.v v16, v0
; ZVFH-NEXT: vsetvli zero, a0, e16, m8, ta, ma
; ZVFH-NEXT: vfabs.v v24, v8, v0.t
; ZVFH-NEXT: vsetvli zero, zero, e16, m8, ta, mu
; ZVFH-NEXT: vmflt.vf v16, v24, fa5, v0.t
; ZVFH-NEXT: fsrmi a0, 1
-; ZVFH-NEXT: vsetvli zero, zero, e16, m8, ta, ma
; ZVFH-NEXT: vmv1r.v v0, v16
+; ZVFH-NEXT: vsetvli zero, zero, e16, m8, ta, ma
; ZVFH-NEXT: vfcvt.x.f.v v24, v8, v0.t
; ZVFH-NEXT: fsrm a0
; ZVFH-NEXT: vfcvt.f.x.v v24, v24, v0.t
@@ -458,7 +458,6 @@ define <vscale x 32 x half> @vp_roundtozero_nxv32f16(<vscale x 32 x half> %va, <
; ZVFHMIN-NEXT: slli a1, a1, 3
; ZVFHMIN-NEXT: sub sp, sp, a1
; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
-; ZVFHMIN-NEXT: vmv1r.v v16, v0
; ZVFHMIN-NEXT: csrr a2, vlenb
; ZVFHMIN-NEXT: slli a1, a2, 1
; ZVFHMIN-NEXT: sub a3, a0, a1
@@ -466,52 +465,60 @@ define <vscale x 32 x half> @vp_roundtozero_nxv32f16(<vscale x 32 x half> %va, <
; ZVFHMIN-NEXT: addi a4, a4, -1
; ZVFHMIN-NEXT: and a3, a4, a3
; ZVFHMIN-NEXT: srli a2, a2, 2
+; ZVFHMIN-NEXT: vmv1r.v v16, v0
; ZVFHMIN-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
; ZVFHMIN-NEXT: vslidedown.vx v17, v0, a2
-; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
-; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v17
+; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfabs.v v8, v24, v0.t
; ZVFHMIN-NEXT: lui a2, 307200
; ZVFHMIN-NEXT: fmv.w.x fa5, a2
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, mu
; ZVFHMIN-NEXT: vmflt.vf v17, v8, fa5, v0.t
; ZVFHMIN-NEXT: fsrmi a2, 1
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v17
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfcvt.x.f.v v8, v24, v0.t
; ZVFHMIN-NEXT: fsrm a2
; ZVFHMIN-NEXT: vfcvt.f.x.v v8, v8, v0.t
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, mu
; ZVFHMIN-NEXT: vfsgnj.vv v24, v8, v24, v0.t
; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v20, v24
+; ZVFHMIN-NEXT: vfncvt.f.f.w v12, v24
; ZVFHMIN-NEXT: bltu a0, a1, .LBB10_2
; ZVFHMIN-NEXT: # %bb.1:
; ZVFHMIN-NEXT: mv a0, a1
; ZVFHMIN-NEXT: .LBB10_2:
; ZVFHMIN-NEXT: addi a1, sp, 16
-; ZVFHMIN-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v8
-; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
+; ZVFHMIN-NEXT: vl8r.v v0, (a1) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v0
+; ZVFHMIN-NEXT: vmv1r.v v8, v16
; ZVFHMIN-NEXT: vmv1r.v v0, v16
-; ZVFHMIN-NEXT: vfabs.v v8, v24, v0.t
+; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
+; ZVFHMIN-NEXT: vfabs.v v16, v24, v0.t
+; ZVFHMIN-NEXT: addi a0, sp, 16
+; ZVFHMIN-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vmv1r.v v0, v8
+; ZVFHMIN-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, mu
-; ZVFHMIN-NEXT: vmflt.vf v16, v8, fa5, v0.t
+; ZVFHMIN-NEXT: vmflt.vf v8, v16, fa5, v0.t
; ZVFHMIN-NEXT: fsrmi a0, 1
+; ZVFHMIN-NEXT: vmv1r.v v0, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vmv1r.v v0, v16
-; ZVFHMIN-NEXT: vfcvt.x.f.v v8, v24, v0.t
+; ZVFHMIN-NEXT: vfcvt.x.f.v v16, v24, v0.t
+; ZVFHMIN-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: fsrm a0
-; ZVFHMIN-NEXT: vfcvt.f.x.v v8, v8, v0.t
+; ZVFHMIN-NEXT: addi a0, sp, 16
+; ZVFHMIN-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vfcvt.f.x.v v16, v16, v0.t
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, mu
-; ZVFHMIN-NEXT: vfsgnj.vv v24, v8, v24, v0.t
+; ZVFHMIN-NEXT: vfsgnj.vv v24, v16, v24, v0.t
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v16, v24
-; ZVFHMIN-NEXT: vmv8r.v v8, v16
+; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v24
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: add sp, sp, a0
@@ -556,20 +563,20 @@ define <vscale x 32 x half> @vp_roundtozero_nxv32f16_unmasked(<vscale x 32 x hal
; ZVFHMIN-NEXT: vmset.m v16
; ZVFHMIN-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
; ZVFHMIN-NEXT: vslidedown.vx v16, v16, a2
-; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
-; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v16
+; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfabs.v v8, v24, v0.t
; ZVFHMIN-NEXT: lui a2, 307200
; ZVFHMIN-NEXT: fmv.w.x fa5, a2
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, mu
; ZVFHMIN-NEXT: vmflt.vf v16, v8, fa5, v0.t
; ZVFHMIN-NEXT: fsrmi a2, 1
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v16
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfcvt.x.f.v v8, v24, v0.t
; ZVFHMIN-NEXT: fsrm a2
; ZVFHMIN-NEXT: vfcvt.f.x.v v8, v8, v0.t
@@ -701,8 +708,8 @@ define <vscale x 4 x float> @vp_roundtozero_nxv4f32(<vscale x 4 x float> %va, <v
; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, mu
; CHECK-NEXT: vmflt.vf v10, v12, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 1
-; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
@@ -745,8 +752,8 @@ define <vscale x 8 x float> @vp_roundtozero_nxv8f32(<vscale x 8 x float> %va, <v
; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, mu
; CHECK-NEXT: vmflt.vf v12, v16, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 1
-; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
@@ -789,8 +796,8 @@ define <vscale x 16 x float> @vp_roundtozero_nxv16f32(<vscale x 16 x float> %va,
; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, mu
; CHECK-NEXT: vmflt.vf v16, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 1
-; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
@@ -867,16 +874,16 @@ declare <vscale x 2 x double> @llvm.vp.roundtozero.nxv2f64(<vscale x 2 x double>
define <vscale x 2 x double> @vp_roundtozero_nxv2f64(<vscale x 2 x double> %va, <vscale x 2 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_roundtozero_nxv2f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v10, v0
; CHECK-NEXT: lui a1, %hi(.LCPI24_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI24_0)(a1)
+; CHECK-NEXT: vmv1r.v v10, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, ma
; CHECK-NEXT: vfabs.v v12, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, mu
; CHECK-NEXT: vmflt.vf v10, v12, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 1
-; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v12, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
@@ -911,16 +918,16 @@ declare <vscale x 4 x double> @llvm.vp.roundtozero.nxv4f64(<vscale x 4 x double>
define <vscale x 4 x double> @vp_roundtozero_nxv4f64(<vscale x 4 x double> %va, <vscale x 4 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_roundtozero_nxv4f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v12, v0
; CHECK-NEXT: lui a1, %hi(.LCPI26_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI26_0)(a1)
+; CHECK-NEXT: vmv1r.v v12, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, ma
; CHECK-NEXT: vfabs.v v16, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, mu
; CHECK-NEXT: vmflt.vf v12, v16, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 1
-; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
@@ -955,16 +962,16 @@ declare <vscale x 7 x double> @llvm.vp.roundtozero.nxv7f64(<vscale x 7 x double>
define <vscale x 7 x double> @vp_roundtozero_nxv7f64(<vscale x 7 x double> %va, <vscale x 7 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_roundtozero_nxv7f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: lui a1, %hi(.LCPI28_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI28_0)(a1)
+; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
; CHECK-NEXT: vmflt.vf v16, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 1
-; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
@@ -999,16 +1006,16 @@ declare <vscale x 8 x double> @llvm.vp.roundtozero.nxv8f64(<vscale x 8 x double>
define <vscale x 8 x double> @vp_roundtozero_nxv8f64(<vscale x 8 x double> %va, <vscale x 8 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_roundtozero_nxv8f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: lui a1, %hi(.LCPI30_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI30_0)(a1)
+; CHECK-NEXT: vmv1r.v v16, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
; CHECK-NEXT: vmflt.vf v16, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 1
-; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
@@ -1047,66 +1054,56 @@ define <vscale x 16 x double> @vp_roundtozero_nxv16f64(<vscale x 16 x double> %v
; CHECK-NEXT: addi sp, sp, -16
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a1, a1, 4
+; CHECK-NEXT: slli a1, a1, 3
; CHECK-NEXT: sub sp, sp, a1
-; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
-; CHECK-NEXT: vmv1r.v v24, v0
-; CHECK-NEXT: addi a1, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
+; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; CHECK-NEXT: vmv1r.v v7, v0
; CHECK-NEXT: csrr a1, vlenb
; CHECK-NEXT: srli a2, a1, 3
; CHECK-NEXT: vsetvli a3, zero, e8, mf4, ta, ma
-; CHECK-NEXT: vslidedown.vx v25, v0, a2
+; CHECK-NEXT: vslidedown.vx v6, v0, a2
; CHECK-NEXT: sub a2, a0, a1
+; CHECK-NEXT: lui a3, %hi(.LCPI32_0)
+; CHECK-NEXT: fld fa5, %lo(.LCPI32_0)(a3)
; CHECK-NEXT: sltu a3, a0, a2
; CHECK-NEXT: addi a3, a3, -1
; CHECK-NEXT: and a2, a3, a2
-; CHECK-NEXT: lui a3, %hi(.LCPI32_0)
-; CHECK-NEXT: fld fa5, %lo(.LCPI32_0)(a3)
+; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: vsetvli zero, a2, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v25
-; CHECK-NEXT: vfabs.v v8, v16, v0.t
+; CHECK-NEXT: vfabs.v v24, v16, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vmflt.vf v25, v8, fa5, v0.t
+; CHECK-NEXT: vmflt.vf v6, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a2, 1
+; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v25
-; CHECK-NEXT: vfcvt.x.f.v v8, v16, v0.t
+; CHECK-NEXT: vfcvt.x.f.v v24, v16, v0.t
+; CHECK-NEXT: addi a3, sp, 16
+; CHECK-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
; CHECK-NEXT: fsrm a2
-; CHECK-NEXT: vfcvt.f.x.v v8, v8, v0.t
+; CHECK-NEXT: addi a2, sp, 16
+; CHECK-NEXT: vl8r.v v24, (a2) # Unknown-size Folded Reload
+; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vfsgnj.vv v16, v8, v16, v0.t
-; CHECK-NEXT: csrr a2, vlenb
-; CHECK-NEXT: slli a2, a2, 3
-; CHECK-NEXT: add a2, sp, a2
-; CHECK-NEXT: addi a2, a2, 16
-; CHECK-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
+; CHECK-NEXT: vfsgnj.vv v16, v24, v16, v0.t
; CHECK-NEXT: bltu a0, a1, .LBB32_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB32_2:
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v24
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vfabs.v v16, v8, v0.t
+; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vmflt.vf v24, v16, fa5, v0.t
+; CHECK-NEXT: vmflt.vf v7, v24, fa5, v0.t
; CHECK-NEXT: fsrmi a0, 1
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v24
-; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
+; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: fsrm a0
-; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
+; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; CHECK-NEXT: vfsgnj.vv v8, v16, v8, v0.t
+; CHECK-NEXT: vfsgnj.vv v8, v24, v8, v0.t
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 4
; CHECK-NEXT: add sp, sp, a0
; CHECK-NEXT: addi sp, sp, 16
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/rv32-spill-vector-csr.ll b/llvm/test/CodeGen/RISCV/rvv/rv32-spill-vector-csr.ll
index e73415ac0085e..8210ea22a6ee9 100644
--- a/llvm/test/CodeGen/RISCV/rvv/rv32-spill-vector-csr.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/rv32-spill-vector-csr.ll
@@ -67,13 +67,13 @@ define <vscale x 1 x double> @foo(<vscale x 1 x double> %a, <vscale x 1 x double
; SPILL-O2-NEXT: lui a0, %hi(.L.str)
; SPILL-O2-NEXT: addi a0, a0, %lo(.L.str)
; SPILL-O2-NEXT: call puts
-; SPILL-O2-NEXT: vsetvli zero, s0, e64, m1, ta, ma
; SPILL-O2-NEXT: csrr a0, vlenb
; SPILL-O2-NEXT: add a0, sp, a0
; SPILL-O2-NEXT: addi a0, a0, 16
; SPILL-O2-NEXT: vl1r.v v8, (a0) # Unknown-size Folded Reload
; SPILL-O2-NEXT: addi a0, sp, 16
; SPILL-O2-NEXT: vl1r.v v9, (a0) # Unknown-size Folded Reload
+; SPILL-O2-NEXT: vsetvli zero, s0, e64, m1, ta, ma
; SPILL-O2-NEXT: vfadd.vv v8, v9, v8
; SPILL-O2-NEXT: csrr a0, vlenb
; SPILL-O2-NEXT: slli a0, a0, 1
diff --git a/llvm/test/CodeGen/RISCV/rvv/rv64-spill-vector-csr.ll b/llvm/test/CodeGen/RISCV/rvv/rv64-spill-vector-csr.ll
index 483f689cf633d..3523629088983 100644
--- a/llvm/test/CodeGen/RISCV/rvv/rv64-spill-vector-csr.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/rv64-spill-vector-csr.ll
@@ -70,13 +70,13 @@ define <vscale x 1 x double> @foo(<vscale x 1 x double> %a, <vscale x 1 x double
; SPILL-O2-NEXT: lui a0, %hi(.L.str)
; SPILL-O2-NEXT: addi a0, a0, %lo(.L.str)
; SPILL-O2-NEXT: call puts
-; SPILL-O2-NEXT: vsetvli zero, s0, e64, m1, ta, ma
; SPILL-O2-NEXT: csrr a0, vlenb
; SPILL-O2-NEXT: add a0, sp, a0
; SPILL-O2-NEXT: addi a0, a0, 16
; SPILL-O2-NEXT: vl1r.v v8, (a0) # Unknown-size Folded Reload
; SPILL-O2-NEXT: addi a0, sp, 16
; SPILL-O2-NEXT: vl1r.v v9, (a0) # Unknown-size Folded Reload
+; SPILL-O2-NEXT: vsetvli zero, s0, e64, m1, ta, ma
; SPILL-O2-NEXT: vfadd.vv v8, v9, v8
; SPILL-O2-NEXT: csrr a0, vlenb
; SPILL-O2-NEXT: slli a0, a0, 1
@@ -102,11 +102,11 @@ define <vscale x 1 x double> @foo(<vscale x 1 x double> %a, <vscale x 1 x double
; SPILL-O2-VLEN128-NEXT: lui a0, %hi(.L.str)
; SPILL-O2-VLEN128-NEXT: addi a0, a0, %lo(.L.str)
; SPILL-O2-VLEN128-NEXT: call puts
-; SPILL-O2-VLEN128-NEXT: vsetvli zero, s0, e64, m1, ta, ma
; SPILL-O2-VLEN128-NEXT: addi a0, sp, 32
; SPILL-O2-VLEN128-NEXT: vl1r.v v8, (a0) # Unknown-size Folded Reload
; SPILL-O2-VLEN128-NEXT: addi a0, sp, 16
; SPILL-O2-VLEN128-NEXT: vl1r.v v9, (a0) # Unknown-size Folded Reload
+; SPILL-O2-VLEN128-NEXT: vsetvli zero, s0, e64, m1, ta, ma
; SPILL-O2-VLEN128-NEXT: vfadd.vv v8, v9, v8
; SPILL-O2-VLEN128-NEXT: addi sp, sp, 32
; SPILL-O2-VLEN128-NEXT: ld ra, 24(sp) # 8-byte Folded Reload
diff --git a/llvm/test/CodeGen/RISCV/rvv/rvv-peephole-vmerge-vops.ll b/llvm/test/CodeGen/RISCV/rvv/rvv-peephole-vmerge-vops.ll
index 743016a7cbcdd..b9ede8d68e3cb 100644
--- a/llvm/test/CodeGen/RISCV/rvv/rvv-peephole-vmerge-vops.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/rvv-peephole-vmerge-vops.ll
@@ -893,8 +893,8 @@ define void @test_dag_loop() {
; CHECK-LABEL: test_dag_loop:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, zero, e8, m4, ta, ma
-; CHECK-NEXT: vmv.v.i v8, 0
; CHECK-NEXT: vmclr.m v0
+; CHECK-NEXT: vmv.v.i v8, 0
; CHECK-NEXT: vmv.v.i v12, 0
; CHECK-NEXT: vsetivli zero, 0, e8, m4, tu, mu
; CHECK-NEXT: vssubu.vx v12, v8, zero, v0.t
diff --git a/llvm/test/CodeGen/RISCV/rvv/setcc-fp-vp.ll b/llvm/test/CodeGen/RISCV/rvv/setcc-fp-vp.ll
index 897bfdea69f1f..3dba881363063 100644
--- a/llvm/test/CodeGen/RISCV/rvv/setcc-fp-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/setcc-fp-vp.ll
@@ -2203,28 +2203,27 @@ define <vscale x 64 x i1> @fcmp_oeq_vv_nxv64f16(<vscale x 64 x half> %va, <vscal
; ZVFH-NEXT: add a1, sp, a1
; ZVFH-NEXT: addi a1, a1, 16
; ZVFH-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
-; ZVFH-NEXT: csrr a1, vlenb
-; ZVFH-NEXT: vsetvli a3, zero, e8, m1, ta, ma
-; ZVFH-NEXT: slli a3, a1, 3
-; ZVFH-NEXT: add a3, a0, a3
-; ZVFH-NEXT: vl8re16.v v8, (a3)
-; ZVFH-NEXT: slli a3, a1, 2
+; ZVFH-NEXT: csrr a3, vlenb
+; ZVFH-NEXT: srli a1, a3, 1
+; ZVFH-NEXT: slli a4, a3, 3
+; ZVFH-NEXT: add a4, a0, a4
+; ZVFH-NEXT: vl8re16.v v8, (a4)
+; ZVFH-NEXT: slli a3, a3, 2
; ZVFH-NEXT: sub a4, a2, a3
; ZVFH-NEXT: sltu a5, a2, a4
; ZVFH-NEXT: addi a5, a5, -1
-; ZVFH-NEXT: and a4, a5, a4
-; ZVFH-NEXT: srli a1, a1, 1
; ZVFH-NEXT: vl8re16.v v0, (a0)
; ZVFH-NEXT: addi a0, sp, 16
; ZVFH-NEXT: vs8r.v v0, (a0) # Unknown-size Folded Spill
+; ZVFH-NEXT: vsetvli a0, zero, e8, m1, ta, ma
; ZVFH-NEXT: vslidedown.vx v0, v24, a1
+; ZVFH-NEXT: and a4, a5, a4
; ZVFH-NEXT: vsetvli zero, a4, e16, m8, ta, ma
; ZVFH-NEXT: vmfeq.vv v7, v16, v8, v0.t
; ZVFH-NEXT: bltu a2, a3, .LBB85_2
; ZVFH-NEXT: # %bb.1:
; ZVFH-NEXT: mv a2, a3
; ZVFH-NEXT: .LBB85_2:
-; ZVFH-NEXT: vsetvli zero, a2, e16, m8, ta, ma
; ZVFH-NEXT: vmv1r.v v0, v24
; ZVFH-NEXT: csrr a0, vlenb
; ZVFH-NEXT: slli a0, a0, 3
@@ -2233,6 +2232,7 @@ define <vscale x 64 x i1> @fcmp_oeq_vv_nxv64f16(<vscale x 64 x half> %va, <vscal
; ZVFH-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; ZVFH-NEXT: addi a0, sp, 16
; ZVFH-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; ZVFH-NEXT: vsetvli zero, a2, e16, m8, ta, ma
; ZVFH-NEXT: vmfeq.vv v16, v8, v24, v0.t
; ZVFH-NEXT: add a0, a1, a1
; ZVFH-NEXT: vsetvli zero, a0, e8, m1, ta, ma
@@ -2261,7 +2261,6 @@ define <vscale x 64 x i1> @fcmp_oeq_vv_nxv64f16(<vscale x 64 x half> %va, <vscal
; ZVFHMIN-NEXT: addi a1, a1, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: vsetvli a1, zero, e8, m1, ta, ma
; ZVFHMIN-NEXT: slli a1, a3, 3
; ZVFHMIN-NEXT: add a1, a0, a1
; ZVFHMIN-NEXT: vl8re16.v v16, (a1)
@@ -2280,8 +2279,13 @@ define <vscale x 64 x i1> @fcmp_oeq_vv_nxv64f16(<vscale x 64 x half> %va, <vscal
; ZVFHMIN-NEXT: add t0, sp, t0
; ZVFHMIN-NEXT: addi t0, t0, 16
; ZVFHMIN-NEXT: vs1r.v v0, (t0) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vslidedown.vx v7, v0, a1
+; ZVFHMIN-NEXT: vsetvli t0, zero, e8, m1, ta, ma
+; ZVFHMIN-NEXT: vslidedown.vx v8, v0, a1
; ZVFHMIN-NEXT: srli a3, a3, 2
+; ZVFHMIN-NEXT: addi t0, sp, 16
+; ZVFHMIN-NEXT: vs1r.v v8, (t0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vsetvli t0, zero, e8, mf2, ta, ma
+; ZVFHMIN-NEXT: vslidedown.vx v0, v8, a3
; ZVFHMIN-NEXT: vl8re16.v v8, (a0)
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: li t0, 26
@@ -2289,16 +2293,12 @@ define <vscale x 64 x i1> @fcmp_oeq_vv_nxv64f16(<vscale x 64 x half> %va, <vscal
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vsetvli a0, zero, e8, mf2, ta, ma
-; ZVFHMIN-NEXT: addi a0, sp, 16
-; ZVFHMIN-NEXT: vs1r.v v7, (a0) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vslidedown.vx v0, v7, a3
-; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 1
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v20
; ZVFHMIN-NEXT: vmv4r.v v16, v24
; ZVFHMIN-NEXT: csrr a0, vlenb
@@ -2314,12 +2314,12 @@ define <vscale x 64 x i1> @fcmp_oeq_vv_nxv64f16(<vscale x 64 x half> %va, <vscal
; ZVFHMIN-NEXT: # %bb.1:
; ZVFHMIN-NEXT: mv a6, a4
; ZVFHMIN-NEXT: .LBB85_2:
-; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 1
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v16
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: li a7, 10
@@ -2328,9 +2328,9 @@ define <vscale x 64 x i1> @fcmp_oeq_vv_nxv64f16(<vscale x 64 x half> %va, <vscal
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vl8r.v v0, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v0
-; ZVFHMIN-NEXT: vsetvli zero, a6, e32, m8, ta, ma
; ZVFHMIN-NEXT: addi a0, sp, 16
; ZVFHMIN-NEXT: vl1r.v v0, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vsetvli zero, a6, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmfeq.vv v6, v16, v8, v0.t
; ZVFHMIN-NEXT: add a0, a3, a3
; ZVFHMIN-NEXT: bltu a2, a5, .LBB85_4
@@ -2343,20 +2343,20 @@ define <vscale x 64 x i1> @fcmp_oeq_vv_nxv64f16(<vscale x 64 x half> %va, <vscal
; ZVFHMIN-NEXT: sltu a6, a2, a5
; ZVFHMIN-NEXT: addi a6, a6, -1
; ZVFHMIN-NEXT: and a5, a6, a5
-; ZVFHMIN-NEXT: vsetvli a6, zero, e8, mf2, ta, ma
; ZVFHMIN-NEXT: csrr a6, vlenb
; ZVFHMIN-NEXT: add a6, sp, a6
; ZVFHMIN-NEXT: addi a6, a6, 16
; ZVFHMIN-NEXT: vl1r.v v8, (a6) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vmv1r.v v7, v8
+; ZVFHMIN-NEXT: vsetvli a6, zero, e8, mf2, ta, ma
; ZVFHMIN-NEXT: vslidedown.vx v0, v8, a3
-; ZVFHMIN-NEXT: vsetvli a6, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: csrr a6, vlenb
; ZVFHMIN-NEXT: li a7, 18
; ZVFHMIN-NEXT: mul a6, a6, a7
; ZVFHMIN-NEXT: add a6, sp, a6
; ZVFHMIN-NEXT: addi a6, a6, 16
; ZVFHMIN-NEXT: vl8r.v v24, (a6) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vsetvli a6, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v28
; ZVFHMIN-NEXT: csrr a6, vlenb
; ZVFHMIN-NEXT: li a7, 10
@@ -2371,13 +2371,13 @@ define <vscale x 64 x i1> @fcmp_oeq_vv_nxv64f16(<vscale x 64 x half> %va, <vscal
; ZVFHMIN-NEXT: addi a6, a6, 16
; ZVFHMIN-NEXT: vl8r.v v16, (a6) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v20
+; ZVFHMIN-NEXT: csrr a6, vlenb
+; ZVFHMIN-NEXT: li a7, 10
+; ZVFHMIN-NEXT: mul a6, a6, a7
+; ZVFHMIN-NEXT: add a6, sp, a6
+; ZVFHMIN-NEXT: addi a6, a6, 16
+; ZVFHMIN-NEXT: vl8r.v v16, (a6) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, a5, e32, m8, ta, ma
-; ZVFHMIN-NEXT: csrr a5, vlenb
-; ZVFHMIN-NEXT: li a6, 10
-; ZVFHMIN-NEXT: mul a5, a5, a6
-; ZVFHMIN-NEXT: add a5, sp, a5
-; ZVFHMIN-NEXT: addi a5, a5, 16
-; ZVFHMIN-NEXT: vl8r.v v16, (a5) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vmfeq.vv v5, v16, v8, v0.t
; ZVFHMIN-NEXT: bltu a2, a4, .LBB85_6
; ZVFHMIN-NEXT: # %bb.5:
@@ -2392,8 +2392,8 @@ define <vscale x 64 x i1> @fcmp_oeq_vv_nxv64f16(<vscale x 64 x half> %va, <vscal
; ZVFHMIN-NEXT: addi a4, a4, 16
; ZVFHMIN-NEXT: vl8r.v v8, (a4) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v8
-; ZVFHMIN-NEXT: vsetvli zero, a2, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v7
+; ZVFHMIN-NEXT: vsetvli zero, a2, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmfeq.vv v8, v16, v24, v0.t
; ZVFHMIN-NEXT: vsetvli zero, a0, e8, mf2, ta, ma
; ZVFHMIN-NEXT: vslideup.vx v8, v5, a3
@@ -3510,48 +3510,48 @@ define <vscale x 32 x i1> @fcmp_oeq_vv_nxv32f64(<vscale x 32 x double> %va, <vsc
; CHECK-NEXT: add a1, sp, a1
; CHECK-NEXT: addi a1, a1, 16
; CHECK-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
-; CHECK-NEXT: csrr a3, vlenb
+; CHECK-NEXT: csrr a4, vlenb
; CHECK-NEXT: li a1, 24
-; CHECK-NEXT: mul t2, a3, a1
-; CHECK-NEXT: slli t1, a3, 3
-; CHECK-NEXT: srli a4, a3, 2
-; CHECK-NEXT: vsetvli a1, zero, e8, mf2, ta, ma
-; CHECK-NEXT: vslidedown.vx v7, v0, a4
-; CHECK-NEXT: srli a1, a3, 3
-; CHECK-NEXT: vsetvli a5, zero, e8, mf4, ta, ma
+; CHECK-NEXT: mul t0, a4, a1
+; CHECK-NEXT: slli t1, a4, 3
+; CHECK-NEXT: srli a1, a4, 2
+; CHECK-NEXT: vsetvli a3, zero, e8, mf2, ta, ma
+; CHECK-NEXT: vslidedown.vx v7, v0, a1
+; CHECK-NEXT: srli a3, a4, 3
; CHECK-NEXT: add a5, a2, t1
; CHECK-NEXT: vl8re64.v v8, (a5)
-; CHECK-NEXT: slli t0, a3, 4
-; CHECK-NEXT: slli a5, a3, 1
-; CHECK-NEXT: vslidedown.vx v0, v0, a1
+; CHECK-NEXT: slli t3, a4, 4
+; CHECK-NEXT: slli a5, a4, 1
+; CHECK-NEXT: vsetvli a7, zero, e8, mf4, ta, ma
+; CHECK-NEXT: vslidedown.vx v0, v0, a3
; CHECK-NEXT: mv a7, a6
; CHECK-NEXT: bltu a6, a5, .LBB171_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a7, a5
; CHECK-NEXT: .LBB171_2:
-; CHECK-NEXT: add t2, a2, t2
+; CHECK-NEXT: add t2, a2, t0
; CHECK-NEXT: add t1, a0, t1
-; CHECK-NEXT: add t0, a2, t0
+; CHECK-NEXT: add t0, a2, t3
; CHECK-NEXT: vl8re64.v v16, (a2)
; CHECK-NEXT: csrr a2, vlenb
; CHECK-NEXT: slli a2, a2, 4
; CHECK-NEXT: add a2, sp, a2
; CHECK-NEXT: addi a2, a2, 16
; CHECK-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
-; CHECK-NEXT: sub a2, a7, a3
+; CHECK-NEXT: sub a2, a7, a4
; CHECK-NEXT: sltu t3, a7, a2
; CHECK-NEXT: addi t3, t3, -1
; CHECK-NEXT: and a2, t3, a2
+; CHECK-NEXT: csrr t3, vlenb
+; CHECK-NEXT: slli t3, t3, 5
+; CHECK-NEXT: add t3, sp, t3
+; CHECK-NEXT: addi t3, t3, 16
+; CHECK-NEXT: vl8r.v v16, (t3) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, a2, e64, m8, ta, ma
-; CHECK-NEXT: csrr a2, vlenb
-; CHECK-NEXT: slli a2, a2, 5
-; CHECK-NEXT: add a2, sp, a2
-; CHECK-NEXT: addi a2, a2, 16
-; CHECK-NEXT: vl8r.v v16, (a2) # Unknown-size Folded Reload
; CHECK-NEXT: vmfeq.vv v6, v16, v8, v0.t
-; CHECK-NEXT: bltu a7, a3, .LBB171_4
+; CHECK-NEXT: bltu a7, a4, .LBB171_4
; CHECK-NEXT: # %bb.3:
-; CHECK-NEXT: mv a7, a3
+; CHECK-NEXT: mv a7, a4
; CHECK-NEXT: .LBB171_4:
; CHECK-NEXT: vl8re64.v v8, (t2)
; CHECK-NEXT: csrr a2, vlenb
@@ -3567,7 +3567,7 @@ define <vscale x 32 x i1> @fcmp_oeq_vv_nxv32f64(<vscale x 32 x double> %va, <vsc
; CHECK-NEXT: addi a2, a2, 16
; CHECK-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
; CHECK-NEXT: vsetvli a2, zero, e8, mf4, ta, ma
-; CHECK-NEXT: vslidedown.vx v18, v7, a1
+; CHECK-NEXT: vslidedown.vx v18, v7, a3
; CHECK-NEXT: vl8re64.v v8, (t0)
; CHECK-NEXT: csrr a2, vlenb
; CHECK-NEXT: slli a2, a2, 3
@@ -3577,7 +3577,6 @@ define <vscale x 32 x i1> @fcmp_oeq_vv_nxv32f64(<vscale x 32 x double> %va, <vsc
; CHECK-NEXT: vl8re64.v v8, (a0)
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: vsetvli zero, a7, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: li a2, 40
@@ -3590,55 +3589,56 @@ define <vscale x 32 x i1> @fcmp_oeq_vv_nxv32f64(<vscale x 32 x double> %va, <vsc
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, a7, e64, m8, ta, ma
; CHECK-NEXT: vmfeq.vv v17, v24, v8, v0.t
-; CHECK-NEXT: add a0, a1, a1
-; CHECK-NEXT: vsetvli zero, a0, e8, mf2, tu, ma
+; CHECK-NEXT: add a2, a3, a3
; CHECK-NEXT: sub a0, a6, a5
-; CHECK-NEXT: sltu a2, a6, a0
-; CHECK-NEXT: addi a2, a2, -1
-; CHECK-NEXT: and a0, a2, a0
-; CHECK-NEXT: vslideup.vx v17, v6, a1
+; CHECK-NEXT: sltu a5, a6, a0
+; CHECK-NEXT: addi a5, a5, -1
+; CHECK-NEXT: and a0, a5, a0
+; CHECK-NEXT: vsetvli zero, a2, e8, mf2, tu, ma
+; CHECK-NEXT: vslideup.vx v17, v6, a3
; CHECK-NEXT: mv a2, a0
-; CHECK-NEXT: bltu a0, a3, .LBB171_6
+; CHECK-NEXT: bltu a0, a4, .LBB171_6
; CHECK-NEXT: # %bb.5:
-; CHECK-NEXT: mv a2, a3
+; CHECK-NEXT: mv a2, a4
; CHECK-NEXT: .LBB171_6:
-; CHECK-NEXT: vsetvli zero, a2, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v7
-; CHECK-NEXT: csrr a2, vlenb
-; CHECK-NEXT: slli a2, a2, 3
-; CHECK-NEXT: add a2, sp, a2
-; CHECK-NEXT: addi a2, a2, 16
-; CHECK-NEXT: vl8r.v v8, (a2) # Unknown-size Folded Reload
-; CHECK-NEXT: addi a2, sp, 16
-; CHECK-NEXT: vl8r.v v24, (a2) # Unknown-size Folded Reload
+; CHECK-NEXT: csrr a5, vlenb
+; CHECK-NEXT: slli a5, a5, 3
+; CHECK-NEXT: add a5, sp, a5
+; CHECK-NEXT: addi a5, a5, 16
+; CHECK-NEXT: vl8r.v v8, (a5) # Unknown-size Folded Reload
+; CHECK-NEXT: addi a5, sp, 16
+; CHECK-NEXT: vl8r.v v24, (a5) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, a2, e64, m8, ta, ma
; CHECK-NEXT: vmfeq.vv v16, v24, v8, v0.t
-; CHECK-NEXT: add a2, a4, a1
-; CHECK-NEXT: vsetvli zero, a2, e8, mf2, tu, ma
-; CHECK-NEXT: vslideup.vx v17, v16, a4
-; CHECK-NEXT: sub a2, a0, a3
+; CHECK-NEXT: sub a2, a0, a4
; CHECK-NEXT: sltu a0, a0, a2
; CHECK-NEXT: addi a0, a0, -1
; CHECK-NEXT: and a0, a0, a2
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v18
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 5
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a2, 24
-; CHECK-NEXT: mul a0, a0, a2
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vmfeq.vv v16, v8, v24, v0.t
-; CHECK-NEXT: slli a0, a1, 1
-; CHECK-NEXT: add a0, a0, a1
-; CHECK-NEXT: add a1, a0, a1
-; CHECK-NEXT: vsetvli zero, a1, e8, mf2, ta, ma
-; CHECK-NEXT: vslideup.vx v17, v16, a0
+; CHECK-NEXT: csrr a2, vlenb
+; CHECK-NEXT: slli a2, a2, 5
+; CHECK-NEXT: add a2, sp, a2
+; CHECK-NEXT: addi a2, a2, 16
+; CHECK-NEXT: vl8r.v v24, (a2) # Unknown-size Folded Reload
+; CHECK-NEXT: csrr a2, vlenb
+; CHECK-NEXT: li a4, 24
+; CHECK-NEXT: mul a2, a2, a4
+; CHECK-NEXT: add a2, sp, a2
+; CHECK-NEXT: addi a2, a2, 16
+; CHECK-NEXT: vl8r.v v8, (a2) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT: vmfeq.vv v18, v8, v24, v0.t
+; CHECK-NEXT: add a0, a1, a3
+; CHECK-NEXT: vsetvli zero, a0, e8, mf2, tu, ma
+; CHECK-NEXT: vslideup.vx v17, v16, a1
+; CHECK-NEXT: slli a0, a3, 1
+; CHECK-NEXT: add a0, a0, a3
+; CHECK-NEXT: add a3, a0, a3
+; CHECK-NEXT: vsetvli zero, a3, e8, mf2, ta, ma
+; CHECK-NEXT: vslideup.vx v17, v18, a0
; CHECK-NEXT: vmv1r.v v0, v17
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: li a1, 48
diff --git a/llvm/test/CodeGen/RISCV/rvv/setcc-int-vp.ll b/llvm/test/CodeGen/RISCV/rvv/setcc-int-vp.ll
index 7fd77c050b295..eb8c58d2d3779 100644
--- a/llvm/test/CodeGen/RISCV/rvv/setcc-int-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/setcc-int-vp.ll
@@ -1092,7 +1092,7 @@ define <vscale x 128 x i1> @icmp_eq_vv_nxv128i8(<vscale x 128 x i8> %va, <vscale
; CHECK-NEXT: slli a1, a1, 4
; CHECK-NEXT: sub sp, sp, a1
; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
-; CHECK-NEXT: vmv1r.v v24, v0
+; CHECK-NEXT: vmv1r.v v7, v0
; CHECK-NEXT: csrr a1, vlenb
; CHECK-NEXT: slli a1, a1, 3
; CHECK-NEXT: add a1, sp, a1
@@ -1102,23 +1102,22 @@ define <vscale x 128 x i1> @icmp_eq_vv_nxv128i8(<vscale x 128 x i8> %va, <vscale
; CHECK-NEXT: slli a1, a1, 3
; CHECK-NEXT: add a4, a0, a1
; CHECK-NEXT: vl8r.v v8, (a4)
-; CHECK-NEXT: vl8r.v v0, (a0)
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vs8r.v v0, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: vsetvli a0, zero, e8, m8, ta, ma
-; CHECK-NEXT: sub a0, a3, a1
+; CHECK-NEXT: vsetvli a4, zero, e8, m8, ta, ma
; CHECK-NEXT: vlm.v v0, (a2)
-; CHECK-NEXT: sltu a2, a3, a0
-; CHECK-NEXT: addi a2, a2, -1
-; CHECK-NEXT: and a0, a2, a0
-; CHECK-NEXT: vsetvli zero, a0, e8, m8, ta, ma
-; CHECK-NEXT: vmseq.vv v7, v16, v8, v0.t
+; CHECK-NEXT: sub a2, a3, a1
+; CHECK-NEXT: sltu a4, a3, a2
+; CHECK-NEXT: vl8r.v v24, (a0)
+; CHECK-NEXT: addi a0, sp, 16
+; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: addi a4, a4, -1
+; CHECK-NEXT: and a2, a4, a2
+; CHECK-NEXT: vsetvli zero, a2, e8, m8, ta, ma
+; CHECK-NEXT: vmseq.vv v6, v16, v8, v0.t
; CHECK-NEXT: bltu a3, a1, .LBB96_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a3, a1
; CHECK-NEXT: .LBB96_2:
-; CHECK-NEXT: vsetvli zero, a3, e8, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
@@ -1126,9 +1125,10 @@ define <vscale x 128 x i1> @icmp_eq_vv_nxv128i8(<vscale x 128 x i8> %va, <vscale
; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, a3, e8, m8, ta, ma
; CHECK-NEXT: vmseq.vv v16, v8, v24, v0.t
; CHECK-NEXT: vmv1r.v v0, v16
-; CHECK-NEXT: vmv1r.v v8, v7
+; CHECK-NEXT: vmv1r.v v8, v6
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 4
; CHECK-NEXT: add sp, sp, a0
@@ -1156,8 +1156,8 @@ define <vscale x 128 x i1> @icmp_eq_vx_nxv128i8(<vscale x 128 x i8> %va, i8 %b,
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a2, a1
; CHECK-NEXT: .LBB97_2:
-; CHECK-NEXT: vsetvli zero, a2, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a2, e8, m8, ta, ma
; CHECK-NEXT: vmseq.vx v16, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v16
; CHECK-NEXT: vmv1r.v v8, v25
@@ -1186,8 +1186,8 @@ define <vscale x 128 x i1> @icmp_eq_vx_swap_nxv128i8(<vscale x 128 x i8> %va, i8
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a2, a1
; CHECK-NEXT: .LBB98_2:
-; CHECK-NEXT: vsetvli zero, a2, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a2, e8, m8, ta, ma
; CHECK-NEXT: vmseq.vx v16, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v16
; CHECK-NEXT: vmv1r.v v8, v25
@@ -2248,28 +2248,27 @@ define <vscale x 32 x i1> @icmp_eq_vv_nxv32i32(<vscale x 32 x i32> %va, <vscale
; CHECK-NEXT: add a1, sp, a1
; CHECK-NEXT: addi a1, a1, 16
; CHECK-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
-; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: vsetvli a3, zero, e8, mf2, ta, ma
-; CHECK-NEXT: slli a3, a1, 3
-; CHECK-NEXT: add a3, a0, a3
-; CHECK-NEXT: vl8re32.v v8, (a3)
-; CHECK-NEXT: slli a3, a1, 1
+; CHECK-NEXT: csrr a3, vlenb
+; CHECK-NEXT: srli a1, a3, 2
+; CHECK-NEXT: slli a4, a3, 3
+; CHECK-NEXT: add a4, a0, a4
+; CHECK-NEXT: vl8re32.v v8, (a4)
+; CHECK-NEXT: slli a3, a3, 1
; CHECK-NEXT: sub a4, a2, a3
; CHECK-NEXT: sltu a5, a2, a4
; CHECK-NEXT: addi a5, a5, -1
-; CHECK-NEXT: and a4, a5, a4
-; CHECK-NEXT: srli a1, a1, 2
; CHECK-NEXT: vl8re32.v v0, (a0)
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vs8r.v v0, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vsetvli a0, zero, e8, mf2, ta, ma
; CHECK-NEXT: vslidedown.vx v0, v24, a1
+; CHECK-NEXT: and a4, a5, a4
; CHECK-NEXT: vsetvli zero, a4, e32, m8, ta, ma
; CHECK-NEXT: vmseq.vv v7, v16, v8, v0.t
; CHECK-NEXT: bltu a2, a3, .LBB189_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a2, a3
; CHECK-NEXT: .LBB189_2:
-; CHECK-NEXT: vsetvli zero, a2, e32, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
@@ -2278,6 +2277,7 @@ define <vscale x 32 x i1> @icmp_eq_vv_nxv32i32(<vscale x 32 x i32> %va, <vscale
; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, a2, e32, m8, ta, ma
; CHECK-NEXT: vmseq.vv v16, v8, v24, v0.t
; CHECK-NEXT: add a0, a1, a1
; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, ma
@@ -2311,8 +2311,8 @@ define <vscale x 32 x i1> @icmp_eq_vx_nxv32i32(<vscale x 32 x i32> %va, i32 %b,
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a1, a3
; CHECK-NEXT: .LBB190_2:
-; CHECK-NEXT: vsetvli zero, a1, e32, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a1, e32, m8, ta, ma
; CHECK-NEXT: vmseq.vx v16, v8, a0, v0.t
; CHECK-NEXT: add a0, a2, a2
; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, ma
@@ -2344,8 +2344,8 @@ define <vscale x 32 x i1> @icmp_eq_vx_swap_nxv32i32(<vscale x 32 x i32> %va, i32
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a1, a3
; CHECK-NEXT: .LBB191_2:
-; CHECK-NEXT: vsetvli zero, a1, e32, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a1, e32, m8, ta, ma
; CHECK-NEXT: vmseq.vx v16, v8, a0, v0.t
; CHECK-NEXT: add a0, a2, a2
; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, ma
diff --git a/llvm/test/CodeGen/RISCV/rvv/shuffle-reverse.ll b/llvm/test/CodeGen/RISCV/rvv/shuffle-reverse.ll
index ab7da9e0faf2b..6e327457bebff 100644
--- a/llvm/test/CodeGen/RISCV/rvv/shuffle-reverse.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/shuffle-reverse.ll
@@ -49,8 +49,8 @@ define <8 x i8> @v4i8_2(<4 x i8> %a, <4 x i8> %b) {
; CHECK-NEXT: vid.v v11
; CHECK-NEXT: vrsub.vi v12, v11, 7
; CHECK-NEXT: vrgather.vv v10, v8, v12
-; CHECK-NEXT: vrsub.vi v8, v11, 3
; CHECK-NEXT: vmv.v.i v0, 15
+; CHECK-NEXT: vrsub.vi v8, v11, 3
; CHECK-NEXT: vrgather.vv v10, v9, v8, v0.t
; CHECK-NEXT: vmv1r.v v8, v10
; CHECK-NEXT: ret
@@ -78,11 +78,11 @@ define <16 x i8> @v8i8_2(<8 x i8> %a, <8 x i8> %b) {
; CHECK-NEXT: vid.v v11
; CHECK-NEXT: vrsub.vi v12, v11, 15
; CHECK-NEXT: vrgather.vv v10, v8, v12
-; CHECK-NEXT: vrsub.vi v8, v11, 7
; CHECK-NEXT: li a0, 255
; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; CHECK-NEXT: vmv.s.x v0, a0
; CHECK-NEXT: vsetvli zero, zero, e8, m1, ta, mu
+; CHECK-NEXT: vrsub.vi v8, v11, 7
; CHECK-NEXT: vrgather.vv v10, v9, v8, v0.t
; CHECK-NEXT: vmv.v.v v8, v10
; CHECK-NEXT: ret
@@ -174,8 +174,8 @@ define <8 x i16> @v4i16_2(<4 x i16> %a, <4 x i16> %b) {
; CHECK-NEXT: vid.v v11
; CHECK-NEXT: vrsub.vi v12, v11, 7
; CHECK-NEXT: vrgather.vv v10, v8, v12
-; CHECK-NEXT: vrsub.vi v8, v11, 3
; CHECK-NEXT: vmv.v.i v0, 15
+; CHECK-NEXT: vrsub.vi v8, v11, 3
; CHECK-NEXT: vrgather.vv v10, v9, v8, v0.t
; CHECK-NEXT: vmv.v.v v8, v10
; CHECK-NEXT: ret
@@ -234,10 +234,10 @@ define <32 x i16> @v16i16_2(<16 x i16> %a, <16 x i16> %b) {
; CHECK-NEXT: addi a0, a0, %lo(.LCPI15_0)
; CHECK-NEXT: li a1, 32
; CHECK-NEXT: vsetvli zero, a1, e16, m4, ta, ma
-; CHECK-NEXT: vle16.v v20, (a0)
-; CHECK-NEXT: vmv2r.v v16, v10
+; CHECK-NEXT: vle16.v v16, (a0)
+; CHECK-NEXT: vmv2r.v v20, v10
; CHECK-NEXT: vmv2r.v v12, v8
-; CHECK-NEXT: vrgather.vv v8, v12, v20
+; CHECK-NEXT: vrgather.vv v8, v12, v16
; CHECK-NEXT: vid.v v12
; CHECK-NEXT: vrsub.vi v12, v12, 15
; CHECK-NEXT: lui a0, 16
@@ -245,7 +245,7 @@ define <32 x i16> @v16i16_2(<16 x i16> %a, <16 x i16> %b) {
; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; CHECK-NEXT: vmv.s.x v0, a0
; CHECK-NEXT: vsetvli zero, zero, e16, m4, ta, mu
-; CHECK-NEXT: vrgather.vv v8, v16, v12, v0.t
+; CHECK-NEXT: vrgather.vv v8, v20, v12, v0.t
; CHECK-NEXT: ret
%v32i16 = shufflevector <16 x i16> %a, <16 x i16> %b, <32 x i32> <i32 31, i32 30, i32 29, i32 28, i32 27, i32 26, i32 25, i32 24, i32 23, i32 22, i32 21, i32 20, i32 19, i32 18, i32 17, i32 16, i32 15, i32 14, i32 13, i32 12, i32 11, i32 10, i32 9, i32 8, i32 7, i32 6, i32 5, i32 4, i32 3, i32 2, i32 1, i32 0>
ret <32 x i16> %v32i16
@@ -329,18 +329,18 @@ define <16 x i32> @v8i32_2(<8 x i32> %a, <8 x i32> %b) {
; CHECK-LABEL: v8i32_2:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv2r.v v16, v10
-; CHECK-NEXT: vmv2r.v v12, v8
; CHECK-NEXT: vsetivli zero, 16, e16, m2, ta, ma
-; CHECK-NEXT: vid.v v14
-; CHECK-NEXT: vrsub.vi v18, v14, 15
+; CHECK-NEXT: vid.v v10
+; CHECK-NEXT: vrsub.vi v18, v10, 15
; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, ma
-; CHECK-NEXT: vrgatherei16.vv v8, v12, v18
+; CHECK-NEXT: vrgatherei16.vv v12, v8, v18
; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, ma
-; CHECK-NEXT: vrsub.vi v12, v14, 7
+; CHECK-NEXT: vrsub.vi v8, v10, 7
; CHECK-NEXT: li a0, 255
; CHECK-NEXT: vmv.s.x v0, a0
; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, mu
-; CHECK-NEXT: vrgatherei16.vv v8, v16, v12, v0.t
+; CHECK-NEXT: vrgatherei16.vv v12, v16, v8, v0.t
+; CHECK-NEXT: vmv.v.v v8, v12
; CHECK-NEXT: ret
%v16i32 = shufflevector <8 x i32> %a, <8 x i32> %b, <16 x i32> <i32 15, i32 14, i32 13, i32 12, i32 11, i32 10, i32 9, i32 8, i32 7, i32 6, i32 5, i32 4, i32 3, i32 2, i32 1, i32 0>
ret <16 x i32> %v16i32
@@ -492,8 +492,8 @@ define <8 x half> @v4f16_2(<4 x half> %a, <4 x half> %b) {
; CHECK-NEXT: vid.v v11
; CHECK-NEXT: vrsub.vi v12, v11, 7
; CHECK-NEXT: vrgather.vv v10, v8, v12
-; CHECK-NEXT: vrsub.vi v8, v11, 3
; CHECK-NEXT: vmv.v.i v0, 15
+; CHECK-NEXT: vrsub.vi v8, v11, 3
; CHECK-NEXT: vrgather.vv v10, v9, v8, v0.t
; CHECK-NEXT: vmv.v.v v8, v10
; CHECK-NEXT: ret
@@ -638,18 +638,18 @@ define <16 x float> @v8f32_2(<8 x float> %a, <8 x float> %b) {
; CHECK-LABEL: v8f32_2:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv2r.v v16, v10
-; CHECK-NEXT: vmv2r.v v12, v8
; CHECK-NEXT: vsetivli zero, 16, e16, m2, ta, ma
-; CHECK-NEXT: vid.v v14
-; CHECK-NEXT: vrsub.vi v18, v14, 15
+; CHECK-NEXT: vid.v v10
+; CHECK-NEXT: vrsub.vi v18, v10, 15
; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, ma
-; CHECK-NEXT: vrgatherei16.vv v8, v12, v18
+; CHECK-NEXT: vrgatherei16.vv v12, v8, v18
; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, ma
-; CHECK-NEXT: vrsub.vi v12, v14, 7
+; CHECK-NEXT: vrsub.vi v8, v10, 7
; CHECK-NEXT: li a0, 255
; CHECK-NEXT: vmv.s.x v0, a0
; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, mu
-; CHECK-NEXT: vrgatherei16.vv v8, v16, v12, v0.t
+; CHECK-NEXT: vrgatherei16.vv v12, v16, v8, v0.t
+; CHECK-NEXT: vmv.v.v v8, v12
; CHECK-NEXT: ret
%v16f32 = shufflevector <8 x float> %a, <8 x float> %b, <16 x i32> <i32 15, i32 14, i32 13, i32 12, i32 11, i32 10, i32 9, i32 8, i32 7, i32 6, i32 5, i32 4, i32 3, i32 2, i32 1, i32 0>
ret <16 x float> %v16f32
diff --git a/llvm/test/CodeGen/RISCV/rvv/sink-splat-operands.ll b/llvm/test/CodeGen/RISCV/rvv/sink-splat-operands.ll
index 8a297db7a3b8f..d1c98f828e76d 100644
--- a/llvm/test/CodeGen/RISCV/rvv/sink-splat-operands.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/sink-splat-operands.ll
@@ -4873,8 +4873,8 @@ define void @sink_splat_vp_icmp(ptr nocapture %x, i32 signext %y, <4 x i1> %m, i
; CHECK-NEXT: .LBB102_1: # %vector.body
; CHECK-NEXT: # =>This Inner Loop Header: Depth=1
; CHECK-NEXT: vle32.v v10, (a0)
-; CHECK-NEXT: vsetvli zero, a2, e32, m1, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a2, e32, m1, ta, ma
; CHECK-NEXT: vmseq.vx v0, v10, a1, v0.t
; CHECK-NEXT: vsetivli zero, 4, e32, m1, ta, ma
; CHECK-NEXT: vse32.v v9, (a0), v0.t
@@ -4914,8 +4914,8 @@ define void @sink_splat_vp_fcmp(ptr nocapture %x, float %y, <4 x i1> %m, i32 zer
; CHECK-NEXT: .LBB103_1: # %vector.body
; CHECK-NEXT: # =>This Inner Loop Header: Depth=1
; CHECK-NEXT: vle32.v v10, (a0)
-; CHECK-NEXT: vsetvli zero, a1, e32, m1, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e32, m1, ta, ma
; CHECK-NEXT: vmfeq.vf v0, v10, fa0, v0.t
; CHECK-NEXT: vsetivli zero, 4, e32, m1, ta, ma
; CHECK-NEXT: vse32.v v9, (a0), v0.t
diff --git a/llvm/test/CodeGen/RISCV/rvv/sshl_sat_vec.ll b/llvm/test/CodeGen/RISCV/rvv/sshl_sat_vec.ll
index b7fe722958bfb..9d0234d2ec2fb 100644
--- a/llvm/test/CodeGen/RISCV/rvv/sshl_sat_vec.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/sshl_sat_vec.ll
@@ -33,13 +33,13 @@ define <4 x i32> @vec_v4i32(<4 x i32> %x, <4 x i32> %y) nounwind {
; CHECK-NEXT: vmsle.vi v0, v8, -1
; CHECK-NEXT: lui a0, 524288
; CHECK-NEXT: addi a1, a0, -1
-; CHECK-NEXT: vsll.vv v10, v8, v9
-; CHECK-NEXT: vsra.vv v9, v10, v9
+; CHECK-NEXT: vmv.v.x v10, a1
+; CHECK-NEXT: vsll.vv v11, v8, v9
+; CHECK-NEXT: vsra.vv v9, v11, v9
; CHECK-NEXT: vmsne.vv v8, v8, v9
-; CHECK-NEXT: vmv.v.x v9, a1
-; CHECK-NEXT: vmerge.vxm v9, v9, a0, v0
+; CHECK-NEXT: vmerge.vxm v9, v10, a0, v0
; CHECK-NEXT: vmv.v.v v0, v8
-; CHECK-NEXT: vmerge.vvm v8, v10, v9, v0
+; CHECK-NEXT: vmerge.vvm v8, v11, v9, v0
; CHECK-NEXT: ret
%tmp = call <4 x i32> @llvm.sshl.sat.v4i32(<4 x i32> %x, <4 x i32> %y)
ret <4 x i32> %tmp
@@ -52,13 +52,13 @@ define <8 x i16> @vec_v8i16(<8 x i16> %x, <8 x i16> %y) nounwind {
; CHECK-NEXT: vmsle.vi v0, v8, -1
; CHECK-NEXT: lui a0, 8
; CHECK-NEXT: addi a1, a0, -1
-; CHECK-NEXT: vsll.vv v10, v8, v9
-; CHECK-NEXT: vsra.vv v9, v10, v9
+; CHECK-NEXT: vmv.v.x v10, a1
+; CHECK-NEXT: vsll.vv v11, v8, v9
+; CHECK-NEXT: vsra.vv v9, v11, v9
; CHECK-NEXT: vmsne.vv v8, v8, v9
-; CHECK-NEXT: vmv.v.x v9, a1
-; CHECK-NEXT: vmerge.vxm v9, v9, a0, v0
+; CHECK-NEXT: vmerge.vxm v9, v10, a0, v0
; CHECK-NEXT: vmv.v.v v0, v8
-; CHECK-NEXT: vmerge.vvm v8, v10, v9, v0
+; CHECK-NEXT: vmerge.vvm v8, v11, v9, v0
; CHECK-NEXT: ret
%tmp = call <8 x i16> @llvm.sshl.sat.v8i16(<8 x i16> %x, <8 x i16> %y)
ret <8 x i16> %tmp
@@ -70,14 +70,14 @@ define <16 x i8> @vec_v16i8(<16 x i8> %x, <16 x i8> %y) nounwind {
; CHECK-NEXT: vsetivli zero, 16, e8, m1, ta, ma
; CHECK-NEXT: vmsle.vi v0, v8, -1
; CHECK-NEXT: li a0, 127
-; CHECK-NEXT: vsll.vv v10, v8, v9
-; CHECK-NEXT: vsra.vv v9, v10, v9
-; CHECK-NEXT: vmsne.vv v8, v8, v9
-; CHECK-NEXT: vmv.v.x v9, a0
+; CHECK-NEXT: vmv.v.x v10, a0
; CHECK-NEXT: li a0, 128
-; CHECK-NEXT: vmerge.vxm v9, v9, a0, v0
+; CHECK-NEXT: vsll.vv v11, v8, v9
+; CHECK-NEXT: vsra.vv v9, v11, v9
+; CHECK-NEXT: vmsne.vv v8, v8, v9
+; CHECK-NEXT: vmerge.vxm v9, v10, a0, v0
; CHECK-NEXT: vmv.v.v v0, v8
-; CHECK-NEXT: vmerge.vvm v8, v10, v9, v0
+; CHECK-NEXT: vmerge.vvm v8, v11, v9, v0
; CHECK-NEXT: ret
%tmp = call <16 x i8> @llvm.sshl.sat.v16i8(<16 x i8> %x, <16 x i8> %y)
ret <16 x i8> %tmp
@@ -115,13 +115,13 @@ define <vscale x 4 x i32> @vec_nxv4i32(<vscale x 4 x i32> %x, <vscale x 4 x i32>
; CHECK-NEXT: vmsle.vi v0, v8, -1
; CHECK-NEXT: lui a0, 524288
; CHECK-NEXT: addi a1, a0, -1
-; CHECK-NEXT: vsll.vv v12, v8, v10
-; CHECK-NEXT: vsra.vv v14, v12, v10
-; CHECK-NEXT: vmsne.vv v10, v8, v14
-; CHECK-NEXT: vmv.v.x v8, a1
-; CHECK-NEXT: vmerge.vxm v8, v8, a0, v0
+; CHECK-NEXT: vmv.v.x v12, a1
+; CHECK-NEXT: vsll.vv v14, v8, v10
+; CHECK-NEXT: vsra.vv v16, v14, v10
+; CHECK-NEXT: vmsne.vv v10, v8, v16
+; CHECK-NEXT: vmerge.vxm v8, v12, a0, v0
; CHECK-NEXT: vmv1r.v v0, v10
-; CHECK-NEXT: vmerge.vvm v8, v12, v8, v0
+; CHECK-NEXT: vmerge.vvm v8, v14, v8, v0
; CHECK-NEXT: ret
%tmp = call <vscale x 4 x i32> @llvm.sshl.sat.nxv4i32(<vscale x 4 x i32> %x, <vscale x 4 x i32> %y)
ret <vscale x 4 x i32> %tmp
@@ -134,13 +134,13 @@ define <vscale x 8 x i16> @vec_nxv8i16(<vscale x 8 x i16> %x, <vscale x 8 x i16>
; CHECK-NEXT: vmsle.vi v0, v8, -1
; CHECK-NEXT: lui a0, 8
; CHECK-NEXT: addi a1, a0, -1
-; CHECK-NEXT: vsll.vv v12, v8, v10
-; CHECK-NEXT: vsra.vv v14, v12, v10
-; CHECK-NEXT: vmsne.vv v10, v8, v14
-; CHECK-NEXT: vmv.v.x v8, a1
-; CHECK-NEXT: vmerge.vxm v8, v8, a0, v0
+; CHECK-NEXT: vmv.v.x v12, a1
+; CHECK-NEXT: vsll.vv v14, v8, v10
+; CHECK-NEXT: vsra.vv v16, v14, v10
+; CHECK-NEXT: vmsne.vv v10, v8, v16
+; CHECK-NEXT: vmerge.vxm v8, v12, a0, v0
; CHECK-NEXT: vmv1r.v v0, v10
-; CHECK-NEXT: vmerge.vvm v8, v12, v8, v0
+; CHECK-NEXT: vmerge.vvm v8, v14, v8, v0
; CHECK-NEXT: ret
%tmp = call <vscale x 8 x i16> @llvm.sshl.sat.nxv8i16(<vscale x 8 x i16> %x, <vscale x 8 x i16> %y)
ret <vscale x 8 x i16> %tmp
@@ -152,14 +152,14 @@ define <vscale x 16 x i8> @vec_nxv16i8(<vscale x 16 x i8> %x, <vscale x 16 x i8>
; CHECK-NEXT: vsetvli a0, zero, e8, m2, ta, ma
; CHECK-NEXT: vmsle.vi v0, v8, -1
; CHECK-NEXT: li a0, 127
-; CHECK-NEXT: vsll.vv v12, v8, v10
-; CHECK-NEXT: vsra.vv v14, v12, v10
-; CHECK-NEXT: vmsne.vv v10, v8, v14
-; CHECK-NEXT: vmv.v.x v8, a0
+; CHECK-NEXT: vmv.v.x v12, a0
; CHECK-NEXT: li a0, 128
-; CHECK-NEXT: vmerge.vxm v8, v8, a0, v0
+; CHECK-NEXT: vsll.vv v14, v8, v10
+; CHECK-NEXT: vsra.vv v16, v14, v10
+; CHECK-NEXT: vmsne.vv v10, v8, v16
+; CHECK-NEXT: vmerge.vxm v8, v12, a0, v0
; CHECK-NEXT: vmv1r.v v0, v10
-; CHECK-NEXT: vmerge.vvm v8, v12, v8, v0
+; CHECK-NEXT: vmerge.vvm v8, v14, v8, v0
; CHECK-NEXT: ret
%tmp = call <vscale x 16 x i8> @llvm.sshl.sat.nxv16i8(<vscale x 16 x i8> %x, <vscale x 16 x i8> %y)
ret <vscale x 16 x i8> %tmp
diff --git a/llvm/test/CodeGen/RISCV/rvv/stepvector.ll b/llvm/test/CodeGen/RISCV/rvv/stepvector.ll
index eff8c26d4d061..b3150ecea6c0b 100644
--- a/llvm/test/CodeGen/RISCV/rvv/stepvector.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/stepvector.ll
@@ -562,8 +562,8 @@ define <vscale x 16 x i64> @add_stepvector_nxv16i64() {
; RV32-NEXT: csrr a0, vlenb
; RV32-NEXT: slli a0, a0, 1
; RV32-NEXT: sw a0, 8(sp)
-; RV32-NEXT: vsetvli a0, zero, e64, m8, ta, ma
; RV32-NEXT: addi a0, sp, 8
+; RV32-NEXT: vsetvli a1, zero, e64, m8, ta, ma
; RV32-NEXT: vlse64.v v16, (a0), zero
; RV32-NEXT: vid.v v8
; RV32-NEXT: vadd.vv v8, v8, v8
@@ -597,8 +597,8 @@ define <vscale x 16 x i64> @mul_stepvector_nxv16i64() {
; RV32-NEXT: slli a1, a0, 1
; RV32-NEXT: add a0, a1, a0
; RV32-NEXT: sw a0, 8(sp)
-; RV32-NEXT: vsetvli a0, zero, e64, m8, ta, ma
; RV32-NEXT: addi a0, sp, 8
+; RV32-NEXT: vsetvli a1, zero, e64, m8, ta, ma
; RV32-NEXT: vlse64.v v16, (a0), zero
; RV32-NEXT: vid.v v8
; RV32-NEXT: li a0, 3
@@ -686,8 +686,8 @@ define <vscale x 16 x i64> @shl_stepvector_nxv16i64() {
; RV32-NEXT: csrr a0, vlenb
; RV32-NEXT: slli a0, a0, 2
; RV32-NEXT: sw a0, 8(sp)
-; RV32-NEXT: vsetvli a0, zero, e64, m8, ta, ma
; RV32-NEXT: addi a0, sp, 8
+; RV32-NEXT: vsetvli a1, zero, e64, m8, ta, ma
; RV32-NEXT: vlse64.v v16, (a0), zero
; RV32-NEXT: vid.v v8
; RV32-NEXT: vsll.vi v8, v8, 2
diff --git a/llvm/test/CodeGen/RISCV/rvv/strided-vpload.ll b/llvm/test/CodeGen/RISCV/rvv/strided-vpload.ll
index 0e2105d5cba86..4d3bced0bcb50 100644
--- a/llvm/test/CodeGen/RISCV/rvv/strided-vpload.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/strided-vpload.ll
@@ -588,15 +588,15 @@ define <vscale x 16 x double> @strided_load_nxv16f64(ptr %ptr, i64 %stride, <vsc
; CHECK-RV32-NEXT: mv a3, a4
; CHECK-RV32-NEXT: .LBB49_2:
; CHECK-RV32-NEXT: mul a5, a3, a1
-; CHECK-RV32-NEXT: add a5, a0, a5
; CHECK-RV32-NEXT: srli a4, a4, 3
; CHECK-RV32-NEXT: vsetvli a6, zero, e8, mf4, ta, ma
; CHECK-RV32-NEXT: vslidedown.vx v8, v9, a4
-; CHECK-RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
+; CHECK-RV32-NEXT: add a5, a0, a5
; CHECK-RV32-NEXT: vmv1r.v v0, v8
+; CHECK-RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
; CHECK-RV32-NEXT: vlse64.v v16, (a5), a1, v0.t
-; CHECK-RV32-NEXT: vsetvli zero, a3, e64, m8, ta, ma
; CHECK-RV32-NEXT: vmv1r.v v0, v9
+; CHECK-RV32-NEXT: vsetvli zero, a3, e64, m8, ta, ma
; CHECK-RV32-NEXT: vlse64.v v8, (a0), a1, v0.t
; CHECK-RV32-NEXT: ret
;
@@ -613,15 +613,15 @@ define <vscale x 16 x double> @strided_load_nxv16f64(ptr %ptr, i64 %stride, <vsc
; CHECK-RV64-NEXT: mv a2, a4
; CHECK-RV64-NEXT: .LBB49_2:
; CHECK-RV64-NEXT: mul a5, a2, a1
-; CHECK-RV64-NEXT: add a5, a0, a5
; CHECK-RV64-NEXT: srli a4, a4, 3
; CHECK-RV64-NEXT: vsetvli a6, zero, e8, mf4, ta, ma
; CHECK-RV64-NEXT: vslidedown.vx v8, v9, a4
-; CHECK-RV64-NEXT: vsetvli zero, a3, e64, m8, ta, ma
+; CHECK-RV64-NEXT: add a5, a0, a5
; CHECK-RV64-NEXT: vmv1r.v v0, v8
+; CHECK-RV64-NEXT: vsetvli zero, a3, e64, m8, ta, ma
; CHECK-RV64-NEXT: vlse64.v v16, (a5), a1, v0.t
-; CHECK-RV64-NEXT: vsetvli zero, a2, e64, m8, ta, ma
; CHECK-RV64-NEXT: vmv1r.v v0, v9
+; CHECK-RV64-NEXT: vsetvli zero, a2, e64, m8, ta, ma
; CHECK-RV64-NEXT: vlse64.v v8, (a0), a1, v0.t
; CHECK-RV64-NEXT: ret
%v = call <vscale x 16 x double> @llvm.experimental.vp.strided.load.nxv16f64.p0.i64(ptr %ptr, i64 %stride, <vscale x 16 x i1> %mask, i32 %evl)
@@ -697,10 +697,10 @@ define <vscale x 16 x double> @strided_load_nxv17f64(ptr %ptr, i64 %stride, <vsc
; CHECK-RV32-NEXT: mv a5, a2
; CHECK-RV32-NEXT: .LBB51_4:
; CHECK-RV32-NEXT: mul t1, a5, a1
-; CHECK-RV32-NEXT: add t1, a0, t1
; CHECK-RV32-NEXT: srli t2, a2, 3
; CHECK-RV32-NEXT: vsetvli t3, zero, e8, mf4, ta, ma
; CHECK-RV32-NEXT: vslidedown.vx v0, v8, t2
+; CHECK-RV32-NEXT: add t1, a0, t1
; CHECK-RV32-NEXT: vsetvli zero, t0, e64, m8, ta, ma
; CHECK-RV32-NEXT: vlse64.v v16, (t1), a1, v0.t
; CHECK-RV32-NEXT: sub a7, a3, a7
@@ -712,14 +712,14 @@ define <vscale x 16 x double> @strided_load_nxv17f64(ptr %ptr, i64 %stride, <vsc
; CHECK-RV32-NEXT: mv a3, a2
; CHECK-RV32-NEXT: .LBB51_6:
; CHECK-RV32-NEXT: mul a6, a6, a1
-; CHECK-RV32-NEXT: add a6, a0, a6
; CHECK-RV32-NEXT: srli a2, a2, 2
; CHECK-RV32-NEXT: vsetvli a7, zero, e8, mf2, ta, ma
; CHECK-RV32-NEXT: vslidedown.vx v0, v8, a2
+; CHECK-RV32-NEXT: add a6, a0, a6
; CHECK-RV32-NEXT: vsetvli zero, a3, e64, m8, ta, ma
; CHECK-RV32-NEXT: vlse64.v v24, (a6), a1, v0.t
-; CHECK-RV32-NEXT: vsetvli zero, a5, e64, m8, ta, ma
; CHECK-RV32-NEXT: vmv1r.v v0, v8
+; CHECK-RV32-NEXT: vsetvli zero, a5, e64, m8, ta, ma
; CHECK-RV32-NEXT: vlse64.v v8, (a0), a1, v0.t
; CHECK-RV32-NEXT: vs1r.v v24, (a4)
; CHECK-RV32-NEXT: ret
@@ -744,10 +744,10 @@ define <vscale x 16 x double> @strided_load_nxv17f64(ptr %ptr, i64 %stride, <vsc
; CHECK-RV64-NEXT: mv a5, a4
; CHECK-RV64-NEXT: .LBB51_4:
; CHECK-RV64-NEXT: mul t1, a5, a1
-; CHECK-RV64-NEXT: add t1, a0, t1
; CHECK-RV64-NEXT: srli t2, a4, 3
; CHECK-RV64-NEXT: vsetvli t3, zero, e8, mf4, ta, ma
; CHECK-RV64-NEXT: vslidedown.vx v0, v8, t2
+; CHECK-RV64-NEXT: add t1, a0, t1
; CHECK-RV64-NEXT: vsetvli zero, t0, e64, m8, ta, ma
; CHECK-RV64-NEXT: vlse64.v v16, (t1), a1, v0.t
; CHECK-RV64-NEXT: sub a7, a2, a7
@@ -759,14 +759,14 @@ define <vscale x 16 x double> @strided_load_nxv17f64(ptr %ptr, i64 %stride, <vsc
; CHECK-RV64-NEXT: mv a2, a4
; CHECK-RV64-NEXT: .LBB51_6:
; CHECK-RV64-NEXT: mul a6, a6, a1
-; CHECK-RV64-NEXT: add a6, a0, a6
; CHECK-RV64-NEXT: srli a4, a4, 2
; CHECK-RV64-NEXT: vsetvli a7, zero, e8, mf2, ta, ma
; CHECK-RV64-NEXT: vslidedown.vx v0, v8, a4
+; CHECK-RV64-NEXT: add a6, a0, a6
; CHECK-RV64-NEXT: vsetvli zero, a2, e64, m8, ta, ma
; CHECK-RV64-NEXT: vlse64.v v24, (a6), a1, v0.t
-; CHECK-RV64-NEXT: vsetvli zero, a5, e64, m8, ta, ma
; CHECK-RV64-NEXT: vmv1r.v v0, v8
+; CHECK-RV64-NEXT: vsetvli zero, a5, e64, m8, ta, ma
; CHECK-RV64-NEXT: vlse64.v v8, (a0), a1, v0.t
; CHECK-RV64-NEXT: vs1r.v v24, (a3)
; CHECK-RV64-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/strided-vpstore.ll b/llvm/test/CodeGen/RISCV/rvv/strided-vpstore.ll
index 9378bb3d3ca61..e8704b35f31f7 100644
--- a/llvm/test/CodeGen/RISCV/rvv/strided-vpstore.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/strided-vpstore.ll
@@ -504,10 +504,10 @@ define void @strided_store_nxv16f64(<vscale x 16 x double> %v, ptr %ptr, i32 sig
; CHECK-NEXT: addi a2, a2, -1
; CHECK-NEXT: and a2, a2, a5
; CHECK-NEXT: mul a4, a4, a1
-; CHECK-NEXT: add a0, a0, a4
; CHECK-NEXT: srli a3, a3, 3
-; CHECK-NEXT: vsetvli a4, zero, e8, mf4, ta, ma
+; CHECK-NEXT: vsetvli a5, zero, e8, mf4, ta, ma
; CHECK-NEXT: vslidedown.vx v0, v0, a3
+; CHECK-NEXT: add a0, a0, a4
; CHECK-NEXT: vsetvli zero, a2, e64, m8, ta, ma
; CHECK-NEXT: vsse64.v v16, (a0), a1, v0.t
; CHECK-NEXT: ret
@@ -567,36 +567,36 @@ define void @strided_store_nxv17f64(<vscale x 17 x double> %v, ptr %ptr, i32 sig
; CHECK-NEXT: vl8re64.v v0, (a0)
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vs8r.v v0, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: vsetvli zero, a7, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a7, e64, m8, ta, ma
; CHECK-NEXT: vsse64.v v8, (a1), a2, v0.t
; CHECK-NEXT: sub a0, a5, a4
; CHECK-NEXT: sltu t0, a5, a0
; CHECK-NEXT: addi t0, t0, -1
-; CHECK-NEXT: and a0, t0, a0
-; CHECK-NEXT: mul a7, a7, a2
-; CHECK-NEXT: add a7, a1, a7
-; CHECK-NEXT: srli t0, a4, 3
+; CHECK-NEXT: and t0, t0, a0
+; CHECK-NEXT: mul a0, a7, a2
+; CHECK-NEXT: add a7, a1, a0
+; CHECK-NEXT: srli a0, a4, 3
; CHECK-NEXT: vsetvli t1, zero, e8, mf4, ta, ma
-; CHECK-NEXT: vslidedown.vx v0, v24, t0
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT: vslidedown.vx v0, v24, a0
; CHECK-NEXT: sub a0, a3, a6
; CHECK-NEXT: sltu a3, a3, a0
; CHECK-NEXT: addi a3, a3, -1
; CHECK-NEXT: and a0, a3, a0
+; CHECK-NEXT: vsetvli zero, t0, e64, m8, ta, ma
; CHECK-NEXT: vsse64.v v16, (a7), a2, v0.t
; CHECK-NEXT: bltu a0, a4, .LBB43_6
; CHECK-NEXT: # %bb.5:
; CHECK-NEXT: mv a0, a4
; CHECK-NEXT: .LBB43_6:
; CHECK-NEXT: mul a3, a5, a2
-; CHECK-NEXT: add a1, a1, a3
; CHECK-NEXT: srli a4, a4, 2
-; CHECK-NEXT: vsetvli a3, zero, e8, mf2, ta, ma
+; CHECK-NEXT: vsetvli a5, zero, e8, mf2, ta, ma
; CHECK-NEXT: vslidedown.vx v0, v24, a4
+; CHECK-NEXT: add a1, a1, a3
+; CHECK-NEXT: addi a3, sp, 16
+; CHECK-NEXT: vl8r.v v8, (a3) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vsse64.v v8, (a1), a2, v0.t
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
diff --git a/llvm/test/CodeGen/RISCV/rvv/vadd-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vadd-vp.ll
index 4b5e737d22eb8..ede395f4df8e1 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vadd-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vadd-vp.ll
@@ -580,8 +580,8 @@ define <vscale x 128 x i8> @vadd_vi_nxv128i8(<vscale x 128 x i8> %va, <vscale x
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a1, a0
; CHECK-NEXT: .LBB50_2:
-; CHECK-NEXT: vsetvli zero, a1, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a1, e8, m8, ta, ma
; CHECK-NEXT: vadd.vi v8, v8, -1, v0.t
; CHECK-NEXT: ret
%v = call <vscale x 128 x i8> @llvm.vp.add.nxv128i8(<vscale x 128 x i8> %va, <vscale x 128 x i8> splat (i8 -1), <vscale x 128 x i1> %m, i32 %evl)
@@ -1359,8 +1359,8 @@ define <vscale x 32 x i32> @vadd_vi_nxv32i32(<vscale x 32 x i32> %va, <vscale x
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB118_2:
-; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; CHECK-NEXT: vadd.vi v8, v8, -1, v0.t
; CHECK-NEXT: ret
%v = call <vscale x 32 x i32> @llvm.vp.add.nxv32i32(<vscale x 32 x i32> %va, <vscale x 32 x i32> splat (i32 -1), <vscale x 32 x i1> %m, i32 %evl)
@@ -1415,8 +1415,8 @@ define <vscale x 32 x i32> @vadd_vi_nxv32i32_evl_nx8(<vscale x 32 x i32> %va, <v
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB120_2:
-; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; CHECK-NEXT: vadd.vi v8, v8, -1, v0.t
; CHECK-NEXT: ret
%evl = call i32 @llvm.vscale.i32()
@@ -1451,8 +1451,8 @@ define <vscale x 32 x i32> @vadd_vi_nxv32i32_evl_nx16(<vscale x 32 x i32> %va, <
; RV64-NEXT: slli a0, a0, 1
; RV64-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; RV64-NEXT: vadd.vi v8, v8, -1, v0.t
-; RV64-NEXT: vsetivli zero, 0, e32, m8, ta, ma
; RV64-NEXT: vmv1r.v v0, v24
+; RV64-NEXT: vsetivli zero, 0, e32, m8, ta, ma
; RV64-NEXT: vadd.vi v16, v16, -1, v0.t
; RV64-NEXT: ret
%evl = call i32 @llvm.vscale.i32()
diff --git a/llvm/test/CodeGen/RISCV/rvv/vcopysign-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vcopysign-vp.ll
index 939a45e15c103..6e34d59a2d989 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vcopysign-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vcopysign-vp.ll
@@ -255,9 +255,9 @@ define <vscale x 32 x half> @vfsgnj_vv_nxv32f16(<vscale x 32 x half> %va, <vscal
; ZVFHMIN-NEXT: srli a2, a2, 2
; ZVFHMIN-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
; ZVFHMIN-NEXT: vslidedown.vx v0, v0, a2
-; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
@@ -272,8 +272,8 @@ define <vscale x 32 x half> @vfsgnj_vv_nxv32f16(<vscale x 32 x half> %va, <vscal
; ZVFHMIN-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v24
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v8
-; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v7
+; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfsgnj.vv v16, v24, v16, v0.t
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v16
@@ -312,9 +312,9 @@ define <vscale x 32 x half> @vfsgnj_vv_nxv32f16_unmasked(<vscale x 32 x half> %v
; ZVFHMIN-NEXT: vmset.m v24
; ZVFHMIN-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
; ZVFHMIN-NEXT: vslidedown.vx v0, v24, a2
-; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
diff --git a/llvm/test/CodeGen/RISCV/rvv/vcpop.ll b/llvm/test/CodeGen/RISCV/rvv/vcpop.ll
index 6f06d8e570de0..e59a9174b03d9 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vcpop.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vcpop.ll
@@ -44,8 +44,8 @@ define iXLen @intrinsic_vcpop_mask_m_nxv1i1(<vscale x 1 x i1> %0, <vscale x 1 x
; CHECK-LABEL: intrinsic_vcpop_mask_m_nxv1i1:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, ma
; CHECK-NEXT: vcpop.m a0, v9, v0.t
; CHECK-NEXT: ret
entry:
@@ -98,8 +98,8 @@ define iXLen @intrinsic_vcpop_mask_m_nxv2i1(<vscale x 2 x i1> %0, <vscale x 2 x
; CHECK-LABEL: intrinsic_vcpop_mask_m_nxv2i1:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a0, e8, mf4, ta, ma
; CHECK-NEXT: vcpop.m a0, v9, v0.t
; CHECK-NEXT: ret
entry:
@@ -138,8 +138,8 @@ define iXLen @intrinsic_vcpop_mask_m_nxv4i1(<vscale x 4 x i1> %0, <vscale x 4 x
; CHECK-LABEL: intrinsic_vcpop_mask_m_nxv4i1:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, ma
; CHECK-NEXT: vcpop.m a0, v9, v0.t
; CHECK-NEXT: ret
entry:
@@ -178,8 +178,8 @@ define iXLen @intrinsic_vcpop_mask_m_nxv8i1(<vscale x 8 x i1> %0, <vscale x 8 x
; CHECK-LABEL: intrinsic_vcpop_mask_m_nxv8i1:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m1, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a0, e8, m1, ta, ma
; CHECK-NEXT: vcpop.m a0, v9, v0.t
; CHECK-NEXT: ret
entry:
@@ -218,8 +218,8 @@ define iXLen @intrinsic_vcpop_mask_m_nxv16i1(<vscale x 16 x i1> %0, <vscale x 16
; CHECK-LABEL: intrinsic_vcpop_mask_m_nxv16i1:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a0, e8, m2, ta, ma
; CHECK-NEXT: vcpop.m a0, v9, v0.t
; CHECK-NEXT: ret
entry:
@@ -258,8 +258,8 @@ define iXLen @intrinsic_vcpop_mask_m_nxv32i1(<vscale x 32 x i1> %0, <vscale x 32
; CHECK-LABEL: intrinsic_vcpop_mask_m_nxv32i1:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, ma
; CHECK-NEXT: vcpop.m a0, v9, v0.t
; CHECK-NEXT: ret
entry:
@@ -298,8 +298,8 @@ define iXLen @intrinsic_vcpop_mask_m_nxv64i1(<vscale x 64 x i1> %0, <vscale x 64
; CHECK-LABEL: intrinsic_vcpop_mask_m_nxv64i1:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a0, e8, m8, ta, ma
; CHECK-NEXT: vcpop.m a0, v9, v0.t
; CHECK-NEXT: ret
entry:
diff --git a/llvm/test/CodeGen/RISCV/rvv/vector-deinterleave-fixed.ll b/llvm/test/CodeGen/RISCV/rvv/vector-deinterleave-fixed.ll
index d02fe5b205f7f..13b701df9af7a 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vector-deinterleave-fixed.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vector-deinterleave-fixed.ll
@@ -12,16 +12,16 @@ define {<16 x i1>, <16 x i1>} @vector_deinterleave_v16i1_v32i1(<32 x i1> %vec) {
; CHECK-NEXT: vmerge.vim v10, v8, 1, v0
; CHECK-NEXT: vid.v v9
; CHECK-NEXT: vadd.vv v11, v9, v9
-; CHECK-NEXT: vrgather.vv v9, v10, v11
; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
; CHECK-NEXT: vslidedown.vi v0, v0, 2
; CHECK-NEXT: vsetivli zero, 16, e8, m1, ta, ma
+; CHECK-NEXT: vrgather.vv v9, v10, v11
; CHECK-NEXT: vmerge.vim v8, v8, 1, v0
-; CHECK-NEXT: vadd.vi v12, v11, -16
; CHECK-NEXT: li a0, -256
; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; CHECK-NEXT: vmv.s.x v0, a0
; CHECK-NEXT: vsetvli zero, zero, e8, m1, ta, mu
+; CHECK-NEXT: vadd.vi v12, v11, -16
; CHECK-NEXT: vrgather.vv v9, v8, v12, v0.t
; CHECK-NEXT: vmsne.vi v9, v9, 0
; CHECK-NEXT: vadd.vi v12, v11, 1
diff --git a/llvm/test/CodeGen/RISCV/rvv/vector-deinterleave-load.ll b/llvm/test/CodeGen/RISCV/rvv/vector-deinterleave-load.ll
index 8f4ff37fffb02..f0f847c61f3b0 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vector-deinterleave-load.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vector-deinterleave-load.ll
@@ -110,23 +110,22 @@ define {<vscale x 8 x i64>, <vscale x 8 x i64>} @vector_deinterleave_load_nxv8i6
; CHECK-NEXT: csrr a1, vlenb
; CHECK-NEXT: slli a1, a1, 3
; CHECK-NEXT: add a1, a0, a1
-; CHECK-NEXT: vl8re64.v v8, (a1)
-; CHECK-NEXT: csrr a1, vlenb
+; CHECK-NEXT: vl8re64.v v8, (a0)
+; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: li a2, 24
-; CHECK-NEXT: mul a1, a1, a2
-; CHECK-NEXT: add a1, sp, a1
-; CHECK-NEXT: addi a1, a1, 16
-; CHECK-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
-; CHECK-NEXT: vl8re64.v v0, (a0)
-; CHECK-NEXT: vsetvli a0, zero, e64, m8, ta, ma
-; CHECK-NEXT: vid.v v8
-; CHECK-NEXT: vadd.vv v16, v8, v8
-; CHECK-NEXT: vrgather.vv v8, v0, v16
+; CHECK-NEXT: mul a0, a0, a2
+; CHECK-NEXT: add a0, sp, a0
+; CHECK-NEXT: addi a0, a0, 16
+; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vl8re64.v v0, (a1)
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vs8r.v v0, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vsetvli a0, zero, e64, m8, ta, ma
+; CHECK-NEXT: vid.v v8
+; CHECK-NEXT: vadd.vv v16, v8, v8
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: li a1, 24
; CHECK-NEXT: mul a0, a0, a1
@@ -134,34 +133,47 @@ define {<vscale x 8 x i64>, <vscale x 8 x i64>} @vector_deinterleave_load_nxv8i6
; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vrgather.vv v24, v8, v16
+; CHECK-NEXT: addi a0, sp, 16
+; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vrgather.vv v8, v0, v16
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 4
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: vadd.vi v8, v16, 1
-; CHECK-NEXT: vrgather.vv v16, v0, v8
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: li a1, 24
; CHECK-NEXT: mul a0, a0, a1
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vl8r.v v0, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vrgather.vv v16, v0, v8
+; CHECK-NEXT: csrr a0, vlenb
+; CHECK-NEXT: slli a0, a0, 3
+; CHECK-NEXT: add a0, sp, a0
+; CHECK-NEXT: addi a0, a0, 16
+; CHECK-NEXT: vl8r.v v0, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vrgather.vv v24, v0, v8
-; CHECK-NEXT: addi a0, sp, 16
+; CHECK-NEXT: csrr a0, vlenb
+; CHECK-NEXT: li a1, 24
+; CHECK-NEXT: mul a0, a0, a1
+; CHECK-NEXT: add a0, sp, a0
+; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 4
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: addi a0, sp, 16
+; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vmv4r.v v28, v8
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
+; CHECK-NEXT: li a1, 24
+; CHECK-NEXT: mul a0, a0, a1
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vmv4r.v v28, v8
-; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vmv4r.v v20, v8
; CHECK-NEXT: vmv8r.v v8, v24
diff --git a/llvm/test/CodeGen/RISCV/rvv/vector-deinterleave.ll b/llvm/test/CodeGen/RISCV/rvv/vector-deinterleave.ll
index 7797577362c93..bcb008857ad32 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vector-deinterleave.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vector-deinterleave.ll
@@ -8,14 +8,15 @@ define {<vscale x 16 x i1>, <vscale x 16 x i1>} @vector_deinterleave_nxv16i1_nxv
; CHECK-LABEL: vector_deinterleave_nxv16i1_nxv32i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetvli a0, zero, e8, m2, ta, ma
-; CHECK-NEXT: vmv.v.i v8, 0
-; CHECK-NEXT: vmerge.vim v12, v8, 1, v0
+; CHECK-NEXT: vmv.v.i v10, 0
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: srli a0, a0, 2
; CHECK-NEXT: vsetvli a1, zero, e8, mf2, ta, ma
-; CHECK-NEXT: vslidedown.vx v0, v0, a0
+; CHECK-NEXT: vslidedown.vx v8, v0, a0
; CHECK-NEXT: vsetvli a0, zero, e8, m2, ta, ma
-; CHECK-NEXT: vmerge.vim v14, v8, 1, v0
+; CHECK-NEXT: vmerge.vim v12, v10, 1, v0
+; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vmerge.vim v14, v10, 1, v0
; CHECK-NEXT: vnsrl.wi v10, v12, 0
; CHECK-NEXT: vmsne.vi v8, v10, 0
; CHECK-NEXT: vnsrl.wi v10, v12, 8
@@ -90,25 +91,38 @@ declare {<vscale x 2 x i64>, <vscale x 2 x i64>} @llvm.vector.deinterleave2.nxv4
define {<vscale x 64 x i1>, <vscale x 64 x i1>} @vector_deinterleave_nxv64i1_nxv128i1(<vscale x 128 x i1> %vec) {
; CHECK-LABEL: vector_deinterleave_nxv64i1_nxv128i1:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v28, v8
+; CHECK-NEXT: addi sp, sp, -16
+; CHECK-NEXT: .cfi_def_cfa_offset 16
+; CHECK-NEXT: csrr a0, vlenb
+; CHECK-NEXT: slli a0, a0, 1
+; CHECK-NEXT: sub sp, sp, a0
+; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x02, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 2 * vlenb
+; CHECK-NEXT: vmv1r.v v12, v8
; CHECK-NEXT: vsetvli a0, zero, e8, m8, ta, ma
-; CHECK-NEXT: vmv.v.i v8, 0
-; CHECK-NEXT: vmerge.vim v16, v8, 1, v0
+; CHECK-NEXT: vmv.v.i v24, 0
+; CHECK-NEXT: vmerge.vim v16, v24, 1, v0
; CHECK-NEXT: vsetvli a0, zero, e8, m4, ta, ma
-; CHECK-NEXT: vnsrl.wi v24, v16, 0
+; CHECK-NEXT: vnsrl.wi v8, v16, 0
+; CHECK-NEXT: vmv1r.v v0, v12
; CHECK-NEXT: vsetvli a0, zero, e8, m8, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v28
-; CHECK-NEXT: vmerge.vim v8, v8, 1, v0
+; CHECK-NEXT: vmerge.vim v24, v24, 1, v0
; CHECK-NEXT: vsetvli a0, zero, e8, m4, ta, ma
-; CHECK-NEXT: vnsrl.wi v28, v8, 0
+; CHECK-NEXT: vnsrl.wi v12, v24, 0
; CHECK-NEXT: vsetvli a0, zero, e8, m8, ta, ma
-; CHECK-NEXT: vmsne.vi v7, v24, 0
+; CHECK-NEXT: vmsne.vi v7, v8, 0
+; CHECK-NEXT: addi a0, sp, 16
+; CHECK-NEXT: vs1r.v v7, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: vsetvli a0, zero, e8, m4, ta, ma
-; CHECK-NEXT: vnsrl.wi v24, v16, 8
-; CHECK-NEXT: vnsrl.wi v28, v8, 8
+; CHECK-NEXT: vnsrl.wi v0, v16, 8
+; CHECK-NEXT: vnsrl.wi v4, v24, 8
; CHECK-NEXT: vsetvli a0, zero, e8, m8, ta, ma
-; CHECK-NEXT: vmsne.vi v9, v24, 0
-; CHECK-NEXT: vmv1r.v v8, v7
+; CHECK-NEXT: vmsne.vi v9, v0, 0
+; CHECK-NEXT: addi a0, sp, 16
+; CHECK-NEXT: vl1r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: csrr a0, vlenb
+; CHECK-NEXT: slli a0, a0, 1
+; CHECK-NEXT: add sp, sp, a0
+; CHECK-NEXT: addi sp, sp, 16
; CHECK-NEXT: ret
%retval = call {<vscale x 64 x i1>, <vscale x 64 x i1>} @llvm.vector.deinterleave2.nxv128i1(<vscale x 128 x i1> %vec)
ret {<vscale x 64 x i1>, <vscale x 64 x i1>} %retval
diff --git a/llvm/test/CodeGen/RISCV/rvv/vector-interleave-store.ll b/llvm/test/CodeGen/RISCV/rvv/vector-interleave-store.ll
index 7ade47e60bc69..5ebf63f0a4411 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vector-interleave-store.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vector-interleave-store.ll
@@ -107,14 +107,14 @@ define void @vector_interleave_store_nxv16i64_nxv8i64(<vscale x 8 x i64> %a, <vs
; CHECK-NEXT: srli a2, a1, 1
; CHECK-NEXT: vsetvli a3, zero, e16, m2, ta, mu
; CHECK-NEXT: vid.v v24
-; CHECK-NEXT: vsrl.vi v26, v24, 1
-; CHECK-NEXT: vand.vi v24, v24, 1
-; CHECK-NEXT: vmsne.vi v28, v24, 0
+; CHECK-NEXT: vand.vi v26, v24, 1
+; CHECK-NEXT: vmsne.vi v28, v26, 0
+; CHECK-NEXT: vsrl.vi v24, v24, 1
; CHECK-NEXT: vmv1r.v v0, v28
-; CHECK-NEXT: vadd.vx v26, v26, a2, v0.t
+; CHECK-NEXT: vadd.vx v24, v24, a2, v0.t
; CHECK-NEXT: vmv4r.v v12, v16
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
-; CHECK-NEXT: vrgatherei16.vv v0, v8, v26
+; CHECK-NEXT: vrgatherei16.vv v0, v8, v24
; CHECK-NEXT: csrr a2, vlenb
; CHECK-NEXT: slli a2, a2, 3
; CHECK-NEXT: add a2, sp, a2
@@ -123,7 +123,7 @@ define void @vector_interleave_store_nxv16i64_nxv8i64(<vscale x 8 x i64> %a, <vs
; CHECK-NEXT: addi a2, sp, 16
; CHECK-NEXT: vl8r.v v8, (a2) # Unknown-size Folded Reload
; CHECK-NEXT: vmv4r.v v16, v12
-; CHECK-NEXT: vrgatherei16.vv v8, v16, v26
+; CHECK-NEXT: vrgatherei16.vv v8, v16, v24
; CHECK-NEXT: slli a1, a1, 3
; CHECK-NEXT: add a1, a0, a1
; CHECK-NEXT: vs8r.v v8, (a1)
diff --git a/llvm/test/CodeGen/RISCV/rvv/vector-interleave.ll b/llvm/test/CodeGen/RISCV/rvv/vector-interleave.ll
index a7e0ad6ee5f4d..2e9f62e2f552c 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vector-interleave.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vector-interleave.ll
@@ -122,9 +122,9 @@ define <vscale x 4 x i64> @vector_interleave_nxv4i64_nxv2i64(<vscale x 2 x i64>
; CHECK-NEXT: srli a0, a0, 2
; CHECK-NEXT: vsetvli a1, zero, e16, m1, ta, mu
; CHECK-NEXT: vid.v v12
+; CHECK-NEXT: vand.vi v13, v12, 1
+; CHECK-NEXT: vmsne.vi v0, v13, 0
; CHECK-NEXT: vsrl.vi v16, v12, 1
-; CHECK-NEXT: vand.vi v12, v12, 1
-; CHECK-NEXT: vmsne.vi v0, v12, 0
; CHECK-NEXT: vadd.vx v16, v16, a0, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, ma
; CHECK-NEXT: vrgatherei16.vv v12, v8, v16
@@ -137,9 +137,9 @@ define <vscale x 4 x i64> @vector_interleave_nxv4i64_nxv2i64(<vscale x 2 x i64>
; ZVBB-NEXT: srli a0, a0, 2
; ZVBB-NEXT: vsetvli a1, zero, e16, m1, ta, mu
; ZVBB-NEXT: vid.v v12
+; ZVBB-NEXT: vand.vi v13, v12, 1
+; ZVBB-NEXT: vmsne.vi v0, v13, 0
; ZVBB-NEXT: vsrl.vi v16, v12, 1
-; ZVBB-NEXT: vand.vi v12, v12, 1
-; ZVBB-NEXT: vmsne.vi v0, v12, 0
; ZVBB-NEXT: vadd.vx v16, v16, a0, v0.t
; ZVBB-NEXT: vsetvli zero, zero, e64, m4, ta, ma
; ZVBB-NEXT: vrgatherei16.vv v12, v8, v16
@@ -269,13 +269,13 @@ define <vscale x 32 x i32> @vector_interleave_nxv32i32_nxv16i32(<vscale x 16 x i
;
; ZVBB-LABEL: vector_interleave_nxv32i32_nxv16i32:
; ZVBB: # %bb.0:
+; ZVBB-NEXT: vmv8r.v v24, v8
; ZVBB-NEXT: li a0, 32
; ZVBB-NEXT: vsetvli a1, zero, e32, m4, ta, ma
-; ZVBB-NEXT: vwsll.vx v24, v16, a0
-; ZVBB-NEXT: vwaddu.wv v24, v24, v8
+; ZVBB-NEXT: vwsll.vx v8, v16, a0
+; ZVBB-NEXT: vwaddu.wv v8, v8, v24
; ZVBB-NEXT: vwsll.vx v0, v20, a0
-; ZVBB-NEXT: vwaddu.wv v0, v0, v12
-; ZVBB-NEXT: vmv8r.v v8, v24
+; ZVBB-NEXT: vwaddu.wv v0, v0, v28
; ZVBB-NEXT: vmv8r.v v16, v0
; ZVBB-NEXT: ret
%res = call <vscale x 32 x i32> @llvm.vector.interleave2.nxv32i32(<vscale x 16 x i32> %a, <vscale x 16 x i32> %b)
@@ -288,44 +288,32 @@ define <vscale x 16 x i64> @vector_interleave_nxv16i64_nxv8i64(<vscale x 8 x i64
; CHECK-NEXT: addi sp, sp, -16
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 4
-; CHECK-NEXT: sub sp, sp, a0
-; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
-; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: sub sp, sp, a0
+; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; CHECK-NEXT: vmv8r.v v0, v8
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: srli a0, a0, 1
; CHECK-NEXT: vsetvli a1, zero, e16, m2, ta, mu
; CHECK-NEXT: vid.v v24
-; CHECK-NEXT: vsrl.vi v6, v24, 1
-; CHECK-NEXT: vand.vi v8, v24, 1
-; CHECK-NEXT: vmsne.vi v0, v8, 0
-; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a1, a1, 3
-; CHECK-NEXT: add a1, sp, a1
-; CHECK-NEXT: addi a1, a1, 16
-; CHECK-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
-; CHECK-NEXT: vadd.vx v6, v6, a0, v0.t
+; CHECK-NEXT: vand.vi v26, v24, 1
+; CHECK-NEXT: vmsne.vi v10, v26, 0
+; CHECK-NEXT: vsrl.vi v8, v24, 1
+; CHECK-NEXT: vmv8r.v v24, v0
+; CHECK-NEXT: vmv4r.v v12, v4
+; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vadd.vx v8, v8, a0, v0.t
; CHECK-NEXT: vmv4r.v v28, v16
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
-; CHECK-NEXT: vrgatherei16.vv v8, v24, v6
+; CHECK-NEXT: vrgatherei16.vv v0, v24, v8
; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vs8r.v v0, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: vmv4r.v v16, v12
-; CHECK-NEXT: vrgatherei16.vv v24, v16, v6
-; CHECK-NEXT: addi a0, sp, 16
+; CHECK-NEXT: vrgatherei16.vv v24, v16, v8
; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vmv.v.v v16, v24
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 4
+; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add sp, sp, a0
; CHECK-NEXT: addi sp, sp, 16
; CHECK-NEXT: ret
@@ -335,44 +323,32 @@ define <vscale x 16 x i64> @vector_interleave_nxv16i64_nxv8i64(<vscale x 8 x i64
; ZVBB-NEXT: addi sp, sp, -16
; ZVBB-NEXT: .cfi_def_cfa_offset 16
; ZVBB-NEXT: csrr a0, vlenb
-; ZVBB-NEXT: slli a0, a0, 4
-; ZVBB-NEXT: sub sp, sp, a0
-; ZVBB-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
-; ZVBB-NEXT: csrr a0, vlenb
; ZVBB-NEXT: slli a0, a0, 3
-; ZVBB-NEXT: add a0, sp, a0
-; ZVBB-NEXT: addi a0, a0, 16
-; ZVBB-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; ZVBB-NEXT: sub sp, sp, a0
+; ZVBB-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; ZVBB-NEXT: vmv8r.v v0, v8
; ZVBB-NEXT: csrr a0, vlenb
; ZVBB-NEXT: srli a0, a0, 1
; ZVBB-NEXT: vsetvli a1, zero, e16, m2, ta, mu
; ZVBB-NEXT: vid.v v24
-; ZVBB-NEXT: vsrl.vi v6, v24, 1
-; ZVBB-NEXT: vand.vi v8, v24, 1
-; ZVBB-NEXT: vmsne.vi v0, v8, 0
-; ZVBB-NEXT: csrr a1, vlenb
-; ZVBB-NEXT: slli a1, a1, 3
-; ZVBB-NEXT: add a1, sp, a1
-; ZVBB-NEXT: addi a1, a1, 16
-; ZVBB-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
-; ZVBB-NEXT: vadd.vx v6, v6, a0, v0.t
+; ZVBB-NEXT: vand.vi v26, v24, 1
+; ZVBB-NEXT: vmsne.vi v10, v26, 0
+; ZVBB-NEXT: vsrl.vi v8, v24, 1
+; ZVBB-NEXT: vmv8r.v v24, v0
+; ZVBB-NEXT: vmv4r.v v12, v4
+; ZVBB-NEXT: vmv1r.v v0, v10
+; ZVBB-NEXT: vadd.vx v8, v8, a0, v0.t
; ZVBB-NEXT: vmv4r.v v28, v16
; ZVBB-NEXT: vsetvli zero, zero, e64, m8, ta, ma
-; ZVBB-NEXT: vrgatherei16.vv v8, v24, v6
+; ZVBB-NEXT: vrgatherei16.vv v0, v24, v8
; ZVBB-NEXT: addi a0, sp, 16
-; ZVBB-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
-; ZVBB-NEXT: csrr a0, vlenb
-; ZVBB-NEXT: slli a0, a0, 3
-; ZVBB-NEXT: add a0, sp, a0
-; ZVBB-NEXT: addi a0, a0, 16
-; ZVBB-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; ZVBB-NEXT: vs8r.v v0, (a0) # Unknown-size Folded Spill
; ZVBB-NEXT: vmv4r.v v16, v12
-; ZVBB-NEXT: vrgatherei16.vv v24, v16, v6
-; ZVBB-NEXT: addi a0, sp, 16
+; ZVBB-NEXT: vrgatherei16.vv v24, v16, v8
; ZVBB-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; ZVBB-NEXT: vmv.v.v v16, v24
; ZVBB-NEXT: csrr a0, vlenb
-; ZVBB-NEXT: slli a0, a0, 4
+; ZVBB-NEXT: slli a0, a0, 3
; ZVBB-NEXT: add sp, sp, a0
; ZVBB-NEXT: addi sp, sp, 16
; ZVBB-NEXT: ret
@@ -516,9 +492,9 @@ define <vscale x 4 x double> @vector_interleave_nxv4f64_nxv2f64(<vscale x 2 x do
; CHECK-NEXT: srli a0, a0, 2
; CHECK-NEXT: vsetvli a1, zero, e16, m1, ta, mu
; CHECK-NEXT: vid.v v12
+; CHECK-NEXT: vand.vi v13, v12, 1
+; CHECK-NEXT: vmsne.vi v0, v13, 0
; CHECK-NEXT: vsrl.vi v16, v12, 1
-; CHECK-NEXT: vand.vi v12, v12, 1
-; CHECK-NEXT: vmsne.vi v0, v12, 0
; CHECK-NEXT: vadd.vx v16, v16, a0, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, ma
; CHECK-NEXT: vrgatherei16.vv v12, v8, v16
@@ -531,9 +507,9 @@ define <vscale x 4 x double> @vector_interleave_nxv4f64_nxv2f64(<vscale x 2 x do
; ZVBB-NEXT: srli a0, a0, 2
; ZVBB-NEXT: vsetvli a1, zero, e16, m1, ta, mu
; ZVBB-NEXT: vid.v v12
+; ZVBB-NEXT: vand.vi v13, v12, 1
+; ZVBB-NEXT: vmsne.vi v0, v13, 0
; ZVBB-NEXT: vsrl.vi v16, v12, 1
-; ZVBB-NEXT: vand.vi v12, v12, 1
-; ZVBB-NEXT: vmsne.vi v0, v12, 0
; ZVBB-NEXT: vadd.vx v16, v16, a0, v0.t
; ZVBB-NEXT: vsetvli zero, zero, e64, m4, ta, ma
; ZVBB-NEXT: vrgatherei16.vv v12, v8, v16
@@ -593,13 +569,13 @@ define <vscale x 32 x float> @vector_interleave_nxv32f32_nxv16f32(<vscale x 16 x
;
; ZVBB-LABEL: vector_interleave_nxv32f32_nxv16f32:
; ZVBB: # %bb.0:
+; ZVBB-NEXT: vmv8r.v v24, v8
; ZVBB-NEXT: li a0, 32
; ZVBB-NEXT: vsetvli a1, zero, e32, m4, ta, ma
-; ZVBB-NEXT: vwsll.vx v24, v16, a0
-; ZVBB-NEXT: vwaddu.wv v24, v24, v8
+; ZVBB-NEXT: vwsll.vx v8, v16, a0
+; ZVBB-NEXT: vwaddu.wv v8, v8, v24
; ZVBB-NEXT: vwsll.vx v0, v20, a0
-; ZVBB-NEXT: vwaddu.wv v0, v0, v12
-; ZVBB-NEXT: vmv8r.v v8, v24
+; ZVBB-NEXT: vwaddu.wv v0, v0, v28
; ZVBB-NEXT: vmv8r.v v16, v0
; ZVBB-NEXT: ret
%res = call <vscale x 32 x float> @llvm.vector.interleave2.nxv32f32(<vscale x 16 x float> %a, <vscale x 16 x float> %b)
@@ -612,44 +588,32 @@ define <vscale x 16 x double> @vector_interleave_nxv16f64_nxv8f64(<vscale x 8 x
; CHECK-NEXT: addi sp, sp, -16
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 4
-; CHECK-NEXT: sub sp, sp, a0
-; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
-; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: sub sp, sp, a0
+; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; CHECK-NEXT: vmv8r.v v0, v8
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: srli a0, a0, 1
; CHECK-NEXT: vsetvli a1, zero, e16, m2, ta, mu
; CHECK-NEXT: vid.v v24
-; CHECK-NEXT: vsrl.vi v6, v24, 1
-; CHECK-NEXT: vand.vi v8, v24, 1
-; CHECK-NEXT: vmsne.vi v0, v8, 0
-; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a1, a1, 3
-; CHECK-NEXT: add a1, sp, a1
-; CHECK-NEXT: addi a1, a1, 16
-; CHECK-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
-; CHECK-NEXT: vadd.vx v6, v6, a0, v0.t
+; CHECK-NEXT: vand.vi v26, v24, 1
+; CHECK-NEXT: vmsne.vi v10, v26, 0
+; CHECK-NEXT: vsrl.vi v8, v24, 1
+; CHECK-NEXT: vmv8r.v v24, v0
+; CHECK-NEXT: vmv4r.v v12, v4
+; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vadd.vx v8, v8, a0, v0.t
; CHECK-NEXT: vmv4r.v v28, v16
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
-; CHECK-NEXT: vrgatherei16.vv v8, v24, v6
+; CHECK-NEXT: vrgatherei16.vv v0, v24, v8
; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vs8r.v v0, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: vmv4r.v v16, v12
-; CHECK-NEXT: vrgatherei16.vv v24, v16, v6
-; CHECK-NEXT: addi a0, sp, 16
+; CHECK-NEXT: vrgatherei16.vv v24, v16, v8
; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vmv.v.v v16, v24
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 4
+; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add sp, sp, a0
; CHECK-NEXT: addi sp, sp, 16
; CHECK-NEXT: ret
@@ -659,44 +623,32 @@ define <vscale x 16 x double> @vector_interleave_nxv16f64_nxv8f64(<vscale x 8 x
; ZVBB-NEXT: addi sp, sp, -16
; ZVBB-NEXT: .cfi_def_cfa_offset 16
; ZVBB-NEXT: csrr a0, vlenb
-; ZVBB-NEXT: slli a0, a0, 4
-; ZVBB-NEXT: sub sp, sp, a0
-; ZVBB-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
-; ZVBB-NEXT: csrr a0, vlenb
; ZVBB-NEXT: slli a0, a0, 3
-; ZVBB-NEXT: add a0, sp, a0
-; ZVBB-NEXT: addi a0, a0, 16
-; ZVBB-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; ZVBB-NEXT: sub sp, sp, a0
+; ZVBB-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; ZVBB-NEXT: vmv8r.v v0, v8
; ZVBB-NEXT: csrr a0, vlenb
; ZVBB-NEXT: srli a0, a0, 1
; ZVBB-NEXT: vsetvli a1, zero, e16, m2, ta, mu
; ZVBB-NEXT: vid.v v24
-; ZVBB-NEXT: vsrl.vi v6, v24, 1
-; ZVBB-NEXT: vand.vi v8, v24, 1
-; ZVBB-NEXT: vmsne.vi v0, v8, 0
-; ZVBB-NEXT: csrr a1, vlenb
-; ZVBB-NEXT: slli a1, a1, 3
-; ZVBB-NEXT: add a1, sp, a1
-; ZVBB-NEXT: addi a1, a1, 16
-; ZVBB-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
-; ZVBB-NEXT: vadd.vx v6, v6, a0, v0.t
+; ZVBB-NEXT: vand.vi v26, v24, 1
+; ZVBB-NEXT: vmsne.vi v10, v26, 0
+; ZVBB-NEXT: vsrl.vi v8, v24, 1
+; ZVBB-NEXT: vmv8r.v v24, v0
+; ZVBB-NEXT: vmv4r.v v12, v4
+; ZVBB-NEXT: vmv1r.v v0, v10
+; ZVBB-NEXT: vadd.vx v8, v8, a0, v0.t
; ZVBB-NEXT: vmv4r.v v28, v16
; ZVBB-NEXT: vsetvli zero, zero, e64, m8, ta, ma
-; ZVBB-NEXT: vrgatherei16.vv v8, v24, v6
+; ZVBB-NEXT: vrgatherei16.vv v0, v24, v8
; ZVBB-NEXT: addi a0, sp, 16
-; ZVBB-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
-; ZVBB-NEXT: csrr a0, vlenb
-; ZVBB-NEXT: slli a0, a0, 3
-; ZVBB-NEXT: add a0, sp, a0
-; ZVBB-NEXT: addi a0, a0, 16
-; ZVBB-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; ZVBB-NEXT: vs8r.v v0, (a0) # Unknown-size Folded Spill
; ZVBB-NEXT: vmv4r.v v16, v12
-; ZVBB-NEXT: vrgatherei16.vv v24, v16, v6
-; ZVBB-NEXT: addi a0, sp, 16
+; ZVBB-NEXT: vrgatherei16.vv v24, v16, v8
; ZVBB-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; ZVBB-NEXT: vmv.v.v v16, v24
; ZVBB-NEXT: csrr a0, vlenb
-; ZVBB-NEXT: slli a0, a0, 4
+; ZVBB-NEXT: slli a0, a0, 3
; ZVBB-NEXT: add sp, sp, a0
; ZVBB-NEXT: addi sp, sp, 16
; ZVBB-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vector-splice.ll b/llvm/test/CodeGen/RISCV/rvv/vector-splice.ll
index be56db52e349a..db2c25a80759a 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vector-splice.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vector-splice.ll
@@ -282,8 +282,8 @@ define <vscale x 32 x i1> @splice_nxv32i1_offset_max(<vscale x 32 x i1> %a, <vsc
; CHECK-NEXT: li a1, 63
; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, ma
; CHECK-NEXT: vslidedown.vx v16, v16, a1
-; CHECK-NEXT: vsetvli a1, zero, e8, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli a1, zero, e8, m4, ta, ma
; CHECK-NEXT: vmerge.vim v8, v12, 1, v0
; CHECK-NEXT: vslideup.vx v16, v8, a0
; CHECK-NEXT: vand.vi v8, v16, 1
@@ -331,8 +331,8 @@ define <vscale x 64 x i1> @splice_nxv64i1_offset_max(<vscale x 64 x i1> %a, <vsc
; CHECK-NEXT: li a1, 127
; CHECK-NEXT: vsetvli zero, a0, e8, m8, ta, ma
; CHECK-NEXT: vslidedown.vx v24, v24, a1
-; CHECK-NEXT: vsetvli a1, zero, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli a1, zero, e8, m8, ta, ma
; CHECK-NEXT: vmerge.vim v8, v16, 1, v0
; CHECK-NEXT: vslideup.vx v24, v8, a0
; CHECK-NEXT: vand.vi v8, v24, 1
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfabs-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vfabs-vp.ll
index df2bc523cd7a8..0f7e3f1e0ea5a 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfabs-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfabs-vp.ll
@@ -250,8 +250,8 @@ define <vscale x 32 x half> @vfabs_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
; ZVFHMIN-NEXT: mv a0, a1
; ZVFHMIN-NEXT: .LBB10_2:
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v8
-; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v16
+; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfabs.v v16, v24, v0.t
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v16
@@ -541,8 +541,8 @@ define <vscale x 16 x double> @vfabs_vv_nxv16f64(<vscale x 16 x double> %va, <vs
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB32_2:
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfabs.v v8, v8, v0.t
; CHECK-NEXT: ret
%v = call <vscale x 16 x double> @llvm.vp.fabs.nxv16f64(<vscale x 16 x double> %va, <vscale x 16 x i1> %m, i32 %evl)
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfadd-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vfadd-vp.ll
index c69a7bc5cece4..ad7fb63fec2fc 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfadd-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfadd-vp.ll
@@ -579,9 +579,9 @@ define <vscale x 32 x half> @vfadd_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
; ZVFHMIN-NEXT: srli a2, a2, 2
; ZVFHMIN-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
; ZVFHMIN-NEXT: vslidedown.vx v0, v0, a2
-; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
@@ -596,8 +596,8 @@ define <vscale x 32 x half> @vfadd_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
; ZVFHMIN-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v24
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v8
-; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v7
+; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfadd.vv v16, v24, v16, v0.t
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v16
@@ -636,9 +636,9 @@ define <vscale x 32 x half> @vfadd_vv_nxv32f16_unmasked(<vscale x 32 x half> %va
; ZVFHMIN-NEXT: vmset.m v24
; ZVFHMIN-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
; ZVFHMIN-NEXT: vslidedown.vx v0, v24, a2
-; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
@@ -675,12 +675,20 @@ define <vscale x 32 x half> @vfadd_vf_nxv32f16(<vscale x 32 x half> %va, half %b
;
; ZVFHMIN-LABEL: vfadd_vf_nxv32f16:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: vmv1r.v v3, v0
+; ZVFHMIN-NEXT: addi sp, sp, -16
+; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
+; ZVFHMIN-NEXT: csrr a1, vlenb
+; ZVFHMIN-NEXT: slli a1, a1, 2
+; ZVFHMIN-NEXT: sub sp, sp, a1
+; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x04, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 4 * vlenb
+; ZVFHMIN-NEXT: vmv1r.v v7, v0
; ZVFHMIN-NEXT: fcvt.s.h fa5, fa0
; ZVFHMIN-NEXT: vsetvli a1, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfmv.v.f v24, fa5
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v4, v24
+; ZVFHMIN-NEXT: vfncvt.f.f.w v16, v24
+; ZVFHMIN-NEXT: addi a1, sp, 16
+; ZVFHMIN-NEXT: vs4r.v v16, (a1) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a2, vlenb
; ZVFHMIN-NEXT: slli a1, a2, 1
; ZVFHMIN-NEXT: sub a3, a0, a1
@@ -692,7 +700,9 @@ define <vscale x 32 x half> @vfadd_vf_nxv32f16(<vscale x 32 x half> %va, half %b
; ZVFHMIN-NEXT: vslidedown.vx v0, v0, a2
; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v4
+; ZVFHMIN-NEXT: addi a2, sp, 16
+; ZVFHMIN-NEXT: vl4r.v v12, (a2) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfadd.vv v16, v16, v24, v0.t
; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
@@ -702,11 +712,15 @@ define <vscale x 32 x half> @vfadd_vf_nxv32f16(<vscale x 32 x half> %va, half %b
; ZVFHMIN-NEXT: mv a0, a1
; ZVFHMIN-NEXT: .LBB24_2:
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8
+; ZVFHMIN-NEXT: vmv1r.v v0, v7
; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vmv1r.v v0, v3
; ZVFHMIN-NEXT: vfadd.vv v16, v16, v24, v0.t
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v16
+; ZVFHMIN-NEXT: csrr a0, vlenb
+; ZVFHMIN-NEXT: slli a0, a0, 2
+; ZVFHMIN-NEXT: add sp, sp, a0
+; ZVFHMIN-NEXT: addi sp, sp, 16
; ZVFHMIN-NEXT: ret
%elt.head = insertelement <vscale x 32 x half> poison, half %b, i32 0
%vb = shufflevector <vscale x 32 x half> %elt.head, <vscale x 32 x half> poison, <vscale x 32 x i32> zeroinitializer
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfdiv-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vfdiv-vp.ll
index 3ad17e85570a2..81d844d1950ab 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfdiv-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfdiv-vp.ll
@@ -525,9 +525,9 @@ define <vscale x 32 x half> @vfdiv_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
; ZVFHMIN-NEXT: srli a2, a2, 2
; ZVFHMIN-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
; ZVFHMIN-NEXT: vslidedown.vx v0, v0, a2
-; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
@@ -542,8 +542,8 @@ define <vscale x 32 x half> @vfdiv_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
; ZVFHMIN-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v24
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v8
-; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v7
+; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfdiv.vv v16, v24, v16, v0.t
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v16
@@ -582,9 +582,9 @@ define <vscale x 32 x half> @vfdiv_vv_nxv32f16_unmasked(<vscale x 32 x half> %va
; ZVFHMIN-NEXT: vmset.m v24
; ZVFHMIN-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
; ZVFHMIN-NEXT: vslidedown.vx v0, v24, a2
-; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
@@ -621,12 +621,20 @@ define <vscale x 32 x half> @vfdiv_vf_nxv32f16(<vscale x 32 x half> %va, half %b
;
; ZVFHMIN-LABEL: vfdiv_vf_nxv32f16:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: vmv1r.v v3, v0
+; ZVFHMIN-NEXT: addi sp, sp, -16
+; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
+; ZVFHMIN-NEXT: csrr a1, vlenb
+; ZVFHMIN-NEXT: slli a1, a1, 2
+; ZVFHMIN-NEXT: sub sp, sp, a1
+; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x04, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 4 * vlenb
+; ZVFHMIN-NEXT: vmv1r.v v7, v0
; ZVFHMIN-NEXT: fcvt.s.h fa5, fa0
; ZVFHMIN-NEXT: vsetvli a1, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfmv.v.f v24, fa5
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v4, v24
+; ZVFHMIN-NEXT: vfncvt.f.f.w v16, v24
+; ZVFHMIN-NEXT: addi a1, sp, 16
+; ZVFHMIN-NEXT: vs4r.v v16, (a1) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a2, vlenb
; ZVFHMIN-NEXT: slli a1, a2, 1
; ZVFHMIN-NEXT: sub a3, a0, a1
@@ -638,7 +646,9 @@ define <vscale x 32 x half> @vfdiv_vf_nxv32f16(<vscale x 32 x half> %va, half %b
; ZVFHMIN-NEXT: vslidedown.vx v0, v0, a2
; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v4
+; ZVFHMIN-NEXT: addi a2, sp, 16
+; ZVFHMIN-NEXT: vl4r.v v12, (a2) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfdiv.vv v16, v16, v24, v0.t
; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
@@ -648,11 +658,15 @@ define <vscale x 32 x half> @vfdiv_vf_nxv32f16(<vscale x 32 x half> %va, half %b
; ZVFHMIN-NEXT: mv a0, a1
; ZVFHMIN-NEXT: .LBB22_2:
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8
+; ZVFHMIN-NEXT: vmv1r.v v0, v7
; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vmv1r.v v0, v3
; ZVFHMIN-NEXT: vfdiv.vv v16, v16, v24, v0.t
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v16
+; ZVFHMIN-NEXT: csrr a0, vlenb
+; ZVFHMIN-NEXT: slli a0, a0, 2
+; ZVFHMIN-NEXT: add sp, sp, a0
+; ZVFHMIN-NEXT: addi sp, sp, 16
; ZVFHMIN-NEXT: ret
%elt.head = insertelement <vscale x 32 x half> poison, half %b, i32 0
%vb = shufflevector <vscale x 32 x half> %elt.head, <vscale x 32 x half> poison, <vscale x 32 x i32> zeroinitializer
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfirst.ll b/llvm/test/CodeGen/RISCV/rvv/vfirst.ll
index 3be3f835f3d11..eafd605c6110e 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfirst.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfirst.ll
@@ -44,8 +44,8 @@ define iXLen @intrinsic_vfirst_mask_m_nxv1i1(<vscale x 1 x i1> %0, <vscale x 1 x
; CHECK-LABEL: intrinsic_vfirst_mask_m_nxv1i1:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, ma
; CHECK-NEXT: vfirst.m a0, v9, v0.t
; CHECK-NEXT: ret
entry:
@@ -98,8 +98,8 @@ define iXLen @intrinsic_vfirst_mask_m_nxv2i1(<vscale x 2 x i1> %0, <vscale x 2 x
; CHECK-LABEL: intrinsic_vfirst_mask_m_nxv2i1:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a0, e8, mf4, ta, ma
; CHECK-NEXT: vfirst.m a0, v9, v0.t
; CHECK-NEXT: ret
entry:
@@ -138,8 +138,8 @@ define iXLen @intrinsic_vfirst_mask_m_nxv4i1(<vscale x 4 x i1> %0, <vscale x 4 x
; CHECK-LABEL: intrinsic_vfirst_mask_m_nxv4i1:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, ma
; CHECK-NEXT: vfirst.m a0, v9, v0.t
; CHECK-NEXT: ret
entry:
@@ -178,8 +178,8 @@ define iXLen @intrinsic_vfirst_mask_m_nxv8i1(<vscale x 8 x i1> %0, <vscale x 8 x
; CHECK-LABEL: intrinsic_vfirst_mask_m_nxv8i1:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m1, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a0, e8, m1, ta, ma
; CHECK-NEXT: vfirst.m a0, v9, v0.t
; CHECK-NEXT: ret
entry:
@@ -218,8 +218,8 @@ define iXLen @intrinsic_vfirst_mask_m_nxv16i1(<vscale x 16 x i1> %0, <vscale x 1
; CHECK-LABEL: intrinsic_vfirst_mask_m_nxv16i1:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a0, e8, m2, ta, ma
; CHECK-NEXT: vfirst.m a0, v9, v0.t
; CHECK-NEXT: ret
entry:
@@ -258,8 +258,8 @@ define iXLen @intrinsic_vfirst_mask_m_nxv32i1(<vscale x 32 x i1> %0, <vscale x 3
; CHECK-LABEL: intrinsic_vfirst_mask_m_nxv32i1:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, ma
; CHECK-NEXT: vfirst.m a0, v9, v0.t
; CHECK-NEXT: ret
entry:
@@ -298,8 +298,8 @@ define iXLen @intrinsic_vfirst_mask_m_nxv64i1(<vscale x 64 x i1> %0, <vscale x 6
; CHECK-LABEL: intrinsic_vfirst_mask_m_nxv64i1:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a0, e8, m8, ta, ma
; CHECK-NEXT: vfirst.m a0, v9, v0.t
; CHECK-NEXT: ret
entry:
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfma-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vfma-vp.ll
index a41c262116136..8201f18175e8b 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfma-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfma-vp.ll
@@ -1125,22 +1125,20 @@ define <vscale x 16 x double> @vfma_vv_nxv16f64(<vscale x 16 x double> %va, <vsc
; CHECK-NEXT: addi a1, a1, 16
; CHECK-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: vsetvli a3, zero, e8, mf4, ta, ma
-; CHECK-NEXT: slli a3, a1, 3
-; CHECK-NEXT: add a5, a2, a3
+; CHECK-NEXT: srli a3, a1, 3
+; CHECK-NEXT: slli a5, a1, 3
+; CHECK-NEXT: add a6, a2, a5
+; CHECK-NEXT: vl8re64.v v8, (a6)
+; CHECK-NEXT: csrr a6, vlenb
+; CHECK-NEXT: slli a6, a6, 4
+; CHECK-NEXT: add a6, sp, a6
+; CHECK-NEXT: addi a6, a6, 16
+; CHECK-NEXT: vs8r.v v8, (a6) # Unknown-size Folded Spill
+; CHECK-NEXT: add a5, a0, a5
+; CHECK-NEXT: sub a6, a4, a1
+; CHECK-NEXT: sltu a7, a4, a6
+; CHECK-NEXT: addi a7, a7, -1
; CHECK-NEXT: vl8re64.v v8, (a5)
-; CHECK-NEXT: csrr a5, vlenb
-; CHECK-NEXT: slli a5, a5, 4
-; CHECK-NEXT: add a5, sp, a5
-; CHECK-NEXT: addi a5, a5, 16
-; CHECK-NEXT: vs8r.v v8, (a5) # Unknown-size Folded Spill
-; CHECK-NEXT: sub a5, a4, a1
-; CHECK-NEXT: sltu a6, a4, a5
-; CHECK-NEXT: addi a6, a6, -1
-; CHECK-NEXT: and a5, a6, a5
-; CHECK-NEXT: srli a6, a1, 3
-; CHECK-NEXT: add a3, a0, a3
-; CHECK-NEXT: vl8re64.v v8, (a3)
; CHECK-NEXT: vl8re64.v v16, (a2)
; CHECK-NEXT: csrr a2, vlenb
; CHECK-NEXT: slli a2, a2, 3
@@ -1150,19 +1148,21 @@ define <vscale x 16 x double> @vfma_vv_nxv16f64(<vscale x 16 x double> %va, <vsc
; CHECK-NEXT: vl8re64.v v16, (a0)
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: vslidedown.vx v0, v0, a6
-; CHECK-NEXT: vsetvli zero, a5, e64, m8, ta, ma
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a2, 24
-; CHECK-NEXT: mul a0, a0, a2
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 4
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli a0, zero, e8, mf4, ta, ma
+; CHECK-NEXT: vslidedown.vx v0, v0, a3
+; CHECK-NEXT: and a0, a7, a6
+; CHECK-NEXT: csrr a2, vlenb
+; CHECK-NEXT: li a3, 24
+; CHECK-NEXT: mul a2, a2, a3
+; CHECK-NEXT: add a2, sp, a2
+; CHECK-NEXT: addi a2, a2, 16
+; CHECK-NEXT: vl8r.v v24, (a2) # Unknown-size Folded Reload
+; CHECK-NEXT: csrr a2, vlenb
+; CHECK-NEXT: slli a2, a2, 4
+; CHECK-NEXT: add a2, sp, a2
+; CHECK-NEXT: addi a2, a2, 16
+; CHECK-NEXT: vl8r.v v16, (a2) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfmadd.vv v8, v24, v16, v0.t
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: li a2, 24
@@ -1174,7 +1174,6 @@ define <vscale x 16 x double> @vfma_vv_nxv16f64(<vscale x 16 x double> %va, <vsc
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a4, a1
; CHECK-NEXT: .LBB92_2:
-; CHECK-NEXT: vsetvli zero, a4, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 5
@@ -1188,6 +1187,7 @@ define <vscale x 16 x double> @vfma_vv_nxv16f64(<vscale x 16 x double> %va, <vsc
; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, a4, e64, m8, ta, ma
; CHECK-NEXT: vfmadd.vv v16, v24, v8, v0.t
; CHECK-NEXT: vmv.v.v v8, v16
; CHECK-NEXT: csrr a0, vlenb
@@ -1234,24 +1234,23 @@ define <vscale x 16 x double> @vfma_vv_nxv16f64_unmasked(<vscale x 16 x double>
; CHECK-NEXT: vl8re64.v v16, (a3)
; CHECK-NEXT: sub a3, a4, a1
; CHECK-NEXT: sltu a5, a4, a3
-; CHECK-NEXT: addi a5, a5, -1
; CHECK-NEXT: vl8re64.v v8, (a2)
; CHECK-NEXT: addi a2, sp, 16
; CHECK-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
; CHECK-NEXT: vl8re64.v v0, (a0)
+; CHECK-NEXT: addi a5, a5, -1
; CHECK-NEXT: and a3, a5, a3
-; CHECK-NEXT: vsetvli zero, a3, e64, m8, ta, ma
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, a3, e64, m8, ta, ma
; CHECK-NEXT: vfmadd.vv v16, v8, v24
; CHECK-NEXT: bltu a4, a1, .LBB93_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a4, a1
; CHECK-NEXT: .LBB93_2:
-; CHECK-NEXT: vsetvli zero, a4, e64, m8, ta, ma
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 4
; CHECK-NEXT: add a0, sp, a0
@@ -1259,6 +1258,7 @@ define <vscale x 16 x double> @vfma_vv_nxv16f64_unmasked(<vscale x 16 x double>
; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, a4, e64, m8, ta, ma
; CHECK-NEXT: vfmadd.vv v0, v24, v8
; CHECK-NEXT: vmv.v.v v8, v0
; CHECK-NEXT: csrr a0, vlenb
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfmadd-constrained-sdnode.ll b/llvm/test/CodeGen/RISCV/rvv/vfmadd-constrained-sdnode.ll
index d4ba0f8c90733..c15b875e8f0c4 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfmadd-constrained-sdnode.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfmadd-constrained-sdnode.ll
@@ -239,15 +239,15 @@ define <vscale x 16 x half> @vfmadd_vv_nxv16f16(<vscale x 16 x half> %va, <vscal
; ZVFHMIN-NEXT: addi a0, sp, 16
; ZVFHMIN-NEXT: vs4r.v v16, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v8
; ZVFHMIN-NEXT: addi a0, sp, 16
; ZVFHMIN-NEXT: vl4r.v v8, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfmadd.vv v16, v0, v24
+; ZVFHMIN-NEXT: vfmadd.vv v24, v0, v16
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v16
+; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v24
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 2
; ZVFHMIN-NEXT: add sp, sp, a0
@@ -315,87 +315,88 @@ define <vscale x 32 x half> @vfmadd_vv_nxv32f16(<vscale x 32 x half> %va, <vscal
; ZVFHMIN-NEXT: addi sp, sp, -16
; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: li a2, 40
-; ZVFHMIN-NEXT: mul a1, a1, a2
+; ZVFHMIN-NEXT: slli a1, a1, 5
; ZVFHMIN-NEXT: sub sp, sp, a1
-; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x28, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 40 * vlenb
-; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a1, a1, 3
-; ZVFHMIN-NEXT: add a1, sp, a1
-; ZVFHMIN-NEXT: addi a1, a1, 16
-; ZVFHMIN-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vmv8r.v v0, v8
-; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: li a2, 24
-; ZVFHMIN-NEXT: mul a1, a1, a2
-; ZVFHMIN-NEXT: add a1, sp, a1
-; ZVFHMIN-NEXT: addi a1, a1, 16
-; ZVFHMIN-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vl8re16.v v24, (a0)
+; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x20, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 32 * vlenb
+; ZVFHMIN-NEXT: vl8re16.v v0, (a0)
+; ZVFHMIN-NEXT: vmv8r.v v24, v16
+; ZVFHMIN-NEXT: csrr a0, vlenb
+; ZVFHMIN-NEXT: slli a0, a0, 4
+; ZVFHMIN-NEXT: add a0, sp, a0
+; ZVFHMIN-NEXT: addi a0, a0, 16
+; ZVFHMIN-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vmv8r.v v16, v8
; ZVFHMIN-NEXT: addi a0, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v0
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v16
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 5
+; ZVFHMIN-NEXT: li a1, 24
+; ZVFHMIN-NEXT: mul a0, a0, a1
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v16
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v24
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 4
+; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v24
+; ZVFHMIN-NEXT: vmv8r.v v8, v0
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v8
+; ZVFHMIN-NEXT: csrr a0, vlenb
+; ZVFHMIN-NEXT: li a1, 24
+; ZVFHMIN-NEXT: mul a0, a0, a1
+; ZVFHMIN-NEXT: add a0, sp, a0
+; ZVFHMIN-NEXT: addi a0, a0, 16
+; ZVFHMIN-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: csrr a0, vlenb
+; ZVFHMIN-NEXT: slli a0, a0, 3
+; ZVFHMIN-NEXT: add a0, sp, a0
+; ZVFHMIN-NEXT: addi a0, a0, 16
+; ZVFHMIN-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
+; ZVFHMIN-NEXT: vfmadd.vv v0, v16, v24
+; ZVFHMIN-NEXT: addi a0, sp, 16
+; ZVFHMIN-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 5
+; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vl8r.v v0, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 4
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfmadd.vv v16, v8, v0
-; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v0, v16
+; ZVFHMIN-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: li a1, 24
; ZVFHMIN-NEXT: mul a0, a0, a1
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 5
+; ZVFHMIN-NEXT: li a1, 24
+; ZVFHMIN-NEXT: mul a0, a0, a1
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
-; ZVFHMIN-NEXT: addi a0, sp, 16
-; ZVFHMIN-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v20
+; ZVFHMIN-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 5
-; ZVFHMIN-NEXT: add a0, sp, a0
-; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfmadd.vv v8, v24, v16
+; ZVFHMIN-NEXT: vfmadd.vv v16, v8, v24
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v4, v8
-; ZVFHMIN-NEXT: vmv8r.v v8, v0
+; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v0
+; ZVFHMIN-NEXT: vfncvt.f.f.w v12, v16
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: li a1, 40
-; ZVFHMIN-NEXT: mul a0, a0, a1
+; ZVFHMIN-NEXT: slli a0, a0, 5
; ZVFHMIN-NEXT: add sp, sp, a0
; ZVFHMIN-NEXT: addi sp, sp, 16
; ZVFHMIN-NEXT: ret
@@ -415,77 +416,68 @@ define <vscale x 32 x half> @vfmadd_vf_nxv32f16(<vscale x 32 x half> %va, <vscal
; ZVFHMIN-NEXT: addi sp, sp, -16
; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 5
-; ZVFHMIN-NEXT: sub sp, sp, a0
-; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x20, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 32 * vlenb
-; ZVFHMIN-NEXT: vmv8r.v v0, v16
-; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: li a1, 24
; ZVFHMIN-NEXT: mul a0, a0, a1
-; ZVFHMIN-NEXT: add a0, sp, a0
-; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: sub sp, sp, a0
+; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x18, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 24 * vlenb
; ZVFHMIN-NEXT: addi a0, sp, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: fcvt.s.h fa5, fa0
; ZVFHMIN-NEXT: vsetvli a0, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfmv.v.f v24, fa5
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v12, v24
+; ZVFHMIN-NEXT: vfncvt.f.f.w v4, v24
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v8
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 4
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v0
+; ZVFHMIN-NEXT: vmv8r.v v8, v16
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v12
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v4
+; ZVFHMIN-NEXT: vmv.v.v v8, v4
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 4
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vl8r.v v0, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfmadd.vv v0, v16, v24
-; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v0
+; ZVFHMIN-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
+; ZVFHMIN-NEXT: vfmadd.vv v16, v24, v0
; ZVFHMIN-NEXT: addi a0, sp, 16
-; ZVFHMIN-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20
+; ZVFHMIN-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v28
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 4
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
-; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: li a1, 24
-; ZVFHMIN-NEXT: mul a0, a0, a1
-; ZVFHMIN-NEXT: add a0, sp, a0
-; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v28
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
+; ZVFHMIN-NEXT: vs8r.v v0, (a0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v8
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v12
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 4
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfmadd.vv v0, v16, v24
+; ZVFHMIN-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
+; ZVFHMIN-NEXT: vfmadd.vv v0, v24, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
+; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v16
; ZVFHMIN-NEXT: vfncvt.f.f.w v12, v0
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 5
+; ZVFHMIN-NEXT: li a1, 24
+; ZVFHMIN-NEXT: mul a0, a0, a1
; ZVFHMIN-NEXT: add sp, sp, a0
; ZVFHMIN-NEXT: addi sp, sp, 16
; ZVFHMIN-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfmadd-sdnode.ll b/llvm/test/CodeGen/RISCV/rvv/vfmadd-sdnode.ll
index 8a72b2ddafaca..4ce556efb44d0 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfmadd-sdnode.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfmadd-sdnode.ll
@@ -323,87 +323,88 @@ define <vscale x 32 x half> @vfmadd_vv_nxv32f16(<vscale x 32 x half> %va, <vscal
; ZVFHMIN-NEXT: addi sp, sp, -16
; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: li a2, 40
-; ZVFHMIN-NEXT: mul a1, a1, a2
+; ZVFHMIN-NEXT: slli a1, a1, 5
; ZVFHMIN-NEXT: sub sp, sp, a1
-; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x28, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 40 * vlenb
-; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a1, a1, 3
-; ZVFHMIN-NEXT: add a1, sp, a1
-; ZVFHMIN-NEXT: addi a1, a1, 16
-; ZVFHMIN-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vmv8r.v v0, v8
-; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: li a2, 24
-; ZVFHMIN-NEXT: mul a1, a1, a2
-; ZVFHMIN-NEXT: add a1, sp, a1
-; ZVFHMIN-NEXT: addi a1, a1, 16
-; ZVFHMIN-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vl8re16.v v24, (a0)
+; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x20, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 32 * vlenb
+; ZVFHMIN-NEXT: vl8re16.v v0, (a0)
+; ZVFHMIN-NEXT: vmv8r.v v24, v16
+; ZVFHMIN-NEXT: csrr a0, vlenb
+; ZVFHMIN-NEXT: slli a0, a0, 4
+; ZVFHMIN-NEXT: add a0, sp, a0
+; ZVFHMIN-NEXT: addi a0, a0, 16
+; ZVFHMIN-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vmv8r.v v16, v8
; ZVFHMIN-NEXT: addi a0, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v0
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v16
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 5
+; ZVFHMIN-NEXT: li a1, 24
+; ZVFHMIN-NEXT: mul a0, a0, a1
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v16
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v24
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 4
+; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v24
+; ZVFHMIN-NEXT: vmv8r.v v8, v0
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v8
+; ZVFHMIN-NEXT: csrr a0, vlenb
+; ZVFHMIN-NEXT: li a1, 24
+; ZVFHMIN-NEXT: mul a0, a0, a1
+; ZVFHMIN-NEXT: add a0, sp, a0
+; ZVFHMIN-NEXT: addi a0, a0, 16
+; ZVFHMIN-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: csrr a0, vlenb
+; ZVFHMIN-NEXT: slli a0, a0, 3
+; ZVFHMIN-NEXT: add a0, sp, a0
+; ZVFHMIN-NEXT: addi a0, a0, 16
+; ZVFHMIN-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
+; ZVFHMIN-NEXT: vfmadd.vv v0, v16, v24
+; ZVFHMIN-NEXT: addi a0, sp, 16
+; ZVFHMIN-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 5
+; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vl8r.v v0, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 4
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfmadd.vv v16, v8, v0
-; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v0, v16
+; ZVFHMIN-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: li a1, 24
; ZVFHMIN-NEXT: mul a0, a0, a1
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 5
+; ZVFHMIN-NEXT: li a1, 24
+; ZVFHMIN-NEXT: mul a0, a0, a1
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
-; ZVFHMIN-NEXT: addi a0, sp, 16
-; ZVFHMIN-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v20
+; ZVFHMIN-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 5
-; ZVFHMIN-NEXT: add a0, sp, a0
-; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfmadd.vv v8, v24, v16
+; ZVFHMIN-NEXT: vfmadd.vv v16, v8, v24
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v4, v8
-; ZVFHMIN-NEXT: vmv8r.v v8, v0
+; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v0
+; ZVFHMIN-NEXT: vfncvt.f.f.w v12, v16
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: li a1, 40
-; ZVFHMIN-NEXT: mul a0, a0, a1
+; ZVFHMIN-NEXT: slli a0, a0, 5
; ZVFHMIN-NEXT: add sp, sp, a0
; ZVFHMIN-NEXT: addi sp, sp, 16
; ZVFHMIN-NEXT: ret
@@ -423,10 +424,14 @@ define <vscale x 32 x half> @vfmadd_vf_nxv32f16(<vscale x 32 x half> %va, <vscal
; ZVFHMIN-NEXT: addi sp, sp, -16
; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 4
+; ZVFHMIN-NEXT: li a1, 24
+; ZVFHMIN-NEXT: mul a0, a0, a1
; ZVFHMIN-NEXT: sub sp, sp, a0
-; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
-; ZVFHMIN-NEXT: addi a0, sp, 16
+; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x18, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 24 * vlenb
+; ZVFHMIN-NEXT: csrr a0, vlenb
+; ZVFHMIN-NEXT: slli a0, a0, 3
+; ZVFHMIN-NEXT: add a0, sp, a0
+; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: fcvt.s.h fa5, fa0
; ZVFHMIN-NEXT: vsetvli a0, zero, e32, m8, ta, ma
@@ -435,33 +440,50 @@ define <vscale x 32 x half> @vfmadd_vf_nxv32f16(<vscale x 32 x half> %va, <vscal
; ZVFHMIN-NEXT: vfncvt.f.f.w v4, v24
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v8
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 3
+; ZVFHMIN-NEXT: slli a0, a0, 4
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v4
+; ZVFHMIN-NEXT: addi a0, sp, 16
+; ZVFHMIN-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v16
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 3
+; ZVFHMIN-NEXT: slli a0, a0, 4
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfmadd.vv v0, v8, v24
-; ZVFHMIN-NEXT: vmv8r.v v24, v8
+; ZVFHMIN-NEXT: csrr a0, vlenb
+; ZVFHMIN-NEXT: slli a0, a0, 3
+; ZVFHMIN-NEXT: add a0, sp, a0
+; ZVFHMIN-NEXT: addi a0, a0, 16
+; ZVFHMIN-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v16, v0
-; ZVFHMIN-NEXT: addi a0, sp, 16
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
+; ZVFHMIN-NEXT: csrr a0, vlenb
+; ZVFHMIN-NEXT: slli a0, a0, 4
+; ZVFHMIN-NEXT: add a0, sp, a0
+; ZVFHMIN-NEXT: addi a0, a0, 16
+; ZVFHMIN-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vmv4r.v v12, v20
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
+; ZVFHMIN-NEXT: csrr a0, vlenb
+; ZVFHMIN-NEXT: slli a0, a0, 4
+; ZVFHMIN-NEXT: add a0, sp, a0
+; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v12
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v20
+; ZVFHMIN-NEXT: addi a0, sp, 16
+; ZVFHMIN-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfmadd.vv v8, v24, v0
+; ZVFHMIN-NEXT: vfmadd.vv v16, v24, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v20, v8
-; ZVFHMIN-NEXT: vmv8r.v v8, v16
+; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v0
+; ZVFHMIN-NEXT: vfncvt.f.f.w v12, v16
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 4
+; ZVFHMIN-NEXT: li a1, 24
+; ZVFHMIN-NEXT: mul a0, a0, a1
; ZVFHMIN-NEXT: add sp, sp, a0
; ZVFHMIN-NEXT: addi sp, sp, 16
; ZVFHMIN-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfmax-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vfmax-vp.ll
index 7556b3ace5c6c..7ab999ea4fa7e 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfmax-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfmax-vp.ll
@@ -255,9 +255,9 @@ define <vscale x 32 x half> @vfmax_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
; ZVFHMIN-NEXT: srli a2, a2, 2
; ZVFHMIN-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
; ZVFHMIN-NEXT: vslidedown.vx v0, v0, a2
-; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
@@ -272,8 +272,8 @@ define <vscale x 32 x half> @vfmax_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
; ZVFHMIN-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v24
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v8
-; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v7
+; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfmax.vv v16, v24, v16, v0.t
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v16
@@ -312,9 +312,9 @@ define <vscale x 32 x half> @vfmax_vv_nxv32f16_unmasked(<vscale x 32 x half> %va
; ZVFHMIN-NEXT: vmset.m v24
; ZVFHMIN-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
; ZVFHMIN-NEXT: vslidedown.vx v0, v24, a2
-; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfmin-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vfmin-vp.ll
index 755c665376128..e928df85b5bb5 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfmin-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfmin-vp.ll
@@ -255,9 +255,9 @@ define <vscale x 32 x half> @vfmin_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
; ZVFHMIN-NEXT: srli a2, a2, 2
; ZVFHMIN-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
; ZVFHMIN-NEXT: vslidedown.vx v0, v0, a2
-; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
@@ -272,8 +272,8 @@ define <vscale x 32 x half> @vfmin_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
; ZVFHMIN-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v24
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v8
-; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v7
+; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfmin.vv v16, v24, v16, v0.t
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v16
@@ -312,9 +312,9 @@ define <vscale x 32 x half> @vfmin_vv_nxv32f16_unmasked(<vscale x 32 x half> %va
; ZVFHMIN-NEXT: vmset.m v24
; ZVFHMIN-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
; ZVFHMIN-NEXT: vslidedown.vx v0, v24, a2
-; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfmsub-constrained-sdnode.ll b/llvm/test/CodeGen/RISCV/rvv/vfmsub-constrained-sdnode.ll
index 5114f0a8d1d65..c835dc72268b3 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfmsub-constrained-sdnode.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfmsub-constrained-sdnode.ll
@@ -347,75 +347,64 @@ define <vscale x 32 x half> @vfmsub_vv_nxv32f16(<vscale x 32 x half> %va, <vscal
; ZVFHMIN-NEXT: addi sp, sp, -16
; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: li a2, 28
+; ZVFHMIN-NEXT: li a2, 24
; ZVFHMIN-NEXT: mul a1, a1, a2
; ZVFHMIN-NEXT: sub sp, sp, a1
-; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x1c, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 28 * vlenb
+; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x18, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 24 * vlenb
+; ZVFHMIN-NEXT: vmv8r.v v24, v16
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a1, a1, 3
+; ZVFHMIN-NEXT: slli a1, a1, 4
; ZVFHMIN-NEXT: add a1, sp, a1
; ZVFHMIN-NEXT: addi a1, a1, 16
; ZVFHMIN-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vl8re16.v v24, (a0)
+; ZVFHMIN-NEXT: vl8re16.v v16, (a0)
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: li a1, 20
-; ZVFHMIN-NEXT: mul a0, a0, a1
+; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v12
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v8
+; ZVFHMIN-NEXT: vmv4r.v v20, v12
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfneg.v v0, v0
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v12, v0
-; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 4
-; ZVFHMIN-NEXT: add a0, sp, a0
-; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vs4r.v v12, (a0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v0
; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v8
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v24
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v16
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfneg.v v0, v0
+; ZVFHMIN-NEXT: vfmacc.vv v0, v8, v24
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v24, v0
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v16
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v24
-; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: li a1, 20
-; ZVFHMIN-NEXT: mul a0, a0, a1
-; ZVFHMIN-NEXT: add a0, sp, a0
-; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v24
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v20
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfmacc.vv v8, v0, v16
+; ZVFHMIN-NEXT: vfneg.v v8, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v16, v8
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v16
+; ZVFHMIN-NEXT: addi a0, sp, 16
+; ZVFHMIN-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 4
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vl4r.v v8, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v8
-; ZVFHMIN-NEXT: addi a0, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v0, (a0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v28
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v12
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v28
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
; ZVFHMIN-NEXT: addi a0, sp, 16
-; ZVFHMIN-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfmadd.vv v8, v0, v24
+; ZVFHMIN-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
+; ZVFHMIN-NEXT: vfmadd.vv v24, v16, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v20, v8
-; ZVFHMIN-NEXT: vmv8r.v v8, v16
+; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v0
+; ZVFHMIN-NEXT: vfncvt.f.f.w v12, v24
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: li a1, 28
+; ZVFHMIN-NEXT: li a1, 24
; ZVFHMIN-NEXT: mul a0, a0, a1
; ZVFHMIN-NEXT: add sp, sp, a0
; ZVFHMIN-NEXT: addi sp, sp, 16
@@ -437,55 +426,93 @@ define <vscale x 32 x half> @vfmsub_vf_nxv32f16(<vscale x 32 x half> %va, <vscal
; ZVFHMIN-NEXT: addi sp, sp, -16
; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 4
+; ZVFHMIN-NEXT: li a1, 28
+; ZVFHMIN-NEXT: mul a0, a0, a1
; ZVFHMIN-NEXT: sub sp, sp, a0
-; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
-; ZVFHMIN-NEXT: vmv8r.v v24, v16
+; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x1c, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 28 * vlenb
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 3
+; ZVFHMIN-NEXT: li a1, 12
+; ZVFHMIN-NEXT: mul a0, a0, a1
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vmv8r.v v24, v8
+; ZVFHMIN-NEXT: addi a0, sp, 16
+; ZVFHMIN-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: fcvt.s.h fa5, fa0
; ZVFHMIN-NEXT: vsetvli a0, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfmv.v.f v0, fa5
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v16, v0
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v12
+; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v0
+; ZVFHMIN-NEXT: csrr a0, vlenb
+; ZVFHMIN-NEXT: slli a0, a0, 3
+; ZVFHMIN-NEXT: add a0, sp, a0
+; ZVFHMIN-NEXT: addi a0, a0, 16
+; ZVFHMIN-NEXT: vs4r.v v8, (a0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v24
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfneg.v v0, v0
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v20, v0
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v8
+; ZVFHMIN-NEXT: vfncvt.f.f.w v12, v0
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v16
+; ZVFHMIN-NEXT: csrr a0, vlenb
+; ZVFHMIN-NEXT: li a1, 20
+; ZVFHMIN-NEXT: mul a0, a0, a1
+; ZVFHMIN-NEXT: add a0, sp, a0
+; ZVFHMIN-NEXT: addi a0, a0, 16
+; ZVFHMIN-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v12
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8
+; ZVFHMIN-NEXT: csrr a0, vlenb
+; ZVFHMIN-NEXT: li a1, 20
+; ZVFHMIN-NEXT: mul a0, a0, a1
+; ZVFHMIN-NEXT: add a0, sp, a0
+; ZVFHMIN-NEXT: addi a0, a0, 16
+; ZVFHMIN-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfneg.v v0, v0
+; ZVFHMIN-NEXT: vfmacc.vv v0, v24, v16
+; ZVFHMIN-NEXT: addi a0, sp, 16
+; ZVFHMIN-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v0
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v24
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v8
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v16
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfmacc.vv v24, v0, v8
+; ZVFHMIN-NEXT: vfneg.v v8, v16
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v24
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20
-; ZVFHMIN-NEXT: addi a0, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v16
+; ZVFHMIN-NEXT: vfncvt.f.f.w v16, v8
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v16
+; ZVFHMIN-NEXT: csrr a0, vlenb
+; ZVFHMIN-NEXT: li a1, 20
+; ZVFHMIN-NEXT: mul a0, a0, a1
+; ZVFHMIN-NEXT: add a0, sp, a0
+; ZVFHMIN-NEXT: addi a0, a0, 16
+; ZVFHMIN-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
+; ZVFHMIN-NEXT: vl4r.v v16, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v16
+; ZVFHMIN-NEXT: csrr a0, vlenb
+; ZVFHMIN-NEXT: li a1, 12
+; ZVFHMIN-NEXT: mul a0, a0, a1
+; ZVFHMIN-NEXT: add a0, sp, a0
+; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v28
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: addi a0, sp, 16
+; ZVFHMIN-NEXT: csrr a0, vlenb
+; ZVFHMIN-NEXT: li a1, 20
+; ZVFHMIN-NEXT: mul a0, a0, a1
+; ZVFHMIN-NEXT: add a0, sp, a0
+; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfmadd.vv v16, v0, v24
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
+; ZVFHMIN-NEXT: vfmadd.vv v16, v8, v24
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
+; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v0
; ZVFHMIN-NEXT: vfncvt.f.f.w v12, v16
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 4
+; ZVFHMIN-NEXT: li a1, 28
+; ZVFHMIN-NEXT: mul a0, a0, a1
; ZVFHMIN-NEXT: add sp, sp, a0
; ZVFHMIN-NEXT: addi sp, sp, 16
; ZVFHMIN-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfmul-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vfmul-vp.ll
index 30d5919238cfa..61f3e63f246c6 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfmul-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfmul-vp.ll
@@ -525,9 +525,9 @@ define <vscale x 32 x half> @vfmul_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
; ZVFHMIN-NEXT: srli a2, a2, 2
; ZVFHMIN-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
; ZVFHMIN-NEXT: vslidedown.vx v0, v0, a2
-; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
@@ -542,8 +542,8 @@ define <vscale x 32 x half> @vfmul_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
; ZVFHMIN-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v24
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v8
-; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v7
+; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfmul.vv v16, v24, v16, v0.t
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v16
@@ -582,9 +582,9 @@ define <vscale x 32 x half> @vfmul_vv_nxv32f16_unmasked(<vscale x 32 x half> %va
; ZVFHMIN-NEXT: vmset.m v24
; ZVFHMIN-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
; ZVFHMIN-NEXT: vslidedown.vx v0, v24, a2
-; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
@@ -621,12 +621,20 @@ define <vscale x 32 x half> @vfmul_vf_nxv32f16(<vscale x 32 x half> %va, half %b
;
; ZVFHMIN-LABEL: vfmul_vf_nxv32f16:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: vmv1r.v v3, v0
+; ZVFHMIN-NEXT: addi sp, sp, -16
+; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
+; ZVFHMIN-NEXT: csrr a1, vlenb
+; ZVFHMIN-NEXT: slli a1, a1, 2
+; ZVFHMIN-NEXT: sub sp, sp, a1
+; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x04, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 4 * vlenb
+; ZVFHMIN-NEXT: vmv1r.v v7, v0
; ZVFHMIN-NEXT: fcvt.s.h fa5, fa0
; ZVFHMIN-NEXT: vsetvli a1, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfmv.v.f v24, fa5
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v4, v24
+; ZVFHMIN-NEXT: vfncvt.f.f.w v16, v24
+; ZVFHMIN-NEXT: addi a1, sp, 16
+; ZVFHMIN-NEXT: vs4r.v v16, (a1) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a2, vlenb
; ZVFHMIN-NEXT: slli a1, a2, 1
; ZVFHMIN-NEXT: sub a3, a0, a1
@@ -638,7 +646,9 @@ define <vscale x 32 x half> @vfmul_vf_nxv32f16(<vscale x 32 x half> %va, half %b
; ZVFHMIN-NEXT: vslidedown.vx v0, v0, a2
; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v4
+; ZVFHMIN-NEXT: addi a2, sp, 16
+; ZVFHMIN-NEXT: vl4r.v v12, (a2) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfmul.vv v16, v16, v24, v0.t
; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
@@ -648,11 +658,15 @@ define <vscale x 32 x half> @vfmul_vf_nxv32f16(<vscale x 32 x half> %va, half %b
; ZVFHMIN-NEXT: mv a0, a1
; ZVFHMIN-NEXT: .LBB22_2:
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8
+; ZVFHMIN-NEXT: vmv1r.v v0, v7
; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vmv1r.v v0, v3
; ZVFHMIN-NEXT: vfmul.vv v16, v16, v24, v0.t
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v16
+; ZVFHMIN-NEXT: csrr a0, vlenb
+; ZVFHMIN-NEXT: slli a0, a0, 2
+; ZVFHMIN-NEXT: add sp, sp, a0
+; ZVFHMIN-NEXT: addi sp, sp, 16
; ZVFHMIN-NEXT: ret
%elt.head = insertelement <vscale x 32 x half> poison, half %b, i32 0
%vb = shufflevector <vscale x 32 x half> %elt.head, <vscale x 32 x half> poison, <vscale x 32 x i32> zeroinitializer
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfmuladd-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vfmuladd-vp.ll
index 292f27794f378..abda6750e5a8a 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfmuladd-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfmuladd-vp.ll
@@ -1125,22 +1125,20 @@ define <vscale x 16 x double> @vfma_vv_nxv16f64(<vscale x 16 x double> %va, <vsc
; CHECK-NEXT: addi a1, a1, 16
; CHECK-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: vsetvli a3, zero, e8, mf4, ta, ma
-; CHECK-NEXT: slli a3, a1, 3
-; CHECK-NEXT: add a5, a2, a3
+; CHECK-NEXT: srli a3, a1, 3
+; CHECK-NEXT: slli a5, a1, 3
+; CHECK-NEXT: add a6, a2, a5
+; CHECK-NEXT: vl8re64.v v8, (a6)
+; CHECK-NEXT: csrr a6, vlenb
+; CHECK-NEXT: slli a6, a6, 4
+; CHECK-NEXT: add a6, sp, a6
+; CHECK-NEXT: addi a6, a6, 16
+; CHECK-NEXT: vs8r.v v8, (a6) # Unknown-size Folded Spill
+; CHECK-NEXT: add a5, a0, a5
+; CHECK-NEXT: sub a6, a4, a1
+; CHECK-NEXT: sltu a7, a4, a6
+; CHECK-NEXT: addi a7, a7, -1
; CHECK-NEXT: vl8re64.v v8, (a5)
-; CHECK-NEXT: csrr a5, vlenb
-; CHECK-NEXT: slli a5, a5, 4
-; CHECK-NEXT: add a5, sp, a5
-; CHECK-NEXT: addi a5, a5, 16
-; CHECK-NEXT: vs8r.v v8, (a5) # Unknown-size Folded Spill
-; CHECK-NEXT: sub a5, a4, a1
-; CHECK-NEXT: sltu a6, a4, a5
-; CHECK-NEXT: addi a6, a6, -1
-; CHECK-NEXT: and a5, a6, a5
-; CHECK-NEXT: srli a6, a1, 3
-; CHECK-NEXT: add a3, a0, a3
-; CHECK-NEXT: vl8re64.v v8, (a3)
; CHECK-NEXT: vl8re64.v v16, (a2)
; CHECK-NEXT: csrr a2, vlenb
; CHECK-NEXT: slli a2, a2, 3
@@ -1150,19 +1148,21 @@ define <vscale x 16 x double> @vfma_vv_nxv16f64(<vscale x 16 x double> %va, <vsc
; CHECK-NEXT: vl8re64.v v16, (a0)
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: vslidedown.vx v0, v0, a6
-; CHECK-NEXT: vsetvli zero, a5, e64, m8, ta, ma
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a2, 24
-; CHECK-NEXT: mul a0, a0, a2
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 4
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli a0, zero, e8, mf4, ta, ma
+; CHECK-NEXT: vslidedown.vx v0, v0, a3
+; CHECK-NEXT: and a0, a7, a6
+; CHECK-NEXT: csrr a2, vlenb
+; CHECK-NEXT: li a3, 24
+; CHECK-NEXT: mul a2, a2, a3
+; CHECK-NEXT: add a2, sp, a2
+; CHECK-NEXT: addi a2, a2, 16
+; CHECK-NEXT: vl8r.v v24, (a2) # Unknown-size Folded Reload
+; CHECK-NEXT: csrr a2, vlenb
+; CHECK-NEXT: slli a2, a2, 4
+; CHECK-NEXT: add a2, sp, a2
+; CHECK-NEXT: addi a2, a2, 16
+; CHECK-NEXT: vl8r.v v16, (a2) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfmadd.vv v8, v24, v16, v0.t
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: li a2, 24
@@ -1174,7 +1174,6 @@ define <vscale x 16 x double> @vfma_vv_nxv16f64(<vscale x 16 x double> %va, <vsc
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a4, a1
; CHECK-NEXT: .LBB92_2:
-; CHECK-NEXT: vsetvli zero, a4, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 5
@@ -1188,6 +1187,7 @@ define <vscale x 16 x double> @vfma_vv_nxv16f64(<vscale x 16 x double> %va, <vsc
; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, a4, e64, m8, ta, ma
; CHECK-NEXT: vfmadd.vv v16, v24, v8, v0.t
; CHECK-NEXT: vmv.v.v v8, v16
; CHECK-NEXT: csrr a0, vlenb
@@ -1234,24 +1234,23 @@ define <vscale x 16 x double> @vfma_vv_nxv16f64_unmasked(<vscale x 16 x double>
; CHECK-NEXT: vl8re64.v v16, (a3)
; CHECK-NEXT: sub a3, a4, a1
; CHECK-NEXT: sltu a5, a4, a3
-; CHECK-NEXT: addi a5, a5, -1
; CHECK-NEXT: vl8re64.v v8, (a2)
; CHECK-NEXT: addi a2, sp, 16
; CHECK-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
; CHECK-NEXT: vl8re64.v v0, (a0)
+; CHECK-NEXT: addi a5, a5, -1
; CHECK-NEXT: and a3, a5, a3
-; CHECK-NEXT: vsetvli zero, a3, e64, m8, ta, ma
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, a3, e64, m8, ta, ma
; CHECK-NEXT: vfmadd.vv v16, v8, v24
; CHECK-NEXT: bltu a4, a1, .LBB93_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a4, a1
; CHECK-NEXT: .LBB93_2:
-; CHECK-NEXT: vsetvli zero, a4, e64, m8, ta, ma
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 4
; CHECK-NEXT: add a0, sp, a0
@@ -1259,6 +1258,7 @@ define <vscale x 16 x double> @vfma_vv_nxv16f64_unmasked(<vscale x 16 x double>
; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, a4, e64, m8, ta, ma
; CHECK-NEXT: vfmadd.vv v0, v24, v8
; CHECK-NEXT: vmv.v.v v8, v0
; CHECK-NEXT: csrr a0, vlenb
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfneg-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vfneg-vp.ll
index 1db5fa1720a27..69ea7ce33cf6b 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfneg-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfneg-vp.ll
@@ -250,8 +250,8 @@ define <vscale x 32 x half> @vfneg_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
; ZVFHMIN-NEXT: mv a0, a1
; ZVFHMIN-NEXT: .LBB10_2:
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v8
-; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v16
+; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfneg.v v16, v24, v0.t
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v16
@@ -541,8 +541,8 @@ define <vscale x 16 x double> @vfneg_vv_nxv16f64(<vscale x 16 x double> %va, <vs
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB32_2:
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfneg.v v8, v8, v0.t
; CHECK-NEXT: ret
%v = call <vscale x 16 x double> @llvm.vp.fneg.nxv16f64(<vscale x 16 x double> %va, <vscale x 16 x i1> %m, i32 %evl)
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfnmadd-constrained-sdnode.ll b/llvm/test/CodeGen/RISCV/rvv/vfnmadd-constrained-sdnode.ll
index 785f60ad1d39c..b54590cd9d844 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfnmadd-constrained-sdnode.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfnmadd-constrained-sdnode.ll
@@ -412,85 +412,85 @@ define <vscale x 32 x half> @vfnmsub_vv_nxv32f16(<vscale x 32 x half> %va, <vsca
; ZVFHMIN-NEXT: mul a1, a1, a2
; ZVFHMIN-NEXT: sub sp, sp, a1
; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x18, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 24 * vlenb
-; ZVFHMIN-NEXT: vl8re16.v v0, (a0)
+; ZVFHMIN-NEXT: vl8re16.v v24, (a0)
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: li a1, 12
-; ZVFHMIN-NEXT: mul a0, a0, a1
+; ZVFHMIN-NEXT: slli a0, a0, 3
+; ZVFHMIN-NEXT: add a0, sp, a0
+; ZVFHMIN-NEXT: addi a0, a0, 16
+; ZVFHMIN-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: csrr a0, vlenb
+; ZVFHMIN-NEXT: slli a0, a0, 4
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v4
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v24
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfneg.v v24, v24
+; ZVFHMIN-NEXT: vfneg.v v0, v0
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v24
+; ZVFHMIN-NEXT: vfncvt.f.f.w v24, v0
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v16
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
+; ZVFHMIN-NEXT: vfneg.v v0, v0
+; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
+; ZVFHMIN-NEXT: vfncvt.f.f.w v16, v0
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v16
+; ZVFHMIN-NEXT: addi a0, sp, 16
+; ZVFHMIN-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v24
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: li a1, 20
-; ZVFHMIN-NEXT: mul a0, a0, a1
+; ZVFHMIN-NEXT: slli a0, a0, 4
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vs4r.v v8, (a0) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v0
+; ZVFHMIN-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v8
+; ZVFHMIN-NEXT: addi a0, sp, 16
+; ZVFHMIN-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfneg.v v24, v24
+; ZVFHMIN-NEXT: vfmadd.vv v0, v24, v8
+; ZVFHMIN-NEXT: csrr a0, vlenb
+; ZVFHMIN-NEXT: slli a0, a0, 3
+; ZVFHMIN-NEXT: add a0, sp, a0
+; ZVFHMIN-NEXT: addi a0, a0, 16
+; ZVFHMIN-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v24
-; ZVFHMIN-NEXT: addi a0, sp, 16
-; ZVFHMIN-NEXT: vs4r.v v8, (a0) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v28
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfneg.v v24, v24
+; ZVFHMIN-NEXT: vfneg.v v8, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v24
+; ZVFHMIN-NEXT: vfncvt.f.f.w v16, v8
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vs4r.v v8, (a0) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v16
+; ZVFHMIN-NEXT: vs4r.v v16, (a0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v20
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfneg.v v24, v24
+; ZVFHMIN-NEXT: vfneg.v v8, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v16, v24
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v16
+; ZVFHMIN-NEXT: vfncvt.f.f.w v20, v8
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v20
; ZVFHMIN-NEXT: addi a0, sp, 16
-; ZVFHMIN-NEXT: vl4r.v v16, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v16
+; ZVFHMIN-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: li a1, 12
-; ZVFHMIN-NEXT: mul a0, a0, a1
+; ZVFHMIN-NEXT: slli a0, a0, 4
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v24
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfmadd.vv v8, v16, v0
-; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v16, v8
+; ZVFHMIN-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vl4r.v v8, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v8
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8
; ZVFHMIN-NEXT: addi a0, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v0, (a0) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v28
-; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: li a1, 20
-; ZVFHMIN-NEXT: mul a0, a0, a1
-; ZVFHMIN-NEXT: add a0, sp, a0
-; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vl4r.v v20, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v20
+; ZVFHMIN-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: addi a0, sp, 16
-; ZVFHMIN-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfmadd.vv v0, v8, v24
+; ZVFHMIN-NEXT: vfmadd.vv v16, v24, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v20, v0
-; ZVFHMIN-NEXT: vmv8r.v v8, v16
+; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v0
+; ZVFHMIN-NEXT: vfncvt.f.f.w v12, v16
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: li a1, 24
; ZVFHMIN-NEXT: mul a0, a0, a1
@@ -515,16 +515,16 @@ define <vscale x 32 x half> @vfnmsub_vf_nxv32f16(<vscale x 32 x half> %va, <vsca
; ZVFHMIN-NEXT: addi sp, sp, -16
; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: li a1, 12
+; ZVFHMIN-NEXT: li a1, 20
; ZVFHMIN-NEXT: mul a0, a0, a1
; ZVFHMIN-NEXT: sub sp, sp, a0
-; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x0c, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 12 * vlenb
+; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x14, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 20 * vlenb
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 2
+; ZVFHMIN-NEXT: li a1, 12
+; ZVFHMIN-NEXT: mul a0, a0, a1
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vmv8r.v v16, v8
; ZVFHMIN-NEXT: fcvt.s.h fa5, fa0
; ZVFHMIN-NEXT: vsetvli a0, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfmv.v.f v24, fa5
@@ -534,48 +534,61 @@ define <vscale x 32 x half> @vfnmsub_vf_nxv32f16(<vscale x 32 x half> %va, <vsca
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfneg.v v0, v24
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v0
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v20
+; ZVFHMIN-NEXT: vfncvt.f.f.w v16, v0
+; ZVFHMIN-NEXT: csrr a0, vlenb
+; ZVFHMIN-NEXT: slli a0, a0, 3
+; ZVFHMIN-NEXT: add a0, sp, a0
+; ZVFHMIN-NEXT: addi a0, a0, 16
+; ZVFHMIN-NEXT: vs4r.v v16, (a0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfneg.v v0, v0
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v12, v0
+; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v0
+; ZVFHMIN-NEXT: csrr a0, vlenb
+; ZVFHMIN-NEXT: li a1, 12
+; ZVFHMIN-NEXT: mul a0, a0, a1
+; ZVFHMIN-NEXT: add a0, sp, a0
+; ZVFHMIN-NEXT: addi a0, a0, 16
+; ZVFHMIN-NEXT: vl8r.v v0, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v0
; ZVFHMIN-NEXT: addi a0, sp, 16
-; ZVFHMIN-NEXT: vs4r.v v12, (a0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v8
; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v16
+; ZVFHMIN-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfneg.v v0, v0
+; ZVFHMIN-NEXT: vfmadd.vv v0, v16, v24
+; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
+; ZVFHMIN-NEXT: vfneg.v v16, v16
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v12, v0
+; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v16
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8
+; ZVFHMIN-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 2
+; ZVFHMIN-NEXT: li a1, 12
+; ZVFHMIN-NEXT: mul a0, a0, a1
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v16
+; ZVFHMIN-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8
-; ZVFHMIN-NEXT: vmv.v.v v12, v8
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfmadd.vv v16, v0, v24
-; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v16
-; ZVFHMIN-NEXT: addi a0, sp, 16
-; ZVFHMIN-NEXT: vl4r.v v24, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v24
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 2
+; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vl8r.v v0, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v4
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v12
+; ZVFHMIN-NEXT: vl4r.v v8, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8
+; ZVFHMIN-NEXT: addi a0, sp, 16
+; ZVFHMIN-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfmadd.vv v0, v24, v16
+; ZVFHMIN-NEXT: vfmadd.vv v16, v24, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v12, v0
+; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v0
+; ZVFHMIN-NEXT: vfncvt.f.f.w v12, v16
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: li a1, 12
+; ZVFHMIN-NEXT: li a1, 20
; ZVFHMIN-NEXT: mul a0, a0, a1
; ZVFHMIN-NEXT: add sp, sp, a0
; ZVFHMIN-NEXT: addi sp, sp, 16
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfnmsub-constrained-sdnode.ll b/llvm/test/CodeGen/RISCV/rvv/vfnmsub-constrained-sdnode.ll
index 1a2da051c962e..2f7e693a8a6f9 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfnmsub-constrained-sdnode.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfnmsub-constrained-sdnode.ll
@@ -376,85 +376,76 @@ define <vscale x 32 x half> @vfnmsub_vv_nxv32f16(<vscale x 32 x half> %va, <vsca
; ZVFHMIN-NEXT: addi sp, sp, -16
; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: li a2, 28
+; ZVFHMIN-NEXT: li a2, 24
; ZVFHMIN-NEXT: mul a1, a1, a2
; ZVFHMIN-NEXT: sub sp, sp, a1
-; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x1c, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 28 * vlenb
-; ZVFHMIN-NEXT: vl8re16.v v0, (a0)
-; ZVFHMIN-NEXT: addi a0, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x18, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 24 * vlenb
+; ZVFHMIN-NEXT: vl8re16.v v24, (a0)
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 4
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v4
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfneg.v v24, v24
-; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v24
+; ZVFHMIN-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: li a1, 24
-; ZVFHMIN-NEXT: mul a0, a0, a1
+; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vs4r.v v8, (a0) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v0
+; ZVFHMIN-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v24
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfneg.v v24, v24
+; ZVFHMIN-NEXT: vfneg.v v0, v0
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v0, v24
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v16
-; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 3
-; ZVFHMIN-NEXT: add a0, sp, a0
-; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v0
+; ZVFHMIN-NEXT: vfncvt.f.f.w v24, v0
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v24
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v16
+; ZVFHMIN-NEXT: addi a0, sp, 16
+; ZVFHMIN-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v8
+; ZVFHMIN-NEXT: vmv4r.v v20, v12
+; ZVFHMIN-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
+; ZVFHMIN-NEXT: vfmadd.vv v0, v24, v8
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 4
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v24
+; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v28
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 3
-; ZVFHMIN-NEXT: add a0, sp, a0
-; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vl8r.v v0, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfmadd.vv v8, v16, v0
+; ZVFHMIN-NEXT: vfneg.v v8, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v16, v8
-; ZVFHMIN-NEXT: addi a0, sp, 16
-; ZVFHMIN-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v12
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 3
+; ZVFHMIN-NEXT: slli a0, a0, 4
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vs8r.v v0, (a0) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v28
+; ZVFHMIN-NEXT: vs4r.v v16, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: li a1, 24
-; ZVFHMIN-NEXT: mul a0, a0, a1
+; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vl4r.v v20, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v20
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
+; ZVFHMIN-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v28
+; ZVFHMIN-NEXT: addi a0, sp, 16
+; ZVFHMIN-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 3
+; ZVFHMIN-NEXT: slli a0, a0, 4
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfmadd.vv v8, v0, v24
+; ZVFHMIN-NEXT: vl4r.v v8, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8
+; ZVFHMIN-NEXT: addi a0, sp, 16
+; ZVFHMIN-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
+; ZVFHMIN-NEXT: vfmadd.vv v16, v24, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v20, v8
-; ZVFHMIN-NEXT: vmv8r.v v8, v16
+; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v0
+; ZVFHMIN-NEXT: vfncvt.f.f.w v12, v16
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: li a1, 28
+; ZVFHMIN-NEXT: li a1, 24
; ZVFHMIN-NEXT: mul a0, a0, a1
; ZVFHMIN-NEXT: add sp, sp, a0
; ZVFHMIN-NEXT: addi sp, sp, 16
@@ -480,7 +471,6 @@ define <vscale x 32 x half> @vfnmsub_vf_nxv32f16(<vscale x 32 x half> %va, <vsca
; ZVFHMIN-NEXT: mul a0, a0, a1
; ZVFHMIN-NEXT: sub sp, sp, a0
; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x18, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 24 * vlenb
-; ZVFHMIN-NEXT: vmv8r.v v24, v8
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 4
; ZVFHMIN-NEXT: add a0, sp, a0
@@ -488,58 +478,59 @@ define <vscale x 32 x half> @vfnmsub_vf_nxv32f16(<vscale x 32 x half> %va, <vsca
; ZVFHMIN-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: fcvt.s.h fa5, fa0
; ZVFHMIN-NEXT: vsetvli a0, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfmv.v.f v8, fa5
+; ZVFHMIN-NEXT: vfmv.v.f v24, fa5
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v4, v8
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v4
+; ZVFHMIN-NEXT: vfncvt.f.f.w v4, v24
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v4
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfneg.v v0, v8
+; ZVFHMIN-NEXT: vfneg.v v24, v24
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v28, v0
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v24
+; ZVFHMIN-NEXT: vfncvt.f.f.w v4, v24
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v8
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vs8r.v v0, (a0) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v16
+; ZVFHMIN-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vmv8r.v v8, v16
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8
; ZVFHMIN-NEXT: addi a0, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v28
-; ZVFHMIN-NEXT: vmv.v.v v16, v28
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
+; ZVFHMIN-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v4
+; ZVFHMIN-NEXT: vmv.v.v v8, v4
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: addi a0, sp, 16
-; ZVFHMIN-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfmadd.vv v0, v8, v24
-; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v0
+; ZVFHMIN-NEXT: vl8r.v v0, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
+; ZVFHMIN-NEXT: vfmadd.vv v24, v0, v16
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 4
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v28
+; ZVFHMIN-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v20
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vs8r.v v0, (a0) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v16
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v12
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfmadd.vv v0, v24, v16
+; ZVFHMIN-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
+; ZVFHMIN-NEXT: vfmadd.vv v16, v0, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v12, v0
+; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v24
+; ZVFHMIN-NEXT: vfncvt.f.f.w v12, v16
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: li a1, 24
; ZVFHMIN-NEXT: mul a0, a0, a1
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfpext-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vfpext-vp.ll
index aaaf4ad460711..28dc3e765dc3e 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfpext-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfpext-vp.ll
@@ -112,8 +112,8 @@ define <vscale x 32 x float> @vfpext_nxv32f16_nxv32f32(<vscale x 32 x half> %a,
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB7_2:
-; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, ma
; CHECK-NEXT: vfwcvt.f.f.v v24, v8, v0.t
; CHECK-NEXT: vmv8r.v v8, v24
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfptosi-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vfptosi-vp.ll
index 15c4bf255e6dc..f3544589407d8 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfptosi-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfptosi-vp.ll
@@ -394,7 +394,15 @@ declare <vscale x 32 x i16> @llvm.vp.fptosi.nxv32i16.nxv32f32(<vscale x 32 x flo
define <vscale x 32 x i16> @vfptosi_nxv32i16_nxv32f32(<vscale x 32 x float> %va, <vscale x 32 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfptosi_nxv32i16_nxv32f32:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v24, v0
+; CHECK-NEXT: addi sp, sp, -16
+; CHECK-NEXT: .cfi_def_cfa_offset 16
+; CHECK-NEXT: csrr a1, vlenb
+; CHECK-NEXT: slli a1, a1, 3
+; CHECK-NEXT: sub sp, sp, a1
+; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; CHECK-NEXT: vmv1r.v v7, v0
+; CHECK-NEXT: addi a1, sp, 16
+; CHECK-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
; CHECK-NEXT: csrr a1, vlenb
; CHECK-NEXT: srli a2, a1, 2
; CHECK-NEXT: vsetvli a3, zero, e8, mf2, ta, ma
@@ -404,16 +412,22 @@ define <vscale x 32 x i16> @vfptosi_nxv32i16_nxv32f32(<vscale x 32 x float> %va,
; CHECK-NEXT: sltu a3, a0, a2
; CHECK-NEXT: addi a3, a3, -1
; CHECK-NEXT: and a2, a3, a2
+; CHECK-NEXT: addi a3, sp, 16
+; CHECK-NEXT: vl8r.v v24, (a3) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, a2, e16, m4, ta, ma
-; CHECK-NEXT: vfncvt.rtz.x.f.w v28, v16, v0.t
+; CHECK-NEXT: vfncvt.rtz.x.f.w v20, v24, v0.t
; CHECK-NEXT: bltu a0, a1, .LBB25_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB25_2:
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v24
-; CHECK-NEXT: vfncvt.rtz.x.f.w v24, v8, v0.t
-; CHECK-NEXT: vmv8r.v v8, v24
+; CHECK-NEXT: vfncvt.rtz.x.f.w v16, v8, v0.t
+; CHECK-NEXT: vmv8r.v v8, v16
+; CHECK-NEXT: csrr a0, vlenb
+; CHECK-NEXT: slli a0, a0, 3
+; CHECK-NEXT: add sp, sp, a0
+; CHECK-NEXT: addi sp, sp, 16
; CHECK-NEXT: ret
%v = call <vscale x 32 x i16> @llvm.vp.fptosi.nxv32i16.nxv32f32(<vscale x 32 x float> %va, <vscale x 32 x i1> %m, i32 %evl)
ret <vscale x 32 x i16> %v
@@ -440,8 +454,8 @@ define <vscale x 32 x i32> @vfptosi_nxv32i32_nxv32f32(<vscale x 32 x float> %va,
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB26_2:
-; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; CHECK-NEXT: vfcvt.rtz.x.f.v v8, v8, v0.t
; CHECK-NEXT: ret
%v = call <vscale x 32 x i32> @llvm.vp.fptosi.nxv32i32.nxv32f32(<vscale x 32 x float> %va, <vscale x 32 x i1> %m, i32 %evl)
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfptoui-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vfptoui-vp.ll
index a2591e7dc35f0..9fd2d8edb2203 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfptoui-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfptoui-vp.ll
@@ -394,7 +394,15 @@ declare <vscale x 32 x i16> @llvm.vp.fptoui.nxv32i16.nxv32f32(<vscale x 32 x flo
define <vscale x 32 x i16> @vfptoui_nxv32i16_nxv32f32(<vscale x 32 x float> %va, <vscale x 32 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfptoui_nxv32i16_nxv32f32:
; CHECK: # %bb.0:
-; CHECK-NEXT: vmv1r.v v24, v0
+; CHECK-NEXT: addi sp, sp, -16
+; CHECK-NEXT: .cfi_def_cfa_offset 16
+; CHECK-NEXT: csrr a1, vlenb
+; CHECK-NEXT: slli a1, a1, 3
+; CHECK-NEXT: sub sp, sp, a1
+; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; CHECK-NEXT: vmv1r.v v7, v0
+; CHECK-NEXT: addi a1, sp, 16
+; CHECK-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
; CHECK-NEXT: csrr a1, vlenb
; CHECK-NEXT: srli a2, a1, 2
; CHECK-NEXT: vsetvli a3, zero, e8, mf2, ta, ma
@@ -404,16 +412,22 @@ define <vscale x 32 x i16> @vfptoui_nxv32i16_nxv32f32(<vscale x 32 x float> %va,
; CHECK-NEXT: sltu a3, a0, a2
; CHECK-NEXT: addi a3, a3, -1
; CHECK-NEXT: and a2, a3, a2
+; CHECK-NEXT: addi a3, sp, 16
+; CHECK-NEXT: vl8r.v v24, (a3) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, a2, e16, m4, ta, ma
-; CHECK-NEXT: vfncvt.rtz.xu.f.w v28, v16, v0.t
+; CHECK-NEXT: vfncvt.rtz.xu.f.w v20, v24, v0.t
; CHECK-NEXT: bltu a0, a1, .LBB25_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB25_2:
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, ma
-; CHECK-NEXT: vmv1r.v v0, v24
-; CHECK-NEXT: vfncvt.rtz.xu.f.w v24, v8, v0.t
-; CHECK-NEXT: vmv8r.v v8, v24
+; CHECK-NEXT: vfncvt.rtz.xu.f.w v16, v8, v0.t
+; CHECK-NEXT: vmv8r.v v8, v16
+; CHECK-NEXT: csrr a0, vlenb
+; CHECK-NEXT: slli a0, a0, 3
+; CHECK-NEXT: add sp, sp, a0
+; CHECK-NEXT: addi sp, sp, 16
; CHECK-NEXT: ret
%v = call <vscale x 32 x i16> @llvm.vp.fptoui.nxv32i16.nxv32f32(<vscale x 32 x float> %va, <vscale x 32 x i1> %m, i32 %evl)
ret <vscale x 32 x i16> %v
@@ -440,8 +454,8 @@ define <vscale x 32 x i32> @vfptoui_nxv32i32_nxv32f32(<vscale x 32 x float> %va,
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB26_2:
-; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; CHECK-NEXT: vfcvt.rtz.xu.f.v v8, v8, v0.t
; CHECK-NEXT: ret
%v = call <vscale x 32 x i32> @llvm.vp.fptoui.nxv32i32.nxv32f32(<vscale x 32 x float> %va, <vscale x 32 x i1> %m, i32 %evl)
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfptrunc-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vfptrunc-vp.ll
index 0c3abe37af27a..d2219cf963596 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfptrunc-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfptrunc-vp.ll
@@ -113,16 +113,16 @@ define <vscale x 16 x float> @vfptrunc_nxv16f32_nxv16f64(<vscale x 16 x double>
; CHECK-NEXT: sltu a3, a0, a2
; CHECK-NEXT: addi a3, a3, -1
; CHECK-NEXT: and a2, a3, a2
+; CHECK-NEXT: addi a3, sp, 16
+; CHECK-NEXT: vl8r.v v24, (a3) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, a2, e32, m4, ta, ma
-; CHECK-NEXT: addi a2, sp, 16
-; CHECK-NEXT: vl8r.v v24, (a2) # Unknown-size Folded Reload
; CHECK-NEXT: vfncvt.f.f.w v20, v24, v0.t
; CHECK-NEXT: bltu a0, a1, .LBB7_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB7_2:
-; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v7
+; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, ma
; CHECK-NEXT: vfncvt.f.f.w v16, v8, v0.t
; CHECK-NEXT: vmv8r.v v8, v16
; CHECK-NEXT: csrr a0, vlenb
@@ -169,11 +169,11 @@ define <vscale x 32 x float> @vfptrunc_nxv32f32_nxv32f64(<vscale x 32 x double>
; CHECK-NEXT: sub a6, a5, a1
; CHECK-NEXT: sltu a7, a5, a6
; CHECK-NEXT: addi a7, a7, -1
-; CHECK-NEXT: and a6, a7, a6
-; CHECK-NEXT: vsetvli a7, zero, e8, mf4, ta, ma
; CHECK-NEXT: vl8re64.v v24, (a0)
+; CHECK-NEXT: vsetvli a0, zero, e8, mf4, ta, ma
; CHECK-NEXT: vslidedown.vx v0, v16, a3
-; CHECK-NEXT: vsetvli zero, a6, e32, m4, ta, ma
+; CHECK-NEXT: and a0, a7, a6
+; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, ma
; CHECK-NEXT: vfncvt.f.f.w v20, v8, v0.t
; CHECK-NEXT: bltu a5, a1, .LBB8_2
; CHECK-NEXT: # %bb.1:
@@ -181,8 +181,8 @@ define <vscale x 32 x float> @vfptrunc_nxv32f32_nxv32f64(<vscale x 32 x double>
; CHECK-NEXT: .LBB8_2:
; CHECK-NEXT: vsetvli a0, zero, e8, mf4, ta, ma
; CHECK-NEXT: vslidedown.vx v6, v7, a3
-; CHECK-NEXT: vsetvli zero, a5, e32, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vsetvli zero, a5, e32, m4, ta, ma
; CHECK-NEXT: vfncvt.f.f.w v16, v24, v0.t
; CHECK-NEXT: bltu a2, a4, .LBB8_4
; CHECK-NEXT: # %bb.3:
@@ -192,22 +192,22 @@ define <vscale x 32 x float> @vfptrunc_nxv32f32_nxv32f64(<vscale x 32 x double>
; CHECK-NEXT: sltu a3, a2, a0
; CHECK-NEXT: addi a3, a3, -1
; CHECK-NEXT: and a0, a3, a0
-; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v6
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: addi a3, sp, 16
+; CHECK-NEXT: vl8r.v v8, (a3) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, ma
; CHECK-NEXT: vfncvt.f.f.w v28, v8, v0.t
; CHECK-NEXT: bltu a2, a1, .LBB8_6
; CHECK-NEXT: # %bb.5:
; CHECK-NEXT: mv a2, a1
; CHECK-NEXT: .LBB8_6:
-; CHECK-NEXT: vsetvli zero, a2, e32, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, a2, e32, m4, ta, ma
; CHECK-NEXT: vfncvt.f.f.w v24, v8, v0.t
; CHECK-NEXT: vmv8r.v v8, v24
; CHECK-NEXT: csrr a0, vlenb
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfsqrt-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vfsqrt-vp.ll
index d6caad15e40a2..bd229e0220a4b 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfsqrt-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfsqrt-vp.ll
@@ -250,8 +250,8 @@ define <vscale x 32 x half> @vfsqrt_vv_nxv32f16(<vscale x 32 x half> %va, <vscal
; ZVFHMIN-NEXT: mv a0, a1
; ZVFHMIN-NEXT: .LBB10_2:
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v8
-; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v16
+; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfsqrt.v v16, v24, v0.t
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v16
@@ -541,8 +541,8 @@ define <vscale x 16 x double> @vfsqrt_vv_nxv16f64(<vscale x 16 x double> %va, <v
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB32_2:
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; CHECK-NEXT: vfsqrt.v v8, v8, v0.t
; CHECK-NEXT: ret
%v = call <vscale x 16 x double> @llvm.vp.sqrt.nxv16f64(<vscale x 16 x double> %va, <vscale x 16 x i1> %m, i32 %evl)
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfsub-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vfsub-vp.ll
index 2eae18d7cc493..c833f8048fe32 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfsub-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfsub-vp.ll
@@ -525,9 +525,9 @@ define <vscale x 32 x half> @vfsub_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
; ZVFHMIN-NEXT: srli a2, a2, 2
; ZVFHMIN-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
; ZVFHMIN-NEXT: vslidedown.vx v0, v0, a2
-; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
@@ -542,8 +542,8 @@ define <vscale x 32 x half> @vfsub_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
; ZVFHMIN-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v24
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v8
-; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v7
+; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfsub.vv v16, v24, v16, v0.t
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v16
@@ -582,9 +582,9 @@ define <vscale x 32 x half> @vfsub_vv_nxv32f16_unmasked(<vscale x 32 x half> %va
; ZVFHMIN-NEXT: vmset.m v24
; ZVFHMIN-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
; ZVFHMIN-NEXT: vslidedown.vx v0, v24, a2
-; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
@@ -621,12 +621,20 @@ define <vscale x 32 x half> @vfsub_vf_nxv32f16(<vscale x 32 x half> %va, half %b
;
; ZVFHMIN-LABEL: vfsub_vf_nxv32f16:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: vmv1r.v v3, v0
+; ZVFHMIN-NEXT: addi sp, sp, -16
+; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
+; ZVFHMIN-NEXT: csrr a1, vlenb
+; ZVFHMIN-NEXT: slli a1, a1, 2
+; ZVFHMIN-NEXT: sub sp, sp, a1
+; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x04, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 4 * vlenb
+; ZVFHMIN-NEXT: vmv1r.v v7, v0
; ZVFHMIN-NEXT: fcvt.s.h fa5, fa0
; ZVFHMIN-NEXT: vsetvli a1, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfmv.v.f v24, fa5
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v4, v24
+; ZVFHMIN-NEXT: vfncvt.f.f.w v16, v24
+; ZVFHMIN-NEXT: addi a1, sp, 16
+; ZVFHMIN-NEXT: vs4r.v v16, (a1) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a2, vlenb
; ZVFHMIN-NEXT: slli a1, a2, 1
; ZVFHMIN-NEXT: sub a3, a0, a1
@@ -638,7 +646,9 @@ define <vscale x 32 x half> @vfsub_vf_nxv32f16(<vscale x 32 x half> %va, half %b
; ZVFHMIN-NEXT: vslidedown.vx v0, v0, a2
; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v4
+; ZVFHMIN-NEXT: addi a2, sp, 16
+; ZVFHMIN-NEXT: vl4r.v v12, (a2) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfsub.vv v16, v16, v24, v0.t
; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
@@ -648,11 +658,15 @@ define <vscale x 32 x half> @vfsub_vf_nxv32f16(<vscale x 32 x half> %va, half %b
; ZVFHMIN-NEXT: mv a0, a1
; ZVFHMIN-NEXT: .LBB22_2:
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8
+; ZVFHMIN-NEXT: vmv1r.v v0, v7
; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vmv1r.v v0, v3
; ZVFHMIN-NEXT: vfsub.vv v16, v16, v24, v0.t
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v16
+; ZVFHMIN-NEXT: csrr a0, vlenb
+; ZVFHMIN-NEXT: slli a0, a0, 2
+; ZVFHMIN-NEXT: add sp, sp, a0
+; ZVFHMIN-NEXT: addi sp, sp, 16
; ZVFHMIN-NEXT: ret
%elt.head = insertelement <vscale x 32 x half> poison, half %b, i32 0
%vb = shufflevector <vscale x 32 x half> %elt.head, <vscale x 32 x half> poison, <vscale x 32 x i32> zeroinitializer
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfwmacc-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vfwmacc-vp.ll
index 78f3792dbaf06..f9d992a40299c 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfwmacc-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfwmacc-vp.ll
@@ -671,9 +671,9 @@ define <vscale x 16 x float> @vfmacc_vv_nxv16f32(<vscale x 16 x half> %a, <vscal
; ZVFHMIN-NEXT: vsetvli zero, a0, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8, v0.t
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12, v0.t
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: addi a0, sp, 16
; ZVFHMIN-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfmadd.vv v24, v16, v8, v0.t
; ZVFHMIN-NEXT: vmv.v.v v8, v24
; ZVFHMIN-NEXT: csrr a0, vlenb
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfwnmacc-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vfwnmacc-vp.ll
index ca0bbfd65ca29..c11867d55ba0c 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfwnmacc-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfwnmacc-vp.ll
@@ -625,9 +625,9 @@ define <vscale x 16 x float> @vfnmacc_vv_nxv16f32(<vscale x 16 x half> %a, <vsca
; ZVFHMIN-NEXT: vsetvli zero, a0, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8, v0.t
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12, v0.t
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: addi a0, sp, 16
; ZVFHMIN-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfnmadd.vv v24, v16, v8, v0.t
; ZVFHMIN-NEXT: vmv.v.v v8, v24
; ZVFHMIN-NEXT: csrr a0, vlenb
@@ -710,17 +710,17 @@ define <vscale x 16 x float> @vfnmacc_vf_nxv16f32_commute(<vscale x 16 x half> %
;
; ZVFHMIN-LABEL: vfnmacc_vf_nxv16f32_commute:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: vmv4r.v v24, v8
; ZVFHMIN-NEXT: fcvt.s.h fa5, fa0
; ZVFHMIN-NEXT: vsetvli a1, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfmv.v.f v8, fa5
+; ZVFHMIN-NEXT: vfmv.v.f v24, fa5
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v4, v8
+; ZVFHMIN-NEXT: vfncvt.f.f.w v4, v24
; ZVFHMIN-NEXT: vsetvli zero, a0, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v24, v0.t
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v4, v0.t
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v8, v0.t
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v4, v0.t
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfnmadd.vv v8, v24, v16, v0.t
+; ZVFHMIN-NEXT: vfnmadd.vv v24, v8, v16, v0.t
+; ZVFHMIN-NEXT: vmv.v.v v8, v24
; ZVFHMIN-NEXT: ret
%elt.head = insertelement <vscale x 16 x half> poison, half %b, i32 0
%vb = shufflevector <vscale x 16 x half> %elt.head, <vscale x 16 x half> poison, <vscale x 16 x i32> zeroinitializer
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfwnmsac-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vfwnmsac-vp.ll
index 2797ca2eb3163..0ad7be47bcc8e 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfwnmsac-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfwnmsac-vp.ll
@@ -601,9 +601,9 @@ define <vscale x 16 x float> @vfnmsac_vv_nxv16f32(<vscale x 16 x half> %a, <vsca
; ZVFHMIN-NEXT: vsetvli zero, a0, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8, v0.t
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12, v0.t
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: addi a0, sp, 16
; ZVFHMIN-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfnmsub.vv v24, v16, v8, v0.t
; ZVFHMIN-NEXT: vmv.v.v v8, v24
; ZVFHMIN-NEXT: csrr a0, vlenb
@@ -682,17 +682,17 @@ define <vscale x 16 x float> @vfnmsac_vf_nxv16f32_commute(<vscale x 16 x half> %
;
; ZVFHMIN-LABEL: vfnmsac_vf_nxv16f32_commute:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: vmv4r.v v24, v8
; ZVFHMIN-NEXT: fcvt.s.h fa5, fa0
; ZVFHMIN-NEXT: vsetvli a1, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfmv.v.f v8, fa5
+; ZVFHMIN-NEXT: vfmv.v.f v24, fa5
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v4, v8
+; ZVFHMIN-NEXT: vfncvt.f.f.w v4, v24
; ZVFHMIN-NEXT: vsetvli zero, a0, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v24, v0.t
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v4, v0.t
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v8, v0.t
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v4, v0.t
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfnmsub.vv v8, v24, v16, v0.t
+; ZVFHMIN-NEXT: vfnmsub.vv v24, v8, v16, v0.t
+; ZVFHMIN-NEXT: vmv.v.v v8, v24
; ZVFHMIN-NEXT: ret
%elt.head = insertelement <vscale x 16 x half> poison, half %b, i32 0
%vb = shufflevector <vscale x 16 x half> %elt.head, <vscale x 16 x half> poison, <vscale x 16 x i32> zeroinitializer
diff --git a/llvm/test/CodeGen/RISCV/rvv/vitofp-sdnode.ll b/llvm/test/CodeGen/RISCV/rvv/vitofp-sdnode.ll
index 5d0172430d15c..77ef0a340270f 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vitofp-sdnode.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vitofp-sdnode.ll
@@ -449,12 +449,12 @@ define <vscale x 32 x half> @vsitofp_nxv32i1_nxv32f16(<vscale x 32 x i1> %va) {
; ZVFHMIN-NEXT: vmv.v.i v12, 0
; ZVFHMIN-NEXT: vmerge.vim v8, v12, -1, v0
; ZVFHMIN-NEXT: vfwcvt.f.x.v v16, v8
-; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v16
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: srli a0, a0, 2
; ZVFHMIN-NEXT: vsetvli a1, zero, e8, mf2, ta, ma
; ZVFHMIN-NEXT: vslidedown.vx v0, v0, a0
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
+; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v16
; ZVFHMIN-NEXT: vmerge.vim v12, v12, -1, v0
; ZVFHMIN-NEXT: vfwcvt.f.x.v v16, v12
; ZVFHMIN-NEXT: vfncvt.f.f.w v12, v16
@@ -478,12 +478,12 @@ define <vscale x 32 x half> @vuitofp_nxv32i1_nxv32f16(<vscale x 32 x i1> %va) {
; ZVFHMIN-NEXT: vmv.v.i v12, 0
; ZVFHMIN-NEXT: vmerge.vim v8, v12, 1, v0
; ZVFHMIN-NEXT: vfwcvt.f.xu.v v16, v8
-; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v16
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: srli a0, a0, 2
; ZVFHMIN-NEXT: vsetvli a1, zero, e8, mf2, ta, ma
; ZVFHMIN-NEXT: vslidedown.vx v0, v0, a0
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
+; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v16
; ZVFHMIN-NEXT: vmerge.vim v12, v12, 1, v0
; ZVFHMIN-NEXT: vfwcvt.f.xu.v v16, v12
; ZVFHMIN-NEXT: vfncvt.f.f.w v12, v16
diff --git a/llvm/test/CodeGen/RISCV/rvv/vmax-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vmax-vp.ll
index a35fc874065a7..8a76467986620 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vmax-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vmax-vp.ll
@@ -423,8 +423,8 @@ define <vscale x 128 x i8> @vmax_vx_nxv128i8(<vscale x 128 x i8> %va, i8 %b, <vs
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a2, a1
; CHECK-NEXT: .LBB34_2:
-; CHECK-NEXT: vsetvli zero, a2, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a2, e8, m8, ta, ma
; CHECK-NEXT: vmax.vx v8, v8, a0, v0.t
; CHECK-NEXT: ret
%elt.head = insertelement <vscale x 128 x i8> poison, i8 %b, i32 0
@@ -986,8 +986,8 @@ define <vscale x 32 x i32> @vmax_vx_nxv32i32(<vscale x 32 x i32> %va, i32 %b, <v
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a1, a2
; CHECK-NEXT: .LBB80_2:
-; CHECK-NEXT: vsetvli zero, a1, e32, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a1, e32, m8, ta, ma
; CHECK-NEXT: vmax.vx v8, v8, a0, v0.t
; CHECK-NEXT: ret
%elt.head = insertelement <vscale x 32 x i32> poison, i32 %b, i32 0
@@ -1046,8 +1046,8 @@ define <vscale x 32 x i32> @vmax_vx_nxv32i32_evl_nx8(<vscale x 32 x i32> %va, i3
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a1, a2
; CHECK-NEXT: .LBB82_2:
-; CHECK-NEXT: vsetvli zero, a1, e32, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a1, e32, m8, ta, ma
; CHECK-NEXT: vmax.vx v8, v8, a0, v0.t
; CHECK-NEXT: ret
%elt.head = insertelement <vscale x 32 x i32> poison, i32 %b, i32 0
@@ -1084,8 +1084,8 @@ define <vscale x 32 x i32> @vmax_vx_nxv32i32_evl_nx16(<vscale x 32 x i32> %va, i
; RV64-NEXT: slli a1, a1, 1
; RV64-NEXT: vsetvli zero, a1, e32, m8, ta, ma
; RV64-NEXT: vmax.vx v8, v8, a0, v0.t
-; RV64-NEXT: vsetivli zero, 0, e32, m8, ta, ma
; RV64-NEXT: vmv1r.v v0, v24
+; RV64-NEXT: vsetivli zero, 0, e32, m8, ta, ma
; RV64-NEXT: vmax.vx v16, v16, a0, v0.t
; RV64-NEXT: ret
%elt.head = insertelement <vscale x 32 x i32> poison, i32 %b, i32 0
diff --git a/llvm/test/CodeGen/RISCV/rvv/vmaxu-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vmaxu-vp.ll
index 1f620a44dbbc8..1c74887c1b20f 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vmaxu-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vmaxu-vp.ll
@@ -425,8 +425,8 @@ define <vscale x 128 x i8> @vmaxu_vx_nxv128i8(<vscale x 128 x i8> %va, i8 %b, <v
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a2, a1
; CHECK-NEXT: .LBB34_2:
-; CHECK-NEXT: vsetvli zero, a2, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a2, e8, m8, ta, ma
; CHECK-NEXT: vmaxu.vx v8, v8, a0, v0.t
; CHECK-NEXT: ret
%elt.head = insertelement <vscale x 128 x i8> poison, i8 %b, i32 0
@@ -988,8 +988,8 @@ define <vscale x 32 x i32> @vmaxu_vx_nxv32i32(<vscale x 32 x i32> %va, i32 %b, <
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a1, a2
; CHECK-NEXT: .LBB80_2:
-; CHECK-NEXT: vsetvli zero, a1, e32, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a1, e32, m8, ta, ma
; CHECK-NEXT: vmaxu.vx v8, v8, a0, v0.t
; CHECK-NEXT: ret
%elt.head = insertelement <vscale x 32 x i32> poison, i32 %b, i32 0
@@ -1048,8 +1048,8 @@ define <vscale x 32 x i32> @vmaxu_vx_nxv32i32_evl_nx8(<vscale x 32 x i32> %va, i
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a1, a2
; CHECK-NEXT: .LBB82_2:
-; CHECK-NEXT: vsetvli zero, a1, e32, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a1, e32, m8, ta, ma
; CHECK-NEXT: vmaxu.vx v8, v8, a0, v0.t
; CHECK-NEXT: ret
%elt.head = insertelement <vscale x 32 x i32> poison, i32 %b, i32 0
@@ -1086,8 +1086,8 @@ define <vscale x 32 x i32> @vmaxu_vx_nxv32i32_evl_nx16(<vscale x 32 x i32> %va,
; RV64-NEXT: slli a1, a1, 1
; RV64-NEXT: vsetvli zero, a1, e32, m8, ta, ma
; RV64-NEXT: vmaxu.vx v8, v8, a0, v0.t
-; RV64-NEXT: vsetivli zero, 0, e32, m8, ta, ma
; RV64-NEXT: vmv1r.v v0, v24
+; RV64-NEXT: vsetivli zero, 0, e32, m8, ta, ma
; RV64-NEXT: vmaxu.vx v16, v16, a0, v0.t
; RV64-NEXT: ret
%elt.head = insertelement <vscale x 32 x i32> poison, i32 %b, i32 0
diff --git a/llvm/test/CodeGen/RISCV/rvv/vmfeq.ll b/llvm/test/CodeGen/RISCV/rvv/vmfeq.ll
index e7184921d87a0..2e5b67c93fce1 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vmfeq.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vmfeq.ll
@@ -34,9 +34,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmfeq.mask.nxv1f16(
define <vscale x 1 x i1> @intrinsic_vmfeq_mask_vv_nxv1f16_nxv1f16(<vscale x 1 x i1> %0, <vscale x 1 x half> %1, <vscale x 1 x half> %2, <vscale x 1 x half> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfeq_mask_vv_nxv1f16_nxv1f16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
-; CHECK-NEXT: vmfeq.vv v0, v8, v9
+; CHECK-NEXT: vmfeq.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmfeq.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -85,9 +86,10 @@ declare <vscale x 2 x i1> @llvm.riscv.vmfeq.mask.nxv2f16(
define <vscale x 2 x i1> @intrinsic_vmfeq_mask_vv_nxv2f16_nxv2f16(<vscale x 2 x i1> %0, <vscale x 2 x half> %1, <vscale x 2 x half> %2, <vscale x 2 x half> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfeq_mask_vv_nxv2f16_nxv2f16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
-; CHECK-NEXT: vmfeq.vv v0, v8, v9
+; CHECK-NEXT: vmfeq.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmfeq.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -136,9 +138,10 @@ declare <vscale x 4 x i1> @llvm.riscv.vmfeq.mask.nxv4f16(
define <vscale x 4 x i1> @intrinsic_vmfeq_mask_vv_nxv4f16_nxv4f16(<vscale x 4 x i1> %0, <vscale x 4 x half> %1, <vscale x 4 x half> %2, <vscale x 4 x half> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfeq_mask_vv_nxv4f16_nxv4f16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
-; CHECK-NEXT: vmfeq.vv v0, v8, v9
+; CHECK-NEXT: vmfeq.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmfeq.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -187,11 +190,12 @@ declare <vscale x 8 x i1> @llvm.riscv.vmfeq.mask.nxv8f16(
define <vscale x 8 x i1> @intrinsic_vmfeq_mask_vv_nxv8f16_nxv8f16(<vscale x 8 x i1> %0, <vscale x 8 x half> %1, <vscale x 8 x half> %2, <vscale x 8 x half> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfeq_mask_vv_nxv8f16_nxv8f16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
-; CHECK-NEXT: vmfeq.vv v0, v8, v10
-; CHECK-NEXT: vmfeq.vv v14, v10, v12, v0.t
+; CHECK-NEXT: vmfeq.vv v14, v8, v10
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmfeq.vv v8, v10, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 8 x i1> @llvm.riscv.vmfeq.nxv8f16(
@@ -238,11 +242,12 @@ declare <vscale x 16 x i1> @llvm.riscv.vmfeq.mask.nxv16f16(
define <vscale x 16 x i1> @intrinsic_vmfeq_mask_vv_nxv16f16_nxv16f16(<vscale x 16 x i1> %0, <vscale x 16 x half> %1, <vscale x 16 x half> %2, <vscale x 16 x half> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfeq_mask_vv_nxv16f16_nxv16f16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
-; CHECK-NEXT: vmfeq.vv v0, v8, v12
-; CHECK-NEXT: vmfeq.vv v20, v12, v16, v0.t
+; CHECK-NEXT: vmfeq.vv v20, v8, v12
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmfeq.vv v8, v12, v16, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 16 x i1> @llvm.riscv.vmfeq.nxv16f16(
@@ -289,9 +294,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmfeq.mask.nxv1f32(
define <vscale x 1 x i1> @intrinsic_vmfeq_mask_vv_nxv1f32_nxv1f32(<vscale x 1 x i1> %0, <vscale x 1 x float> %1, <vscale x 1 x float> %2, <vscale x 1 x float> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfeq_mask_vv_nxv1f32_nxv1f32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
-; CHECK-NEXT: vmfeq.vv v0, v8, v9
+; CHECK-NEXT: vmfeq.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmfeq.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -340,9 +346,10 @@ declare <vscale x 2 x i1> @llvm.riscv.vmfeq.mask.nxv2f32(
define <vscale x 2 x i1> @intrinsic_vmfeq_mask_vv_nxv2f32_nxv2f32(<vscale x 2 x i1> %0, <vscale x 2 x float> %1, <vscale x 2 x float> %2, <vscale x 2 x float> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfeq_mask_vv_nxv2f32_nxv2f32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
-; CHECK-NEXT: vmfeq.vv v0, v8, v9
+; CHECK-NEXT: vmfeq.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmfeq.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -391,11 +398,12 @@ declare <vscale x 4 x i1> @llvm.riscv.vmfeq.mask.nxv4f32(
define <vscale x 4 x i1> @intrinsic_vmfeq_mask_vv_nxv4f32_nxv4f32(<vscale x 4 x i1> %0, <vscale x 4 x float> %1, <vscale x 4 x float> %2, <vscale x 4 x float> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfeq_mask_vv_nxv4f32_nxv4f32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
-; CHECK-NEXT: vmfeq.vv v0, v8, v10
-; CHECK-NEXT: vmfeq.vv v14, v10, v12, v0.t
+; CHECK-NEXT: vmfeq.vv v14, v8, v10
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmfeq.vv v8, v10, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 4 x i1> @llvm.riscv.vmfeq.nxv4f32(
@@ -442,11 +450,12 @@ declare <vscale x 8 x i1> @llvm.riscv.vmfeq.mask.nxv8f32(
define <vscale x 8 x i1> @intrinsic_vmfeq_mask_vv_nxv8f32_nxv8f32(<vscale x 8 x i1> %0, <vscale x 8 x float> %1, <vscale x 8 x float> %2, <vscale x 8 x float> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfeq_mask_vv_nxv8f32_nxv8f32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
-; CHECK-NEXT: vmfeq.vv v0, v8, v12
-; CHECK-NEXT: vmfeq.vv v20, v12, v16, v0.t
+; CHECK-NEXT: vmfeq.vv v20, v8, v12
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmfeq.vv v8, v12, v16, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 8 x i1> @llvm.riscv.vmfeq.nxv8f32(
@@ -493,9 +502,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmfeq.mask.nxv1f64(
define <vscale x 1 x i1> @intrinsic_vmfeq_mask_vv_nxv1f64_nxv1f64(<vscale x 1 x i1> %0, <vscale x 1 x double> %1, <vscale x 1 x double> %2, <vscale x 1 x double> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfeq_mask_vv_nxv1f64_nxv1f64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
-; CHECK-NEXT: vmfeq.vv v0, v8, v9
+; CHECK-NEXT: vmfeq.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmfeq.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -544,11 +554,12 @@ declare <vscale x 2 x i1> @llvm.riscv.vmfeq.mask.nxv2f64(
define <vscale x 2 x i1> @intrinsic_vmfeq_mask_vv_nxv2f64_nxv2f64(<vscale x 2 x i1> %0, <vscale x 2 x double> %1, <vscale x 2 x double> %2, <vscale x 2 x double> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfeq_mask_vv_nxv2f64_nxv2f64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
-; CHECK-NEXT: vmfeq.vv v0, v8, v10
-; CHECK-NEXT: vmfeq.vv v14, v10, v12, v0.t
+; CHECK-NEXT: vmfeq.vv v14, v8, v10
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmfeq.vv v8, v10, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 2 x i1> @llvm.riscv.vmfeq.nxv2f64(
@@ -595,11 +606,12 @@ declare <vscale x 4 x i1> @llvm.riscv.vmfeq.mask.nxv4f64(
define <vscale x 4 x i1> @intrinsic_vmfeq_mask_vv_nxv4f64_nxv4f64(<vscale x 4 x i1> %0, <vscale x 4 x double> %1, <vscale x 4 x double> %2, <vscale x 4 x double> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfeq_mask_vv_nxv4f64_nxv4f64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
-; CHECK-NEXT: vmfeq.vv v0, v8, v12
-; CHECK-NEXT: vmfeq.vv v20, v12, v16, v0.t
+; CHECK-NEXT: vmfeq.vv v20, v8, v12
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmfeq.vv v8, v12, v16, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 4 x i1> @llvm.riscv.vmfeq.nxv4f64(
@@ -647,8 +659,8 @@ define <vscale x 1 x i1> @intrinsic_vmfeq_mask_vf_nxv1f16_f16(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmfeq_mask_vf_nxv1f16_f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
; CHECK-NEXT: vmfeq.vf v10, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -694,8 +706,8 @@ define <vscale x 2 x i1> @intrinsic_vmfeq_mask_vf_nxv2f16_f16(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmfeq_mask_vf_nxv2f16_f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
; CHECK-NEXT: vmfeq.vf v10, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -741,8 +753,8 @@ define <vscale x 4 x i1> @intrinsic_vmfeq_mask_vf_nxv4f16_f16(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmfeq_mask_vf_nxv4f16_f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
; CHECK-NEXT: vmfeq.vf v10, v8, fa0, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -788,8 +800,8 @@ define <vscale x 8 x i1> @intrinsic_vmfeq_mask_vf_nxv8f16_f16(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmfeq_mask_vf_nxv8f16_f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
; CHECK-NEXT: vmfeq.vf v11, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -835,8 +847,8 @@ define <vscale x 16 x i1> @intrinsic_vmfeq_mask_vf_nxv16f16_f16(<vscale x 16 x i
; CHECK-LABEL: intrinsic_vmfeq_mask_vf_nxv16f16_f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
; CHECK-NEXT: vmfeq.vf v13, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -882,8 +894,8 @@ define <vscale x 1 x i1> @intrinsic_vmfeq_mask_vf_nxv1f32_f32(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmfeq_mask_vf_nxv1f32_f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
; CHECK-NEXT: vmfeq.vf v10, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -929,8 +941,8 @@ define <vscale x 2 x i1> @intrinsic_vmfeq_mask_vf_nxv2f32_f32(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmfeq_mask_vf_nxv2f32_f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
; CHECK-NEXT: vmfeq.vf v10, v8, fa0, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -976,8 +988,8 @@ define <vscale x 4 x i1> @intrinsic_vmfeq_mask_vf_nxv4f32_f32(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmfeq_mask_vf_nxv4f32_f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
; CHECK-NEXT: vmfeq.vf v11, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -1023,8 +1035,8 @@ define <vscale x 8 x i1> @intrinsic_vmfeq_mask_vf_nxv8f32_f32(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmfeq_mask_vf_nxv8f32_f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
; CHECK-NEXT: vmfeq.vf v13, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -1070,8 +1082,8 @@ define <vscale x 1 x i1> @intrinsic_vmfeq_mask_vf_nxv1f64_f64(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmfeq_mask_vf_nxv1f64_f64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
; CHECK-NEXT: vmfeq.vf v10, v8, fa0, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -1117,8 +1129,8 @@ define <vscale x 2 x i1> @intrinsic_vmfeq_mask_vf_nxv2f64_f64(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmfeq_mask_vf_nxv2f64_f64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
; CHECK-NEXT: vmfeq.vf v11, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -1164,8 +1176,8 @@ define <vscale x 4 x i1> @intrinsic_vmfeq_mask_vf_nxv4f64_f64(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmfeq_mask_vf_nxv4f64_f64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
; CHECK-NEXT: vmfeq.vf v13, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vmfge.ll b/llvm/test/CodeGen/RISCV/rvv/vmfge.ll
index a6dad9eaa4f35..b5ca47707c8a8 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vmfge.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vmfge.ll
@@ -34,9 +34,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmfge.mask.nxv1f16(
define <vscale x 1 x i1> @intrinsic_vmfge_mask_vv_nxv1f16_nxv1f16(<vscale x 1 x i1> %0, <vscale x 1 x half> %1, <vscale x 1 x half> %2, <vscale x 1 x half> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfge_mask_vv_nxv1f16_nxv1f16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
-; CHECK-NEXT: vmfle.vv v0, v9, v8
+; CHECK-NEXT: vmfle.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmfle.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -85,9 +86,10 @@ declare <vscale x 2 x i1> @llvm.riscv.vmfge.mask.nxv2f16(
define <vscale x 2 x i1> @intrinsic_vmfge_mask_vv_nxv2f16_nxv2f16(<vscale x 2 x i1> %0, <vscale x 2 x half> %1, <vscale x 2 x half> %2, <vscale x 2 x half> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfge_mask_vv_nxv2f16_nxv2f16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
-; CHECK-NEXT: vmfle.vv v0, v9, v8
+; CHECK-NEXT: vmfle.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmfle.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -136,9 +138,10 @@ declare <vscale x 4 x i1> @llvm.riscv.vmfge.mask.nxv4f16(
define <vscale x 4 x i1> @intrinsic_vmfge_mask_vv_nxv4f16_nxv4f16(<vscale x 4 x i1> %0, <vscale x 4 x half> %1, <vscale x 4 x half> %2, <vscale x 4 x half> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfge_mask_vv_nxv4f16_nxv4f16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
-; CHECK-NEXT: vmfle.vv v0, v9, v8
+; CHECK-NEXT: vmfle.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmfle.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -187,11 +190,12 @@ declare <vscale x 8 x i1> @llvm.riscv.vmfge.mask.nxv8f16(
define <vscale x 8 x i1> @intrinsic_vmfge_mask_vv_nxv8f16_nxv8f16(<vscale x 8 x i1> %0, <vscale x 8 x half> %1, <vscale x 8 x half> %2, <vscale x 8 x half> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfge_mask_vv_nxv8f16_nxv8f16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
-; CHECK-NEXT: vmfle.vv v0, v10, v8
-; CHECK-NEXT: vmfle.vv v14, v12, v10, v0.t
+; CHECK-NEXT: vmfle.vv v14, v10, v8
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmfle.vv v8, v12, v10, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 8 x i1> @llvm.riscv.vmfge.nxv8f16(
@@ -238,11 +242,12 @@ declare <vscale x 16 x i1> @llvm.riscv.vmfge.mask.nxv16f16(
define <vscale x 16 x i1> @intrinsic_vmfge_mask_vv_nxv16f16_nxv16f16(<vscale x 16 x i1> %0, <vscale x 16 x half> %1, <vscale x 16 x half> %2, <vscale x 16 x half> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfge_mask_vv_nxv16f16_nxv16f16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
-; CHECK-NEXT: vmfle.vv v0, v12, v8
-; CHECK-NEXT: vmfle.vv v20, v16, v12, v0.t
+; CHECK-NEXT: vmfle.vv v20, v12, v8
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmfle.vv v8, v16, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 16 x i1> @llvm.riscv.vmfge.nxv16f16(
@@ -289,9 +294,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmfge.mask.nxv1f32(
define <vscale x 1 x i1> @intrinsic_vmfge_mask_vv_nxv1f32_nxv1f32(<vscale x 1 x i1> %0, <vscale x 1 x float> %1, <vscale x 1 x float> %2, <vscale x 1 x float> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfge_mask_vv_nxv1f32_nxv1f32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
-; CHECK-NEXT: vmfle.vv v0, v9, v8
+; CHECK-NEXT: vmfle.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmfle.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -340,9 +346,10 @@ declare <vscale x 2 x i1> @llvm.riscv.vmfge.mask.nxv2f32(
define <vscale x 2 x i1> @intrinsic_vmfge_mask_vv_nxv2f32_nxv2f32(<vscale x 2 x i1> %0, <vscale x 2 x float> %1, <vscale x 2 x float> %2, <vscale x 2 x float> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfge_mask_vv_nxv2f32_nxv2f32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
-; CHECK-NEXT: vmfle.vv v0, v9, v8
+; CHECK-NEXT: vmfle.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmfle.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -391,11 +398,12 @@ declare <vscale x 4 x i1> @llvm.riscv.vmfge.mask.nxv4f32(
define <vscale x 4 x i1> @intrinsic_vmfge_mask_vv_nxv4f32_nxv4f32(<vscale x 4 x i1> %0, <vscale x 4 x float> %1, <vscale x 4 x float> %2, <vscale x 4 x float> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfge_mask_vv_nxv4f32_nxv4f32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
-; CHECK-NEXT: vmfle.vv v0, v10, v8
-; CHECK-NEXT: vmfle.vv v14, v12, v10, v0.t
+; CHECK-NEXT: vmfle.vv v14, v10, v8
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmfle.vv v8, v12, v10, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 4 x i1> @llvm.riscv.vmfge.nxv4f32(
@@ -442,11 +450,12 @@ declare <vscale x 8 x i1> @llvm.riscv.vmfge.mask.nxv8f32(
define <vscale x 8 x i1> @intrinsic_vmfge_mask_vv_nxv8f32_nxv8f32(<vscale x 8 x i1> %0, <vscale x 8 x float> %1, <vscale x 8 x float> %2, <vscale x 8 x float> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfge_mask_vv_nxv8f32_nxv8f32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
-; CHECK-NEXT: vmfle.vv v0, v12, v8
-; CHECK-NEXT: vmfle.vv v20, v16, v12, v0.t
+; CHECK-NEXT: vmfle.vv v20, v12, v8
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmfle.vv v8, v16, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 8 x i1> @llvm.riscv.vmfge.nxv8f32(
@@ -493,9 +502,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmfge.mask.nxv1f64(
define <vscale x 1 x i1> @intrinsic_vmfge_mask_vv_nxv1f64_nxv1f64(<vscale x 1 x i1> %0, <vscale x 1 x double> %1, <vscale x 1 x double> %2, <vscale x 1 x double> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfge_mask_vv_nxv1f64_nxv1f64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
-; CHECK-NEXT: vmfle.vv v0, v9, v8
+; CHECK-NEXT: vmfle.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmfle.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -544,11 +554,12 @@ declare <vscale x 2 x i1> @llvm.riscv.vmfge.mask.nxv2f64(
define <vscale x 2 x i1> @intrinsic_vmfge_mask_vv_nxv2f64_nxv2f64(<vscale x 2 x i1> %0, <vscale x 2 x double> %1, <vscale x 2 x double> %2, <vscale x 2 x double> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfge_mask_vv_nxv2f64_nxv2f64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
-; CHECK-NEXT: vmfle.vv v0, v10, v8
-; CHECK-NEXT: vmfle.vv v14, v12, v10, v0.t
+; CHECK-NEXT: vmfle.vv v14, v10, v8
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmfle.vv v8, v12, v10, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 2 x i1> @llvm.riscv.vmfge.nxv2f64(
@@ -595,11 +606,12 @@ declare <vscale x 4 x i1> @llvm.riscv.vmfge.mask.nxv4f64(
define <vscale x 4 x i1> @intrinsic_vmfge_mask_vv_nxv4f64_nxv4f64(<vscale x 4 x i1> %0, <vscale x 4 x double> %1, <vscale x 4 x double> %2, <vscale x 4 x double> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfge_mask_vv_nxv4f64_nxv4f64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
-; CHECK-NEXT: vmfle.vv v0, v12, v8
-; CHECK-NEXT: vmfle.vv v20, v16, v12, v0.t
+; CHECK-NEXT: vmfle.vv v20, v12, v8
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmfle.vv v8, v16, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 4 x i1> @llvm.riscv.vmfge.nxv4f64(
@@ -647,8 +659,8 @@ define <vscale x 1 x i1> @intrinsic_vmfge_mask_vf_nxv1f16_f16(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmfge_mask_vf_nxv1f16_f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
; CHECK-NEXT: vmfge.vf v10, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -694,8 +706,8 @@ define <vscale x 2 x i1> @intrinsic_vmfge_mask_vf_nxv2f16_f16(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmfge_mask_vf_nxv2f16_f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
; CHECK-NEXT: vmfge.vf v10, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -741,8 +753,8 @@ define <vscale x 4 x i1> @intrinsic_vmfge_mask_vf_nxv4f16_f16(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmfge_mask_vf_nxv4f16_f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
; CHECK-NEXT: vmfge.vf v10, v8, fa0, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -788,8 +800,8 @@ define <vscale x 8 x i1> @intrinsic_vmfge_mask_vf_nxv8f16_f16(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmfge_mask_vf_nxv8f16_f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
; CHECK-NEXT: vmfge.vf v11, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -835,8 +847,8 @@ define <vscale x 16 x i1> @intrinsic_vmfge_mask_vf_nxv16f16_f16(<vscale x 16 x i
; CHECK-LABEL: intrinsic_vmfge_mask_vf_nxv16f16_f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
; CHECK-NEXT: vmfge.vf v13, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -882,8 +894,8 @@ define <vscale x 1 x i1> @intrinsic_vmfge_mask_vf_nxv1f32_f32(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmfge_mask_vf_nxv1f32_f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
; CHECK-NEXT: vmfge.vf v10, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -929,8 +941,8 @@ define <vscale x 2 x i1> @intrinsic_vmfge_mask_vf_nxv2f32_f32(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmfge_mask_vf_nxv2f32_f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
; CHECK-NEXT: vmfge.vf v10, v8, fa0, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -976,8 +988,8 @@ define <vscale x 4 x i1> @intrinsic_vmfge_mask_vf_nxv4f32_f32(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmfge_mask_vf_nxv4f32_f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
; CHECK-NEXT: vmfge.vf v11, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -1023,8 +1035,8 @@ define <vscale x 8 x i1> @intrinsic_vmfge_mask_vf_nxv8f32_f32(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmfge_mask_vf_nxv8f32_f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
; CHECK-NEXT: vmfge.vf v13, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -1070,8 +1082,8 @@ define <vscale x 1 x i1> @intrinsic_vmfge_mask_vf_nxv1f64_f64(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmfge_mask_vf_nxv1f64_f64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
; CHECK-NEXT: vmfge.vf v10, v8, fa0, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -1117,8 +1129,8 @@ define <vscale x 2 x i1> @intrinsic_vmfge_mask_vf_nxv2f64_f64(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmfge_mask_vf_nxv2f64_f64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
; CHECK-NEXT: vmfge.vf v11, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -1164,8 +1176,8 @@ define <vscale x 4 x i1> @intrinsic_vmfge_mask_vf_nxv4f64_f64(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmfge_mask_vf_nxv4f64_f64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
; CHECK-NEXT: vmfge.vf v13, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vmfgt.ll b/llvm/test/CodeGen/RISCV/rvv/vmfgt.ll
index f643a4036381c..971249d38d1b2 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vmfgt.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vmfgt.ll
@@ -34,9 +34,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmfgt.mask.nxv1f16(
define <vscale x 1 x i1> @intrinsic_vmfgt_mask_vv_nxv1f16_nxv1f16(<vscale x 1 x i1> %0, <vscale x 1 x half> %1, <vscale x 1 x half> %2, <vscale x 1 x half> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfgt_mask_vv_nxv1f16_nxv1f16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
-; CHECK-NEXT: vmflt.vv v0, v9, v8
+; CHECK-NEXT: vmflt.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmflt.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -85,9 +86,10 @@ declare <vscale x 2 x i1> @llvm.riscv.vmfgt.mask.nxv2f16(
define <vscale x 2 x i1> @intrinsic_vmfgt_mask_vv_nxv2f16_nxv2f16(<vscale x 2 x i1> %0, <vscale x 2 x half> %1, <vscale x 2 x half> %2, <vscale x 2 x half> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfgt_mask_vv_nxv2f16_nxv2f16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
-; CHECK-NEXT: vmflt.vv v0, v9, v8
+; CHECK-NEXT: vmflt.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmflt.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -136,9 +138,10 @@ declare <vscale x 4 x i1> @llvm.riscv.vmfgt.mask.nxv4f16(
define <vscale x 4 x i1> @intrinsic_vmfgt_mask_vv_nxv4f16_nxv4f16(<vscale x 4 x i1> %0, <vscale x 4 x half> %1, <vscale x 4 x half> %2, <vscale x 4 x half> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfgt_mask_vv_nxv4f16_nxv4f16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
-; CHECK-NEXT: vmflt.vv v0, v9, v8
+; CHECK-NEXT: vmflt.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmflt.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -187,11 +190,12 @@ declare <vscale x 8 x i1> @llvm.riscv.vmfgt.mask.nxv8f16(
define <vscale x 8 x i1> @intrinsic_vmfgt_mask_vv_nxv8f16_nxv8f16(<vscale x 8 x i1> %0, <vscale x 8 x half> %1, <vscale x 8 x half> %2, <vscale x 8 x half> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfgt_mask_vv_nxv8f16_nxv8f16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
-; CHECK-NEXT: vmflt.vv v0, v10, v8
-; CHECK-NEXT: vmflt.vv v14, v12, v10, v0.t
+; CHECK-NEXT: vmflt.vv v14, v10, v8
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmflt.vv v8, v12, v10, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 8 x i1> @llvm.riscv.vmfgt.nxv8f16(
@@ -238,11 +242,12 @@ declare <vscale x 16 x i1> @llvm.riscv.vmfgt.mask.nxv16f16(
define <vscale x 16 x i1> @intrinsic_vmfgt_mask_vv_nxv16f16_nxv16f16(<vscale x 16 x i1> %0, <vscale x 16 x half> %1, <vscale x 16 x half> %2, <vscale x 16 x half> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfgt_mask_vv_nxv16f16_nxv16f16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
-; CHECK-NEXT: vmflt.vv v0, v12, v8
-; CHECK-NEXT: vmflt.vv v20, v16, v12, v0.t
+; CHECK-NEXT: vmflt.vv v20, v12, v8
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmflt.vv v8, v16, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 16 x i1> @llvm.riscv.vmfgt.nxv16f16(
@@ -289,9 +294,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmfgt.mask.nxv1f32(
define <vscale x 1 x i1> @intrinsic_vmfgt_mask_vv_nxv1f32_nxv1f32(<vscale x 1 x i1> %0, <vscale x 1 x float> %1, <vscale x 1 x float> %2, <vscale x 1 x float> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfgt_mask_vv_nxv1f32_nxv1f32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
-; CHECK-NEXT: vmflt.vv v0, v9, v8
+; CHECK-NEXT: vmflt.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmflt.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -340,9 +346,10 @@ declare <vscale x 2 x i1> @llvm.riscv.vmfgt.mask.nxv2f32(
define <vscale x 2 x i1> @intrinsic_vmfgt_mask_vv_nxv2f32_nxv2f32(<vscale x 2 x i1> %0, <vscale x 2 x float> %1, <vscale x 2 x float> %2, <vscale x 2 x float> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfgt_mask_vv_nxv2f32_nxv2f32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
-; CHECK-NEXT: vmflt.vv v0, v9, v8
+; CHECK-NEXT: vmflt.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmflt.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -391,11 +398,12 @@ declare <vscale x 4 x i1> @llvm.riscv.vmfgt.mask.nxv4f32(
define <vscale x 4 x i1> @intrinsic_vmfgt_mask_vv_nxv4f32_nxv4f32(<vscale x 4 x i1> %0, <vscale x 4 x float> %1, <vscale x 4 x float> %2, <vscale x 4 x float> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfgt_mask_vv_nxv4f32_nxv4f32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
-; CHECK-NEXT: vmflt.vv v0, v10, v8
-; CHECK-NEXT: vmflt.vv v14, v12, v10, v0.t
+; CHECK-NEXT: vmflt.vv v14, v10, v8
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmflt.vv v8, v12, v10, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 4 x i1> @llvm.riscv.vmfgt.nxv4f32(
@@ -442,11 +450,12 @@ declare <vscale x 8 x i1> @llvm.riscv.vmfgt.mask.nxv8f32(
define <vscale x 8 x i1> @intrinsic_vmfgt_mask_vv_nxv8f32_nxv8f32(<vscale x 8 x i1> %0, <vscale x 8 x float> %1, <vscale x 8 x float> %2, <vscale x 8 x float> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfgt_mask_vv_nxv8f32_nxv8f32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
-; CHECK-NEXT: vmflt.vv v0, v12, v8
-; CHECK-NEXT: vmflt.vv v20, v16, v12, v0.t
+; CHECK-NEXT: vmflt.vv v20, v12, v8
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmflt.vv v8, v16, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 8 x i1> @llvm.riscv.vmfgt.nxv8f32(
@@ -493,9 +502,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmfgt.mask.nxv1f64(
define <vscale x 1 x i1> @intrinsic_vmfgt_mask_vv_nxv1f64_nxv1f64(<vscale x 1 x i1> %0, <vscale x 1 x double> %1, <vscale x 1 x double> %2, <vscale x 1 x double> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfgt_mask_vv_nxv1f64_nxv1f64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
-; CHECK-NEXT: vmflt.vv v0, v9, v8
+; CHECK-NEXT: vmflt.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmflt.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -544,11 +554,12 @@ declare <vscale x 2 x i1> @llvm.riscv.vmfgt.mask.nxv2f64(
define <vscale x 2 x i1> @intrinsic_vmfgt_mask_vv_nxv2f64_nxv2f64(<vscale x 2 x i1> %0, <vscale x 2 x double> %1, <vscale x 2 x double> %2, <vscale x 2 x double> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfgt_mask_vv_nxv2f64_nxv2f64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
-; CHECK-NEXT: vmflt.vv v0, v10, v8
-; CHECK-NEXT: vmflt.vv v14, v12, v10, v0.t
+; CHECK-NEXT: vmflt.vv v14, v10, v8
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmflt.vv v8, v12, v10, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 2 x i1> @llvm.riscv.vmfgt.nxv2f64(
@@ -595,11 +606,12 @@ declare <vscale x 4 x i1> @llvm.riscv.vmfgt.mask.nxv4f64(
define <vscale x 4 x i1> @intrinsic_vmfgt_mask_vv_nxv4f64_nxv4f64(<vscale x 4 x i1> %0, <vscale x 4 x double> %1, <vscale x 4 x double> %2, <vscale x 4 x double> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfgt_mask_vv_nxv4f64_nxv4f64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
-; CHECK-NEXT: vmflt.vv v0, v12, v8
-; CHECK-NEXT: vmflt.vv v20, v16, v12, v0.t
+; CHECK-NEXT: vmflt.vv v20, v12, v8
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmflt.vv v8, v16, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 4 x i1> @llvm.riscv.vmfgt.nxv4f64(
@@ -647,8 +659,8 @@ define <vscale x 1 x i1> @intrinsic_vmfgt_mask_vf_nxv1f16_f16(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmfgt_mask_vf_nxv1f16_f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
; CHECK-NEXT: vmfgt.vf v10, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -694,8 +706,8 @@ define <vscale x 2 x i1> @intrinsic_vmfgt_mask_vf_nxv2f16_f16(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmfgt_mask_vf_nxv2f16_f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
; CHECK-NEXT: vmfgt.vf v10, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -741,8 +753,8 @@ define <vscale x 4 x i1> @intrinsic_vmfgt_mask_vf_nxv4f16_f16(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmfgt_mask_vf_nxv4f16_f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
; CHECK-NEXT: vmfgt.vf v10, v8, fa0, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -788,8 +800,8 @@ define <vscale x 8 x i1> @intrinsic_vmfgt_mask_vf_nxv8f16_f16(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmfgt_mask_vf_nxv8f16_f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
; CHECK-NEXT: vmfgt.vf v11, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -835,8 +847,8 @@ define <vscale x 16 x i1> @intrinsic_vmfgt_mask_vf_nxv16f16_f16(<vscale x 16 x i
; CHECK-LABEL: intrinsic_vmfgt_mask_vf_nxv16f16_f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
; CHECK-NEXT: vmfgt.vf v13, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -882,8 +894,8 @@ define <vscale x 1 x i1> @intrinsic_vmfgt_mask_vf_nxv1f32_f32(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmfgt_mask_vf_nxv1f32_f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
; CHECK-NEXT: vmfgt.vf v10, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -929,8 +941,8 @@ define <vscale x 2 x i1> @intrinsic_vmfgt_mask_vf_nxv2f32_f32(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmfgt_mask_vf_nxv2f32_f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
; CHECK-NEXT: vmfgt.vf v10, v8, fa0, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -976,8 +988,8 @@ define <vscale x 4 x i1> @intrinsic_vmfgt_mask_vf_nxv4f32_f32(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmfgt_mask_vf_nxv4f32_f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
; CHECK-NEXT: vmfgt.vf v11, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -1023,8 +1035,8 @@ define <vscale x 8 x i1> @intrinsic_vmfgt_mask_vf_nxv8f32_f32(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmfgt_mask_vf_nxv8f32_f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
; CHECK-NEXT: vmfgt.vf v13, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -1070,8 +1082,8 @@ define <vscale x 1 x i1> @intrinsic_vmfgt_mask_vf_nxv1f64_f64(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmfgt_mask_vf_nxv1f64_f64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
; CHECK-NEXT: vmfgt.vf v10, v8, fa0, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -1117,8 +1129,8 @@ define <vscale x 2 x i1> @intrinsic_vmfgt_mask_vf_nxv2f64_f64(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmfgt_mask_vf_nxv2f64_f64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
; CHECK-NEXT: vmfgt.vf v11, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -1164,8 +1176,8 @@ define <vscale x 4 x i1> @intrinsic_vmfgt_mask_vf_nxv4f64_f64(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmfgt_mask_vf_nxv4f64_f64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
; CHECK-NEXT: vmfgt.vf v13, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vmfle.ll b/llvm/test/CodeGen/RISCV/rvv/vmfle.ll
index 6c52364c1fbd5..f19a181a365af 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vmfle.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vmfle.ll
@@ -34,9 +34,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmfle.mask.nxv1f16(
define <vscale x 1 x i1> @intrinsic_vmfle_mask_vv_nxv1f16_nxv1f16(<vscale x 1 x i1> %0, <vscale x 1 x half> %1, <vscale x 1 x half> %2, <vscale x 1 x half> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfle_mask_vv_nxv1f16_nxv1f16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
-; CHECK-NEXT: vmfle.vv v0, v8, v9
+; CHECK-NEXT: vmfle.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmfle.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -85,9 +86,10 @@ declare <vscale x 2 x i1> @llvm.riscv.vmfle.mask.nxv2f16(
define <vscale x 2 x i1> @intrinsic_vmfle_mask_vv_nxv2f16_nxv2f16(<vscale x 2 x i1> %0, <vscale x 2 x half> %1, <vscale x 2 x half> %2, <vscale x 2 x half> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfle_mask_vv_nxv2f16_nxv2f16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
-; CHECK-NEXT: vmfle.vv v0, v8, v9
+; CHECK-NEXT: vmfle.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmfle.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -136,9 +138,10 @@ declare <vscale x 4 x i1> @llvm.riscv.vmfle.mask.nxv4f16(
define <vscale x 4 x i1> @intrinsic_vmfle_mask_vv_nxv4f16_nxv4f16(<vscale x 4 x i1> %0, <vscale x 4 x half> %1, <vscale x 4 x half> %2, <vscale x 4 x half> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfle_mask_vv_nxv4f16_nxv4f16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
-; CHECK-NEXT: vmfle.vv v0, v8, v9
+; CHECK-NEXT: vmfle.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmfle.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -187,11 +190,12 @@ declare <vscale x 8 x i1> @llvm.riscv.vmfle.mask.nxv8f16(
define <vscale x 8 x i1> @intrinsic_vmfle_mask_vv_nxv8f16_nxv8f16(<vscale x 8 x i1> %0, <vscale x 8 x half> %1, <vscale x 8 x half> %2, <vscale x 8 x half> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfle_mask_vv_nxv8f16_nxv8f16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
-; CHECK-NEXT: vmfle.vv v0, v8, v10
-; CHECK-NEXT: vmfle.vv v14, v10, v12, v0.t
+; CHECK-NEXT: vmfle.vv v14, v8, v10
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmfle.vv v8, v10, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 8 x i1> @llvm.riscv.vmfle.nxv8f16(
@@ -238,11 +242,12 @@ declare <vscale x 16 x i1> @llvm.riscv.vmfle.mask.nxv16f16(
define <vscale x 16 x i1> @intrinsic_vmfle_mask_vv_nxv16f16_nxv16f16(<vscale x 16 x i1> %0, <vscale x 16 x half> %1, <vscale x 16 x half> %2, <vscale x 16 x half> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfle_mask_vv_nxv16f16_nxv16f16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
-; CHECK-NEXT: vmfle.vv v0, v8, v12
-; CHECK-NEXT: vmfle.vv v20, v12, v16, v0.t
+; CHECK-NEXT: vmfle.vv v20, v8, v12
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmfle.vv v8, v12, v16, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 16 x i1> @llvm.riscv.vmfle.nxv16f16(
@@ -289,9 +294,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmfle.mask.nxv1f32(
define <vscale x 1 x i1> @intrinsic_vmfle_mask_vv_nxv1f32_nxv1f32(<vscale x 1 x i1> %0, <vscale x 1 x float> %1, <vscale x 1 x float> %2, <vscale x 1 x float> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfle_mask_vv_nxv1f32_nxv1f32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
-; CHECK-NEXT: vmfle.vv v0, v8, v9
+; CHECK-NEXT: vmfle.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmfle.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -340,9 +346,10 @@ declare <vscale x 2 x i1> @llvm.riscv.vmfle.mask.nxv2f32(
define <vscale x 2 x i1> @intrinsic_vmfle_mask_vv_nxv2f32_nxv2f32(<vscale x 2 x i1> %0, <vscale x 2 x float> %1, <vscale x 2 x float> %2, <vscale x 2 x float> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfle_mask_vv_nxv2f32_nxv2f32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
-; CHECK-NEXT: vmfle.vv v0, v8, v9
+; CHECK-NEXT: vmfle.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmfle.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -391,11 +398,12 @@ declare <vscale x 4 x i1> @llvm.riscv.vmfle.mask.nxv4f32(
define <vscale x 4 x i1> @intrinsic_vmfle_mask_vv_nxv4f32_nxv4f32(<vscale x 4 x i1> %0, <vscale x 4 x float> %1, <vscale x 4 x float> %2, <vscale x 4 x float> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfle_mask_vv_nxv4f32_nxv4f32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
-; CHECK-NEXT: vmfle.vv v0, v8, v10
-; CHECK-NEXT: vmfle.vv v14, v10, v12, v0.t
+; CHECK-NEXT: vmfle.vv v14, v8, v10
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmfle.vv v8, v10, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 4 x i1> @llvm.riscv.vmfle.nxv4f32(
@@ -442,11 +450,12 @@ declare <vscale x 8 x i1> @llvm.riscv.vmfle.mask.nxv8f32(
define <vscale x 8 x i1> @intrinsic_vmfle_mask_vv_nxv8f32_nxv8f32(<vscale x 8 x i1> %0, <vscale x 8 x float> %1, <vscale x 8 x float> %2, <vscale x 8 x float> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfle_mask_vv_nxv8f32_nxv8f32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
-; CHECK-NEXT: vmfle.vv v0, v8, v12
-; CHECK-NEXT: vmfle.vv v20, v12, v16, v0.t
+; CHECK-NEXT: vmfle.vv v20, v8, v12
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmfle.vv v8, v12, v16, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 8 x i1> @llvm.riscv.vmfle.nxv8f32(
@@ -493,9 +502,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmfle.mask.nxv1f64(
define <vscale x 1 x i1> @intrinsic_vmfle_mask_vv_nxv1f64_nxv1f64(<vscale x 1 x i1> %0, <vscale x 1 x double> %1, <vscale x 1 x double> %2, <vscale x 1 x double> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfle_mask_vv_nxv1f64_nxv1f64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
-; CHECK-NEXT: vmfle.vv v0, v8, v9
+; CHECK-NEXT: vmfle.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmfle.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -544,11 +554,12 @@ declare <vscale x 2 x i1> @llvm.riscv.vmfle.mask.nxv2f64(
define <vscale x 2 x i1> @intrinsic_vmfle_mask_vv_nxv2f64_nxv2f64(<vscale x 2 x i1> %0, <vscale x 2 x double> %1, <vscale x 2 x double> %2, <vscale x 2 x double> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfle_mask_vv_nxv2f64_nxv2f64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
-; CHECK-NEXT: vmfle.vv v0, v8, v10
-; CHECK-NEXT: vmfle.vv v14, v10, v12, v0.t
+; CHECK-NEXT: vmfle.vv v14, v8, v10
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmfle.vv v8, v10, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 2 x i1> @llvm.riscv.vmfle.nxv2f64(
@@ -595,11 +606,12 @@ declare <vscale x 4 x i1> @llvm.riscv.vmfle.mask.nxv4f64(
define <vscale x 4 x i1> @intrinsic_vmfle_mask_vv_nxv4f64_nxv4f64(<vscale x 4 x i1> %0, <vscale x 4 x double> %1, <vscale x 4 x double> %2, <vscale x 4 x double> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfle_mask_vv_nxv4f64_nxv4f64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
-; CHECK-NEXT: vmfle.vv v0, v8, v12
-; CHECK-NEXT: vmfle.vv v20, v12, v16, v0.t
+; CHECK-NEXT: vmfle.vv v20, v8, v12
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmfle.vv v8, v12, v16, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 4 x i1> @llvm.riscv.vmfle.nxv4f64(
@@ -647,8 +659,8 @@ define <vscale x 1 x i1> @intrinsic_vmfle_mask_vf_nxv1f16_f16(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmfle_mask_vf_nxv1f16_f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
; CHECK-NEXT: vmfle.vf v10, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -694,8 +706,8 @@ define <vscale x 2 x i1> @intrinsic_vmfle_mask_vf_nxv2f16_f16(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmfle_mask_vf_nxv2f16_f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
; CHECK-NEXT: vmfle.vf v10, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -741,8 +753,8 @@ define <vscale x 4 x i1> @intrinsic_vmfle_mask_vf_nxv4f16_f16(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmfle_mask_vf_nxv4f16_f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
; CHECK-NEXT: vmfle.vf v10, v8, fa0, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -788,8 +800,8 @@ define <vscale x 8 x i1> @intrinsic_vmfle_mask_vf_nxv8f16_f16(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmfle_mask_vf_nxv8f16_f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
; CHECK-NEXT: vmfle.vf v11, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -835,8 +847,8 @@ define <vscale x 16 x i1> @intrinsic_vmfle_mask_vf_nxv16f16_f16(<vscale x 16 x i
; CHECK-LABEL: intrinsic_vmfle_mask_vf_nxv16f16_f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
; CHECK-NEXT: vmfle.vf v13, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -882,8 +894,8 @@ define <vscale x 1 x i1> @intrinsic_vmfle_mask_vf_nxv1f32_f32(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmfle_mask_vf_nxv1f32_f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
; CHECK-NEXT: vmfle.vf v10, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -929,8 +941,8 @@ define <vscale x 2 x i1> @intrinsic_vmfle_mask_vf_nxv2f32_f32(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmfle_mask_vf_nxv2f32_f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
; CHECK-NEXT: vmfle.vf v10, v8, fa0, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -976,8 +988,8 @@ define <vscale x 4 x i1> @intrinsic_vmfle_mask_vf_nxv4f32_f32(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmfle_mask_vf_nxv4f32_f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
; CHECK-NEXT: vmfle.vf v11, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -1023,8 +1035,8 @@ define <vscale x 8 x i1> @intrinsic_vmfle_mask_vf_nxv8f32_f32(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmfle_mask_vf_nxv8f32_f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
; CHECK-NEXT: vmfle.vf v13, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -1070,8 +1082,8 @@ define <vscale x 1 x i1> @intrinsic_vmfle_mask_vf_nxv1f64_f64(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmfle_mask_vf_nxv1f64_f64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
; CHECK-NEXT: vmfle.vf v10, v8, fa0, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -1117,8 +1129,8 @@ define <vscale x 2 x i1> @intrinsic_vmfle_mask_vf_nxv2f64_f64(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmfle_mask_vf_nxv2f64_f64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
; CHECK-NEXT: vmfle.vf v11, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -1164,8 +1176,8 @@ define <vscale x 4 x i1> @intrinsic_vmfle_mask_vf_nxv4f64_f64(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmfle_mask_vf_nxv4f64_f64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
; CHECK-NEXT: vmfle.vf v13, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vmflt.ll b/llvm/test/CodeGen/RISCV/rvv/vmflt.ll
index 37a9c6b081a1d..0a04642219334 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vmflt.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vmflt.ll
@@ -34,9 +34,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmflt.mask.nxv1f16(
define <vscale x 1 x i1> @intrinsic_vmflt_mask_vv_nxv1f16_nxv1f16(<vscale x 1 x i1> %0, <vscale x 1 x half> %1, <vscale x 1 x half> %2, <vscale x 1 x half> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmflt_mask_vv_nxv1f16_nxv1f16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
-; CHECK-NEXT: vmflt.vv v0, v8, v9
+; CHECK-NEXT: vmflt.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmflt.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -85,9 +86,10 @@ declare <vscale x 2 x i1> @llvm.riscv.vmflt.mask.nxv2f16(
define <vscale x 2 x i1> @intrinsic_vmflt_mask_vv_nxv2f16_nxv2f16(<vscale x 2 x i1> %0, <vscale x 2 x half> %1, <vscale x 2 x half> %2, <vscale x 2 x half> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmflt_mask_vv_nxv2f16_nxv2f16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
-; CHECK-NEXT: vmflt.vv v0, v8, v9
+; CHECK-NEXT: vmflt.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmflt.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -136,9 +138,10 @@ declare <vscale x 4 x i1> @llvm.riscv.vmflt.mask.nxv4f16(
define <vscale x 4 x i1> @intrinsic_vmflt_mask_vv_nxv4f16_nxv4f16(<vscale x 4 x i1> %0, <vscale x 4 x half> %1, <vscale x 4 x half> %2, <vscale x 4 x half> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmflt_mask_vv_nxv4f16_nxv4f16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
-; CHECK-NEXT: vmflt.vv v0, v8, v9
+; CHECK-NEXT: vmflt.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmflt.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -187,11 +190,12 @@ declare <vscale x 8 x i1> @llvm.riscv.vmflt.mask.nxv8f16(
define <vscale x 8 x i1> @intrinsic_vmflt_mask_vv_nxv8f16_nxv8f16(<vscale x 8 x i1> %0, <vscale x 8 x half> %1, <vscale x 8 x half> %2, <vscale x 8 x half> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmflt_mask_vv_nxv8f16_nxv8f16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
-; CHECK-NEXT: vmflt.vv v0, v8, v10
-; CHECK-NEXT: vmflt.vv v14, v10, v12, v0.t
+; CHECK-NEXT: vmflt.vv v14, v8, v10
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmflt.vv v8, v10, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 8 x i1> @llvm.riscv.vmflt.nxv8f16(
@@ -238,11 +242,12 @@ declare <vscale x 16 x i1> @llvm.riscv.vmflt.mask.nxv16f16(
define <vscale x 16 x i1> @intrinsic_vmflt_mask_vv_nxv16f16_nxv16f16(<vscale x 16 x i1> %0, <vscale x 16 x half> %1, <vscale x 16 x half> %2, <vscale x 16 x half> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmflt_mask_vv_nxv16f16_nxv16f16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
-; CHECK-NEXT: vmflt.vv v0, v8, v12
-; CHECK-NEXT: vmflt.vv v20, v12, v16, v0.t
+; CHECK-NEXT: vmflt.vv v20, v8, v12
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmflt.vv v8, v12, v16, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 16 x i1> @llvm.riscv.vmflt.nxv16f16(
@@ -289,9 +294,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmflt.mask.nxv1f32(
define <vscale x 1 x i1> @intrinsic_vmflt_mask_vv_nxv1f32_nxv1f32(<vscale x 1 x i1> %0, <vscale x 1 x float> %1, <vscale x 1 x float> %2, <vscale x 1 x float> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmflt_mask_vv_nxv1f32_nxv1f32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
-; CHECK-NEXT: vmflt.vv v0, v8, v9
+; CHECK-NEXT: vmflt.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmflt.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -340,9 +346,10 @@ declare <vscale x 2 x i1> @llvm.riscv.vmflt.mask.nxv2f32(
define <vscale x 2 x i1> @intrinsic_vmflt_mask_vv_nxv2f32_nxv2f32(<vscale x 2 x i1> %0, <vscale x 2 x float> %1, <vscale x 2 x float> %2, <vscale x 2 x float> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmflt_mask_vv_nxv2f32_nxv2f32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
-; CHECK-NEXT: vmflt.vv v0, v8, v9
+; CHECK-NEXT: vmflt.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmflt.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -391,11 +398,12 @@ declare <vscale x 4 x i1> @llvm.riscv.vmflt.mask.nxv4f32(
define <vscale x 4 x i1> @intrinsic_vmflt_mask_vv_nxv4f32_nxv4f32(<vscale x 4 x i1> %0, <vscale x 4 x float> %1, <vscale x 4 x float> %2, <vscale x 4 x float> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmflt_mask_vv_nxv4f32_nxv4f32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
-; CHECK-NEXT: vmflt.vv v0, v8, v10
-; CHECK-NEXT: vmflt.vv v14, v10, v12, v0.t
+; CHECK-NEXT: vmflt.vv v14, v8, v10
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmflt.vv v8, v10, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 4 x i1> @llvm.riscv.vmflt.nxv4f32(
@@ -442,11 +450,12 @@ declare <vscale x 8 x i1> @llvm.riscv.vmflt.mask.nxv8f32(
define <vscale x 8 x i1> @intrinsic_vmflt_mask_vv_nxv8f32_nxv8f32(<vscale x 8 x i1> %0, <vscale x 8 x float> %1, <vscale x 8 x float> %2, <vscale x 8 x float> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmflt_mask_vv_nxv8f32_nxv8f32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
-; CHECK-NEXT: vmflt.vv v0, v8, v12
-; CHECK-NEXT: vmflt.vv v20, v12, v16, v0.t
+; CHECK-NEXT: vmflt.vv v20, v8, v12
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmflt.vv v8, v12, v16, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 8 x i1> @llvm.riscv.vmflt.nxv8f32(
@@ -493,9 +502,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmflt.mask.nxv1f64(
define <vscale x 1 x i1> @intrinsic_vmflt_mask_vv_nxv1f64_nxv1f64(<vscale x 1 x i1> %0, <vscale x 1 x double> %1, <vscale x 1 x double> %2, <vscale x 1 x double> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmflt_mask_vv_nxv1f64_nxv1f64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
-; CHECK-NEXT: vmflt.vv v0, v8, v9
+; CHECK-NEXT: vmflt.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmflt.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -544,11 +554,12 @@ declare <vscale x 2 x i1> @llvm.riscv.vmflt.mask.nxv2f64(
define <vscale x 2 x i1> @intrinsic_vmflt_mask_vv_nxv2f64_nxv2f64(<vscale x 2 x i1> %0, <vscale x 2 x double> %1, <vscale x 2 x double> %2, <vscale x 2 x double> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmflt_mask_vv_nxv2f64_nxv2f64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
-; CHECK-NEXT: vmflt.vv v0, v8, v10
-; CHECK-NEXT: vmflt.vv v14, v10, v12, v0.t
+; CHECK-NEXT: vmflt.vv v14, v8, v10
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmflt.vv v8, v10, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 2 x i1> @llvm.riscv.vmflt.nxv2f64(
@@ -595,11 +606,12 @@ declare <vscale x 4 x i1> @llvm.riscv.vmflt.mask.nxv4f64(
define <vscale x 4 x i1> @intrinsic_vmflt_mask_vv_nxv4f64_nxv4f64(<vscale x 4 x i1> %0, <vscale x 4 x double> %1, <vscale x 4 x double> %2, <vscale x 4 x double> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmflt_mask_vv_nxv4f64_nxv4f64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
-; CHECK-NEXT: vmflt.vv v0, v8, v12
-; CHECK-NEXT: vmflt.vv v20, v12, v16, v0.t
+; CHECK-NEXT: vmflt.vv v20, v8, v12
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmflt.vv v8, v12, v16, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 4 x i1> @llvm.riscv.vmflt.nxv4f64(
@@ -647,8 +659,8 @@ define <vscale x 1 x i1> @intrinsic_vmflt_mask_vf_nxv1f16_f16(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmflt_mask_vf_nxv1f16_f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
; CHECK-NEXT: vmflt.vf v10, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -694,8 +706,8 @@ define <vscale x 2 x i1> @intrinsic_vmflt_mask_vf_nxv2f16_f16(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmflt_mask_vf_nxv2f16_f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
; CHECK-NEXT: vmflt.vf v10, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -741,8 +753,8 @@ define <vscale x 4 x i1> @intrinsic_vmflt_mask_vf_nxv4f16_f16(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmflt_mask_vf_nxv4f16_f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
; CHECK-NEXT: vmflt.vf v10, v8, fa0, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -788,8 +800,8 @@ define <vscale x 8 x i1> @intrinsic_vmflt_mask_vf_nxv8f16_f16(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmflt_mask_vf_nxv8f16_f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
; CHECK-NEXT: vmflt.vf v11, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -835,8 +847,8 @@ define <vscale x 16 x i1> @intrinsic_vmflt_mask_vf_nxv16f16_f16(<vscale x 16 x i
; CHECK-LABEL: intrinsic_vmflt_mask_vf_nxv16f16_f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
; CHECK-NEXT: vmflt.vf v13, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -882,8 +894,8 @@ define <vscale x 1 x i1> @intrinsic_vmflt_mask_vf_nxv1f32_f32(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmflt_mask_vf_nxv1f32_f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
; CHECK-NEXT: vmflt.vf v10, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -929,8 +941,8 @@ define <vscale x 2 x i1> @intrinsic_vmflt_mask_vf_nxv2f32_f32(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmflt_mask_vf_nxv2f32_f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
; CHECK-NEXT: vmflt.vf v10, v8, fa0, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -976,8 +988,8 @@ define <vscale x 4 x i1> @intrinsic_vmflt_mask_vf_nxv4f32_f32(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmflt_mask_vf_nxv4f32_f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
; CHECK-NEXT: vmflt.vf v11, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -1023,8 +1035,8 @@ define <vscale x 8 x i1> @intrinsic_vmflt_mask_vf_nxv8f32_f32(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmflt_mask_vf_nxv8f32_f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
; CHECK-NEXT: vmflt.vf v13, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -1070,8 +1082,8 @@ define <vscale x 1 x i1> @intrinsic_vmflt_mask_vf_nxv1f64_f64(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmflt_mask_vf_nxv1f64_f64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
; CHECK-NEXT: vmflt.vf v10, v8, fa0, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -1117,8 +1129,8 @@ define <vscale x 2 x i1> @intrinsic_vmflt_mask_vf_nxv2f64_f64(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmflt_mask_vf_nxv2f64_f64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
; CHECK-NEXT: vmflt.vf v11, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -1164,8 +1176,8 @@ define <vscale x 4 x i1> @intrinsic_vmflt_mask_vf_nxv4f64_f64(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmflt_mask_vf_nxv4f64_f64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
; CHECK-NEXT: vmflt.vf v13, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vmfne.ll b/llvm/test/CodeGen/RISCV/rvv/vmfne.ll
index 5defce42091e5..520099247e0f3 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vmfne.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vmfne.ll
@@ -34,9 +34,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmfne.mask.nxv1f16(
define <vscale x 1 x i1> @intrinsic_vmfne_mask_vv_nxv1f16_nxv1f16(<vscale x 1 x i1> %0, <vscale x 1 x half> %1, <vscale x 1 x half> %2, <vscale x 1 x half> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfne_mask_vv_nxv1f16_nxv1f16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
-; CHECK-NEXT: vmfne.vv v0, v8, v9
+; CHECK-NEXT: vmfne.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmfne.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -85,9 +86,10 @@ declare <vscale x 2 x i1> @llvm.riscv.vmfne.mask.nxv2f16(
define <vscale x 2 x i1> @intrinsic_vmfne_mask_vv_nxv2f16_nxv2f16(<vscale x 2 x i1> %0, <vscale x 2 x half> %1, <vscale x 2 x half> %2, <vscale x 2 x half> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfne_mask_vv_nxv2f16_nxv2f16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
-; CHECK-NEXT: vmfne.vv v0, v8, v9
+; CHECK-NEXT: vmfne.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmfne.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -136,9 +138,10 @@ declare <vscale x 4 x i1> @llvm.riscv.vmfne.mask.nxv4f16(
define <vscale x 4 x i1> @intrinsic_vmfne_mask_vv_nxv4f16_nxv4f16(<vscale x 4 x i1> %0, <vscale x 4 x half> %1, <vscale x 4 x half> %2, <vscale x 4 x half> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfne_mask_vv_nxv4f16_nxv4f16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
-; CHECK-NEXT: vmfne.vv v0, v8, v9
+; CHECK-NEXT: vmfne.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmfne.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -187,11 +190,12 @@ declare <vscale x 8 x i1> @llvm.riscv.vmfne.mask.nxv8f16(
define <vscale x 8 x i1> @intrinsic_vmfne_mask_vv_nxv8f16_nxv8f16(<vscale x 8 x i1> %0, <vscale x 8 x half> %1, <vscale x 8 x half> %2, <vscale x 8 x half> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfne_mask_vv_nxv8f16_nxv8f16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
-; CHECK-NEXT: vmfne.vv v0, v8, v10
-; CHECK-NEXT: vmfne.vv v14, v10, v12, v0.t
+; CHECK-NEXT: vmfne.vv v14, v8, v10
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmfne.vv v8, v10, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 8 x i1> @llvm.riscv.vmfne.nxv8f16(
@@ -238,11 +242,12 @@ declare <vscale x 16 x i1> @llvm.riscv.vmfne.mask.nxv16f16(
define <vscale x 16 x i1> @intrinsic_vmfne_mask_vv_nxv16f16_nxv16f16(<vscale x 16 x i1> %0, <vscale x 16 x half> %1, <vscale x 16 x half> %2, <vscale x 16 x half> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfne_mask_vv_nxv16f16_nxv16f16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
-; CHECK-NEXT: vmfne.vv v0, v8, v12
-; CHECK-NEXT: vmfne.vv v20, v12, v16, v0.t
+; CHECK-NEXT: vmfne.vv v20, v8, v12
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmfne.vv v8, v12, v16, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 16 x i1> @llvm.riscv.vmfne.nxv16f16(
@@ -289,9 +294,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmfne.mask.nxv1f32(
define <vscale x 1 x i1> @intrinsic_vmfne_mask_vv_nxv1f32_nxv1f32(<vscale x 1 x i1> %0, <vscale x 1 x float> %1, <vscale x 1 x float> %2, <vscale x 1 x float> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfne_mask_vv_nxv1f32_nxv1f32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
-; CHECK-NEXT: vmfne.vv v0, v8, v9
+; CHECK-NEXT: vmfne.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmfne.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -340,9 +346,10 @@ declare <vscale x 2 x i1> @llvm.riscv.vmfne.mask.nxv2f32(
define <vscale x 2 x i1> @intrinsic_vmfne_mask_vv_nxv2f32_nxv2f32(<vscale x 2 x i1> %0, <vscale x 2 x float> %1, <vscale x 2 x float> %2, <vscale x 2 x float> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfne_mask_vv_nxv2f32_nxv2f32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
-; CHECK-NEXT: vmfne.vv v0, v8, v9
+; CHECK-NEXT: vmfne.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmfne.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -391,11 +398,12 @@ declare <vscale x 4 x i1> @llvm.riscv.vmfne.mask.nxv4f32(
define <vscale x 4 x i1> @intrinsic_vmfne_mask_vv_nxv4f32_nxv4f32(<vscale x 4 x i1> %0, <vscale x 4 x float> %1, <vscale x 4 x float> %2, <vscale x 4 x float> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfne_mask_vv_nxv4f32_nxv4f32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
-; CHECK-NEXT: vmfne.vv v0, v8, v10
-; CHECK-NEXT: vmfne.vv v14, v10, v12, v0.t
+; CHECK-NEXT: vmfne.vv v14, v8, v10
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmfne.vv v8, v10, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 4 x i1> @llvm.riscv.vmfne.nxv4f32(
@@ -442,11 +450,12 @@ declare <vscale x 8 x i1> @llvm.riscv.vmfne.mask.nxv8f32(
define <vscale x 8 x i1> @intrinsic_vmfne_mask_vv_nxv8f32_nxv8f32(<vscale x 8 x i1> %0, <vscale x 8 x float> %1, <vscale x 8 x float> %2, <vscale x 8 x float> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfne_mask_vv_nxv8f32_nxv8f32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
-; CHECK-NEXT: vmfne.vv v0, v8, v12
-; CHECK-NEXT: vmfne.vv v20, v12, v16, v0.t
+; CHECK-NEXT: vmfne.vv v20, v8, v12
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmfne.vv v8, v12, v16, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 8 x i1> @llvm.riscv.vmfne.nxv8f32(
@@ -493,9 +502,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmfne.mask.nxv1f64(
define <vscale x 1 x i1> @intrinsic_vmfne_mask_vv_nxv1f64_nxv1f64(<vscale x 1 x i1> %0, <vscale x 1 x double> %1, <vscale x 1 x double> %2, <vscale x 1 x double> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfne_mask_vv_nxv1f64_nxv1f64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
-; CHECK-NEXT: vmfne.vv v0, v8, v9
+; CHECK-NEXT: vmfne.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmfne.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -544,11 +554,12 @@ declare <vscale x 2 x i1> @llvm.riscv.vmfne.mask.nxv2f64(
define <vscale x 2 x i1> @intrinsic_vmfne_mask_vv_nxv2f64_nxv2f64(<vscale x 2 x i1> %0, <vscale x 2 x double> %1, <vscale x 2 x double> %2, <vscale x 2 x double> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfne_mask_vv_nxv2f64_nxv2f64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
-; CHECK-NEXT: vmfne.vv v0, v8, v10
-; CHECK-NEXT: vmfne.vv v14, v10, v12, v0.t
+; CHECK-NEXT: vmfne.vv v14, v8, v10
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmfne.vv v8, v10, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 2 x i1> @llvm.riscv.vmfne.nxv2f64(
@@ -595,11 +606,12 @@ declare <vscale x 4 x i1> @llvm.riscv.vmfne.mask.nxv4f64(
define <vscale x 4 x i1> @intrinsic_vmfne_mask_vv_nxv4f64_nxv4f64(<vscale x 4 x i1> %0, <vscale x 4 x double> %1, <vscale x 4 x double> %2, <vscale x 4 x double> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmfne_mask_vv_nxv4f64_nxv4f64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
-; CHECK-NEXT: vmfne.vv v0, v8, v12
-; CHECK-NEXT: vmfne.vv v20, v12, v16, v0.t
+; CHECK-NEXT: vmfne.vv v20, v8, v12
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmfne.vv v8, v12, v16, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 4 x i1> @llvm.riscv.vmfne.nxv4f64(
@@ -647,8 +659,8 @@ define <vscale x 1 x i1> @intrinsic_vmfne_mask_vf_nxv1f16_f16(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmfne_mask_vf_nxv1f16_f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
; CHECK-NEXT: vmfne.vf v10, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -694,8 +706,8 @@ define <vscale x 2 x i1> @intrinsic_vmfne_mask_vf_nxv2f16_f16(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmfne_mask_vf_nxv2f16_f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
; CHECK-NEXT: vmfne.vf v10, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -741,8 +753,8 @@ define <vscale x 4 x i1> @intrinsic_vmfne_mask_vf_nxv4f16_f16(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmfne_mask_vf_nxv4f16_f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
; CHECK-NEXT: vmfne.vf v10, v8, fa0, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -788,8 +800,8 @@ define <vscale x 8 x i1> @intrinsic_vmfne_mask_vf_nxv8f16_f16(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmfne_mask_vf_nxv8f16_f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
; CHECK-NEXT: vmfne.vf v11, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -835,8 +847,8 @@ define <vscale x 16 x i1> @intrinsic_vmfne_mask_vf_nxv16f16_f16(<vscale x 16 x i
; CHECK-LABEL: intrinsic_vmfne_mask_vf_nxv16f16_f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
; CHECK-NEXT: vmfne.vf v13, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -882,8 +894,8 @@ define <vscale x 1 x i1> @intrinsic_vmfne_mask_vf_nxv1f32_f32(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmfne_mask_vf_nxv1f32_f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
; CHECK-NEXT: vmfne.vf v10, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -929,8 +941,8 @@ define <vscale x 2 x i1> @intrinsic_vmfne_mask_vf_nxv2f32_f32(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmfne_mask_vf_nxv2f32_f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
; CHECK-NEXT: vmfne.vf v10, v8, fa0, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -976,8 +988,8 @@ define <vscale x 4 x i1> @intrinsic_vmfne_mask_vf_nxv4f32_f32(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmfne_mask_vf_nxv4f32_f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
; CHECK-NEXT: vmfne.vf v11, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -1023,8 +1035,8 @@ define <vscale x 8 x i1> @intrinsic_vmfne_mask_vf_nxv8f32_f32(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmfne_mask_vf_nxv8f32_f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
; CHECK-NEXT: vmfne.vf v13, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -1070,8 +1082,8 @@ define <vscale x 1 x i1> @intrinsic_vmfne_mask_vf_nxv1f64_f64(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmfne_mask_vf_nxv1f64_f64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
; CHECK-NEXT: vmfne.vf v10, v8, fa0, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -1117,8 +1129,8 @@ define <vscale x 2 x i1> @intrinsic_vmfne_mask_vf_nxv2f64_f64(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmfne_mask_vf_nxv2f64_f64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
; CHECK-NEXT: vmfne.vf v11, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -1164,8 +1176,8 @@ define <vscale x 4 x i1> @intrinsic_vmfne_mask_vf_nxv4f64_f64(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmfne_mask_vf_nxv4f64_f64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
; CHECK-NEXT: vmfne.vf v13, v8, fa0, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vmin-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vmin-vp.ll
index 8fabf93356aeb..1c71242c3c7d7 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vmin-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vmin-vp.ll
@@ -423,8 +423,8 @@ define <vscale x 128 x i8> @vmin_vx_nxv128i8(<vscale x 128 x i8> %va, i8 %b, <vs
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a2, a1
; CHECK-NEXT: .LBB34_2:
-; CHECK-NEXT: vsetvli zero, a2, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a2, e8, m8, ta, ma
; CHECK-NEXT: vmin.vx v8, v8, a0, v0.t
; CHECK-NEXT: ret
%elt.head = insertelement <vscale x 128 x i8> poison, i8 %b, i32 0
@@ -986,8 +986,8 @@ define <vscale x 32 x i32> @vmin_vx_nxv32i32(<vscale x 32 x i32> %va, i32 %b, <v
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a1, a2
; CHECK-NEXT: .LBB80_2:
-; CHECK-NEXT: vsetvli zero, a1, e32, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a1, e32, m8, ta, ma
; CHECK-NEXT: vmin.vx v8, v8, a0, v0.t
; CHECK-NEXT: ret
%elt.head = insertelement <vscale x 32 x i32> poison, i32 %b, i32 0
@@ -1046,8 +1046,8 @@ define <vscale x 32 x i32> @vmin_vx_nxv32i32_evl_nx8(<vscale x 32 x i32> %va, i3
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a1, a2
; CHECK-NEXT: .LBB82_2:
-; CHECK-NEXT: vsetvli zero, a1, e32, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a1, e32, m8, ta, ma
; CHECK-NEXT: vmin.vx v8, v8, a0, v0.t
; CHECK-NEXT: ret
%elt.head = insertelement <vscale x 32 x i32> poison, i32 %b, i32 0
@@ -1084,8 +1084,8 @@ define <vscale x 32 x i32> @vmin_vx_nxv32i32_evl_nx16(<vscale x 32 x i32> %va, i
; RV64-NEXT: slli a1, a1, 1
; RV64-NEXT: vsetvli zero, a1, e32, m8, ta, ma
; RV64-NEXT: vmin.vx v8, v8, a0, v0.t
-; RV64-NEXT: vsetivli zero, 0, e32, m8, ta, ma
; RV64-NEXT: vmv1r.v v0, v24
+; RV64-NEXT: vsetivli zero, 0, e32, m8, ta, ma
; RV64-NEXT: vmin.vx v16, v16, a0, v0.t
; RV64-NEXT: ret
%elt.head = insertelement <vscale x 32 x i32> poison, i32 %b, i32 0
diff --git a/llvm/test/CodeGen/RISCV/rvv/vminu-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vminu-vp.ll
index 8ec85e545a0f8..6d89a9777cf91 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vminu-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vminu-vp.ll
@@ -425,8 +425,8 @@ define <vscale x 128 x i8> @vminu_vx_nxv128i8(<vscale x 128 x i8> %va, i8 %b, <v
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a2, a1
; CHECK-NEXT: .LBB34_2:
-; CHECK-NEXT: vsetvli zero, a2, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a2, e8, m8, ta, ma
; CHECK-NEXT: vminu.vx v8, v8, a0, v0.t
; CHECK-NEXT: ret
%elt.head = insertelement <vscale x 128 x i8> poison, i8 %b, i32 0
@@ -988,8 +988,8 @@ define <vscale x 32 x i32> @vminu_vx_nxv32i32(<vscale x 32 x i32> %va, i32 %b, <
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a1, a2
; CHECK-NEXT: .LBB80_2:
-; CHECK-NEXT: vsetvli zero, a1, e32, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a1, e32, m8, ta, ma
; CHECK-NEXT: vminu.vx v8, v8, a0, v0.t
; CHECK-NEXT: ret
%elt.head = insertelement <vscale x 32 x i32> poison, i32 %b, i32 0
@@ -1048,8 +1048,8 @@ define <vscale x 32 x i32> @vminu_vx_nxv32i32_evl_nx8(<vscale x 32 x i32> %va, i
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a1, a2
; CHECK-NEXT: .LBB82_2:
-; CHECK-NEXT: vsetvli zero, a1, e32, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a1, e32, m8, ta, ma
; CHECK-NEXT: vminu.vx v8, v8, a0, v0.t
; CHECK-NEXT: ret
%elt.head = insertelement <vscale x 32 x i32> poison, i32 %b, i32 0
@@ -1086,8 +1086,8 @@ define <vscale x 32 x i32> @vminu_vx_nxv32i32_evl_nx16(<vscale x 32 x i32> %va,
; RV64-NEXT: slli a1, a1, 1
; RV64-NEXT: vsetvli zero, a1, e32, m8, ta, ma
; RV64-NEXT: vminu.vx v8, v8, a0, v0.t
-; RV64-NEXT: vsetivli zero, 0, e32, m8, ta, ma
; RV64-NEXT: vmv1r.v v0, v24
+; RV64-NEXT: vsetivli zero, 0, e32, m8, ta, ma
; RV64-NEXT: vminu.vx v16, v16, a0, v0.t
; RV64-NEXT: ret
%elt.head = insertelement <vscale x 32 x i32> poison, i32 %b, i32 0
diff --git a/llvm/test/CodeGen/RISCV/rvv/vmsbf.ll b/llvm/test/CodeGen/RISCV/rvv/vmsbf.ll
index 2d6e958fcd0ba..14a1f084c3985 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vmsbf.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vmsbf.ll
@@ -32,8 +32,8 @@ define <vscale x 1 x i1> @intrinsic_vmsbf_mask_m_nxv1i1_nxv1i1(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmsbf_mask_m_nxv1i1_nxv1i1:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf8, tu, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, mf8, tu, mu
; CHECK-NEXT: vmsbf.m v10, v8, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -74,8 +74,8 @@ define <vscale x 2 x i1> @intrinsic_vmsbf_mask_m_nxv2i1_nxv2i1(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmsbf_mask_m_nxv2i1_nxv2i1:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf4, tu, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, mf4, tu, mu
; CHECK-NEXT: vmsbf.m v10, v8, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -116,8 +116,8 @@ define <vscale x 4 x i1> @intrinsic_vmsbf_mask_m_nxv4i1_nxv4i1(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmsbf_mask_m_nxv4i1_nxv4i1:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf2, tu, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, mf2, tu, mu
; CHECK-NEXT: vmsbf.m v10, v8, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -158,8 +158,8 @@ define <vscale x 8 x i1> @intrinsic_vmsbf_mask_m_nxv8i1_nxv8i1(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmsbf_mask_m_nxv8i1_nxv8i1:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m1, tu, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, m1, tu, mu
; CHECK-NEXT: vmsbf.m v10, v8, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -200,8 +200,8 @@ define <vscale x 16 x i1> @intrinsic_vmsbf_mask_m_nxv16i1_nxv16i1(<vscale x 16 x
; CHECK-LABEL: intrinsic_vmsbf_mask_m_nxv16i1_nxv16i1:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m2, tu, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, m2, tu, mu
; CHECK-NEXT: vmsbf.m v10, v8, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -242,8 +242,8 @@ define <vscale x 32 x i1> @intrinsic_vmsbf_mask_m_nxv32i1_nxv32i1(<vscale x 32 x
; CHECK-LABEL: intrinsic_vmsbf_mask_m_nxv32i1_nxv32i1:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m4, tu, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, m4, tu, mu
; CHECK-NEXT: vmsbf.m v10, v8, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -284,8 +284,8 @@ define <vscale x 64 x i1> @intrinsic_vmsbf_mask_m_nxv64i1_nxv64i1(<vscale x 64 x
; CHECK-LABEL: intrinsic_vmsbf_mask_m_nxv64i1_nxv64i1:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m8, tu, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, m8, tu, mu
; CHECK-NEXT: vmsbf.m v10, v8, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vmseq.ll b/llvm/test/CodeGen/RISCV/rvv/vmseq.ll
index cc6c1f585bb7d..9f181f7a30ebe 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vmseq.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vmseq.ll
@@ -34,9 +34,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmseq.mask.nxv1i8(
define <vscale x 1 x i1> @intrinsic_vmseq_mask_vv_nxv1i8_nxv1i8(<vscale x 1 x i1> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, <vscale x 1 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmseq_mask_vv_nxv1i8_nxv1i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, mu
-; CHECK-NEXT: vmseq.vv v0, v8, v9
+; CHECK-NEXT: vmseq.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmseq.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -85,9 +86,10 @@ declare <vscale x 2 x i1> @llvm.riscv.vmseq.mask.nxv2i8(
define <vscale x 2 x i1> @intrinsic_vmseq_mask_vv_nxv2i8_nxv2i8(<vscale x 2 x i1> %0, <vscale x 2 x i8> %1, <vscale x 2 x i8> %2, <vscale x 2 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmseq_mask_vv_nxv2i8_nxv2i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e8, mf4, ta, mu
-; CHECK-NEXT: vmseq.vv v0, v8, v9
+; CHECK-NEXT: vmseq.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmseq.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -136,9 +138,10 @@ declare <vscale x 4 x i1> @llvm.riscv.vmseq.mask.nxv4i8(
define <vscale x 4 x i1> @intrinsic_vmseq_mask_vv_nxv4i8_nxv4i8(<vscale x 4 x i1> %0, <vscale x 4 x i8> %1, <vscale x 4 x i8> %2, <vscale x 4 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmseq_mask_vv_nxv4i8_nxv4i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, mu
-; CHECK-NEXT: vmseq.vv v0, v8, v9
+; CHECK-NEXT: vmseq.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmseq.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -187,9 +190,10 @@ declare <vscale x 8 x i1> @llvm.riscv.vmseq.mask.nxv8i8(
define <vscale x 8 x i1> @intrinsic_vmseq_mask_vv_nxv8i8_nxv8i8(<vscale x 8 x i1> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, <vscale x 8 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmseq_mask_vv_nxv8i8_nxv8i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e8, m1, ta, mu
-; CHECK-NEXT: vmseq.vv v0, v8, v9
+; CHECK-NEXT: vmseq.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmseq.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -238,11 +242,12 @@ declare <vscale x 16 x i1> @llvm.riscv.vmseq.mask.nxv16i8(
define <vscale x 16 x i1> @intrinsic_vmseq_mask_vv_nxv16i8_nxv16i8(<vscale x 16 x i1> %0, <vscale x 16 x i8> %1, <vscale x 16 x i8> %2, <vscale x 16 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmseq_mask_vv_nxv16i8_nxv16i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e8, m2, ta, mu
-; CHECK-NEXT: vmseq.vv v0, v8, v10
-; CHECK-NEXT: vmseq.vv v14, v10, v12, v0.t
+; CHECK-NEXT: vmseq.vv v14, v8, v10
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmseq.vv v8, v10, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 16 x i1> @llvm.riscv.vmseq.nxv16i8(
@@ -289,11 +294,12 @@ declare <vscale x 32 x i1> @llvm.riscv.vmseq.mask.nxv32i8(
define <vscale x 32 x i1> @intrinsic_vmseq_mask_vv_nxv32i8_nxv32i8(<vscale x 32 x i1> %0, <vscale x 32 x i8> %1, <vscale x 32 x i8> %2, <vscale x 32 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmseq_mask_vv_nxv32i8_nxv32i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, mu
-; CHECK-NEXT: vmseq.vv v0, v8, v12
-; CHECK-NEXT: vmseq.vv v20, v12, v16, v0.t
+; CHECK-NEXT: vmseq.vv v20, v8, v12
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmseq.vv v8, v12, v16, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 32 x i1> @llvm.riscv.vmseq.nxv32i8(
@@ -340,9 +346,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmseq.mask.nxv1i16(
define <vscale x 1 x i1> @intrinsic_vmseq_mask_vv_nxv1i16_nxv1i16(<vscale x 1 x i1> %0, <vscale x 1 x i16> %1, <vscale x 1 x i16> %2, <vscale x 1 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmseq_mask_vv_nxv1i16_nxv1i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
-; CHECK-NEXT: vmseq.vv v0, v8, v9
+; CHECK-NEXT: vmseq.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmseq.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -391,9 +398,10 @@ declare <vscale x 2 x i1> @llvm.riscv.vmseq.mask.nxv2i16(
define <vscale x 2 x i1> @intrinsic_vmseq_mask_vv_nxv2i16_nxv2i16(<vscale x 2 x i1> %0, <vscale x 2 x i16> %1, <vscale x 2 x i16> %2, <vscale x 2 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmseq_mask_vv_nxv2i16_nxv2i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
-; CHECK-NEXT: vmseq.vv v0, v8, v9
+; CHECK-NEXT: vmseq.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmseq.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -442,9 +450,10 @@ declare <vscale x 4 x i1> @llvm.riscv.vmseq.mask.nxv4i16(
define <vscale x 4 x i1> @intrinsic_vmseq_mask_vv_nxv4i16_nxv4i16(<vscale x 4 x i1> %0, <vscale x 4 x i16> %1, <vscale x 4 x i16> %2, <vscale x 4 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmseq_mask_vv_nxv4i16_nxv4i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
-; CHECK-NEXT: vmseq.vv v0, v8, v9
+; CHECK-NEXT: vmseq.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmseq.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -493,11 +502,12 @@ declare <vscale x 8 x i1> @llvm.riscv.vmseq.mask.nxv8i16(
define <vscale x 8 x i1> @intrinsic_vmseq_mask_vv_nxv8i16_nxv8i16(<vscale x 8 x i1> %0, <vscale x 8 x i16> %1, <vscale x 8 x i16> %2, <vscale x 8 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmseq_mask_vv_nxv8i16_nxv8i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
-; CHECK-NEXT: vmseq.vv v0, v8, v10
-; CHECK-NEXT: vmseq.vv v14, v10, v12, v0.t
+; CHECK-NEXT: vmseq.vv v14, v8, v10
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmseq.vv v8, v10, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 8 x i1> @llvm.riscv.vmseq.nxv8i16(
@@ -544,11 +554,12 @@ declare <vscale x 16 x i1> @llvm.riscv.vmseq.mask.nxv16i16(
define <vscale x 16 x i1> @intrinsic_vmseq_mask_vv_nxv16i16_nxv16i16(<vscale x 16 x i1> %0, <vscale x 16 x i16> %1, <vscale x 16 x i16> %2, <vscale x 16 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmseq_mask_vv_nxv16i16_nxv16i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
-; CHECK-NEXT: vmseq.vv v0, v8, v12
-; CHECK-NEXT: vmseq.vv v20, v12, v16, v0.t
+; CHECK-NEXT: vmseq.vv v20, v8, v12
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmseq.vv v8, v12, v16, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 16 x i1> @llvm.riscv.vmseq.nxv16i16(
@@ -595,9 +606,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmseq.mask.nxv1i32(
define <vscale x 1 x i1> @intrinsic_vmseq_mask_vv_nxv1i32_nxv1i32(<vscale x 1 x i1> %0, <vscale x 1 x i32> %1, <vscale x 1 x i32> %2, <vscale x 1 x i32> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmseq_mask_vv_nxv1i32_nxv1i32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
-; CHECK-NEXT: vmseq.vv v0, v8, v9
+; CHECK-NEXT: vmseq.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmseq.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -646,9 +658,10 @@ declare <vscale x 2 x i1> @llvm.riscv.vmseq.mask.nxv2i32(
define <vscale x 2 x i1> @intrinsic_vmseq_mask_vv_nxv2i32_nxv2i32(<vscale x 2 x i1> %0, <vscale x 2 x i32> %1, <vscale x 2 x i32> %2, <vscale x 2 x i32> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmseq_mask_vv_nxv2i32_nxv2i32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
-; CHECK-NEXT: vmseq.vv v0, v8, v9
+; CHECK-NEXT: vmseq.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmseq.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -697,11 +710,12 @@ declare <vscale x 4 x i1> @llvm.riscv.vmseq.mask.nxv4i32(
define <vscale x 4 x i1> @intrinsic_vmseq_mask_vv_nxv4i32_nxv4i32(<vscale x 4 x i1> %0, <vscale x 4 x i32> %1, <vscale x 4 x i32> %2, <vscale x 4 x i32> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmseq_mask_vv_nxv4i32_nxv4i32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
-; CHECK-NEXT: vmseq.vv v0, v8, v10
-; CHECK-NEXT: vmseq.vv v14, v10, v12, v0.t
+; CHECK-NEXT: vmseq.vv v14, v8, v10
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmseq.vv v8, v10, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 4 x i1> @llvm.riscv.vmseq.nxv4i32(
@@ -748,11 +762,12 @@ declare <vscale x 8 x i1> @llvm.riscv.vmseq.mask.nxv8i32(
define <vscale x 8 x i1> @intrinsic_vmseq_mask_vv_nxv8i32_nxv8i32(<vscale x 8 x i1> %0, <vscale x 8 x i32> %1, <vscale x 8 x i32> %2, <vscale x 8 x i32> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmseq_mask_vv_nxv8i32_nxv8i32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
-; CHECK-NEXT: vmseq.vv v0, v8, v12
-; CHECK-NEXT: vmseq.vv v20, v12, v16, v0.t
+; CHECK-NEXT: vmseq.vv v20, v8, v12
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmseq.vv v8, v12, v16, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 8 x i1> @llvm.riscv.vmseq.nxv8i32(
@@ -799,9 +814,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmseq.mask.nxv1i64(
define <vscale x 1 x i1> @intrinsic_vmseq_mask_vv_nxv1i64_nxv1i64(<vscale x 1 x i1> %0, <vscale x 1 x i64> %1, <vscale x 1 x i64> %2, <vscale x 1 x i64> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmseq_mask_vv_nxv1i64_nxv1i64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
-; CHECK-NEXT: vmseq.vv v0, v8, v9
+; CHECK-NEXT: vmseq.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmseq.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -850,11 +866,12 @@ declare <vscale x 2 x i1> @llvm.riscv.vmseq.mask.nxv2i64(
define <vscale x 2 x i1> @intrinsic_vmseq_mask_vv_nxv2i64_nxv2i64(<vscale x 2 x i1> %0, <vscale x 2 x i64> %1, <vscale x 2 x i64> %2, <vscale x 2 x i64> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmseq_mask_vv_nxv2i64_nxv2i64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
-; CHECK-NEXT: vmseq.vv v0, v8, v10
-; CHECK-NEXT: vmseq.vv v14, v10, v12, v0.t
+; CHECK-NEXT: vmseq.vv v14, v8, v10
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmseq.vv v8, v10, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 2 x i1> @llvm.riscv.vmseq.nxv2i64(
@@ -901,11 +918,12 @@ declare <vscale x 4 x i1> @llvm.riscv.vmseq.mask.nxv4i64(
define <vscale x 4 x i1> @intrinsic_vmseq_mask_vv_nxv4i64_nxv4i64(<vscale x 4 x i1> %0, <vscale x 4 x i64> %1, <vscale x 4 x i64> %2, <vscale x 4 x i64> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmseq_mask_vv_nxv4i64_nxv4i64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
-; CHECK-NEXT: vmseq.vv v0, v8, v12
-; CHECK-NEXT: vmseq.vv v20, v12, v16, v0.t
+; CHECK-NEXT: vmseq.vv v20, v8, v12
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmseq.vv v8, v12, v16, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 4 x i1> @llvm.riscv.vmseq.nxv4i64(
@@ -953,8 +971,8 @@ define <vscale x 1 x i1> @intrinsic_vmseq_mask_vx_nxv1i8_i8(<vscale x 1 x i1> %0
; CHECK-LABEL: intrinsic_vmseq_mask_vx_nxv1i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, mu
; CHECK-NEXT: vmseq.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1000,8 +1018,8 @@ define <vscale x 2 x i1> @intrinsic_vmseq_mask_vx_nxv2i8_i8(<vscale x 2 x i1> %0
; CHECK-LABEL: intrinsic_vmseq_mask_vx_nxv2i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e8, mf4, ta, mu
; CHECK-NEXT: vmseq.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1047,8 +1065,8 @@ define <vscale x 4 x i1> @intrinsic_vmseq_mask_vx_nxv4i8_i8(<vscale x 4 x i1> %0
; CHECK-LABEL: intrinsic_vmseq_mask_vx_nxv4i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e8, mf2, ta, mu
; CHECK-NEXT: vmseq.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1094,8 +1112,8 @@ define <vscale x 8 x i1> @intrinsic_vmseq_mask_vx_nxv8i8_i8(<vscale x 8 x i1> %0
; CHECK-LABEL: intrinsic_vmseq_mask_vx_nxv8i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e8, m1, ta, mu
; CHECK-NEXT: vmseq.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -1141,8 +1159,8 @@ define <vscale x 16 x i1> @intrinsic_vmseq_mask_vx_nxv16i8_i8(<vscale x 16 x i1>
; CHECK-LABEL: intrinsic_vmseq_mask_vx_nxv16i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a1, e8, m2, ta, mu
; CHECK-NEXT: vmseq.vx v11, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -1188,8 +1206,8 @@ define <vscale x 32 x i1> @intrinsic_vmseq_mask_vx_nxv32i8_i8(<vscale x 32 x i1>
; CHECK-LABEL: intrinsic_vmseq_mask_vx_nxv32i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a1, e8, m4, ta, mu
; CHECK-NEXT: vmseq.vx v13, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -1235,8 +1253,8 @@ define <vscale x 1 x i1> @intrinsic_vmseq_mask_vx_nxv1i16_i16(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmseq_mask_vx_nxv1i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, mf4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e16, mf4, ta, mu
; CHECK-NEXT: vmseq.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1282,8 +1300,8 @@ define <vscale x 2 x i1> @intrinsic_vmseq_mask_vx_nxv2i16_i16(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmseq_mask_vx_nxv2i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e16, mf2, ta, mu
; CHECK-NEXT: vmseq.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1329,8 +1347,8 @@ define <vscale x 4 x i1> @intrinsic_vmseq_mask_vx_nxv4i16_i16(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmseq_mask_vx_nxv4i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e16, m1, ta, mu
; CHECK-NEXT: vmseq.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -1376,8 +1394,8 @@ define <vscale x 8 x i1> @intrinsic_vmseq_mask_vx_nxv8i16_i16(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmseq_mask_vx_nxv8i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a1, e16, m2, ta, mu
; CHECK-NEXT: vmseq.vx v11, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -1423,8 +1441,8 @@ define <vscale x 16 x i1> @intrinsic_vmseq_mask_vx_nxv16i16_i16(<vscale x 16 x i
; CHECK-LABEL: intrinsic_vmseq_mask_vx_nxv16i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a1, e16, m4, ta, mu
; CHECK-NEXT: vmseq.vx v13, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -1470,8 +1488,8 @@ define <vscale x 1 x i1> @intrinsic_vmseq_mask_vx_nxv1i32_i32(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmseq_mask_vx_nxv1i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e32, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e32, mf2, ta, mu
; CHECK-NEXT: vmseq.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1517,8 +1535,8 @@ define <vscale x 2 x i1> @intrinsic_vmseq_mask_vx_nxv2i32_i32(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmseq_mask_vx_nxv2i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e32, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e32, m1, ta, mu
; CHECK-NEXT: vmseq.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -1564,8 +1582,8 @@ define <vscale x 4 x i1> @intrinsic_vmseq_mask_vx_nxv4i32_i32(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmseq_mask_vx_nxv4i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a1, e32, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a1, e32, m2, ta, mu
; CHECK-NEXT: vmseq.vx v11, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -1611,8 +1629,8 @@ define <vscale x 8 x i1> @intrinsic_vmseq_mask_vx_nxv8i32_i32(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmseq_mask_vx_nxv8i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a1, e32, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a1, e32, m4, ta, mu
; CHECK-NEXT: vmseq.vx v13, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -1685,8 +1703,8 @@ define <vscale x 1 x i1> @intrinsic_vmseq_mask_vx_nxv1i64_i64(<vscale x 1 x i1>
; RV64-LABEL: intrinsic_vmseq_mask_vx_nxv1i64_i64:
; RV64: # %bb.0: # %entry
; RV64-NEXT: vmv1r.v v10, v0
-; RV64-NEXT: vsetvli zero, a1, e64, m1, ta, mu
; RV64-NEXT: vmv1r.v v0, v9
+; RV64-NEXT: vsetvli zero, a1, e64, m1, ta, mu
; RV64-NEXT: vmseq.vx v10, v8, a0, v0.t
; RV64-NEXT: vmv.v.v v0, v10
; RV64-NEXT: ret
@@ -1759,8 +1777,8 @@ define <vscale x 2 x i1> @intrinsic_vmseq_mask_vx_nxv2i64_i64(<vscale x 2 x i1>
; RV64-LABEL: intrinsic_vmseq_mask_vx_nxv2i64_i64:
; RV64: # %bb.0: # %entry
; RV64-NEXT: vmv1r.v v11, v0
-; RV64-NEXT: vsetvli zero, a1, e64, m2, ta, mu
; RV64-NEXT: vmv1r.v v0, v10
+; RV64-NEXT: vsetvli zero, a1, e64, m2, ta, mu
; RV64-NEXT: vmseq.vx v11, v8, a0, v0.t
; RV64-NEXT: vmv1r.v v0, v11
; RV64-NEXT: ret
@@ -1833,8 +1851,8 @@ define <vscale x 4 x i1> @intrinsic_vmseq_mask_vx_nxv4i64_i64(<vscale x 4 x i1>
; RV64-LABEL: intrinsic_vmseq_mask_vx_nxv4i64_i64:
; RV64: # %bb.0: # %entry
; RV64-NEXT: vmv1r.v v13, v0
-; RV64-NEXT: vsetvli zero, a1, e64, m4, ta, mu
; RV64-NEXT: vmv1r.v v0, v12
+; RV64-NEXT: vsetvli zero, a1, e64, m4, ta, mu
; RV64-NEXT: vmseq.vx v13, v8, a0, v0.t
; RV64-NEXT: vmv1r.v v0, v13
; RV64-NEXT: ret
@@ -1868,8 +1886,8 @@ define <vscale x 1 x i1> @intrinsic_vmseq_mask_vi_nxv1i8_i8(<vscale x 1 x i1> %0
; CHECK-LABEL: intrinsic_vmseq_mask_vi_nxv1i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, mu
; CHECK-NEXT: vmseq.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1903,8 +1921,8 @@ define <vscale x 2 x i1> @intrinsic_vmseq_mask_vi_nxv2i8_i8(<vscale x 2 x i1> %0
; CHECK-LABEL: intrinsic_vmseq_mask_vi_nxv2i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, mf4, ta, mu
; CHECK-NEXT: vmseq.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1938,8 +1956,8 @@ define <vscale x 4 x i1> @intrinsic_vmseq_mask_vi_nxv4i8_i8(<vscale x 4 x i1> %0
; CHECK-LABEL: intrinsic_vmseq_mask_vi_nxv4i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, mu
; CHECK-NEXT: vmseq.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1973,8 +1991,8 @@ define <vscale x 8 x i1> @intrinsic_vmseq_mask_vi_nxv8i8_i8(<vscale x 8 x i1> %0
; CHECK-LABEL: intrinsic_vmseq_mask_vi_nxv8i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, m1, ta, mu
; CHECK-NEXT: vmseq.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -2008,8 +2026,8 @@ define <vscale x 16 x i1> @intrinsic_vmseq_mask_vi_nxv16i8_i8(<vscale x 16 x i1>
; CHECK-LABEL: intrinsic_vmseq_mask_vi_nxv16i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e8, m2, ta, mu
; CHECK-NEXT: vmseq.vi v11, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -2043,8 +2061,8 @@ define <vscale x 32 x i1> @intrinsic_vmseq_mask_vi_nxv32i8_i8(<vscale x 32 x i1>
; CHECK-LABEL: intrinsic_vmseq_mask_vi_nxv32i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, mu
; CHECK-NEXT: vmseq.vi v13, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -2078,8 +2096,8 @@ define <vscale x 1 x i1> @intrinsic_vmseq_mask_vi_nxv1i16_i16(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmseq_mask_vi_nxv1i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
; CHECK-NEXT: vmseq.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -2113,8 +2131,8 @@ define <vscale x 2 x i1> @intrinsic_vmseq_mask_vi_nxv2i16_i16(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmseq_mask_vi_nxv2i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
; CHECK-NEXT: vmseq.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -2148,8 +2166,8 @@ define <vscale x 4 x i1> @intrinsic_vmseq_mask_vi_nxv4i16_i16(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmseq_mask_vi_nxv4i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
; CHECK-NEXT: vmseq.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -2183,8 +2201,8 @@ define <vscale x 8 x i1> @intrinsic_vmseq_mask_vi_nxv8i16_i16(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmseq_mask_vi_nxv8i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
; CHECK-NEXT: vmseq.vi v11, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -2218,8 +2236,8 @@ define <vscale x 16 x i1> @intrinsic_vmseq_mask_vi_nxv16i16_i16(<vscale x 16 x i
; CHECK-LABEL: intrinsic_vmseq_mask_vi_nxv16i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
; CHECK-NEXT: vmseq.vi v13, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -2253,8 +2271,8 @@ define <vscale x 1 x i1> @intrinsic_vmseq_mask_vi_nxv1i32_i32(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmseq_mask_vi_nxv1i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
; CHECK-NEXT: vmseq.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -2288,8 +2306,8 @@ define <vscale x 2 x i1> @intrinsic_vmseq_mask_vi_nxv2i32_i32(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmseq_mask_vi_nxv2i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
; CHECK-NEXT: vmseq.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -2323,8 +2341,8 @@ define <vscale x 4 x i1> @intrinsic_vmseq_mask_vi_nxv4i32_i32(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmseq_mask_vi_nxv4i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
; CHECK-NEXT: vmseq.vi v11, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -2358,8 +2376,8 @@ define <vscale x 8 x i1> @intrinsic_vmseq_mask_vi_nxv8i32_i32(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmseq_mask_vi_nxv8i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
; CHECK-NEXT: vmseq.vi v13, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -2393,8 +2411,8 @@ define <vscale x 1 x i1> @intrinsic_vmseq_mask_vi_nxv1i64_i64(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmseq_mask_vi_nxv1i64_i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
; CHECK-NEXT: vmseq.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -2428,8 +2446,8 @@ define <vscale x 2 x i1> @intrinsic_vmseq_mask_vi_nxv2i64_i64(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmseq_mask_vi_nxv2i64_i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
; CHECK-NEXT: vmseq.vi v11, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -2463,8 +2481,8 @@ define <vscale x 4 x i1> @intrinsic_vmseq_mask_vi_nxv4i64_i64(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmseq_mask_vi_nxv4i64_i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
; CHECK-NEXT: vmseq.vi v13, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vmsge.ll b/llvm/test/CodeGen/RISCV/rvv/vmsge.ll
index c8f9b60a3f2da..75fc407abbc2f 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vmsge.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vmsge.ll
@@ -34,9 +34,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsge.mask.nxv1i8(
define <vscale x 1 x i1> @intrinsic_vmsge_mask_vv_nxv1i8_nxv1i8(<vscale x 1 x i1> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, <vscale x 1 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsge_mask_vv_nxv1i8_nxv1i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, mu
-; CHECK-NEXT: vmsle.vv v0, v9, v8
+; CHECK-NEXT: vmsle.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmsle.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -85,9 +86,10 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsge.mask.nxv2i8(
define <vscale x 2 x i1> @intrinsic_vmsge_mask_vv_nxv2i8_nxv2i8(<vscale x 2 x i1> %0, <vscale x 2 x i8> %1, <vscale x 2 x i8> %2, <vscale x 2 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsge_mask_vv_nxv2i8_nxv2i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e8, mf4, ta, mu
-; CHECK-NEXT: vmsle.vv v0, v9, v8
+; CHECK-NEXT: vmsle.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmsle.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -136,9 +138,10 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsge.mask.nxv4i8(
define <vscale x 4 x i1> @intrinsic_vmsge_mask_vv_nxv4i8_nxv4i8(<vscale x 4 x i1> %0, <vscale x 4 x i8> %1, <vscale x 4 x i8> %2, <vscale x 4 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsge_mask_vv_nxv4i8_nxv4i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, mu
-; CHECK-NEXT: vmsle.vv v0, v9, v8
+; CHECK-NEXT: vmsle.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmsle.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -187,9 +190,10 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsge.mask.nxv8i8(
define <vscale x 8 x i1> @intrinsic_vmsge_mask_vv_nxv8i8_nxv8i8(<vscale x 8 x i1> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, <vscale x 8 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsge_mask_vv_nxv8i8_nxv8i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e8, m1, ta, mu
-; CHECK-NEXT: vmsle.vv v0, v9, v8
+; CHECK-NEXT: vmsle.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmsle.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -238,11 +242,12 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsge.mask.nxv16i8(
define <vscale x 16 x i1> @intrinsic_vmsge_mask_vv_nxv16i8_nxv16i8(<vscale x 16 x i1> %0, <vscale x 16 x i8> %1, <vscale x 16 x i8> %2, <vscale x 16 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsge_mask_vv_nxv16i8_nxv16i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e8, m2, ta, mu
-; CHECK-NEXT: vmsle.vv v0, v10, v8
-; CHECK-NEXT: vmsle.vv v14, v12, v10, v0.t
+; CHECK-NEXT: vmsle.vv v14, v10, v8
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmsle.vv v8, v12, v10, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 16 x i1> @llvm.riscv.vmsge.nxv16i8(
@@ -289,11 +294,12 @@ declare <vscale x 32 x i1> @llvm.riscv.vmsge.mask.nxv32i8(
define <vscale x 32 x i1> @intrinsic_vmsge_mask_vv_nxv32i8_nxv32i8(<vscale x 32 x i1> %0, <vscale x 32 x i8> %1, <vscale x 32 x i8> %2, <vscale x 32 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsge_mask_vv_nxv32i8_nxv32i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, mu
-; CHECK-NEXT: vmsle.vv v0, v12, v8
-; CHECK-NEXT: vmsle.vv v20, v16, v12, v0.t
+; CHECK-NEXT: vmsle.vv v20, v12, v8
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmsle.vv v8, v16, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 32 x i1> @llvm.riscv.vmsge.nxv32i8(
@@ -340,9 +346,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsge.mask.nxv1i16(
define <vscale x 1 x i1> @intrinsic_vmsge_mask_vv_nxv1i16_nxv1i16(<vscale x 1 x i1> %0, <vscale x 1 x i16> %1, <vscale x 1 x i16> %2, <vscale x 1 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsge_mask_vv_nxv1i16_nxv1i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
-; CHECK-NEXT: vmsle.vv v0, v9, v8
+; CHECK-NEXT: vmsle.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmsle.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -391,9 +398,10 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsge.mask.nxv2i16(
define <vscale x 2 x i1> @intrinsic_vmsge_mask_vv_nxv2i16_nxv2i16(<vscale x 2 x i1> %0, <vscale x 2 x i16> %1, <vscale x 2 x i16> %2, <vscale x 2 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsge_mask_vv_nxv2i16_nxv2i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
-; CHECK-NEXT: vmsle.vv v0, v9, v8
+; CHECK-NEXT: vmsle.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmsle.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -442,9 +450,10 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsge.mask.nxv4i16(
define <vscale x 4 x i1> @intrinsic_vmsge_mask_vv_nxv4i16_nxv4i16(<vscale x 4 x i1> %0, <vscale x 4 x i16> %1, <vscale x 4 x i16> %2, <vscale x 4 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsge_mask_vv_nxv4i16_nxv4i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
-; CHECK-NEXT: vmsle.vv v0, v9, v8
+; CHECK-NEXT: vmsle.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmsle.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -493,11 +502,12 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsge.mask.nxv8i16(
define <vscale x 8 x i1> @intrinsic_vmsge_mask_vv_nxv8i16_nxv8i16(<vscale x 8 x i1> %0, <vscale x 8 x i16> %1, <vscale x 8 x i16> %2, <vscale x 8 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsge_mask_vv_nxv8i16_nxv8i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
-; CHECK-NEXT: vmsle.vv v0, v10, v8
-; CHECK-NEXT: vmsle.vv v14, v12, v10, v0.t
+; CHECK-NEXT: vmsle.vv v14, v10, v8
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmsle.vv v8, v12, v10, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 8 x i1> @llvm.riscv.vmsge.nxv8i16(
@@ -544,11 +554,12 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsge.mask.nxv16i16(
define <vscale x 16 x i1> @intrinsic_vmsge_mask_vv_nxv16i16_nxv16i16(<vscale x 16 x i1> %0, <vscale x 16 x i16> %1, <vscale x 16 x i16> %2, <vscale x 16 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsge_mask_vv_nxv16i16_nxv16i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
-; CHECK-NEXT: vmsle.vv v0, v12, v8
-; CHECK-NEXT: vmsle.vv v20, v16, v12, v0.t
+; CHECK-NEXT: vmsle.vv v20, v12, v8
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmsle.vv v8, v16, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 16 x i1> @llvm.riscv.vmsge.nxv16i16(
@@ -595,9 +606,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsge.mask.nxv1i32(
define <vscale x 1 x i1> @intrinsic_vmsge_mask_vv_nxv1i32_nxv1i32(<vscale x 1 x i1> %0, <vscale x 1 x i32> %1, <vscale x 1 x i32> %2, <vscale x 1 x i32> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsge_mask_vv_nxv1i32_nxv1i32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
-; CHECK-NEXT: vmsle.vv v0, v9, v8
+; CHECK-NEXT: vmsle.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmsle.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -646,9 +658,10 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsge.mask.nxv2i32(
define <vscale x 2 x i1> @intrinsic_vmsge_mask_vv_nxv2i32_nxv2i32(<vscale x 2 x i1> %0, <vscale x 2 x i32> %1, <vscale x 2 x i32> %2, <vscale x 2 x i32> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsge_mask_vv_nxv2i32_nxv2i32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
-; CHECK-NEXT: vmsle.vv v0, v9, v8
+; CHECK-NEXT: vmsle.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmsle.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -697,11 +710,12 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsge.mask.nxv4i32(
define <vscale x 4 x i1> @intrinsic_vmsge_mask_vv_nxv4i32_nxv4i32(<vscale x 4 x i1> %0, <vscale x 4 x i32> %1, <vscale x 4 x i32> %2, <vscale x 4 x i32> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsge_mask_vv_nxv4i32_nxv4i32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
-; CHECK-NEXT: vmsle.vv v0, v10, v8
-; CHECK-NEXT: vmsle.vv v14, v12, v10, v0.t
+; CHECK-NEXT: vmsle.vv v14, v10, v8
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmsle.vv v8, v12, v10, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 4 x i1> @llvm.riscv.vmsge.nxv4i32(
@@ -748,11 +762,12 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsge.mask.nxv8i32(
define <vscale x 8 x i1> @intrinsic_vmsge_mask_vv_nxv8i32_nxv8i32(<vscale x 8 x i1> %0, <vscale x 8 x i32> %1, <vscale x 8 x i32> %2, <vscale x 8 x i32> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsge_mask_vv_nxv8i32_nxv8i32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
-; CHECK-NEXT: vmsle.vv v0, v12, v8
-; CHECK-NEXT: vmsle.vv v20, v16, v12, v0.t
+; CHECK-NEXT: vmsle.vv v20, v12, v8
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmsle.vv v8, v16, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 8 x i1> @llvm.riscv.vmsge.nxv8i32(
@@ -799,9 +814,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsge.mask.nxv1i64(
define <vscale x 1 x i1> @intrinsic_vmsge_mask_vv_nxv1i64_nxv1i64(<vscale x 1 x i1> %0, <vscale x 1 x i64> %1, <vscale x 1 x i64> %2, <vscale x 1 x i64> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsge_mask_vv_nxv1i64_nxv1i64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
-; CHECK-NEXT: vmsle.vv v0, v9, v8
+; CHECK-NEXT: vmsle.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmsle.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -850,11 +866,12 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsge.mask.nxv2i64(
define <vscale x 2 x i1> @intrinsic_vmsge_mask_vv_nxv2i64_nxv2i64(<vscale x 2 x i1> %0, <vscale x 2 x i64> %1, <vscale x 2 x i64> %2, <vscale x 2 x i64> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsge_mask_vv_nxv2i64_nxv2i64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
-; CHECK-NEXT: vmsle.vv v0, v10, v8
-; CHECK-NEXT: vmsle.vv v14, v12, v10, v0.t
+; CHECK-NEXT: vmsle.vv v14, v10, v8
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmsle.vv v8, v12, v10, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 2 x i1> @llvm.riscv.vmsge.nxv2i64(
@@ -901,11 +918,12 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsge.mask.nxv4i64(
define <vscale x 4 x i1> @intrinsic_vmsge_mask_vv_nxv4i64_nxv4i64(<vscale x 4 x i1> %0, <vscale x 4 x i64> %1, <vscale x 4 x i64> %2, <vscale x 4 x i64> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsge_mask_vv_nxv4i64_nxv4i64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
-; CHECK-NEXT: vmsle.vv v0, v12, v8
-; CHECK-NEXT: vmsle.vv v20, v16, v12, v0.t
+; CHECK-NEXT: vmsle.vv v20, v12, v8
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmsle.vv v8, v16, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 4 x i1> @llvm.riscv.vmsge.nxv4i64(
@@ -954,8 +972,8 @@ define <vscale x 1 x i1> @intrinsic_vmsge_mask_vx_nxv1i8_i8(<vscale x 1 x i1> %0
; CHECK-LABEL: intrinsic_vmsge_mask_vx_nxv1i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, mu
; CHECK-NEXT: vmslt.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmxor.mm v0, v10, v9
; CHECK-NEXT: ret
@@ -1002,8 +1020,8 @@ define <vscale x 2 x i1> @intrinsic_vmsge_mask_vx_nxv2i8_i8(<vscale x 2 x i1> %0
; CHECK-LABEL: intrinsic_vmsge_mask_vx_nxv2i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e8, mf4, ta, mu
; CHECK-NEXT: vmslt.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmxor.mm v0, v10, v9
; CHECK-NEXT: ret
@@ -1050,8 +1068,8 @@ define <vscale x 4 x i1> @intrinsic_vmsge_mask_vx_nxv4i8_i8(<vscale x 4 x i1> %0
; CHECK-LABEL: intrinsic_vmsge_mask_vx_nxv4i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e8, mf2, ta, mu
; CHECK-NEXT: vmslt.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmxor.mm v0, v10, v9
; CHECK-NEXT: ret
@@ -1098,8 +1116,8 @@ define <vscale x 8 x i1> @intrinsic_vmsge_mask_vx_nxv8i8_i8(<vscale x 8 x i1> %0
; CHECK-LABEL: intrinsic_vmsge_mask_vx_nxv8i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e8, m1, ta, mu
; CHECK-NEXT: vmslt.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmxor.mm v0, v10, v9
; CHECK-NEXT: ret
@@ -1146,8 +1164,8 @@ define <vscale x 16 x i1> @intrinsic_vmsge_mask_vx_nxv16i8_i8(<vscale x 16 x i1>
; CHECK-LABEL: intrinsic_vmsge_mask_vx_nxv16i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a1, e8, m2, ta, mu
; CHECK-NEXT: vmslt.vx v11, v8, a0, v0.t
; CHECK-NEXT: vmxor.mm v0, v11, v10
; CHECK-NEXT: ret
@@ -1194,8 +1212,8 @@ define <vscale x 32 x i1> @intrinsic_vmsge_mask_vx_nxv32i8_i8(<vscale x 32 x i1>
; CHECK-LABEL: intrinsic_vmsge_mask_vx_nxv32i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a1, e8, m4, ta, mu
; CHECK-NEXT: vmslt.vx v13, v8, a0, v0.t
; CHECK-NEXT: vmxor.mm v0, v13, v12
; CHECK-NEXT: ret
@@ -1242,8 +1260,8 @@ define <vscale x 1 x i1> @intrinsic_vmsge_mask_vx_nxv1i16_i16(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmsge_mask_vx_nxv1i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, mf4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e16, mf4, ta, mu
; CHECK-NEXT: vmslt.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmxor.mm v0, v10, v9
; CHECK-NEXT: ret
@@ -1290,8 +1308,8 @@ define <vscale x 2 x i1> @intrinsic_vmsge_mask_vx_nxv2i16_i16(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmsge_mask_vx_nxv2i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e16, mf2, ta, mu
; CHECK-NEXT: vmslt.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmxor.mm v0, v10, v9
; CHECK-NEXT: ret
@@ -1338,8 +1356,8 @@ define <vscale x 4 x i1> @intrinsic_vmsge_mask_vx_nxv4i16_i16(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmsge_mask_vx_nxv4i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e16, m1, ta, mu
; CHECK-NEXT: vmslt.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmxor.mm v0, v10, v9
; CHECK-NEXT: ret
@@ -1386,8 +1404,8 @@ define <vscale x 8 x i1> @intrinsic_vmsge_mask_vx_nxv8i16_i16(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmsge_mask_vx_nxv8i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a1, e16, m2, ta, mu
; CHECK-NEXT: vmslt.vx v11, v8, a0, v0.t
; CHECK-NEXT: vmxor.mm v0, v11, v10
; CHECK-NEXT: ret
@@ -1434,8 +1452,8 @@ define <vscale x 16 x i1> @intrinsic_vmsge_mask_vx_nxv16i16_i16(<vscale x 16 x i
; CHECK-LABEL: intrinsic_vmsge_mask_vx_nxv16i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a1, e16, m4, ta, mu
; CHECK-NEXT: vmslt.vx v13, v8, a0, v0.t
; CHECK-NEXT: vmxor.mm v0, v13, v12
; CHECK-NEXT: ret
@@ -1482,8 +1500,8 @@ define <vscale x 1 x i1> @intrinsic_vmsge_mask_vx_nxv1i32_i32(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmsge_mask_vx_nxv1i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e32, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e32, mf2, ta, mu
; CHECK-NEXT: vmslt.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmxor.mm v0, v10, v9
; CHECK-NEXT: ret
@@ -1530,8 +1548,8 @@ define <vscale x 2 x i1> @intrinsic_vmsge_mask_vx_nxv2i32_i32(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmsge_mask_vx_nxv2i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e32, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e32, m1, ta, mu
; CHECK-NEXT: vmslt.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmxor.mm v0, v10, v9
; CHECK-NEXT: ret
@@ -1578,8 +1596,8 @@ define <vscale x 4 x i1> @intrinsic_vmsge_mask_vx_nxv4i32_i32(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmsge_mask_vx_nxv4i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a1, e32, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a1, e32, m2, ta, mu
; CHECK-NEXT: vmslt.vx v11, v8, a0, v0.t
; CHECK-NEXT: vmxor.mm v0, v11, v10
; CHECK-NEXT: ret
@@ -1626,8 +1644,8 @@ define <vscale x 8 x i1> @intrinsic_vmsge_mask_vx_nxv8i32_i32(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmsge_mask_vx_nxv8i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a1, e32, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a1, e32, m4, ta, mu
; CHECK-NEXT: vmslt.vx v13, v8, a0, v0.t
; CHECK-NEXT: vmxor.mm v0, v13, v12
; CHECK-NEXT: ret
@@ -1701,8 +1719,8 @@ define <vscale x 1 x i1> @intrinsic_vmsge_mask_vx_nxv1i64_i64(<vscale x 1 x i1>
; RV64-LABEL: intrinsic_vmsge_mask_vx_nxv1i64_i64:
; RV64: # %bb.0: # %entry
; RV64-NEXT: vmv1r.v v10, v0
-; RV64-NEXT: vsetvli zero, a1, e64, m1, ta, mu
; RV64-NEXT: vmv1r.v v0, v9
+; RV64-NEXT: vsetvli zero, a1, e64, m1, ta, mu
; RV64-NEXT: vmslt.vx v10, v8, a0, v0.t
; RV64-NEXT: vmxor.mm v0, v10, v9
; RV64-NEXT: ret
@@ -1776,8 +1794,8 @@ define <vscale x 2 x i1> @intrinsic_vmsge_mask_vx_nxv2i64_i64(<vscale x 2 x i1>
; RV64-LABEL: intrinsic_vmsge_mask_vx_nxv2i64_i64:
; RV64: # %bb.0: # %entry
; RV64-NEXT: vmv1r.v v11, v0
-; RV64-NEXT: vsetvli zero, a1, e64, m2, ta, mu
; RV64-NEXT: vmv1r.v v0, v10
+; RV64-NEXT: vsetvli zero, a1, e64, m2, ta, mu
; RV64-NEXT: vmslt.vx v11, v8, a0, v0.t
; RV64-NEXT: vmxor.mm v0, v11, v10
; RV64-NEXT: ret
@@ -1851,8 +1869,8 @@ define <vscale x 4 x i1> @intrinsic_vmsge_mask_vx_nxv4i64_i64(<vscale x 4 x i1>
; RV64-LABEL: intrinsic_vmsge_mask_vx_nxv4i64_i64:
; RV64: # %bb.0: # %entry
; RV64-NEXT: vmv1r.v v13, v0
-; RV64-NEXT: vsetvli zero, a1, e64, m4, ta, mu
; RV64-NEXT: vmv1r.v v0, v12
+; RV64-NEXT: vsetvli zero, a1, e64, m4, ta, mu
; RV64-NEXT: vmslt.vx v13, v8, a0, v0.t
; RV64-NEXT: vmxor.mm v0, v13, v12
; RV64-NEXT: ret
@@ -1886,8 +1904,8 @@ define <vscale x 1 x i1> @intrinsic_vmsge_mask_vi_nxv1i8_i8(<vscale x 1 x i1> %0
; CHECK-LABEL: intrinsic_vmsge_mask_vi_nxv1i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, mu
; CHECK-NEXT: vmsgt.vi v10, v8, -15, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1921,8 +1939,8 @@ define <vscale x 2 x i1> @intrinsic_vmsge_mask_vi_nxv2i8_i8(<vscale x 2 x i1> %0
; CHECK-LABEL: intrinsic_vmsge_mask_vi_nxv2i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, mf4, ta, mu
; CHECK-NEXT: vmsgt.vi v10, v8, -13, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1956,8 +1974,8 @@ define <vscale x 4 x i1> @intrinsic_vmsge_mask_vi_nxv4i8_i8(<vscale x 4 x i1> %0
; CHECK-LABEL: intrinsic_vmsge_mask_vi_nxv4i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, mu
; CHECK-NEXT: vmsgt.vi v10, v8, -11, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1991,8 +2009,8 @@ define <vscale x 8 x i1> @intrinsic_vmsge_mask_vi_nxv8i8_i8(<vscale x 8 x i1> %0
; CHECK-LABEL: intrinsic_vmsge_mask_vi_nxv8i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, m1, ta, mu
; CHECK-NEXT: vmsgt.vi v10, v8, -9, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -2026,8 +2044,8 @@ define <vscale x 16 x i1> @intrinsic_vmsge_mask_vi_nxv16i8_i8(<vscale x 16 x i1>
; CHECK-LABEL: intrinsic_vmsge_mask_vi_nxv16i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e8, m2, ta, mu
; CHECK-NEXT: vmsgt.vi v11, v8, -7, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -2061,8 +2079,8 @@ define <vscale x 32 x i1> @intrinsic_vmsge_mask_vi_nxv32i8_i8(<vscale x 32 x i1>
; CHECK-LABEL: intrinsic_vmsge_mask_vi_nxv32i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, mu
; CHECK-NEXT: vmsgt.vi v13, v8, -5, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -2096,8 +2114,8 @@ define <vscale x 1 x i1> @intrinsic_vmsge_mask_vi_nxv1i16_i16(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmsge_mask_vi_nxv1i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
; CHECK-NEXT: vmsgt.vi v10, v8, -3, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -2131,8 +2149,8 @@ define <vscale x 2 x i1> @intrinsic_vmsge_mask_vi_nxv2i16_i16(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmsge_mask_vi_nxv2i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
; CHECK-NEXT: vmsgt.vi v10, v8, -1, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -2166,8 +2184,8 @@ define <vscale x 4 x i1> @intrinsic_vmsge_mask_vi_nxv4i16_i16(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmsge_mask_vi_nxv4i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
; CHECK-NEXT: vmsgt.vi v10, v8, 0, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -2201,8 +2219,8 @@ define <vscale x 8 x i1> @intrinsic_vmsge_mask_vi_nxv8i16_i16(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmsge_mask_vi_nxv8i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
; CHECK-NEXT: vmsgt.vi v11, v8, 2, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -2236,8 +2254,8 @@ define <vscale x 16 x i1> @intrinsic_vmsge_mask_vi_nxv16i16_i16(<vscale x 16 x i
; CHECK-LABEL: intrinsic_vmsge_mask_vi_nxv16i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
; CHECK-NEXT: vmsgt.vi v13, v8, 4, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -2271,8 +2289,8 @@ define <vscale x 1 x i1> @intrinsic_vmsge_mask_vi_nxv1i32_i32(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmsge_mask_vi_nxv1i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
; CHECK-NEXT: vmsgt.vi v10, v8, 6, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -2306,8 +2324,8 @@ define <vscale x 2 x i1> @intrinsic_vmsge_mask_vi_nxv2i32_i32(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmsge_mask_vi_nxv2i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
; CHECK-NEXT: vmsgt.vi v10, v8, 8, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -2341,8 +2359,8 @@ define <vscale x 4 x i1> @intrinsic_vmsge_mask_vi_nxv4i32_i32(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmsge_mask_vi_nxv4i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
; CHECK-NEXT: vmsgt.vi v11, v8, 10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -2376,8 +2394,8 @@ define <vscale x 8 x i1> @intrinsic_vmsge_mask_vi_nxv8i32_i32(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmsge_mask_vi_nxv8i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
; CHECK-NEXT: vmsgt.vi v13, v8, 12, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -2411,8 +2429,8 @@ define <vscale x 1 x i1> @intrinsic_vmsge_mask_vi_nxv1i64_i64(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmsge_mask_vi_nxv1i64_i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
; CHECK-NEXT: vmsgt.vi v10, v8, 8, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -2446,8 +2464,8 @@ define <vscale x 2 x i1> @intrinsic_vmsge_mask_vi_nxv2i64_i64(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmsge_mask_vi_nxv2i64_i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
; CHECK-NEXT: vmsgt.vi v11, v8, 8, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -2481,8 +2499,8 @@ define <vscale x 4 x i1> @intrinsic_vmsge_mask_vi_nxv4i64_i64(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmsge_mask_vi_nxv4i64_i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
; CHECK-NEXT: vmsgt.vi v13, v8, 8, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vmsgeu.ll b/llvm/test/CodeGen/RISCV/rvv/vmsgeu.ll
index b6c6d9e90f610..5568c1e9b1cfb 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vmsgeu.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vmsgeu.ll
@@ -34,9 +34,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsgeu.mask.nxv1i8(
define <vscale x 1 x i1> @intrinsic_vmsgeu_mask_vv_nxv1i8_nxv1i8(<vscale x 1 x i1> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, <vscale x 1 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgeu_mask_vv_nxv1i8_nxv1i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, mu
-; CHECK-NEXT: vmsleu.vv v0, v9, v8
+; CHECK-NEXT: vmsleu.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmsleu.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -85,9 +86,10 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsgeu.mask.nxv2i8(
define <vscale x 2 x i1> @intrinsic_vmsgeu_mask_vv_nxv2i8_nxv2i8(<vscale x 2 x i1> %0, <vscale x 2 x i8> %1, <vscale x 2 x i8> %2, <vscale x 2 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgeu_mask_vv_nxv2i8_nxv2i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e8, mf4, ta, mu
-; CHECK-NEXT: vmsleu.vv v0, v9, v8
+; CHECK-NEXT: vmsleu.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmsleu.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -136,9 +138,10 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsgeu.mask.nxv4i8(
define <vscale x 4 x i1> @intrinsic_vmsgeu_mask_vv_nxv4i8_nxv4i8(<vscale x 4 x i1> %0, <vscale x 4 x i8> %1, <vscale x 4 x i8> %2, <vscale x 4 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgeu_mask_vv_nxv4i8_nxv4i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, mu
-; CHECK-NEXT: vmsleu.vv v0, v9, v8
+; CHECK-NEXT: vmsleu.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmsleu.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -187,9 +190,10 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsgeu.mask.nxv8i8(
define <vscale x 8 x i1> @intrinsic_vmsgeu_mask_vv_nxv8i8_nxv8i8(<vscale x 8 x i1> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, <vscale x 8 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgeu_mask_vv_nxv8i8_nxv8i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e8, m1, ta, mu
-; CHECK-NEXT: vmsleu.vv v0, v9, v8
+; CHECK-NEXT: vmsleu.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmsleu.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -238,11 +242,12 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsgeu.mask.nxv16i8(
define <vscale x 16 x i1> @intrinsic_vmsgeu_mask_vv_nxv16i8_nxv16i8(<vscale x 16 x i1> %0, <vscale x 16 x i8> %1, <vscale x 16 x i8> %2, <vscale x 16 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgeu_mask_vv_nxv16i8_nxv16i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e8, m2, ta, mu
-; CHECK-NEXT: vmsleu.vv v0, v10, v8
-; CHECK-NEXT: vmsleu.vv v14, v12, v10, v0.t
+; CHECK-NEXT: vmsleu.vv v14, v10, v8
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmsleu.vv v8, v12, v10, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 16 x i1> @llvm.riscv.vmsgeu.nxv16i8(
@@ -289,11 +294,12 @@ declare <vscale x 32 x i1> @llvm.riscv.vmsgeu.mask.nxv32i8(
define <vscale x 32 x i1> @intrinsic_vmsgeu_mask_vv_nxv32i8_nxv32i8(<vscale x 32 x i1> %0, <vscale x 32 x i8> %1, <vscale x 32 x i8> %2, <vscale x 32 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgeu_mask_vv_nxv32i8_nxv32i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, mu
-; CHECK-NEXT: vmsleu.vv v0, v12, v8
-; CHECK-NEXT: vmsleu.vv v20, v16, v12, v0.t
+; CHECK-NEXT: vmsleu.vv v20, v12, v8
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmsleu.vv v8, v16, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 32 x i1> @llvm.riscv.vmsgeu.nxv32i8(
@@ -340,9 +346,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsgeu.mask.nxv1i16(
define <vscale x 1 x i1> @intrinsic_vmsgeu_mask_vv_nxv1i16_nxv1i16(<vscale x 1 x i1> %0, <vscale x 1 x i16> %1, <vscale x 1 x i16> %2, <vscale x 1 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgeu_mask_vv_nxv1i16_nxv1i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
-; CHECK-NEXT: vmsleu.vv v0, v9, v8
+; CHECK-NEXT: vmsleu.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmsleu.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -391,9 +398,10 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsgeu.mask.nxv2i16(
define <vscale x 2 x i1> @intrinsic_vmsgeu_mask_vv_nxv2i16_nxv2i16(<vscale x 2 x i1> %0, <vscale x 2 x i16> %1, <vscale x 2 x i16> %2, <vscale x 2 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgeu_mask_vv_nxv2i16_nxv2i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
-; CHECK-NEXT: vmsleu.vv v0, v9, v8
+; CHECK-NEXT: vmsleu.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmsleu.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -442,9 +450,10 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsgeu.mask.nxv4i16(
define <vscale x 4 x i1> @intrinsic_vmsgeu_mask_vv_nxv4i16_nxv4i16(<vscale x 4 x i1> %0, <vscale x 4 x i16> %1, <vscale x 4 x i16> %2, <vscale x 4 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgeu_mask_vv_nxv4i16_nxv4i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
-; CHECK-NEXT: vmsleu.vv v0, v9, v8
+; CHECK-NEXT: vmsleu.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmsleu.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -493,11 +502,12 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsgeu.mask.nxv8i16(
define <vscale x 8 x i1> @intrinsic_vmsgeu_mask_vv_nxv8i16_nxv8i16(<vscale x 8 x i1> %0, <vscale x 8 x i16> %1, <vscale x 8 x i16> %2, <vscale x 8 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgeu_mask_vv_nxv8i16_nxv8i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
-; CHECK-NEXT: vmsleu.vv v0, v10, v8
-; CHECK-NEXT: vmsleu.vv v14, v12, v10, v0.t
+; CHECK-NEXT: vmsleu.vv v14, v10, v8
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmsleu.vv v8, v12, v10, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 8 x i1> @llvm.riscv.vmsgeu.nxv8i16(
@@ -544,11 +554,12 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsgeu.mask.nxv16i16(
define <vscale x 16 x i1> @intrinsic_vmsgeu_mask_vv_nxv16i16_nxv16i16(<vscale x 16 x i1> %0, <vscale x 16 x i16> %1, <vscale x 16 x i16> %2, <vscale x 16 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgeu_mask_vv_nxv16i16_nxv16i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
-; CHECK-NEXT: vmsleu.vv v0, v12, v8
-; CHECK-NEXT: vmsleu.vv v20, v16, v12, v0.t
+; CHECK-NEXT: vmsleu.vv v20, v12, v8
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmsleu.vv v8, v16, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 16 x i1> @llvm.riscv.vmsgeu.nxv16i16(
@@ -595,9 +606,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsgeu.mask.nxv1i32(
define <vscale x 1 x i1> @intrinsic_vmsgeu_mask_vv_nxv1i32_nxv1i32(<vscale x 1 x i1> %0, <vscale x 1 x i32> %1, <vscale x 1 x i32> %2, <vscale x 1 x i32> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgeu_mask_vv_nxv1i32_nxv1i32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
-; CHECK-NEXT: vmsleu.vv v0, v9, v8
+; CHECK-NEXT: vmsleu.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmsleu.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -646,9 +658,10 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsgeu.mask.nxv2i32(
define <vscale x 2 x i1> @intrinsic_vmsgeu_mask_vv_nxv2i32_nxv2i32(<vscale x 2 x i1> %0, <vscale x 2 x i32> %1, <vscale x 2 x i32> %2, <vscale x 2 x i32> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgeu_mask_vv_nxv2i32_nxv2i32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
-; CHECK-NEXT: vmsleu.vv v0, v9, v8
+; CHECK-NEXT: vmsleu.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmsleu.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -697,11 +710,12 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsgeu.mask.nxv4i32(
define <vscale x 4 x i1> @intrinsic_vmsgeu_mask_vv_nxv4i32_nxv4i32(<vscale x 4 x i1> %0, <vscale x 4 x i32> %1, <vscale x 4 x i32> %2, <vscale x 4 x i32> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgeu_mask_vv_nxv4i32_nxv4i32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
-; CHECK-NEXT: vmsleu.vv v0, v10, v8
-; CHECK-NEXT: vmsleu.vv v14, v12, v10, v0.t
+; CHECK-NEXT: vmsleu.vv v14, v10, v8
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmsleu.vv v8, v12, v10, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 4 x i1> @llvm.riscv.vmsgeu.nxv4i32(
@@ -748,11 +762,12 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsgeu.mask.nxv8i32(
define <vscale x 8 x i1> @intrinsic_vmsgeu_mask_vv_nxv8i32_nxv8i32(<vscale x 8 x i1> %0, <vscale x 8 x i32> %1, <vscale x 8 x i32> %2, <vscale x 8 x i32> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgeu_mask_vv_nxv8i32_nxv8i32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
-; CHECK-NEXT: vmsleu.vv v0, v12, v8
-; CHECK-NEXT: vmsleu.vv v20, v16, v12, v0.t
+; CHECK-NEXT: vmsleu.vv v20, v12, v8
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmsleu.vv v8, v16, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 8 x i1> @llvm.riscv.vmsgeu.nxv8i32(
@@ -799,9 +814,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsgeu.mask.nxv1i64(
define <vscale x 1 x i1> @intrinsic_vmsgeu_mask_vv_nxv1i64_nxv1i64(<vscale x 1 x i1> %0, <vscale x 1 x i64> %1, <vscale x 1 x i64> %2, <vscale x 1 x i64> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgeu_mask_vv_nxv1i64_nxv1i64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
-; CHECK-NEXT: vmsleu.vv v0, v9, v8
+; CHECK-NEXT: vmsleu.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmsleu.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -850,11 +866,12 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsgeu.mask.nxv2i64(
define <vscale x 2 x i1> @intrinsic_vmsgeu_mask_vv_nxv2i64_nxv2i64(<vscale x 2 x i1> %0, <vscale x 2 x i64> %1, <vscale x 2 x i64> %2, <vscale x 2 x i64> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgeu_mask_vv_nxv2i64_nxv2i64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
-; CHECK-NEXT: vmsleu.vv v0, v10, v8
-; CHECK-NEXT: vmsleu.vv v14, v12, v10, v0.t
+; CHECK-NEXT: vmsleu.vv v14, v10, v8
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmsleu.vv v8, v12, v10, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 2 x i1> @llvm.riscv.vmsgeu.nxv2i64(
@@ -901,11 +918,12 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsgeu.mask.nxv4i64(
define <vscale x 4 x i1> @intrinsic_vmsgeu_mask_vv_nxv4i64_nxv4i64(<vscale x 4 x i1> %0, <vscale x 4 x i64> %1, <vscale x 4 x i64> %2, <vscale x 4 x i64> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgeu_mask_vv_nxv4i64_nxv4i64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
-; CHECK-NEXT: vmsleu.vv v0, v12, v8
-; CHECK-NEXT: vmsleu.vv v20, v16, v12, v0.t
+; CHECK-NEXT: vmsleu.vv v20, v12, v8
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmsleu.vv v8, v16, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 4 x i1> @llvm.riscv.vmsgeu.nxv4i64(
@@ -954,8 +972,8 @@ define <vscale x 1 x i1> @intrinsic_vmsgeu_mask_vx_nxv1i8_i8(<vscale x 1 x i1> %
; CHECK-LABEL: intrinsic_vmsgeu_mask_vx_nxv1i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, mu
; CHECK-NEXT: vmsltu.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmxor.mm v0, v10, v9
; CHECK-NEXT: ret
@@ -1002,8 +1020,8 @@ define <vscale x 2 x i1> @intrinsic_vmsgeu_mask_vx_nxv2i8_i8(<vscale x 2 x i1> %
; CHECK-LABEL: intrinsic_vmsgeu_mask_vx_nxv2i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e8, mf4, ta, mu
; CHECK-NEXT: vmsltu.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmxor.mm v0, v10, v9
; CHECK-NEXT: ret
@@ -1050,8 +1068,8 @@ define <vscale x 4 x i1> @intrinsic_vmsgeu_mask_vx_nxv4i8_i8(<vscale x 4 x i1> %
; CHECK-LABEL: intrinsic_vmsgeu_mask_vx_nxv4i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e8, mf2, ta, mu
; CHECK-NEXT: vmsltu.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmxor.mm v0, v10, v9
; CHECK-NEXT: ret
@@ -1098,8 +1116,8 @@ define <vscale x 8 x i1> @intrinsic_vmsgeu_mask_vx_nxv8i8_i8(<vscale x 8 x i1> %
; CHECK-LABEL: intrinsic_vmsgeu_mask_vx_nxv8i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e8, m1, ta, mu
; CHECK-NEXT: vmsltu.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmxor.mm v0, v10, v9
; CHECK-NEXT: ret
@@ -1146,8 +1164,8 @@ define <vscale x 16 x i1> @intrinsic_vmsgeu_mask_vx_nxv16i8_i8(<vscale x 16 x i1
; CHECK-LABEL: intrinsic_vmsgeu_mask_vx_nxv16i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a1, e8, m2, ta, mu
; CHECK-NEXT: vmsltu.vx v11, v8, a0, v0.t
; CHECK-NEXT: vmxor.mm v0, v11, v10
; CHECK-NEXT: ret
@@ -1194,8 +1212,8 @@ define <vscale x 32 x i1> @intrinsic_vmsgeu_mask_vx_nxv32i8_i8(<vscale x 32 x i1
; CHECK-LABEL: intrinsic_vmsgeu_mask_vx_nxv32i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a1, e8, m4, ta, mu
; CHECK-NEXT: vmsltu.vx v13, v8, a0, v0.t
; CHECK-NEXT: vmxor.mm v0, v13, v12
; CHECK-NEXT: ret
@@ -1242,8 +1260,8 @@ define <vscale x 1 x i1> @intrinsic_vmsgeu_mask_vx_nxv1i16_i16(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmsgeu_mask_vx_nxv1i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, mf4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e16, mf4, ta, mu
; CHECK-NEXT: vmsltu.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmxor.mm v0, v10, v9
; CHECK-NEXT: ret
@@ -1290,8 +1308,8 @@ define <vscale x 2 x i1> @intrinsic_vmsgeu_mask_vx_nxv2i16_i16(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmsgeu_mask_vx_nxv2i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e16, mf2, ta, mu
; CHECK-NEXT: vmsltu.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmxor.mm v0, v10, v9
; CHECK-NEXT: ret
@@ -1338,8 +1356,8 @@ define <vscale x 4 x i1> @intrinsic_vmsgeu_mask_vx_nxv4i16_i16(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmsgeu_mask_vx_nxv4i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e16, m1, ta, mu
; CHECK-NEXT: vmsltu.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmxor.mm v0, v10, v9
; CHECK-NEXT: ret
@@ -1386,8 +1404,8 @@ define <vscale x 8 x i1> @intrinsic_vmsgeu_mask_vx_nxv8i16_i16(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmsgeu_mask_vx_nxv8i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a1, e16, m2, ta, mu
; CHECK-NEXT: vmsltu.vx v11, v8, a0, v0.t
; CHECK-NEXT: vmxor.mm v0, v11, v10
; CHECK-NEXT: ret
@@ -1434,8 +1452,8 @@ define <vscale x 16 x i1> @intrinsic_vmsgeu_mask_vx_nxv16i16_i16(<vscale x 16 x
; CHECK-LABEL: intrinsic_vmsgeu_mask_vx_nxv16i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a1, e16, m4, ta, mu
; CHECK-NEXT: vmsltu.vx v13, v8, a0, v0.t
; CHECK-NEXT: vmxor.mm v0, v13, v12
; CHECK-NEXT: ret
@@ -1482,8 +1500,8 @@ define <vscale x 1 x i1> @intrinsic_vmsgeu_mask_vx_nxv1i32_i32(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmsgeu_mask_vx_nxv1i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e32, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e32, mf2, ta, mu
; CHECK-NEXT: vmsltu.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmxor.mm v0, v10, v9
; CHECK-NEXT: ret
@@ -1530,8 +1548,8 @@ define <vscale x 2 x i1> @intrinsic_vmsgeu_mask_vx_nxv2i32_i32(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmsgeu_mask_vx_nxv2i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e32, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e32, m1, ta, mu
; CHECK-NEXT: vmsltu.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmxor.mm v0, v10, v9
; CHECK-NEXT: ret
@@ -1578,8 +1596,8 @@ define <vscale x 4 x i1> @intrinsic_vmsgeu_mask_vx_nxv4i32_i32(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmsgeu_mask_vx_nxv4i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a1, e32, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a1, e32, m2, ta, mu
; CHECK-NEXT: vmsltu.vx v11, v8, a0, v0.t
; CHECK-NEXT: vmxor.mm v0, v11, v10
; CHECK-NEXT: ret
@@ -1626,8 +1644,8 @@ define <vscale x 8 x i1> @intrinsic_vmsgeu_mask_vx_nxv8i32_i32(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmsgeu_mask_vx_nxv8i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a1, e32, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a1, e32, m4, ta, mu
; CHECK-NEXT: vmsltu.vx v13, v8, a0, v0.t
; CHECK-NEXT: vmxor.mm v0, v13, v12
; CHECK-NEXT: ret
@@ -1701,8 +1719,8 @@ define <vscale x 1 x i1> @intrinsic_vmsgeu_mask_vx_nxv1i64_i64(<vscale x 1 x i1>
; RV64-LABEL: intrinsic_vmsgeu_mask_vx_nxv1i64_i64:
; RV64: # %bb.0: # %entry
; RV64-NEXT: vmv1r.v v10, v0
-; RV64-NEXT: vsetvli zero, a1, e64, m1, ta, mu
; RV64-NEXT: vmv1r.v v0, v9
+; RV64-NEXT: vsetvli zero, a1, e64, m1, ta, mu
; RV64-NEXT: vmsltu.vx v10, v8, a0, v0.t
; RV64-NEXT: vmxor.mm v0, v10, v9
; RV64-NEXT: ret
@@ -1776,8 +1794,8 @@ define <vscale x 2 x i1> @intrinsic_vmsgeu_mask_vx_nxv2i64_i64(<vscale x 2 x i1>
; RV64-LABEL: intrinsic_vmsgeu_mask_vx_nxv2i64_i64:
; RV64: # %bb.0: # %entry
; RV64-NEXT: vmv1r.v v11, v0
-; RV64-NEXT: vsetvli zero, a1, e64, m2, ta, mu
; RV64-NEXT: vmv1r.v v0, v10
+; RV64-NEXT: vsetvli zero, a1, e64, m2, ta, mu
; RV64-NEXT: vmsltu.vx v11, v8, a0, v0.t
; RV64-NEXT: vmxor.mm v0, v11, v10
; RV64-NEXT: ret
@@ -1851,8 +1869,8 @@ define <vscale x 4 x i1> @intrinsic_vmsgeu_mask_vx_nxv4i64_i64(<vscale x 4 x i1>
; RV64-LABEL: intrinsic_vmsgeu_mask_vx_nxv4i64_i64:
; RV64: # %bb.0: # %entry
; RV64-NEXT: vmv1r.v v13, v0
-; RV64-NEXT: vsetvli zero, a1, e64, m4, ta, mu
; RV64-NEXT: vmv1r.v v0, v12
+; RV64-NEXT: vsetvli zero, a1, e64, m4, ta, mu
; RV64-NEXT: vmsltu.vx v13, v8, a0, v0.t
; RV64-NEXT: vmxor.mm v0, v13, v12
; RV64-NEXT: ret
@@ -1886,8 +1904,8 @@ define <vscale x 1 x i1> @intrinsic_vmsgeu_mask_vi_nxv1i8_i8(<vscale x 1 x i1> %
; CHECK-LABEL: intrinsic_vmsgeu_mask_vi_nxv1i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, mu
; CHECK-NEXT: vmsgtu.vi v10, v8, -15, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1921,8 +1939,8 @@ define <vscale x 2 x i1> @intrinsic_vmsgeu_mask_vi_nxv2i8_i8(<vscale x 2 x i1> %
; CHECK-LABEL: intrinsic_vmsgeu_mask_vi_nxv2i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, mf4, ta, mu
; CHECK-NEXT: vmsgtu.vi v10, v8, -13, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1956,8 +1974,8 @@ define <vscale x 4 x i1> @intrinsic_vmsgeu_mask_vi_nxv4i8_i8(<vscale x 4 x i1> %
; CHECK-LABEL: intrinsic_vmsgeu_mask_vi_nxv4i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, mu
; CHECK-NEXT: vmsgtu.vi v10, v8, -11, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1991,8 +2009,8 @@ define <vscale x 8 x i1> @intrinsic_vmsgeu_mask_vi_nxv8i8_i8(<vscale x 8 x i1> %
; CHECK-LABEL: intrinsic_vmsgeu_mask_vi_nxv8i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, m1, ta, mu
; CHECK-NEXT: vmsgtu.vi v10, v8, -9, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -2026,8 +2044,8 @@ define <vscale x 16 x i1> @intrinsic_vmsgeu_mask_vi_nxv16i8_i8(<vscale x 16 x i1
; CHECK-LABEL: intrinsic_vmsgeu_mask_vi_nxv16i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e8, m2, ta, mu
; CHECK-NEXT: vmsgtu.vi v11, v8, -7, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -2061,8 +2079,8 @@ define <vscale x 32 x i1> @intrinsic_vmsgeu_mask_vi_nxv32i8_i8(<vscale x 32 x i1
; CHECK-LABEL: intrinsic_vmsgeu_mask_vi_nxv32i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, mu
; CHECK-NEXT: vmsgtu.vi v13, v8, -5, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -2096,8 +2114,8 @@ define <vscale x 1 x i1> @intrinsic_vmsgeu_mask_vi_nxv1i16_i16(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmsgeu_mask_vi_nxv1i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
; CHECK-NEXT: vmsgtu.vi v10, v8, -3, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -2178,8 +2196,8 @@ define <vscale x 4 x i1> @intrinsic_vmsgeu_mask_vi_nxv4i16_i16(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmsgeu_mask_vi_nxv4i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
; CHECK-NEXT: vmsgtu.vi v10, v8, 0, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -2213,8 +2231,8 @@ define <vscale x 8 x i1> @intrinsic_vmsgeu_mask_vi_nxv8i16_i16(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmsgeu_mask_vi_nxv8i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
; CHECK-NEXT: vmsgtu.vi v11, v8, 2, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -2248,8 +2266,8 @@ define <vscale x 16 x i1> @intrinsic_vmsgeu_mask_vi_nxv16i16_i16(<vscale x 16 x
; CHECK-LABEL: intrinsic_vmsgeu_mask_vi_nxv16i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
; CHECK-NEXT: vmsgtu.vi v13, v8, 4, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -2283,8 +2301,8 @@ define <vscale x 1 x i1> @intrinsic_vmsgeu_mask_vi_nxv1i32_i32(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmsgeu_mask_vi_nxv1i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
; CHECK-NEXT: vmsgtu.vi v10, v8, 6, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -2318,8 +2336,8 @@ define <vscale x 2 x i1> @intrinsic_vmsgeu_mask_vi_nxv2i32_i32(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmsgeu_mask_vi_nxv2i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
; CHECK-NEXT: vmsgtu.vi v10, v8, 8, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -2353,8 +2371,8 @@ define <vscale x 4 x i1> @intrinsic_vmsgeu_mask_vi_nxv4i32_i32(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmsgeu_mask_vi_nxv4i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
; CHECK-NEXT: vmsgtu.vi v11, v8, 10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -2388,8 +2406,8 @@ define <vscale x 8 x i1> @intrinsic_vmsgeu_mask_vi_nxv8i32_i32(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmsgeu_mask_vi_nxv8i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
; CHECK-NEXT: vmsgtu.vi v13, v8, 12, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -2423,8 +2441,8 @@ define <vscale x 1 x i1> @intrinsic_vmsgeu_mask_vi_nxv1i64_i64(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmsgeu_mask_vi_nxv1i64_i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
; CHECK-NEXT: vmsgtu.vi v10, v8, 14, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -2458,8 +2476,8 @@ define <vscale x 2 x i1> @intrinsic_vmsgeu_mask_vi_nxv2i64_i64(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmsgeu_mask_vi_nxv2i64_i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
; CHECK-NEXT: vmsgtu.vi v11, v8, -16, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -2493,8 +2511,8 @@ define <vscale x 4 x i1> @intrinsic_vmsgeu_mask_vi_nxv4i64_i64(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmsgeu_mask_vi_nxv4i64_i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
; CHECK-NEXT: vmsgtu.vi v13, v8, -14, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vmsgt.ll b/llvm/test/CodeGen/RISCV/rvv/vmsgt.ll
index dfd7096a65ebb..f1fa6484d976b 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vmsgt.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vmsgt.ll
@@ -34,9 +34,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsgt.mask.nxv1i8(
define <vscale x 1 x i1> @intrinsic_vmsgt_mask_vv_nxv1i8_nxv1i8(<vscale x 1 x i1> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, <vscale x 1 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgt_mask_vv_nxv1i8_nxv1i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, mu
-; CHECK-NEXT: vmslt.vv v0, v9, v8
+; CHECK-NEXT: vmslt.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmslt.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -85,9 +86,10 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsgt.mask.nxv2i8(
define <vscale x 2 x i1> @intrinsic_vmsgt_mask_vv_nxv2i8_nxv2i8(<vscale x 2 x i1> %0, <vscale x 2 x i8> %1, <vscale x 2 x i8> %2, <vscale x 2 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgt_mask_vv_nxv2i8_nxv2i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e8, mf4, ta, mu
-; CHECK-NEXT: vmslt.vv v0, v9, v8
+; CHECK-NEXT: vmslt.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmslt.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -136,9 +138,10 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsgt.mask.nxv4i8(
define <vscale x 4 x i1> @intrinsic_vmsgt_mask_vv_nxv4i8_nxv4i8(<vscale x 4 x i1> %0, <vscale x 4 x i8> %1, <vscale x 4 x i8> %2, <vscale x 4 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgt_mask_vv_nxv4i8_nxv4i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, mu
-; CHECK-NEXT: vmslt.vv v0, v9, v8
+; CHECK-NEXT: vmslt.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmslt.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -187,9 +190,10 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsgt.mask.nxv8i8(
define <vscale x 8 x i1> @intrinsic_vmsgt_mask_vv_nxv8i8_nxv8i8(<vscale x 8 x i1> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, <vscale x 8 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgt_mask_vv_nxv8i8_nxv8i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e8, m1, ta, mu
-; CHECK-NEXT: vmslt.vv v0, v9, v8
+; CHECK-NEXT: vmslt.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmslt.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -238,11 +242,12 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsgt.mask.nxv16i8(
define <vscale x 16 x i1> @intrinsic_vmsgt_mask_vv_nxv16i8_nxv16i8(<vscale x 16 x i1> %0, <vscale x 16 x i8> %1, <vscale x 16 x i8> %2, <vscale x 16 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgt_mask_vv_nxv16i8_nxv16i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e8, m2, ta, mu
-; CHECK-NEXT: vmslt.vv v0, v10, v8
-; CHECK-NEXT: vmslt.vv v14, v12, v10, v0.t
+; CHECK-NEXT: vmslt.vv v14, v10, v8
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmslt.vv v8, v12, v10, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 16 x i1> @llvm.riscv.vmsgt.nxv16i8(
@@ -289,11 +294,12 @@ declare <vscale x 32 x i1> @llvm.riscv.vmsgt.mask.nxv32i8(
define <vscale x 32 x i1> @intrinsic_vmsgt_mask_vv_nxv32i8_nxv32i8(<vscale x 32 x i1> %0, <vscale x 32 x i8> %1, <vscale x 32 x i8> %2, <vscale x 32 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgt_mask_vv_nxv32i8_nxv32i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, mu
-; CHECK-NEXT: vmslt.vv v0, v12, v8
-; CHECK-NEXT: vmslt.vv v20, v16, v12, v0.t
+; CHECK-NEXT: vmslt.vv v20, v12, v8
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmslt.vv v8, v16, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 32 x i1> @llvm.riscv.vmsgt.nxv32i8(
@@ -340,9 +346,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsgt.mask.nxv1i16(
define <vscale x 1 x i1> @intrinsic_vmsgt_mask_vv_nxv1i16_nxv1i16(<vscale x 1 x i1> %0, <vscale x 1 x i16> %1, <vscale x 1 x i16> %2, <vscale x 1 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgt_mask_vv_nxv1i16_nxv1i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
-; CHECK-NEXT: vmslt.vv v0, v9, v8
+; CHECK-NEXT: vmslt.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmslt.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -391,9 +398,10 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsgt.mask.nxv2i16(
define <vscale x 2 x i1> @intrinsic_vmsgt_mask_vv_nxv2i16_nxv2i16(<vscale x 2 x i1> %0, <vscale x 2 x i16> %1, <vscale x 2 x i16> %2, <vscale x 2 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgt_mask_vv_nxv2i16_nxv2i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
-; CHECK-NEXT: vmslt.vv v0, v9, v8
+; CHECK-NEXT: vmslt.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmslt.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -442,9 +450,10 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsgt.mask.nxv4i16(
define <vscale x 4 x i1> @intrinsic_vmsgt_mask_vv_nxv4i16_nxv4i16(<vscale x 4 x i1> %0, <vscale x 4 x i16> %1, <vscale x 4 x i16> %2, <vscale x 4 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgt_mask_vv_nxv4i16_nxv4i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
-; CHECK-NEXT: vmslt.vv v0, v9, v8
+; CHECK-NEXT: vmslt.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmslt.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -493,11 +502,12 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsgt.mask.nxv8i16(
define <vscale x 8 x i1> @intrinsic_vmsgt_mask_vv_nxv8i16_nxv8i16(<vscale x 8 x i1> %0, <vscale x 8 x i16> %1, <vscale x 8 x i16> %2, <vscale x 8 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgt_mask_vv_nxv8i16_nxv8i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
-; CHECK-NEXT: vmslt.vv v0, v10, v8
-; CHECK-NEXT: vmslt.vv v14, v12, v10, v0.t
+; CHECK-NEXT: vmslt.vv v14, v10, v8
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmslt.vv v8, v12, v10, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 8 x i1> @llvm.riscv.vmsgt.nxv8i16(
@@ -544,11 +554,12 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsgt.mask.nxv16i16(
define <vscale x 16 x i1> @intrinsic_vmsgt_mask_vv_nxv16i16_nxv16i16(<vscale x 16 x i1> %0, <vscale x 16 x i16> %1, <vscale x 16 x i16> %2, <vscale x 16 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgt_mask_vv_nxv16i16_nxv16i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
-; CHECK-NEXT: vmslt.vv v0, v12, v8
-; CHECK-NEXT: vmslt.vv v20, v16, v12, v0.t
+; CHECK-NEXT: vmslt.vv v20, v12, v8
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmslt.vv v8, v16, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 16 x i1> @llvm.riscv.vmsgt.nxv16i16(
@@ -595,9 +606,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsgt.mask.nxv1i32(
define <vscale x 1 x i1> @intrinsic_vmsgt_mask_vv_nxv1i32_nxv1i32(<vscale x 1 x i1> %0, <vscale x 1 x i32> %1, <vscale x 1 x i32> %2, <vscale x 1 x i32> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgt_mask_vv_nxv1i32_nxv1i32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
-; CHECK-NEXT: vmslt.vv v0, v9, v8
+; CHECK-NEXT: vmslt.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmslt.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -646,9 +658,10 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsgt.mask.nxv2i32(
define <vscale x 2 x i1> @intrinsic_vmsgt_mask_vv_nxv2i32_nxv2i32(<vscale x 2 x i1> %0, <vscale x 2 x i32> %1, <vscale x 2 x i32> %2, <vscale x 2 x i32> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgt_mask_vv_nxv2i32_nxv2i32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
-; CHECK-NEXT: vmslt.vv v0, v9, v8
+; CHECK-NEXT: vmslt.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmslt.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -697,11 +710,12 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsgt.mask.nxv4i32(
define <vscale x 4 x i1> @intrinsic_vmsgt_mask_vv_nxv4i32_nxv4i32(<vscale x 4 x i1> %0, <vscale x 4 x i32> %1, <vscale x 4 x i32> %2, <vscale x 4 x i32> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgt_mask_vv_nxv4i32_nxv4i32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
-; CHECK-NEXT: vmslt.vv v0, v10, v8
-; CHECK-NEXT: vmslt.vv v14, v12, v10, v0.t
+; CHECK-NEXT: vmslt.vv v14, v10, v8
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmslt.vv v8, v12, v10, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 4 x i1> @llvm.riscv.vmsgt.nxv4i32(
@@ -748,11 +762,12 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsgt.mask.nxv8i32(
define <vscale x 8 x i1> @intrinsic_vmsgt_mask_vv_nxv8i32_nxv8i32(<vscale x 8 x i1> %0, <vscale x 8 x i32> %1, <vscale x 8 x i32> %2, <vscale x 8 x i32> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgt_mask_vv_nxv8i32_nxv8i32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
-; CHECK-NEXT: vmslt.vv v0, v12, v8
-; CHECK-NEXT: vmslt.vv v20, v16, v12, v0.t
+; CHECK-NEXT: vmslt.vv v20, v12, v8
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmslt.vv v8, v16, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 8 x i1> @llvm.riscv.vmsgt.nxv8i32(
@@ -799,9 +814,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsgt.mask.nxv1i64(
define <vscale x 1 x i1> @intrinsic_vmsgt_mask_vv_nxv1i64_nxv1i64(<vscale x 1 x i1> %0, <vscale x 1 x i64> %1, <vscale x 1 x i64> %2, <vscale x 1 x i64> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgt_mask_vv_nxv1i64_nxv1i64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
-; CHECK-NEXT: vmslt.vv v0, v9, v8
+; CHECK-NEXT: vmslt.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmslt.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -850,11 +866,12 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsgt.mask.nxv2i64(
define <vscale x 2 x i1> @intrinsic_vmsgt_mask_vv_nxv2i64_nxv2i64(<vscale x 2 x i1> %0, <vscale x 2 x i64> %1, <vscale x 2 x i64> %2, <vscale x 2 x i64> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgt_mask_vv_nxv2i64_nxv2i64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
-; CHECK-NEXT: vmslt.vv v0, v10, v8
-; CHECK-NEXT: vmslt.vv v14, v12, v10, v0.t
+; CHECK-NEXT: vmslt.vv v14, v10, v8
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmslt.vv v8, v12, v10, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 2 x i1> @llvm.riscv.vmsgt.nxv2i64(
@@ -901,11 +918,12 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsgt.mask.nxv4i64(
define <vscale x 4 x i1> @intrinsic_vmsgt_mask_vv_nxv4i64_nxv4i64(<vscale x 4 x i1> %0, <vscale x 4 x i64> %1, <vscale x 4 x i64> %2, <vscale x 4 x i64> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgt_mask_vv_nxv4i64_nxv4i64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
-; CHECK-NEXT: vmslt.vv v0, v12, v8
-; CHECK-NEXT: vmslt.vv v20, v16, v12, v0.t
+; CHECK-NEXT: vmslt.vv v20, v12, v8
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmslt.vv v8, v16, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 4 x i1> @llvm.riscv.vmsgt.nxv4i64(
@@ -953,8 +971,8 @@ define <vscale x 1 x i1> @intrinsic_vmsgt_mask_vx_nxv1i8_i8(<vscale x 1 x i1> %0
; CHECK-LABEL: intrinsic_vmsgt_mask_vx_nxv1i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, mu
; CHECK-NEXT: vmsgt.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1000,8 +1018,8 @@ define <vscale x 2 x i1> @intrinsic_vmsgt_mask_vx_nxv2i8_i8(<vscale x 2 x i1> %0
; CHECK-LABEL: intrinsic_vmsgt_mask_vx_nxv2i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e8, mf4, ta, mu
; CHECK-NEXT: vmsgt.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1047,8 +1065,8 @@ define <vscale x 4 x i1> @intrinsic_vmsgt_mask_vx_nxv4i8_i8(<vscale x 4 x i1> %0
; CHECK-LABEL: intrinsic_vmsgt_mask_vx_nxv4i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e8, mf2, ta, mu
; CHECK-NEXT: vmsgt.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1094,8 +1112,8 @@ define <vscale x 8 x i1> @intrinsic_vmsgt_mask_vx_nxv8i8_i8(<vscale x 8 x i1> %0
; CHECK-LABEL: intrinsic_vmsgt_mask_vx_nxv8i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e8, m1, ta, mu
; CHECK-NEXT: vmsgt.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -1141,8 +1159,8 @@ define <vscale x 16 x i1> @intrinsic_vmsgt_mask_vx_nxv16i8_i8(<vscale x 16 x i1>
; CHECK-LABEL: intrinsic_vmsgt_mask_vx_nxv16i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a1, e8, m2, ta, mu
; CHECK-NEXT: vmsgt.vx v11, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -1188,8 +1206,8 @@ define <vscale x 32 x i1> @intrinsic_vmsgt_mask_vx_nxv32i8_i8(<vscale x 32 x i1>
; CHECK-LABEL: intrinsic_vmsgt_mask_vx_nxv32i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a1, e8, m4, ta, mu
; CHECK-NEXT: vmsgt.vx v13, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -1235,8 +1253,8 @@ define <vscale x 1 x i1> @intrinsic_vmsgt_mask_vx_nxv1i16_i16(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmsgt_mask_vx_nxv1i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, mf4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e16, mf4, ta, mu
; CHECK-NEXT: vmsgt.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1282,8 +1300,8 @@ define <vscale x 2 x i1> @intrinsic_vmsgt_mask_vx_nxv2i16_i16(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmsgt_mask_vx_nxv2i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e16, mf2, ta, mu
; CHECK-NEXT: vmsgt.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1329,8 +1347,8 @@ define <vscale x 4 x i1> @intrinsic_vmsgt_mask_vx_nxv4i16_i16(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmsgt_mask_vx_nxv4i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e16, m1, ta, mu
; CHECK-NEXT: vmsgt.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -1376,8 +1394,8 @@ define <vscale x 8 x i1> @intrinsic_vmsgt_mask_vx_nxv8i16_i16(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmsgt_mask_vx_nxv8i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a1, e16, m2, ta, mu
; CHECK-NEXT: vmsgt.vx v11, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -1423,8 +1441,8 @@ define <vscale x 16 x i1> @intrinsic_vmsgt_mask_vx_nxv16i16_i16(<vscale x 16 x i
; CHECK-LABEL: intrinsic_vmsgt_mask_vx_nxv16i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a1, e16, m4, ta, mu
; CHECK-NEXT: vmsgt.vx v13, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -1470,8 +1488,8 @@ define <vscale x 1 x i1> @intrinsic_vmsgt_mask_vx_nxv1i32_i32(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmsgt_mask_vx_nxv1i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e32, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e32, mf2, ta, mu
; CHECK-NEXT: vmsgt.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1517,8 +1535,8 @@ define <vscale x 2 x i1> @intrinsic_vmsgt_mask_vx_nxv2i32_i32(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmsgt_mask_vx_nxv2i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e32, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e32, m1, ta, mu
; CHECK-NEXT: vmsgt.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -1564,8 +1582,8 @@ define <vscale x 4 x i1> @intrinsic_vmsgt_mask_vx_nxv4i32_i32(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmsgt_mask_vx_nxv4i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a1, e32, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a1, e32, m2, ta, mu
; CHECK-NEXT: vmsgt.vx v11, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -1611,8 +1629,8 @@ define <vscale x 8 x i1> @intrinsic_vmsgt_mask_vx_nxv8i32_i32(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmsgt_mask_vx_nxv8i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a1, e32, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a1, e32, m4, ta, mu
; CHECK-NEXT: vmsgt.vx v13, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -1685,8 +1703,8 @@ define <vscale x 1 x i1> @intrinsic_vmsgt_mask_vx_nxv1i64_i64(<vscale x 1 x i1>
; RV64-LABEL: intrinsic_vmsgt_mask_vx_nxv1i64_i64:
; RV64: # %bb.0: # %entry
; RV64-NEXT: vmv1r.v v10, v0
-; RV64-NEXT: vsetvli zero, a1, e64, m1, ta, mu
; RV64-NEXT: vmv1r.v v0, v9
+; RV64-NEXT: vsetvli zero, a1, e64, m1, ta, mu
; RV64-NEXT: vmsgt.vx v10, v8, a0, v0.t
; RV64-NEXT: vmv.v.v v0, v10
; RV64-NEXT: ret
@@ -1759,8 +1777,8 @@ define <vscale x 2 x i1> @intrinsic_vmsgt_mask_vx_nxv2i64_i64(<vscale x 2 x i1>
; RV64-LABEL: intrinsic_vmsgt_mask_vx_nxv2i64_i64:
; RV64: # %bb.0: # %entry
; RV64-NEXT: vmv1r.v v11, v0
-; RV64-NEXT: vsetvli zero, a1, e64, m2, ta, mu
; RV64-NEXT: vmv1r.v v0, v10
+; RV64-NEXT: vsetvli zero, a1, e64, m2, ta, mu
; RV64-NEXT: vmsgt.vx v11, v8, a0, v0.t
; RV64-NEXT: vmv1r.v v0, v11
; RV64-NEXT: ret
@@ -1833,8 +1851,8 @@ define <vscale x 4 x i1> @intrinsic_vmsgt_mask_vx_nxv4i64_i64(<vscale x 4 x i1>
; RV64-LABEL: intrinsic_vmsgt_mask_vx_nxv4i64_i64:
; RV64: # %bb.0: # %entry
; RV64-NEXT: vmv1r.v v13, v0
-; RV64-NEXT: vsetvli zero, a1, e64, m4, ta, mu
; RV64-NEXT: vmv1r.v v0, v12
+; RV64-NEXT: vsetvli zero, a1, e64, m4, ta, mu
; RV64-NEXT: vmsgt.vx v13, v8, a0, v0.t
; RV64-NEXT: vmv1r.v v0, v13
; RV64-NEXT: ret
@@ -1868,8 +1886,8 @@ define <vscale x 1 x i1> @intrinsic_vmsgt_mask_vi_nxv1i8_i8(<vscale x 1 x i1> %0
; CHECK-LABEL: intrinsic_vmsgt_mask_vi_nxv1i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, mu
; CHECK-NEXT: vmsgt.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1903,8 +1921,8 @@ define <vscale x 2 x i1> @intrinsic_vmsgt_mask_vi_nxv2i8_i8(<vscale x 2 x i1> %0
; CHECK-LABEL: intrinsic_vmsgt_mask_vi_nxv2i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, mf4, ta, mu
; CHECK-NEXT: vmsgt.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1938,8 +1956,8 @@ define <vscale x 4 x i1> @intrinsic_vmsgt_mask_vi_nxv4i8_i8(<vscale x 4 x i1> %0
; CHECK-LABEL: intrinsic_vmsgt_mask_vi_nxv4i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, mu
; CHECK-NEXT: vmsgt.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1973,8 +1991,8 @@ define <vscale x 8 x i1> @intrinsic_vmsgt_mask_vi_nxv8i8_i8(<vscale x 8 x i1> %0
; CHECK-LABEL: intrinsic_vmsgt_mask_vi_nxv8i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, m1, ta, mu
; CHECK-NEXT: vmsgt.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -2008,8 +2026,8 @@ define <vscale x 16 x i1> @intrinsic_vmsgt_mask_vi_nxv16i8_i8(<vscale x 16 x i1>
; CHECK-LABEL: intrinsic_vmsgt_mask_vi_nxv16i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e8, m2, ta, mu
; CHECK-NEXT: vmsgt.vi v11, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -2043,8 +2061,8 @@ define <vscale x 32 x i1> @intrinsic_vmsgt_mask_vi_nxv32i8_i8(<vscale x 32 x i1>
; CHECK-LABEL: intrinsic_vmsgt_mask_vi_nxv32i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, mu
; CHECK-NEXT: vmsgt.vi v13, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -2078,8 +2096,8 @@ define <vscale x 1 x i1> @intrinsic_vmsgt_mask_vi_nxv1i16_i16(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmsgt_mask_vi_nxv1i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
; CHECK-NEXT: vmsgt.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -2113,8 +2131,8 @@ define <vscale x 2 x i1> @intrinsic_vmsgt_mask_vi_nxv2i16_i16(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmsgt_mask_vi_nxv2i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
; CHECK-NEXT: vmsgt.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -2148,8 +2166,8 @@ define <vscale x 4 x i1> @intrinsic_vmsgt_mask_vi_nxv4i16_i16(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmsgt_mask_vi_nxv4i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
; CHECK-NEXT: vmsgt.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -2183,8 +2201,8 @@ define <vscale x 8 x i1> @intrinsic_vmsgt_mask_vi_nxv8i16_i16(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmsgt_mask_vi_nxv8i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
; CHECK-NEXT: vmsgt.vi v11, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -2218,8 +2236,8 @@ define <vscale x 16 x i1> @intrinsic_vmsgt_mask_vi_nxv16i16_i16(<vscale x 16 x i
; CHECK-LABEL: intrinsic_vmsgt_mask_vi_nxv16i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
; CHECK-NEXT: vmsgt.vi v13, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -2253,8 +2271,8 @@ define <vscale x 1 x i1> @intrinsic_vmsgt_mask_vi_nxv1i32_i32(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmsgt_mask_vi_nxv1i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
; CHECK-NEXT: vmsgt.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -2288,8 +2306,8 @@ define <vscale x 2 x i1> @intrinsic_vmsgt_mask_vi_nxv2i32_i32(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmsgt_mask_vi_nxv2i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
; CHECK-NEXT: vmsgt.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -2323,8 +2341,8 @@ define <vscale x 4 x i1> @intrinsic_vmsgt_mask_vi_nxv4i32_i32(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmsgt_mask_vi_nxv4i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
; CHECK-NEXT: vmsgt.vi v11, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -2358,8 +2376,8 @@ define <vscale x 8 x i1> @intrinsic_vmsgt_mask_vi_nxv8i32_i32(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmsgt_mask_vi_nxv8i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
; CHECK-NEXT: vmsgt.vi v13, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -2393,8 +2411,8 @@ define <vscale x 1 x i1> @intrinsic_vmsgt_mask_vi_nxv1i64_i64(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmsgt_mask_vi_nxv1i64_i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
; CHECK-NEXT: vmsgt.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -2428,8 +2446,8 @@ define <vscale x 2 x i1> @intrinsic_vmsgt_mask_vi_nxv2i64_i64(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmsgt_mask_vi_nxv2i64_i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
; CHECK-NEXT: vmsgt.vi v11, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -2463,8 +2481,8 @@ define <vscale x 4 x i1> @intrinsic_vmsgt_mask_vi_nxv4i64_i64(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmsgt_mask_vi_nxv4i64_i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
; CHECK-NEXT: vmsgt.vi v13, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vmsgtu.ll b/llvm/test/CodeGen/RISCV/rvv/vmsgtu.ll
index 8826be03bbebb..de7a0ad87be27 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vmsgtu.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vmsgtu.ll
@@ -34,9 +34,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsgtu.mask.nxv1i8(
define <vscale x 1 x i1> @intrinsic_vmsgtu_mask_vv_nxv1i8_nxv1i8(<vscale x 1 x i1> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, <vscale x 1 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgtu_mask_vv_nxv1i8_nxv1i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, mu
-; CHECK-NEXT: vmsltu.vv v0, v9, v8
+; CHECK-NEXT: vmsltu.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmsltu.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -85,9 +86,10 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsgtu.mask.nxv2i8(
define <vscale x 2 x i1> @intrinsic_vmsgtu_mask_vv_nxv2i8_nxv2i8(<vscale x 2 x i1> %0, <vscale x 2 x i8> %1, <vscale x 2 x i8> %2, <vscale x 2 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgtu_mask_vv_nxv2i8_nxv2i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e8, mf4, ta, mu
-; CHECK-NEXT: vmsltu.vv v0, v9, v8
+; CHECK-NEXT: vmsltu.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmsltu.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -136,9 +138,10 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsgtu.mask.nxv4i8(
define <vscale x 4 x i1> @intrinsic_vmsgtu_mask_vv_nxv4i8_nxv4i8(<vscale x 4 x i1> %0, <vscale x 4 x i8> %1, <vscale x 4 x i8> %2, <vscale x 4 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgtu_mask_vv_nxv4i8_nxv4i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, mu
-; CHECK-NEXT: vmsltu.vv v0, v9, v8
+; CHECK-NEXT: vmsltu.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmsltu.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -187,9 +190,10 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsgtu.mask.nxv8i8(
define <vscale x 8 x i1> @intrinsic_vmsgtu_mask_vv_nxv8i8_nxv8i8(<vscale x 8 x i1> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, <vscale x 8 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgtu_mask_vv_nxv8i8_nxv8i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e8, m1, ta, mu
-; CHECK-NEXT: vmsltu.vv v0, v9, v8
+; CHECK-NEXT: vmsltu.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmsltu.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -238,11 +242,12 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsgtu.mask.nxv16i8(
define <vscale x 16 x i1> @intrinsic_vmsgtu_mask_vv_nxv16i8_nxv16i8(<vscale x 16 x i1> %0, <vscale x 16 x i8> %1, <vscale x 16 x i8> %2, <vscale x 16 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgtu_mask_vv_nxv16i8_nxv16i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e8, m2, ta, mu
-; CHECK-NEXT: vmsltu.vv v0, v10, v8
-; CHECK-NEXT: vmsltu.vv v14, v12, v10, v0.t
+; CHECK-NEXT: vmsltu.vv v14, v10, v8
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmsltu.vv v8, v12, v10, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 16 x i1> @llvm.riscv.vmsgtu.nxv16i8(
@@ -289,11 +294,12 @@ declare <vscale x 32 x i1> @llvm.riscv.vmsgtu.mask.nxv32i8(
define <vscale x 32 x i1> @intrinsic_vmsgtu_mask_vv_nxv32i8_nxv32i8(<vscale x 32 x i1> %0, <vscale x 32 x i8> %1, <vscale x 32 x i8> %2, <vscale x 32 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgtu_mask_vv_nxv32i8_nxv32i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, mu
-; CHECK-NEXT: vmsltu.vv v0, v12, v8
-; CHECK-NEXT: vmsltu.vv v20, v16, v12, v0.t
+; CHECK-NEXT: vmsltu.vv v20, v12, v8
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmsltu.vv v8, v16, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 32 x i1> @llvm.riscv.vmsgtu.nxv32i8(
@@ -340,9 +346,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsgtu.mask.nxv1i16(
define <vscale x 1 x i1> @intrinsic_vmsgtu_mask_vv_nxv1i16_nxv1i16(<vscale x 1 x i1> %0, <vscale x 1 x i16> %1, <vscale x 1 x i16> %2, <vscale x 1 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgtu_mask_vv_nxv1i16_nxv1i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
-; CHECK-NEXT: vmsltu.vv v0, v9, v8
+; CHECK-NEXT: vmsltu.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmsltu.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -391,9 +398,10 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsgtu.mask.nxv2i16(
define <vscale x 2 x i1> @intrinsic_vmsgtu_mask_vv_nxv2i16_nxv2i16(<vscale x 2 x i1> %0, <vscale x 2 x i16> %1, <vscale x 2 x i16> %2, <vscale x 2 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgtu_mask_vv_nxv2i16_nxv2i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
-; CHECK-NEXT: vmsltu.vv v0, v9, v8
+; CHECK-NEXT: vmsltu.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmsltu.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -442,9 +450,10 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsgtu.mask.nxv4i16(
define <vscale x 4 x i1> @intrinsic_vmsgtu_mask_vv_nxv4i16_nxv4i16(<vscale x 4 x i1> %0, <vscale x 4 x i16> %1, <vscale x 4 x i16> %2, <vscale x 4 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgtu_mask_vv_nxv4i16_nxv4i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
-; CHECK-NEXT: vmsltu.vv v0, v9, v8
+; CHECK-NEXT: vmsltu.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmsltu.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -493,11 +502,12 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsgtu.mask.nxv8i16(
define <vscale x 8 x i1> @intrinsic_vmsgtu_mask_vv_nxv8i16_nxv8i16(<vscale x 8 x i1> %0, <vscale x 8 x i16> %1, <vscale x 8 x i16> %2, <vscale x 8 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgtu_mask_vv_nxv8i16_nxv8i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
-; CHECK-NEXT: vmsltu.vv v0, v10, v8
-; CHECK-NEXT: vmsltu.vv v14, v12, v10, v0.t
+; CHECK-NEXT: vmsltu.vv v14, v10, v8
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmsltu.vv v8, v12, v10, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 8 x i1> @llvm.riscv.vmsgtu.nxv8i16(
@@ -544,11 +554,12 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsgtu.mask.nxv16i16(
define <vscale x 16 x i1> @intrinsic_vmsgtu_mask_vv_nxv16i16_nxv16i16(<vscale x 16 x i1> %0, <vscale x 16 x i16> %1, <vscale x 16 x i16> %2, <vscale x 16 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgtu_mask_vv_nxv16i16_nxv16i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
-; CHECK-NEXT: vmsltu.vv v0, v12, v8
-; CHECK-NEXT: vmsltu.vv v20, v16, v12, v0.t
+; CHECK-NEXT: vmsltu.vv v20, v12, v8
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmsltu.vv v8, v16, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 16 x i1> @llvm.riscv.vmsgtu.nxv16i16(
@@ -595,9 +606,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsgtu.mask.nxv1i32(
define <vscale x 1 x i1> @intrinsic_vmsgtu_mask_vv_nxv1i32_nxv1i32(<vscale x 1 x i1> %0, <vscale x 1 x i32> %1, <vscale x 1 x i32> %2, <vscale x 1 x i32> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgtu_mask_vv_nxv1i32_nxv1i32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
-; CHECK-NEXT: vmsltu.vv v0, v9, v8
+; CHECK-NEXT: vmsltu.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmsltu.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -646,9 +658,10 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsgtu.mask.nxv2i32(
define <vscale x 2 x i1> @intrinsic_vmsgtu_mask_vv_nxv2i32_nxv2i32(<vscale x 2 x i1> %0, <vscale x 2 x i32> %1, <vscale x 2 x i32> %2, <vscale x 2 x i32> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgtu_mask_vv_nxv2i32_nxv2i32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
-; CHECK-NEXT: vmsltu.vv v0, v9, v8
+; CHECK-NEXT: vmsltu.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmsltu.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -697,11 +710,12 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsgtu.mask.nxv4i32(
define <vscale x 4 x i1> @intrinsic_vmsgtu_mask_vv_nxv4i32_nxv4i32(<vscale x 4 x i1> %0, <vscale x 4 x i32> %1, <vscale x 4 x i32> %2, <vscale x 4 x i32> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgtu_mask_vv_nxv4i32_nxv4i32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
-; CHECK-NEXT: vmsltu.vv v0, v10, v8
-; CHECK-NEXT: vmsltu.vv v14, v12, v10, v0.t
+; CHECK-NEXT: vmsltu.vv v14, v10, v8
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmsltu.vv v8, v12, v10, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 4 x i1> @llvm.riscv.vmsgtu.nxv4i32(
@@ -748,11 +762,12 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsgtu.mask.nxv8i32(
define <vscale x 8 x i1> @intrinsic_vmsgtu_mask_vv_nxv8i32_nxv8i32(<vscale x 8 x i1> %0, <vscale x 8 x i32> %1, <vscale x 8 x i32> %2, <vscale x 8 x i32> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgtu_mask_vv_nxv8i32_nxv8i32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
-; CHECK-NEXT: vmsltu.vv v0, v12, v8
-; CHECK-NEXT: vmsltu.vv v20, v16, v12, v0.t
+; CHECK-NEXT: vmsltu.vv v20, v12, v8
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmsltu.vv v8, v16, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 8 x i1> @llvm.riscv.vmsgtu.nxv8i32(
@@ -799,9 +814,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsgtu.mask.nxv1i64(
define <vscale x 1 x i1> @intrinsic_vmsgtu_mask_vv_nxv1i64_nxv1i64(<vscale x 1 x i1> %0, <vscale x 1 x i64> %1, <vscale x 1 x i64> %2, <vscale x 1 x i64> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgtu_mask_vv_nxv1i64_nxv1i64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
-; CHECK-NEXT: vmsltu.vv v0, v9, v8
+; CHECK-NEXT: vmsltu.vv v8, v9, v8
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmsltu.vv v11, v10, v9, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -850,11 +866,12 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsgtu.mask.nxv2i64(
define <vscale x 2 x i1> @intrinsic_vmsgtu_mask_vv_nxv2i64_nxv2i64(<vscale x 2 x i1> %0, <vscale x 2 x i64> %1, <vscale x 2 x i64> %2, <vscale x 2 x i64> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgtu_mask_vv_nxv2i64_nxv2i64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
-; CHECK-NEXT: vmsltu.vv v0, v10, v8
-; CHECK-NEXT: vmsltu.vv v14, v12, v10, v0.t
+; CHECK-NEXT: vmsltu.vv v14, v10, v8
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmsltu.vv v8, v12, v10, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 2 x i1> @llvm.riscv.vmsgtu.nxv2i64(
@@ -901,11 +918,12 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsgtu.mask.nxv4i64(
define <vscale x 4 x i1> @intrinsic_vmsgtu_mask_vv_nxv4i64_nxv4i64(<vscale x 4 x i1> %0, <vscale x 4 x i64> %1, <vscale x 4 x i64> %2, <vscale x 4 x i64> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsgtu_mask_vv_nxv4i64_nxv4i64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
-; CHECK-NEXT: vmsltu.vv v0, v12, v8
-; CHECK-NEXT: vmsltu.vv v20, v16, v12, v0.t
+; CHECK-NEXT: vmsltu.vv v20, v12, v8
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmsltu.vv v8, v16, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 4 x i1> @llvm.riscv.vmsgtu.nxv4i64(
@@ -953,8 +971,8 @@ define <vscale x 1 x i1> @intrinsic_vmsgtu_mask_vx_nxv1i8_i8(<vscale x 1 x i1> %
; CHECK-LABEL: intrinsic_vmsgtu_mask_vx_nxv1i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, mu
; CHECK-NEXT: vmsgtu.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1000,8 +1018,8 @@ define <vscale x 2 x i1> @intrinsic_vmsgtu_mask_vx_nxv2i8_i8(<vscale x 2 x i1> %
; CHECK-LABEL: intrinsic_vmsgtu_mask_vx_nxv2i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e8, mf4, ta, mu
; CHECK-NEXT: vmsgtu.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1047,8 +1065,8 @@ define <vscale x 4 x i1> @intrinsic_vmsgtu_mask_vx_nxv4i8_i8(<vscale x 4 x i1> %
; CHECK-LABEL: intrinsic_vmsgtu_mask_vx_nxv4i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e8, mf2, ta, mu
; CHECK-NEXT: vmsgtu.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1094,8 +1112,8 @@ define <vscale x 8 x i1> @intrinsic_vmsgtu_mask_vx_nxv8i8_i8(<vscale x 8 x i1> %
; CHECK-LABEL: intrinsic_vmsgtu_mask_vx_nxv8i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e8, m1, ta, mu
; CHECK-NEXT: vmsgtu.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -1141,8 +1159,8 @@ define <vscale x 16 x i1> @intrinsic_vmsgtu_mask_vx_nxv16i8_i8(<vscale x 16 x i1
; CHECK-LABEL: intrinsic_vmsgtu_mask_vx_nxv16i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a1, e8, m2, ta, mu
; CHECK-NEXT: vmsgtu.vx v11, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -1188,8 +1206,8 @@ define <vscale x 32 x i1> @intrinsic_vmsgtu_mask_vx_nxv32i8_i8(<vscale x 32 x i1
; CHECK-LABEL: intrinsic_vmsgtu_mask_vx_nxv32i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a1, e8, m4, ta, mu
; CHECK-NEXT: vmsgtu.vx v13, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -1235,8 +1253,8 @@ define <vscale x 1 x i1> @intrinsic_vmsgtu_mask_vx_nxv1i16_i16(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmsgtu_mask_vx_nxv1i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, mf4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e16, mf4, ta, mu
; CHECK-NEXT: vmsgtu.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1282,8 +1300,8 @@ define <vscale x 2 x i1> @intrinsic_vmsgtu_mask_vx_nxv2i16_i16(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmsgtu_mask_vx_nxv2i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e16, mf2, ta, mu
; CHECK-NEXT: vmsgtu.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1329,8 +1347,8 @@ define <vscale x 4 x i1> @intrinsic_vmsgtu_mask_vx_nxv4i16_i16(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmsgtu_mask_vx_nxv4i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e16, m1, ta, mu
; CHECK-NEXT: vmsgtu.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -1376,8 +1394,8 @@ define <vscale x 8 x i1> @intrinsic_vmsgtu_mask_vx_nxv8i16_i16(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmsgtu_mask_vx_nxv8i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a1, e16, m2, ta, mu
; CHECK-NEXT: vmsgtu.vx v11, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -1423,8 +1441,8 @@ define <vscale x 16 x i1> @intrinsic_vmsgtu_mask_vx_nxv16i16_i16(<vscale x 16 x
; CHECK-LABEL: intrinsic_vmsgtu_mask_vx_nxv16i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a1, e16, m4, ta, mu
; CHECK-NEXT: vmsgtu.vx v13, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -1470,8 +1488,8 @@ define <vscale x 1 x i1> @intrinsic_vmsgtu_mask_vx_nxv1i32_i32(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmsgtu_mask_vx_nxv1i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e32, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e32, mf2, ta, mu
; CHECK-NEXT: vmsgtu.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1517,8 +1535,8 @@ define <vscale x 2 x i1> @intrinsic_vmsgtu_mask_vx_nxv2i32_i32(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmsgtu_mask_vx_nxv2i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e32, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e32, m1, ta, mu
; CHECK-NEXT: vmsgtu.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -1564,8 +1582,8 @@ define <vscale x 4 x i1> @intrinsic_vmsgtu_mask_vx_nxv4i32_i32(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmsgtu_mask_vx_nxv4i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a1, e32, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a1, e32, m2, ta, mu
; CHECK-NEXT: vmsgtu.vx v11, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -1611,8 +1629,8 @@ define <vscale x 8 x i1> @intrinsic_vmsgtu_mask_vx_nxv8i32_i32(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmsgtu_mask_vx_nxv8i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a1, e32, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a1, e32, m4, ta, mu
; CHECK-NEXT: vmsgtu.vx v13, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -1685,8 +1703,8 @@ define <vscale x 1 x i1> @intrinsic_vmsgtu_mask_vx_nxv1i64_i64(<vscale x 1 x i1>
; RV64-LABEL: intrinsic_vmsgtu_mask_vx_nxv1i64_i64:
; RV64: # %bb.0: # %entry
; RV64-NEXT: vmv1r.v v10, v0
-; RV64-NEXT: vsetvli zero, a1, e64, m1, ta, mu
; RV64-NEXT: vmv1r.v v0, v9
+; RV64-NEXT: vsetvli zero, a1, e64, m1, ta, mu
; RV64-NEXT: vmsgtu.vx v10, v8, a0, v0.t
; RV64-NEXT: vmv.v.v v0, v10
; RV64-NEXT: ret
@@ -1759,8 +1777,8 @@ define <vscale x 2 x i1> @intrinsic_vmsgtu_mask_vx_nxv2i64_i64(<vscale x 2 x i1>
; RV64-LABEL: intrinsic_vmsgtu_mask_vx_nxv2i64_i64:
; RV64: # %bb.0: # %entry
; RV64-NEXT: vmv1r.v v11, v0
-; RV64-NEXT: vsetvli zero, a1, e64, m2, ta, mu
; RV64-NEXT: vmv1r.v v0, v10
+; RV64-NEXT: vsetvli zero, a1, e64, m2, ta, mu
; RV64-NEXT: vmsgtu.vx v11, v8, a0, v0.t
; RV64-NEXT: vmv1r.v v0, v11
; RV64-NEXT: ret
@@ -1833,8 +1851,8 @@ define <vscale x 4 x i1> @intrinsic_vmsgtu_mask_vx_nxv4i64_i64(<vscale x 4 x i1>
; RV64-LABEL: intrinsic_vmsgtu_mask_vx_nxv4i64_i64:
; RV64: # %bb.0: # %entry
; RV64-NEXT: vmv1r.v v13, v0
-; RV64-NEXT: vsetvli zero, a1, e64, m4, ta, mu
; RV64-NEXT: vmv1r.v v0, v12
+; RV64-NEXT: vsetvli zero, a1, e64, m4, ta, mu
; RV64-NEXT: vmsgtu.vx v13, v8, a0, v0.t
; RV64-NEXT: vmv1r.v v0, v13
; RV64-NEXT: ret
@@ -1868,8 +1886,8 @@ define <vscale x 1 x i1> @intrinsic_vmsgtu_mask_vi_nxv1i8_i8(<vscale x 1 x i1> %
; CHECK-LABEL: intrinsic_vmsgtu_mask_vi_nxv1i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, mu
; CHECK-NEXT: vmsgtu.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1903,8 +1921,8 @@ define <vscale x 2 x i1> @intrinsic_vmsgtu_mask_vi_nxv2i8_i8(<vscale x 2 x i1> %
; CHECK-LABEL: intrinsic_vmsgtu_mask_vi_nxv2i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, mf4, ta, mu
; CHECK-NEXT: vmsgtu.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1938,8 +1956,8 @@ define <vscale x 4 x i1> @intrinsic_vmsgtu_mask_vi_nxv4i8_i8(<vscale x 4 x i1> %
; CHECK-LABEL: intrinsic_vmsgtu_mask_vi_nxv4i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, mu
; CHECK-NEXT: vmsgtu.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1973,8 +1991,8 @@ define <vscale x 8 x i1> @intrinsic_vmsgtu_mask_vi_nxv8i8_i8(<vscale x 8 x i1> %
; CHECK-LABEL: intrinsic_vmsgtu_mask_vi_nxv8i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, m1, ta, mu
; CHECK-NEXT: vmsgtu.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -2008,8 +2026,8 @@ define <vscale x 16 x i1> @intrinsic_vmsgtu_mask_vi_nxv16i8_i8(<vscale x 16 x i1
; CHECK-LABEL: intrinsic_vmsgtu_mask_vi_nxv16i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e8, m2, ta, mu
; CHECK-NEXT: vmsgtu.vi v11, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -2043,8 +2061,8 @@ define <vscale x 32 x i1> @intrinsic_vmsgtu_mask_vi_nxv32i8_i8(<vscale x 32 x i1
; CHECK-LABEL: intrinsic_vmsgtu_mask_vi_nxv32i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, mu
; CHECK-NEXT: vmsgtu.vi v13, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -2078,8 +2096,8 @@ define <vscale x 1 x i1> @intrinsic_vmsgtu_mask_vi_nxv1i16_i16(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmsgtu_mask_vi_nxv1i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
; CHECK-NEXT: vmsgtu.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -2113,8 +2131,8 @@ define <vscale x 2 x i1> @intrinsic_vmsgtu_mask_vi_nxv2i16_i16(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmsgtu_mask_vi_nxv2i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
; CHECK-NEXT: vmsgtu.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -2148,8 +2166,8 @@ define <vscale x 4 x i1> @intrinsic_vmsgtu_mask_vi_nxv4i16_i16(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmsgtu_mask_vi_nxv4i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
; CHECK-NEXT: vmsgtu.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -2183,8 +2201,8 @@ define <vscale x 8 x i1> @intrinsic_vmsgtu_mask_vi_nxv8i16_i16(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmsgtu_mask_vi_nxv8i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
; CHECK-NEXT: vmsgtu.vi v11, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -2218,8 +2236,8 @@ define <vscale x 16 x i1> @intrinsic_vmsgtu_mask_vi_nxv16i16_i16(<vscale x 16 x
; CHECK-LABEL: intrinsic_vmsgtu_mask_vi_nxv16i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
; CHECK-NEXT: vmsgtu.vi v13, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -2253,8 +2271,8 @@ define <vscale x 1 x i1> @intrinsic_vmsgtu_mask_vi_nxv1i32_i32(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmsgtu_mask_vi_nxv1i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
; CHECK-NEXT: vmsgtu.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -2288,8 +2306,8 @@ define <vscale x 2 x i1> @intrinsic_vmsgtu_mask_vi_nxv2i32_i32(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmsgtu_mask_vi_nxv2i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
; CHECK-NEXT: vmsgtu.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -2323,8 +2341,8 @@ define <vscale x 4 x i1> @intrinsic_vmsgtu_mask_vi_nxv4i32_i32(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmsgtu_mask_vi_nxv4i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
; CHECK-NEXT: vmsgtu.vi v11, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -2358,8 +2376,8 @@ define <vscale x 8 x i1> @intrinsic_vmsgtu_mask_vi_nxv8i32_i32(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmsgtu_mask_vi_nxv8i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
; CHECK-NEXT: vmsgtu.vi v13, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -2393,8 +2411,8 @@ define <vscale x 1 x i1> @intrinsic_vmsgtu_mask_vi_nxv1i64_i64(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmsgtu_mask_vi_nxv1i64_i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
; CHECK-NEXT: vmsgtu.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -2428,8 +2446,8 @@ define <vscale x 2 x i1> @intrinsic_vmsgtu_mask_vi_nxv2i64_i64(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmsgtu_mask_vi_nxv2i64_i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
; CHECK-NEXT: vmsgtu.vi v11, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -2463,8 +2481,8 @@ define <vscale x 4 x i1> @intrinsic_vmsgtu_mask_vi_nxv4i64_i64(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmsgtu_mask_vi_nxv4i64_i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
; CHECK-NEXT: vmsgtu.vi v13, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vmsif.ll b/llvm/test/CodeGen/RISCV/rvv/vmsif.ll
index 8ce9a3020b7a5..05d402afc934c 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vmsif.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vmsif.ll
@@ -32,8 +32,8 @@ define <vscale x 1 x i1> @intrinsic_vmsif_mask_m_nxv1i1_nxv1i1(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmsif_mask_m_nxv1i1_nxv1i1:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf8, tu, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, mf8, tu, mu
; CHECK-NEXT: vmsif.m v10, v8, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -74,8 +74,8 @@ define <vscale x 2 x i1> @intrinsic_vmsif_mask_m_nxv2i1_nxv2i1(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmsif_mask_m_nxv2i1_nxv2i1:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf4, tu, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, mf4, tu, mu
; CHECK-NEXT: vmsif.m v10, v8, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -116,8 +116,8 @@ define <vscale x 4 x i1> @intrinsic_vmsif_mask_m_nxv4i1_nxv4i1(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmsif_mask_m_nxv4i1_nxv4i1:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf2, tu, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, mf2, tu, mu
; CHECK-NEXT: vmsif.m v10, v8, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -158,8 +158,8 @@ define <vscale x 8 x i1> @intrinsic_vmsif_mask_m_nxv8i1_nxv8i1(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmsif_mask_m_nxv8i1_nxv8i1:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m1, tu, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, m1, tu, mu
; CHECK-NEXT: vmsif.m v10, v8, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -200,8 +200,8 @@ define <vscale x 16 x i1> @intrinsic_vmsif_mask_m_nxv16i1_nxv16i1(<vscale x 16 x
; CHECK-LABEL: intrinsic_vmsif_mask_m_nxv16i1_nxv16i1:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m2, tu, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, m2, tu, mu
; CHECK-NEXT: vmsif.m v10, v8, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -242,8 +242,8 @@ define <vscale x 32 x i1> @intrinsic_vmsif_mask_m_nxv32i1_nxv32i1(<vscale x 32 x
; CHECK-LABEL: intrinsic_vmsif_mask_m_nxv32i1_nxv32i1:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m4, tu, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, m4, tu, mu
; CHECK-NEXT: vmsif.m v10, v8, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -284,8 +284,8 @@ define <vscale x 64 x i1> @intrinsic_vmsif_mask_m_nxv64i1_nxv64i1(<vscale x 64 x
; CHECK-LABEL: intrinsic_vmsif_mask_m_nxv64i1_nxv64i1:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m8, tu, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, m8, tu, mu
; CHECK-NEXT: vmsif.m v10, v8, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vmsle.ll b/llvm/test/CodeGen/RISCV/rvv/vmsle.ll
index 5d5a28edbfe15..f54aef3ed4052 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vmsle.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vmsle.ll
@@ -34,9 +34,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsle.mask.nxv1i8(
define <vscale x 1 x i1> @intrinsic_vmsle_mask_vv_nxv1i8_nxv1i8(<vscale x 1 x i1> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, <vscale x 1 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsle_mask_vv_nxv1i8_nxv1i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, mu
-; CHECK-NEXT: vmsle.vv v0, v8, v9
+; CHECK-NEXT: vmsle.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmsle.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -85,9 +86,10 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsle.mask.nxv2i8(
define <vscale x 2 x i1> @intrinsic_vmsle_mask_vv_nxv2i8_nxv2i8(<vscale x 2 x i1> %0, <vscale x 2 x i8> %1, <vscale x 2 x i8> %2, <vscale x 2 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsle_mask_vv_nxv2i8_nxv2i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e8, mf4, ta, mu
-; CHECK-NEXT: vmsle.vv v0, v8, v9
+; CHECK-NEXT: vmsle.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmsle.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -136,9 +138,10 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsle.mask.nxv4i8(
define <vscale x 4 x i1> @intrinsic_vmsle_mask_vv_nxv4i8_nxv4i8(<vscale x 4 x i1> %0, <vscale x 4 x i8> %1, <vscale x 4 x i8> %2, <vscale x 4 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsle_mask_vv_nxv4i8_nxv4i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, mu
-; CHECK-NEXT: vmsle.vv v0, v8, v9
+; CHECK-NEXT: vmsle.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmsle.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -187,9 +190,10 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsle.mask.nxv8i8(
define <vscale x 8 x i1> @intrinsic_vmsle_mask_vv_nxv8i8_nxv8i8(<vscale x 8 x i1> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, <vscale x 8 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsle_mask_vv_nxv8i8_nxv8i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e8, m1, ta, mu
-; CHECK-NEXT: vmsle.vv v0, v8, v9
+; CHECK-NEXT: vmsle.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmsle.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -238,11 +242,12 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsle.mask.nxv16i8(
define <vscale x 16 x i1> @intrinsic_vmsle_mask_vv_nxv16i8_nxv16i8(<vscale x 16 x i1> %0, <vscale x 16 x i8> %1, <vscale x 16 x i8> %2, <vscale x 16 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsle_mask_vv_nxv16i8_nxv16i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e8, m2, ta, mu
-; CHECK-NEXT: vmsle.vv v0, v8, v10
-; CHECK-NEXT: vmsle.vv v14, v10, v12, v0.t
+; CHECK-NEXT: vmsle.vv v14, v8, v10
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmsle.vv v8, v10, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 16 x i1> @llvm.riscv.vmsle.nxv16i8(
@@ -289,11 +294,12 @@ declare <vscale x 32 x i1> @llvm.riscv.vmsle.mask.nxv32i8(
define <vscale x 32 x i1> @intrinsic_vmsle_mask_vv_nxv32i8_nxv32i8(<vscale x 32 x i1> %0, <vscale x 32 x i8> %1, <vscale x 32 x i8> %2, <vscale x 32 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsle_mask_vv_nxv32i8_nxv32i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, mu
-; CHECK-NEXT: vmsle.vv v0, v8, v12
-; CHECK-NEXT: vmsle.vv v20, v12, v16, v0.t
+; CHECK-NEXT: vmsle.vv v20, v8, v12
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmsle.vv v8, v12, v16, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 32 x i1> @llvm.riscv.vmsle.nxv32i8(
@@ -340,9 +346,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsle.mask.nxv1i16(
define <vscale x 1 x i1> @intrinsic_vmsle_mask_vv_nxv1i16_nxv1i16(<vscale x 1 x i1> %0, <vscale x 1 x i16> %1, <vscale x 1 x i16> %2, <vscale x 1 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsle_mask_vv_nxv1i16_nxv1i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
-; CHECK-NEXT: vmsle.vv v0, v8, v9
+; CHECK-NEXT: vmsle.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmsle.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -391,9 +398,10 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsle.mask.nxv2i16(
define <vscale x 2 x i1> @intrinsic_vmsle_mask_vv_nxv2i16_nxv2i16(<vscale x 2 x i1> %0, <vscale x 2 x i16> %1, <vscale x 2 x i16> %2, <vscale x 2 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsle_mask_vv_nxv2i16_nxv2i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
-; CHECK-NEXT: vmsle.vv v0, v8, v9
+; CHECK-NEXT: vmsle.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmsle.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -442,9 +450,10 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsle.mask.nxv4i16(
define <vscale x 4 x i1> @intrinsic_vmsle_mask_vv_nxv4i16_nxv4i16(<vscale x 4 x i1> %0, <vscale x 4 x i16> %1, <vscale x 4 x i16> %2, <vscale x 4 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsle_mask_vv_nxv4i16_nxv4i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
-; CHECK-NEXT: vmsle.vv v0, v8, v9
+; CHECK-NEXT: vmsle.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmsle.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -493,11 +502,12 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsle.mask.nxv8i16(
define <vscale x 8 x i1> @intrinsic_vmsle_mask_vv_nxv8i16_nxv8i16(<vscale x 8 x i1> %0, <vscale x 8 x i16> %1, <vscale x 8 x i16> %2, <vscale x 8 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsle_mask_vv_nxv8i16_nxv8i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
-; CHECK-NEXT: vmsle.vv v0, v8, v10
-; CHECK-NEXT: vmsle.vv v14, v10, v12, v0.t
+; CHECK-NEXT: vmsle.vv v14, v8, v10
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmsle.vv v8, v10, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 8 x i1> @llvm.riscv.vmsle.nxv8i16(
@@ -544,11 +554,12 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsle.mask.nxv16i16(
define <vscale x 16 x i1> @intrinsic_vmsle_mask_vv_nxv16i16_nxv16i16(<vscale x 16 x i1> %0, <vscale x 16 x i16> %1, <vscale x 16 x i16> %2, <vscale x 16 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsle_mask_vv_nxv16i16_nxv16i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
-; CHECK-NEXT: vmsle.vv v0, v8, v12
-; CHECK-NEXT: vmsle.vv v20, v12, v16, v0.t
+; CHECK-NEXT: vmsle.vv v20, v8, v12
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmsle.vv v8, v12, v16, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 16 x i1> @llvm.riscv.vmsle.nxv16i16(
@@ -595,9 +606,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsle.mask.nxv1i32(
define <vscale x 1 x i1> @intrinsic_vmsle_mask_vv_nxv1i32_nxv1i32(<vscale x 1 x i1> %0, <vscale x 1 x i32> %1, <vscale x 1 x i32> %2, <vscale x 1 x i32> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsle_mask_vv_nxv1i32_nxv1i32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
-; CHECK-NEXT: vmsle.vv v0, v8, v9
+; CHECK-NEXT: vmsle.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmsle.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -646,9 +658,10 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsle.mask.nxv2i32(
define <vscale x 2 x i1> @intrinsic_vmsle_mask_vv_nxv2i32_nxv2i32(<vscale x 2 x i1> %0, <vscale x 2 x i32> %1, <vscale x 2 x i32> %2, <vscale x 2 x i32> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsle_mask_vv_nxv2i32_nxv2i32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
-; CHECK-NEXT: vmsle.vv v0, v8, v9
+; CHECK-NEXT: vmsle.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmsle.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -697,11 +710,12 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsle.mask.nxv4i32(
define <vscale x 4 x i1> @intrinsic_vmsle_mask_vv_nxv4i32_nxv4i32(<vscale x 4 x i1> %0, <vscale x 4 x i32> %1, <vscale x 4 x i32> %2, <vscale x 4 x i32> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsle_mask_vv_nxv4i32_nxv4i32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
-; CHECK-NEXT: vmsle.vv v0, v8, v10
-; CHECK-NEXT: vmsle.vv v14, v10, v12, v0.t
+; CHECK-NEXT: vmsle.vv v14, v8, v10
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmsle.vv v8, v10, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 4 x i1> @llvm.riscv.vmsle.nxv4i32(
@@ -748,11 +762,12 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsle.mask.nxv8i32(
define <vscale x 8 x i1> @intrinsic_vmsle_mask_vv_nxv8i32_nxv8i32(<vscale x 8 x i1> %0, <vscale x 8 x i32> %1, <vscale x 8 x i32> %2, <vscale x 8 x i32> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsle_mask_vv_nxv8i32_nxv8i32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
-; CHECK-NEXT: vmsle.vv v0, v8, v12
-; CHECK-NEXT: vmsle.vv v20, v12, v16, v0.t
+; CHECK-NEXT: vmsle.vv v20, v8, v12
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmsle.vv v8, v12, v16, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 8 x i1> @llvm.riscv.vmsle.nxv8i32(
@@ -799,9 +814,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsle.mask.nxv1i64(
define <vscale x 1 x i1> @intrinsic_vmsle_mask_vv_nxv1i64_nxv1i64(<vscale x 1 x i1> %0, <vscale x 1 x i64> %1, <vscale x 1 x i64> %2, <vscale x 1 x i64> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsle_mask_vv_nxv1i64_nxv1i64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
-; CHECK-NEXT: vmsle.vv v0, v8, v9
+; CHECK-NEXT: vmsle.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmsle.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -850,11 +866,12 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsle.mask.nxv2i64(
define <vscale x 2 x i1> @intrinsic_vmsle_mask_vv_nxv2i64_nxv2i64(<vscale x 2 x i1> %0, <vscale x 2 x i64> %1, <vscale x 2 x i64> %2, <vscale x 2 x i64> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsle_mask_vv_nxv2i64_nxv2i64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
-; CHECK-NEXT: vmsle.vv v0, v8, v10
-; CHECK-NEXT: vmsle.vv v14, v10, v12, v0.t
+; CHECK-NEXT: vmsle.vv v14, v8, v10
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmsle.vv v8, v10, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 2 x i1> @llvm.riscv.vmsle.nxv2i64(
@@ -901,11 +918,12 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsle.mask.nxv4i64(
define <vscale x 4 x i1> @intrinsic_vmsle_mask_vv_nxv4i64_nxv4i64(<vscale x 4 x i1> %0, <vscale x 4 x i64> %1, <vscale x 4 x i64> %2, <vscale x 4 x i64> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsle_mask_vv_nxv4i64_nxv4i64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
-; CHECK-NEXT: vmsle.vv v0, v8, v12
-; CHECK-NEXT: vmsle.vv v20, v12, v16, v0.t
+; CHECK-NEXT: vmsle.vv v20, v8, v12
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmsle.vv v8, v12, v16, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 4 x i1> @llvm.riscv.vmsle.nxv4i64(
@@ -953,8 +971,8 @@ define <vscale x 1 x i1> @intrinsic_vmsle_mask_vx_nxv1i8_i8(<vscale x 1 x i1> %0
; CHECK-LABEL: intrinsic_vmsle_mask_vx_nxv1i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, mu
; CHECK-NEXT: vmsle.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1000,8 +1018,8 @@ define <vscale x 2 x i1> @intrinsic_vmsle_mask_vx_nxv2i8_i8(<vscale x 2 x i1> %0
; CHECK-LABEL: intrinsic_vmsle_mask_vx_nxv2i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e8, mf4, ta, mu
; CHECK-NEXT: vmsle.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1047,8 +1065,8 @@ define <vscale x 4 x i1> @intrinsic_vmsle_mask_vx_nxv4i8_i8(<vscale x 4 x i1> %0
; CHECK-LABEL: intrinsic_vmsle_mask_vx_nxv4i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e8, mf2, ta, mu
; CHECK-NEXT: vmsle.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1094,8 +1112,8 @@ define <vscale x 8 x i1> @intrinsic_vmsle_mask_vx_nxv8i8_i8(<vscale x 8 x i1> %0
; CHECK-LABEL: intrinsic_vmsle_mask_vx_nxv8i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e8, m1, ta, mu
; CHECK-NEXT: vmsle.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -1141,8 +1159,8 @@ define <vscale x 16 x i1> @intrinsic_vmsle_mask_vx_nxv16i8_i8(<vscale x 16 x i1>
; CHECK-LABEL: intrinsic_vmsle_mask_vx_nxv16i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a1, e8, m2, ta, mu
; CHECK-NEXT: vmsle.vx v11, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -1188,8 +1206,8 @@ define <vscale x 32 x i1> @intrinsic_vmsle_mask_vx_nxv32i8_i8(<vscale x 32 x i1>
; CHECK-LABEL: intrinsic_vmsle_mask_vx_nxv32i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a1, e8, m4, ta, mu
; CHECK-NEXT: vmsle.vx v13, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -1235,8 +1253,8 @@ define <vscale x 1 x i1> @intrinsic_vmsle_mask_vx_nxv1i16_i16(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmsle_mask_vx_nxv1i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, mf4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e16, mf4, ta, mu
; CHECK-NEXT: vmsle.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1282,8 +1300,8 @@ define <vscale x 2 x i1> @intrinsic_vmsle_mask_vx_nxv2i16_i16(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmsle_mask_vx_nxv2i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e16, mf2, ta, mu
; CHECK-NEXT: vmsle.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1329,8 +1347,8 @@ define <vscale x 4 x i1> @intrinsic_vmsle_mask_vx_nxv4i16_i16(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmsle_mask_vx_nxv4i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e16, m1, ta, mu
; CHECK-NEXT: vmsle.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -1376,8 +1394,8 @@ define <vscale x 8 x i1> @intrinsic_vmsle_mask_vx_nxv8i16_i16(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmsle_mask_vx_nxv8i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a1, e16, m2, ta, mu
; CHECK-NEXT: vmsle.vx v11, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -1423,8 +1441,8 @@ define <vscale x 16 x i1> @intrinsic_vmsle_mask_vx_nxv16i16_i16(<vscale x 16 x i
; CHECK-LABEL: intrinsic_vmsle_mask_vx_nxv16i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a1, e16, m4, ta, mu
; CHECK-NEXT: vmsle.vx v13, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -1470,8 +1488,8 @@ define <vscale x 1 x i1> @intrinsic_vmsle_mask_vx_nxv1i32_i32(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmsle_mask_vx_nxv1i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e32, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e32, mf2, ta, mu
; CHECK-NEXT: vmsle.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1517,8 +1535,8 @@ define <vscale x 2 x i1> @intrinsic_vmsle_mask_vx_nxv2i32_i32(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmsle_mask_vx_nxv2i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e32, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e32, m1, ta, mu
; CHECK-NEXT: vmsle.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -1564,8 +1582,8 @@ define <vscale x 4 x i1> @intrinsic_vmsle_mask_vx_nxv4i32_i32(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmsle_mask_vx_nxv4i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a1, e32, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a1, e32, m2, ta, mu
; CHECK-NEXT: vmsle.vx v11, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -1611,8 +1629,8 @@ define <vscale x 8 x i1> @intrinsic_vmsle_mask_vx_nxv8i32_i32(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmsle_mask_vx_nxv8i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a1, e32, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a1, e32, m4, ta, mu
; CHECK-NEXT: vmsle.vx v13, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -1685,8 +1703,8 @@ define <vscale x 1 x i1> @intrinsic_vmsle_mask_vx_nxv1i64_i64(<vscale x 1 x i1>
; RV64-LABEL: intrinsic_vmsle_mask_vx_nxv1i64_i64:
; RV64: # %bb.0: # %entry
; RV64-NEXT: vmv1r.v v10, v0
-; RV64-NEXT: vsetvli zero, a1, e64, m1, ta, mu
; RV64-NEXT: vmv1r.v v0, v9
+; RV64-NEXT: vsetvli zero, a1, e64, m1, ta, mu
; RV64-NEXT: vmsle.vx v10, v8, a0, v0.t
; RV64-NEXT: vmv.v.v v0, v10
; RV64-NEXT: ret
@@ -1759,8 +1777,8 @@ define <vscale x 2 x i1> @intrinsic_vmsle_mask_vx_nxv2i64_i64(<vscale x 2 x i1>
; RV64-LABEL: intrinsic_vmsle_mask_vx_nxv2i64_i64:
; RV64: # %bb.0: # %entry
; RV64-NEXT: vmv1r.v v11, v0
-; RV64-NEXT: vsetvli zero, a1, e64, m2, ta, mu
; RV64-NEXT: vmv1r.v v0, v10
+; RV64-NEXT: vsetvli zero, a1, e64, m2, ta, mu
; RV64-NEXT: vmsle.vx v11, v8, a0, v0.t
; RV64-NEXT: vmv1r.v v0, v11
; RV64-NEXT: ret
@@ -1833,8 +1851,8 @@ define <vscale x 4 x i1> @intrinsic_vmsle_mask_vx_nxv4i64_i64(<vscale x 4 x i1>
; RV64-LABEL: intrinsic_vmsle_mask_vx_nxv4i64_i64:
; RV64: # %bb.0: # %entry
; RV64-NEXT: vmv1r.v v13, v0
-; RV64-NEXT: vsetvli zero, a1, e64, m4, ta, mu
; RV64-NEXT: vmv1r.v v0, v12
+; RV64-NEXT: vsetvli zero, a1, e64, m4, ta, mu
; RV64-NEXT: vmsle.vx v13, v8, a0, v0.t
; RV64-NEXT: vmv1r.v v0, v13
; RV64-NEXT: ret
@@ -1868,8 +1886,8 @@ define <vscale x 1 x i1> @intrinsic_vmsle_mask_vi_nxv1i8_i8(<vscale x 1 x i1> %0
; CHECK-LABEL: intrinsic_vmsle_mask_vi_nxv1i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, mu
; CHECK-NEXT: vmsle.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1903,8 +1921,8 @@ define <vscale x 2 x i1> @intrinsic_vmsle_mask_vi_nxv2i8_i8(<vscale x 2 x i1> %0
; CHECK-LABEL: intrinsic_vmsle_mask_vi_nxv2i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, mf4, ta, mu
; CHECK-NEXT: vmsle.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1938,8 +1956,8 @@ define <vscale x 4 x i1> @intrinsic_vmsle_mask_vi_nxv4i8_i8(<vscale x 4 x i1> %0
; CHECK-LABEL: intrinsic_vmsle_mask_vi_nxv4i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, mu
; CHECK-NEXT: vmsle.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1973,8 +1991,8 @@ define <vscale x 8 x i1> @intrinsic_vmsle_mask_vi_nxv8i8_i8(<vscale x 8 x i1> %0
; CHECK-LABEL: intrinsic_vmsle_mask_vi_nxv8i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, m1, ta, mu
; CHECK-NEXT: vmsle.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -2008,8 +2026,8 @@ define <vscale x 16 x i1> @intrinsic_vmsle_mask_vi_nxv16i8_i8(<vscale x 16 x i1>
; CHECK-LABEL: intrinsic_vmsle_mask_vi_nxv16i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e8, m2, ta, mu
; CHECK-NEXT: vmsle.vi v11, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -2043,8 +2061,8 @@ define <vscale x 32 x i1> @intrinsic_vmsle_mask_vi_nxv32i8_i8(<vscale x 32 x i1>
; CHECK-LABEL: intrinsic_vmsle_mask_vi_nxv32i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, mu
; CHECK-NEXT: vmsle.vi v13, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -2078,8 +2096,8 @@ define <vscale x 1 x i1> @intrinsic_vmsle_mask_vi_nxv1i16_i16(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmsle_mask_vi_nxv1i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
; CHECK-NEXT: vmsle.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -2113,8 +2131,8 @@ define <vscale x 2 x i1> @intrinsic_vmsle_mask_vi_nxv2i16_i16(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmsle_mask_vi_nxv2i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
; CHECK-NEXT: vmsle.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -2148,8 +2166,8 @@ define <vscale x 4 x i1> @intrinsic_vmsle_mask_vi_nxv4i16_i16(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmsle_mask_vi_nxv4i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
; CHECK-NEXT: vmsle.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -2183,8 +2201,8 @@ define <vscale x 8 x i1> @intrinsic_vmsle_mask_vi_nxv8i16_i16(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmsle_mask_vi_nxv8i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
; CHECK-NEXT: vmsle.vi v11, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -2218,8 +2236,8 @@ define <vscale x 16 x i1> @intrinsic_vmsle_mask_vi_nxv16i16_i16(<vscale x 16 x i
; CHECK-LABEL: intrinsic_vmsle_mask_vi_nxv16i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
; CHECK-NEXT: vmsle.vi v13, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -2253,8 +2271,8 @@ define <vscale x 1 x i1> @intrinsic_vmsle_mask_vi_nxv1i32_i32(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmsle_mask_vi_nxv1i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
; CHECK-NEXT: vmsle.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -2288,8 +2306,8 @@ define <vscale x 2 x i1> @intrinsic_vmsle_mask_vi_nxv2i32_i32(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmsle_mask_vi_nxv2i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
; CHECK-NEXT: vmsle.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -2323,8 +2341,8 @@ define <vscale x 4 x i1> @intrinsic_vmsle_mask_vi_nxv4i32_i32(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmsle_mask_vi_nxv4i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
; CHECK-NEXT: vmsle.vi v11, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -2358,8 +2376,8 @@ define <vscale x 8 x i1> @intrinsic_vmsle_mask_vi_nxv8i32_i32(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmsle_mask_vi_nxv8i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
; CHECK-NEXT: vmsle.vi v13, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -2393,8 +2411,8 @@ define <vscale x 1 x i1> @intrinsic_vmsle_mask_vi_nxv1i64_i64(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmsle_mask_vi_nxv1i64_i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
; CHECK-NEXT: vmsle.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -2428,8 +2446,8 @@ define <vscale x 2 x i1> @intrinsic_vmsle_mask_vi_nxv2i64_i64(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmsle_mask_vi_nxv2i64_i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
; CHECK-NEXT: vmsle.vi v11, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -2463,8 +2481,8 @@ define <vscale x 4 x i1> @intrinsic_vmsle_mask_vi_nxv4i64_i64(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmsle_mask_vi_nxv4i64_i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
; CHECK-NEXT: vmsle.vi v13, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vmsleu.ll b/llvm/test/CodeGen/RISCV/rvv/vmsleu.ll
index c58ac2d071831..540577247484e 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vmsleu.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vmsleu.ll
@@ -34,9 +34,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsleu.mask.nxv1i8(
define <vscale x 1 x i1> @intrinsic_vmsleu_mask_vv_nxv1i8_nxv1i8(<vscale x 1 x i1> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, <vscale x 1 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsleu_mask_vv_nxv1i8_nxv1i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, mu
-; CHECK-NEXT: vmsleu.vv v0, v8, v9
+; CHECK-NEXT: vmsleu.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmsleu.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -85,9 +86,10 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsleu.mask.nxv2i8(
define <vscale x 2 x i1> @intrinsic_vmsleu_mask_vv_nxv2i8_nxv2i8(<vscale x 2 x i1> %0, <vscale x 2 x i8> %1, <vscale x 2 x i8> %2, <vscale x 2 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsleu_mask_vv_nxv2i8_nxv2i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e8, mf4, ta, mu
-; CHECK-NEXT: vmsleu.vv v0, v8, v9
+; CHECK-NEXT: vmsleu.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmsleu.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -136,9 +138,10 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsleu.mask.nxv4i8(
define <vscale x 4 x i1> @intrinsic_vmsleu_mask_vv_nxv4i8_nxv4i8(<vscale x 4 x i1> %0, <vscale x 4 x i8> %1, <vscale x 4 x i8> %2, <vscale x 4 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsleu_mask_vv_nxv4i8_nxv4i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, mu
-; CHECK-NEXT: vmsleu.vv v0, v8, v9
+; CHECK-NEXT: vmsleu.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmsleu.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -187,9 +190,10 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsleu.mask.nxv8i8(
define <vscale x 8 x i1> @intrinsic_vmsleu_mask_vv_nxv8i8_nxv8i8(<vscale x 8 x i1> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, <vscale x 8 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsleu_mask_vv_nxv8i8_nxv8i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e8, m1, ta, mu
-; CHECK-NEXT: vmsleu.vv v0, v8, v9
+; CHECK-NEXT: vmsleu.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmsleu.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -238,11 +242,12 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsleu.mask.nxv16i8(
define <vscale x 16 x i1> @intrinsic_vmsleu_mask_vv_nxv16i8_nxv16i8(<vscale x 16 x i1> %0, <vscale x 16 x i8> %1, <vscale x 16 x i8> %2, <vscale x 16 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsleu_mask_vv_nxv16i8_nxv16i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e8, m2, ta, mu
-; CHECK-NEXT: vmsleu.vv v0, v8, v10
-; CHECK-NEXT: vmsleu.vv v14, v10, v12, v0.t
+; CHECK-NEXT: vmsleu.vv v14, v8, v10
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmsleu.vv v8, v10, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 16 x i1> @llvm.riscv.vmsleu.nxv16i8(
@@ -289,11 +294,12 @@ declare <vscale x 32 x i1> @llvm.riscv.vmsleu.mask.nxv32i8(
define <vscale x 32 x i1> @intrinsic_vmsleu_mask_vv_nxv32i8_nxv32i8(<vscale x 32 x i1> %0, <vscale x 32 x i8> %1, <vscale x 32 x i8> %2, <vscale x 32 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsleu_mask_vv_nxv32i8_nxv32i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, mu
-; CHECK-NEXT: vmsleu.vv v0, v8, v12
-; CHECK-NEXT: vmsleu.vv v20, v12, v16, v0.t
+; CHECK-NEXT: vmsleu.vv v20, v8, v12
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmsleu.vv v8, v12, v16, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 32 x i1> @llvm.riscv.vmsleu.nxv32i8(
@@ -340,9 +346,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsleu.mask.nxv1i16(
define <vscale x 1 x i1> @intrinsic_vmsleu_mask_vv_nxv1i16_nxv1i16(<vscale x 1 x i1> %0, <vscale x 1 x i16> %1, <vscale x 1 x i16> %2, <vscale x 1 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsleu_mask_vv_nxv1i16_nxv1i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
-; CHECK-NEXT: vmsleu.vv v0, v8, v9
+; CHECK-NEXT: vmsleu.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmsleu.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -391,9 +398,10 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsleu.mask.nxv2i16(
define <vscale x 2 x i1> @intrinsic_vmsleu_mask_vv_nxv2i16_nxv2i16(<vscale x 2 x i1> %0, <vscale x 2 x i16> %1, <vscale x 2 x i16> %2, <vscale x 2 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsleu_mask_vv_nxv2i16_nxv2i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
-; CHECK-NEXT: vmsleu.vv v0, v8, v9
+; CHECK-NEXT: vmsleu.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmsleu.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -442,9 +450,10 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsleu.mask.nxv4i16(
define <vscale x 4 x i1> @intrinsic_vmsleu_mask_vv_nxv4i16_nxv4i16(<vscale x 4 x i1> %0, <vscale x 4 x i16> %1, <vscale x 4 x i16> %2, <vscale x 4 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsleu_mask_vv_nxv4i16_nxv4i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
-; CHECK-NEXT: vmsleu.vv v0, v8, v9
+; CHECK-NEXT: vmsleu.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmsleu.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -493,11 +502,12 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsleu.mask.nxv8i16(
define <vscale x 8 x i1> @intrinsic_vmsleu_mask_vv_nxv8i16_nxv8i16(<vscale x 8 x i1> %0, <vscale x 8 x i16> %1, <vscale x 8 x i16> %2, <vscale x 8 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsleu_mask_vv_nxv8i16_nxv8i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
-; CHECK-NEXT: vmsleu.vv v0, v8, v10
-; CHECK-NEXT: vmsleu.vv v14, v10, v12, v0.t
+; CHECK-NEXT: vmsleu.vv v14, v8, v10
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmsleu.vv v8, v10, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 8 x i1> @llvm.riscv.vmsleu.nxv8i16(
@@ -544,11 +554,12 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsleu.mask.nxv16i16(
define <vscale x 16 x i1> @intrinsic_vmsleu_mask_vv_nxv16i16_nxv16i16(<vscale x 16 x i1> %0, <vscale x 16 x i16> %1, <vscale x 16 x i16> %2, <vscale x 16 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsleu_mask_vv_nxv16i16_nxv16i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
-; CHECK-NEXT: vmsleu.vv v0, v8, v12
-; CHECK-NEXT: vmsleu.vv v20, v12, v16, v0.t
+; CHECK-NEXT: vmsleu.vv v20, v8, v12
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmsleu.vv v8, v12, v16, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 16 x i1> @llvm.riscv.vmsleu.nxv16i16(
@@ -595,9 +606,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsleu.mask.nxv1i32(
define <vscale x 1 x i1> @intrinsic_vmsleu_mask_vv_nxv1i32_nxv1i32(<vscale x 1 x i1> %0, <vscale x 1 x i32> %1, <vscale x 1 x i32> %2, <vscale x 1 x i32> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsleu_mask_vv_nxv1i32_nxv1i32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
-; CHECK-NEXT: vmsleu.vv v0, v8, v9
+; CHECK-NEXT: vmsleu.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmsleu.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -646,9 +658,10 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsleu.mask.nxv2i32(
define <vscale x 2 x i1> @intrinsic_vmsleu_mask_vv_nxv2i32_nxv2i32(<vscale x 2 x i1> %0, <vscale x 2 x i32> %1, <vscale x 2 x i32> %2, <vscale x 2 x i32> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsleu_mask_vv_nxv2i32_nxv2i32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
-; CHECK-NEXT: vmsleu.vv v0, v8, v9
+; CHECK-NEXT: vmsleu.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmsleu.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -697,11 +710,12 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsleu.mask.nxv4i32(
define <vscale x 4 x i1> @intrinsic_vmsleu_mask_vv_nxv4i32_nxv4i32(<vscale x 4 x i1> %0, <vscale x 4 x i32> %1, <vscale x 4 x i32> %2, <vscale x 4 x i32> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsleu_mask_vv_nxv4i32_nxv4i32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
-; CHECK-NEXT: vmsleu.vv v0, v8, v10
-; CHECK-NEXT: vmsleu.vv v14, v10, v12, v0.t
+; CHECK-NEXT: vmsleu.vv v14, v8, v10
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmsleu.vv v8, v10, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 4 x i1> @llvm.riscv.vmsleu.nxv4i32(
@@ -748,11 +762,12 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsleu.mask.nxv8i32(
define <vscale x 8 x i1> @intrinsic_vmsleu_mask_vv_nxv8i32_nxv8i32(<vscale x 8 x i1> %0, <vscale x 8 x i32> %1, <vscale x 8 x i32> %2, <vscale x 8 x i32> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsleu_mask_vv_nxv8i32_nxv8i32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
-; CHECK-NEXT: vmsleu.vv v0, v8, v12
-; CHECK-NEXT: vmsleu.vv v20, v12, v16, v0.t
+; CHECK-NEXT: vmsleu.vv v20, v8, v12
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmsleu.vv v8, v12, v16, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 8 x i1> @llvm.riscv.vmsleu.nxv8i32(
@@ -799,9 +814,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsleu.mask.nxv1i64(
define <vscale x 1 x i1> @intrinsic_vmsleu_mask_vv_nxv1i64_nxv1i64(<vscale x 1 x i1> %0, <vscale x 1 x i64> %1, <vscale x 1 x i64> %2, <vscale x 1 x i64> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsleu_mask_vv_nxv1i64_nxv1i64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
-; CHECK-NEXT: vmsleu.vv v0, v8, v9
+; CHECK-NEXT: vmsleu.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmsleu.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -850,11 +866,12 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsleu.mask.nxv2i64(
define <vscale x 2 x i1> @intrinsic_vmsleu_mask_vv_nxv2i64_nxv2i64(<vscale x 2 x i1> %0, <vscale x 2 x i64> %1, <vscale x 2 x i64> %2, <vscale x 2 x i64> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsleu_mask_vv_nxv2i64_nxv2i64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
-; CHECK-NEXT: vmsleu.vv v0, v8, v10
-; CHECK-NEXT: vmsleu.vv v14, v10, v12, v0.t
+; CHECK-NEXT: vmsleu.vv v14, v8, v10
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmsleu.vv v8, v10, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 2 x i1> @llvm.riscv.vmsleu.nxv2i64(
@@ -901,11 +918,12 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsleu.mask.nxv4i64(
define <vscale x 4 x i1> @intrinsic_vmsleu_mask_vv_nxv4i64_nxv4i64(<vscale x 4 x i1> %0, <vscale x 4 x i64> %1, <vscale x 4 x i64> %2, <vscale x 4 x i64> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsleu_mask_vv_nxv4i64_nxv4i64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
-; CHECK-NEXT: vmsleu.vv v0, v8, v12
-; CHECK-NEXT: vmsleu.vv v20, v12, v16, v0.t
+; CHECK-NEXT: vmsleu.vv v20, v8, v12
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmsleu.vv v8, v12, v16, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 4 x i1> @llvm.riscv.vmsleu.nxv4i64(
@@ -953,8 +971,8 @@ define <vscale x 1 x i1> @intrinsic_vmsleu_mask_vx_nxv1i8_i8(<vscale x 1 x i1> %
; CHECK-LABEL: intrinsic_vmsleu_mask_vx_nxv1i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, mu
; CHECK-NEXT: vmsleu.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1000,8 +1018,8 @@ define <vscale x 2 x i1> @intrinsic_vmsleu_mask_vx_nxv2i8_i8(<vscale x 2 x i1> %
; CHECK-LABEL: intrinsic_vmsleu_mask_vx_nxv2i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e8, mf4, ta, mu
; CHECK-NEXT: vmsleu.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1047,8 +1065,8 @@ define <vscale x 4 x i1> @intrinsic_vmsleu_mask_vx_nxv4i8_i8(<vscale x 4 x i1> %
; CHECK-LABEL: intrinsic_vmsleu_mask_vx_nxv4i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e8, mf2, ta, mu
; CHECK-NEXT: vmsleu.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1094,8 +1112,8 @@ define <vscale x 8 x i1> @intrinsic_vmsleu_mask_vx_nxv8i8_i8(<vscale x 8 x i1> %
; CHECK-LABEL: intrinsic_vmsleu_mask_vx_nxv8i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e8, m1, ta, mu
; CHECK-NEXT: vmsleu.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -1141,8 +1159,8 @@ define <vscale x 16 x i1> @intrinsic_vmsleu_mask_vx_nxv16i8_i8(<vscale x 16 x i1
; CHECK-LABEL: intrinsic_vmsleu_mask_vx_nxv16i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a1, e8, m2, ta, mu
; CHECK-NEXT: vmsleu.vx v11, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -1188,8 +1206,8 @@ define <vscale x 32 x i1> @intrinsic_vmsleu_mask_vx_nxv32i8_i8(<vscale x 32 x i1
; CHECK-LABEL: intrinsic_vmsleu_mask_vx_nxv32i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a1, e8, m4, ta, mu
; CHECK-NEXT: vmsleu.vx v13, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -1235,8 +1253,8 @@ define <vscale x 1 x i1> @intrinsic_vmsleu_mask_vx_nxv1i16_i16(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmsleu_mask_vx_nxv1i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, mf4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e16, mf4, ta, mu
; CHECK-NEXT: vmsleu.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1282,8 +1300,8 @@ define <vscale x 2 x i1> @intrinsic_vmsleu_mask_vx_nxv2i16_i16(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmsleu_mask_vx_nxv2i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e16, mf2, ta, mu
; CHECK-NEXT: vmsleu.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1329,8 +1347,8 @@ define <vscale x 4 x i1> @intrinsic_vmsleu_mask_vx_nxv4i16_i16(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmsleu_mask_vx_nxv4i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e16, m1, ta, mu
; CHECK-NEXT: vmsleu.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -1376,8 +1394,8 @@ define <vscale x 8 x i1> @intrinsic_vmsleu_mask_vx_nxv8i16_i16(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmsleu_mask_vx_nxv8i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a1, e16, m2, ta, mu
; CHECK-NEXT: vmsleu.vx v11, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -1423,8 +1441,8 @@ define <vscale x 16 x i1> @intrinsic_vmsleu_mask_vx_nxv16i16_i16(<vscale x 16 x
; CHECK-LABEL: intrinsic_vmsleu_mask_vx_nxv16i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a1, e16, m4, ta, mu
; CHECK-NEXT: vmsleu.vx v13, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -1470,8 +1488,8 @@ define <vscale x 1 x i1> @intrinsic_vmsleu_mask_vx_nxv1i32_i32(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmsleu_mask_vx_nxv1i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e32, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e32, mf2, ta, mu
; CHECK-NEXT: vmsleu.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1517,8 +1535,8 @@ define <vscale x 2 x i1> @intrinsic_vmsleu_mask_vx_nxv2i32_i32(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmsleu_mask_vx_nxv2i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e32, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e32, m1, ta, mu
; CHECK-NEXT: vmsleu.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -1564,8 +1582,8 @@ define <vscale x 4 x i1> @intrinsic_vmsleu_mask_vx_nxv4i32_i32(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmsleu_mask_vx_nxv4i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a1, e32, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a1, e32, m2, ta, mu
; CHECK-NEXT: vmsleu.vx v11, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -1611,8 +1629,8 @@ define <vscale x 8 x i1> @intrinsic_vmsleu_mask_vx_nxv8i32_i32(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmsleu_mask_vx_nxv8i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a1, e32, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a1, e32, m4, ta, mu
; CHECK-NEXT: vmsleu.vx v13, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -1685,8 +1703,8 @@ define <vscale x 1 x i1> @intrinsic_vmsleu_mask_vx_nxv1i64_i64(<vscale x 1 x i1>
; RV64-LABEL: intrinsic_vmsleu_mask_vx_nxv1i64_i64:
; RV64: # %bb.0: # %entry
; RV64-NEXT: vmv1r.v v10, v0
-; RV64-NEXT: vsetvli zero, a1, e64, m1, ta, mu
; RV64-NEXT: vmv1r.v v0, v9
+; RV64-NEXT: vsetvli zero, a1, e64, m1, ta, mu
; RV64-NEXT: vmsleu.vx v10, v8, a0, v0.t
; RV64-NEXT: vmv.v.v v0, v10
; RV64-NEXT: ret
@@ -1759,8 +1777,8 @@ define <vscale x 2 x i1> @intrinsic_vmsleu_mask_vx_nxv2i64_i64(<vscale x 2 x i1>
; RV64-LABEL: intrinsic_vmsleu_mask_vx_nxv2i64_i64:
; RV64: # %bb.0: # %entry
; RV64-NEXT: vmv1r.v v11, v0
-; RV64-NEXT: vsetvli zero, a1, e64, m2, ta, mu
; RV64-NEXT: vmv1r.v v0, v10
+; RV64-NEXT: vsetvli zero, a1, e64, m2, ta, mu
; RV64-NEXT: vmsleu.vx v11, v8, a0, v0.t
; RV64-NEXT: vmv1r.v v0, v11
; RV64-NEXT: ret
@@ -1833,8 +1851,8 @@ define <vscale x 4 x i1> @intrinsic_vmsleu_mask_vx_nxv4i64_i64(<vscale x 4 x i1>
; RV64-LABEL: intrinsic_vmsleu_mask_vx_nxv4i64_i64:
; RV64: # %bb.0: # %entry
; RV64-NEXT: vmv1r.v v13, v0
-; RV64-NEXT: vsetvli zero, a1, e64, m4, ta, mu
; RV64-NEXT: vmv1r.v v0, v12
+; RV64-NEXT: vsetvli zero, a1, e64, m4, ta, mu
; RV64-NEXT: vmsleu.vx v13, v8, a0, v0.t
; RV64-NEXT: vmv1r.v v0, v13
; RV64-NEXT: ret
@@ -1868,8 +1886,8 @@ define <vscale x 1 x i1> @intrinsic_vmsleu_mask_vi_nxv1i8_i8(<vscale x 1 x i1> %
; CHECK-LABEL: intrinsic_vmsleu_mask_vi_nxv1i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, mu
; CHECK-NEXT: vmsleu.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1903,8 +1921,8 @@ define <vscale x 2 x i1> @intrinsic_vmsleu_mask_vi_nxv2i8_i8(<vscale x 2 x i1> %
; CHECK-LABEL: intrinsic_vmsleu_mask_vi_nxv2i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, mf4, ta, mu
; CHECK-NEXT: vmsleu.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1938,8 +1956,8 @@ define <vscale x 4 x i1> @intrinsic_vmsleu_mask_vi_nxv4i8_i8(<vscale x 4 x i1> %
; CHECK-LABEL: intrinsic_vmsleu_mask_vi_nxv4i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, mu
; CHECK-NEXT: vmsleu.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1973,8 +1991,8 @@ define <vscale x 8 x i1> @intrinsic_vmsleu_mask_vi_nxv8i8_i8(<vscale x 8 x i1> %
; CHECK-LABEL: intrinsic_vmsleu_mask_vi_nxv8i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, m1, ta, mu
; CHECK-NEXT: vmsleu.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -2008,8 +2026,8 @@ define <vscale x 16 x i1> @intrinsic_vmsleu_mask_vi_nxv16i8_i8(<vscale x 16 x i1
; CHECK-LABEL: intrinsic_vmsleu_mask_vi_nxv16i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e8, m2, ta, mu
; CHECK-NEXT: vmsleu.vi v11, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -2043,8 +2061,8 @@ define <vscale x 32 x i1> @intrinsic_vmsleu_mask_vi_nxv32i8_i8(<vscale x 32 x i1
; CHECK-LABEL: intrinsic_vmsleu_mask_vi_nxv32i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, mu
; CHECK-NEXT: vmsleu.vi v13, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -2078,8 +2096,8 @@ define <vscale x 1 x i1> @intrinsic_vmsleu_mask_vi_nxv1i16_i16(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmsleu_mask_vi_nxv1i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
; CHECK-NEXT: vmsleu.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -2113,8 +2131,8 @@ define <vscale x 2 x i1> @intrinsic_vmsleu_mask_vi_nxv2i16_i16(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmsleu_mask_vi_nxv2i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
; CHECK-NEXT: vmsleu.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -2148,8 +2166,8 @@ define <vscale x 4 x i1> @intrinsic_vmsleu_mask_vi_nxv4i16_i16(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmsleu_mask_vi_nxv4i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
; CHECK-NEXT: vmsleu.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -2183,8 +2201,8 @@ define <vscale x 8 x i1> @intrinsic_vmsleu_mask_vi_nxv8i16_i16(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmsleu_mask_vi_nxv8i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
; CHECK-NEXT: vmsleu.vi v11, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -2218,8 +2236,8 @@ define <vscale x 16 x i1> @intrinsic_vmsleu_mask_vi_nxv16i16_i16(<vscale x 16 x
; CHECK-LABEL: intrinsic_vmsleu_mask_vi_nxv16i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
; CHECK-NEXT: vmsleu.vi v13, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -2253,8 +2271,8 @@ define <vscale x 1 x i1> @intrinsic_vmsleu_mask_vi_nxv1i32_i32(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmsleu_mask_vi_nxv1i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
; CHECK-NEXT: vmsleu.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -2288,8 +2306,8 @@ define <vscale x 2 x i1> @intrinsic_vmsleu_mask_vi_nxv2i32_i32(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmsleu_mask_vi_nxv2i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
; CHECK-NEXT: vmsleu.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -2323,8 +2341,8 @@ define <vscale x 4 x i1> @intrinsic_vmsleu_mask_vi_nxv4i32_i32(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmsleu_mask_vi_nxv4i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
; CHECK-NEXT: vmsleu.vi v11, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -2358,8 +2376,8 @@ define <vscale x 8 x i1> @intrinsic_vmsleu_mask_vi_nxv8i32_i32(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmsleu_mask_vi_nxv8i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
; CHECK-NEXT: vmsleu.vi v13, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -2393,8 +2411,8 @@ define <vscale x 1 x i1> @intrinsic_vmsleu_mask_vi_nxv1i64_i64(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmsleu_mask_vi_nxv1i64_i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
; CHECK-NEXT: vmsleu.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -2428,8 +2446,8 @@ define <vscale x 2 x i1> @intrinsic_vmsleu_mask_vi_nxv2i64_i64(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmsleu_mask_vi_nxv2i64_i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
; CHECK-NEXT: vmsleu.vi v11, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -2463,8 +2481,8 @@ define <vscale x 4 x i1> @intrinsic_vmsleu_mask_vi_nxv4i64_i64(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmsleu_mask_vi_nxv4i64_i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
; CHECK-NEXT: vmsleu.vi v13, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vmslt.ll b/llvm/test/CodeGen/RISCV/rvv/vmslt.ll
index 6c6e580b043d1..554d25172d4fd 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vmslt.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vmslt.ll
@@ -34,9 +34,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmslt.mask.nxv1i8(
define <vscale x 1 x i1> @intrinsic_vmslt_mask_vv_nxv1i8_nxv1i8(<vscale x 1 x i1> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, <vscale x 1 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmslt_mask_vv_nxv1i8_nxv1i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, mu
-; CHECK-NEXT: vmslt.vv v0, v8, v9
+; CHECK-NEXT: vmslt.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmslt.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -85,9 +86,10 @@ declare <vscale x 2 x i1> @llvm.riscv.vmslt.mask.nxv2i8(
define <vscale x 2 x i1> @intrinsic_vmslt_mask_vv_nxv2i8_nxv2i8(<vscale x 2 x i1> %0, <vscale x 2 x i8> %1, <vscale x 2 x i8> %2, <vscale x 2 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmslt_mask_vv_nxv2i8_nxv2i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e8, mf4, ta, mu
-; CHECK-NEXT: vmslt.vv v0, v8, v9
+; CHECK-NEXT: vmslt.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmslt.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -136,9 +138,10 @@ declare <vscale x 4 x i1> @llvm.riscv.vmslt.mask.nxv4i8(
define <vscale x 4 x i1> @intrinsic_vmslt_mask_vv_nxv4i8_nxv4i8(<vscale x 4 x i1> %0, <vscale x 4 x i8> %1, <vscale x 4 x i8> %2, <vscale x 4 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmslt_mask_vv_nxv4i8_nxv4i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, mu
-; CHECK-NEXT: vmslt.vv v0, v8, v9
+; CHECK-NEXT: vmslt.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmslt.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -187,9 +190,10 @@ declare <vscale x 8 x i1> @llvm.riscv.vmslt.mask.nxv8i8(
define <vscale x 8 x i1> @intrinsic_vmslt_mask_vv_nxv8i8_nxv8i8(<vscale x 8 x i1> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, <vscale x 8 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmslt_mask_vv_nxv8i8_nxv8i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e8, m1, ta, mu
-; CHECK-NEXT: vmslt.vv v0, v8, v9
+; CHECK-NEXT: vmslt.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmslt.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -238,11 +242,12 @@ declare <vscale x 16 x i1> @llvm.riscv.vmslt.mask.nxv16i8(
define <vscale x 16 x i1> @intrinsic_vmslt_mask_vv_nxv16i8_nxv16i8(<vscale x 16 x i1> %0, <vscale x 16 x i8> %1, <vscale x 16 x i8> %2, <vscale x 16 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmslt_mask_vv_nxv16i8_nxv16i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e8, m2, ta, mu
-; CHECK-NEXT: vmslt.vv v0, v8, v10
-; CHECK-NEXT: vmslt.vv v14, v10, v12, v0.t
+; CHECK-NEXT: vmslt.vv v14, v8, v10
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmslt.vv v8, v10, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 16 x i1> @llvm.riscv.vmslt.nxv16i8(
@@ -289,11 +294,12 @@ declare <vscale x 32 x i1> @llvm.riscv.vmslt.mask.nxv32i8(
define <vscale x 32 x i1> @intrinsic_vmslt_mask_vv_nxv32i8_nxv32i8(<vscale x 32 x i1> %0, <vscale x 32 x i8> %1, <vscale x 32 x i8> %2, <vscale x 32 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmslt_mask_vv_nxv32i8_nxv32i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, mu
-; CHECK-NEXT: vmslt.vv v0, v8, v12
-; CHECK-NEXT: vmslt.vv v20, v12, v16, v0.t
+; CHECK-NEXT: vmslt.vv v20, v8, v12
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmslt.vv v8, v12, v16, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 32 x i1> @llvm.riscv.vmslt.nxv32i8(
@@ -340,9 +346,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmslt.mask.nxv1i16(
define <vscale x 1 x i1> @intrinsic_vmslt_mask_vv_nxv1i16_nxv1i16(<vscale x 1 x i1> %0, <vscale x 1 x i16> %1, <vscale x 1 x i16> %2, <vscale x 1 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmslt_mask_vv_nxv1i16_nxv1i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
-; CHECK-NEXT: vmslt.vv v0, v8, v9
+; CHECK-NEXT: vmslt.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmslt.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -391,9 +398,10 @@ declare <vscale x 2 x i1> @llvm.riscv.vmslt.mask.nxv2i16(
define <vscale x 2 x i1> @intrinsic_vmslt_mask_vv_nxv2i16_nxv2i16(<vscale x 2 x i1> %0, <vscale x 2 x i16> %1, <vscale x 2 x i16> %2, <vscale x 2 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmslt_mask_vv_nxv2i16_nxv2i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
-; CHECK-NEXT: vmslt.vv v0, v8, v9
+; CHECK-NEXT: vmslt.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmslt.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -442,9 +450,10 @@ declare <vscale x 4 x i1> @llvm.riscv.vmslt.mask.nxv4i16(
define <vscale x 4 x i1> @intrinsic_vmslt_mask_vv_nxv4i16_nxv4i16(<vscale x 4 x i1> %0, <vscale x 4 x i16> %1, <vscale x 4 x i16> %2, <vscale x 4 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmslt_mask_vv_nxv4i16_nxv4i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
-; CHECK-NEXT: vmslt.vv v0, v8, v9
+; CHECK-NEXT: vmslt.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmslt.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -493,11 +502,12 @@ declare <vscale x 8 x i1> @llvm.riscv.vmslt.mask.nxv8i16(
define <vscale x 8 x i1> @intrinsic_vmslt_mask_vv_nxv8i16_nxv8i16(<vscale x 8 x i1> %0, <vscale x 8 x i16> %1, <vscale x 8 x i16> %2, <vscale x 8 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmslt_mask_vv_nxv8i16_nxv8i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
-; CHECK-NEXT: vmslt.vv v0, v8, v10
-; CHECK-NEXT: vmslt.vv v14, v10, v12, v0.t
+; CHECK-NEXT: vmslt.vv v14, v8, v10
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmslt.vv v8, v10, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 8 x i1> @llvm.riscv.vmslt.nxv8i16(
@@ -544,11 +554,12 @@ declare <vscale x 16 x i1> @llvm.riscv.vmslt.mask.nxv16i16(
define <vscale x 16 x i1> @intrinsic_vmslt_mask_vv_nxv16i16_nxv16i16(<vscale x 16 x i1> %0, <vscale x 16 x i16> %1, <vscale x 16 x i16> %2, <vscale x 16 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmslt_mask_vv_nxv16i16_nxv16i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
-; CHECK-NEXT: vmslt.vv v0, v8, v12
-; CHECK-NEXT: vmslt.vv v20, v12, v16, v0.t
+; CHECK-NEXT: vmslt.vv v20, v8, v12
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmslt.vv v8, v12, v16, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 16 x i1> @llvm.riscv.vmslt.nxv16i16(
@@ -595,9 +606,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmslt.mask.nxv1i32(
define <vscale x 1 x i1> @intrinsic_vmslt_mask_vv_nxv1i32_nxv1i32(<vscale x 1 x i1> %0, <vscale x 1 x i32> %1, <vscale x 1 x i32> %2, <vscale x 1 x i32> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmslt_mask_vv_nxv1i32_nxv1i32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
-; CHECK-NEXT: vmslt.vv v0, v8, v9
+; CHECK-NEXT: vmslt.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmslt.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -646,9 +658,10 @@ declare <vscale x 2 x i1> @llvm.riscv.vmslt.mask.nxv2i32(
define <vscale x 2 x i1> @intrinsic_vmslt_mask_vv_nxv2i32_nxv2i32(<vscale x 2 x i1> %0, <vscale x 2 x i32> %1, <vscale x 2 x i32> %2, <vscale x 2 x i32> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmslt_mask_vv_nxv2i32_nxv2i32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
-; CHECK-NEXT: vmslt.vv v0, v8, v9
+; CHECK-NEXT: vmslt.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmslt.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -697,11 +710,12 @@ declare <vscale x 4 x i1> @llvm.riscv.vmslt.mask.nxv4i32(
define <vscale x 4 x i1> @intrinsic_vmslt_mask_vv_nxv4i32_nxv4i32(<vscale x 4 x i1> %0, <vscale x 4 x i32> %1, <vscale x 4 x i32> %2, <vscale x 4 x i32> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmslt_mask_vv_nxv4i32_nxv4i32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
-; CHECK-NEXT: vmslt.vv v0, v8, v10
-; CHECK-NEXT: vmslt.vv v14, v10, v12, v0.t
+; CHECK-NEXT: vmslt.vv v14, v8, v10
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmslt.vv v8, v10, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 4 x i1> @llvm.riscv.vmslt.nxv4i32(
@@ -748,11 +762,12 @@ declare <vscale x 8 x i1> @llvm.riscv.vmslt.mask.nxv8i32(
define <vscale x 8 x i1> @intrinsic_vmslt_mask_vv_nxv8i32_nxv8i32(<vscale x 8 x i1> %0, <vscale x 8 x i32> %1, <vscale x 8 x i32> %2, <vscale x 8 x i32> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmslt_mask_vv_nxv8i32_nxv8i32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
-; CHECK-NEXT: vmslt.vv v0, v8, v12
-; CHECK-NEXT: vmslt.vv v20, v12, v16, v0.t
+; CHECK-NEXT: vmslt.vv v20, v8, v12
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmslt.vv v8, v12, v16, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 8 x i1> @llvm.riscv.vmslt.nxv8i32(
@@ -799,9 +814,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmslt.mask.nxv1i64(
define <vscale x 1 x i1> @intrinsic_vmslt_mask_vv_nxv1i64_nxv1i64(<vscale x 1 x i1> %0, <vscale x 1 x i64> %1, <vscale x 1 x i64> %2, <vscale x 1 x i64> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmslt_mask_vv_nxv1i64_nxv1i64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
-; CHECK-NEXT: vmslt.vv v0, v8, v9
+; CHECK-NEXT: vmslt.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmslt.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -850,11 +866,12 @@ declare <vscale x 2 x i1> @llvm.riscv.vmslt.mask.nxv2i64(
define <vscale x 2 x i1> @intrinsic_vmslt_mask_vv_nxv2i64_nxv2i64(<vscale x 2 x i1> %0, <vscale x 2 x i64> %1, <vscale x 2 x i64> %2, <vscale x 2 x i64> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmslt_mask_vv_nxv2i64_nxv2i64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
-; CHECK-NEXT: vmslt.vv v0, v8, v10
-; CHECK-NEXT: vmslt.vv v14, v10, v12, v0.t
+; CHECK-NEXT: vmslt.vv v14, v8, v10
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmslt.vv v8, v10, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 2 x i1> @llvm.riscv.vmslt.nxv2i64(
@@ -901,11 +918,12 @@ declare <vscale x 4 x i1> @llvm.riscv.vmslt.mask.nxv4i64(
define <vscale x 4 x i1> @intrinsic_vmslt_mask_vv_nxv4i64_nxv4i64(<vscale x 4 x i1> %0, <vscale x 4 x i64> %1, <vscale x 4 x i64> %2, <vscale x 4 x i64> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmslt_mask_vv_nxv4i64_nxv4i64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
-; CHECK-NEXT: vmslt.vv v0, v8, v12
-; CHECK-NEXT: vmslt.vv v20, v12, v16, v0.t
+; CHECK-NEXT: vmslt.vv v20, v8, v12
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmslt.vv v8, v12, v16, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 4 x i1> @llvm.riscv.vmslt.nxv4i64(
@@ -953,8 +971,8 @@ define <vscale x 1 x i1> @intrinsic_vmslt_mask_vx_nxv1i8_i8(<vscale x 1 x i1> %0
; CHECK-LABEL: intrinsic_vmslt_mask_vx_nxv1i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, mu
; CHECK-NEXT: vmslt.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1000,8 +1018,8 @@ define <vscale x 2 x i1> @intrinsic_vmslt_mask_vx_nxv2i8_i8(<vscale x 2 x i1> %0
; CHECK-LABEL: intrinsic_vmslt_mask_vx_nxv2i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e8, mf4, ta, mu
; CHECK-NEXT: vmslt.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1047,8 +1065,8 @@ define <vscale x 4 x i1> @intrinsic_vmslt_mask_vx_nxv4i8_i8(<vscale x 4 x i1> %0
; CHECK-LABEL: intrinsic_vmslt_mask_vx_nxv4i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e8, mf2, ta, mu
; CHECK-NEXT: vmslt.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1094,8 +1112,8 @@ define <vscale x 8 x i1> @intrinsic_vmslt_mask_vx_nxv8i8_i8(<vscale x 8 x i1> %0
; CHECK-LABEL: intrinsic_vmslt_mask_vx_nxv8i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e8, m1, ta, mu
; CHECK-NEXT: vmslt.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -1141,8 +1159,8 @@ define <vscale x 16 x i1> @intrinsic_vmslt_mask_vx_nxv16i8_i8(<vscale x 16 x i1>
; CHECK-LABEL: intrinsic_vmslt_mask_vx_nxv16i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a1, e8, m2, ta, mu
; CHECK-NEXT: vmslt.vx v11, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -1188,8 +1206,8 @@ define <vscale x 32 x i1> @intrinsic_vmslt_mask_vx_nxv32i8_i8(<vscale x 32 x i1>
; CHECK-LABEL: intrinsic_vmslt_mask_vx_nxv32i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a1, e8, m4, ta, mu
; CHECK-NEXT: vmslt.vx v13, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -1235,8 +1253,8 @@ define <vscale x 1 x i1> @intrinsic_vmslt_mask_vx_nxv1i16_i16(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmslt_mask_vx_nxv1i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, mf4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e16, mf4, ta, mu
; CHECK-NEXT: vmslt.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1282,8 +1300,8 @@ define <vscale x 2 x i1> @intrinsic_vmslt_mask_vx_nxv2i16_i16(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmslt_mask_vx_nxv2i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e16, mf2, ta, mu
; CHECK-NEXT: vmslt.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1329,8 +1347,8 @@ define <vscale x 4 x i1> @intrinsic_vmslt_mask_vx_nxv4i16_i16(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmslt_mask_vx_nxv4i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e16, m1, ta, mu
; CHECK-NEXT: vmslt.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -1376,8 +1394,8 @@ define <vscale x 8 x i1> @intrinsic_vmslt_mask_vx_nxv8i16_i16(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmslt_mask_vx_nxv8i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a1, e16, m2, ta, mu
; CHECK-NEXT: vmslt.vx v11, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -1423,8 +1441,8 @@ define <vscale x 16 x i1> @intrinsic_vmslt_mask_vx_nxv16i16_i16(<vscale x 16 x i
; CHECK-LABEL: intrinsic_vmslt_mask_vx_nxv16i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a1, e16, m4, ta, mu
; CHECK-NEXT: vmslt.vx v13, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -1470,8 +1488,8 @@ define <vscale x 1 x i1> @intrinsic_vmslt_mask_vx_nxv1i32_i32(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmslt_mask_vx_nxv1i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e32, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e32, mf2, ta, mu
; CHECK-NEXT: vmslt.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1517,8 +1535,8 @@ define <vscale x 2 x i1> @intrinsic_vmslt_mask_vx_nxv2i32_i32(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmslt_mask_vx_nxv2i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e32, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e32, m1, ta, mu
; CHECK-NEXT: vmslt.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -1564,8 +1582,8 @@ define <vscale x 4 x i1> @intrinsic_vmslt_mask_vx_nxv4i32_i32(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmslt_mask_vx_nxv4i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a1, e32, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a1, e32, m2, ta, mu
; CHECK-NEXT: vmslt.vx v11, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -1611,8 +1629,8 @@ define <vscale x 8 x i1> @intrinsic_vmslt_mask_vx_nxv8i32_i32(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmslt_mask_vx_nxv8i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a1, e32, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a1, e32, m4, ta, mu
; CHECK-NEXT: vmslt.vx v13, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -1685,8 +1703,8 @@ define <vscale x 1 x i1> @intrinsic_vmslt_mask_vx_nxv1i64_i64(<vscale x 1 x i1>
; RV64-LABEL: intrinsic_vmslt_mask_vx_nxv1i64_i64:
; RV64: # %bb.0: # %entry
; RV64-NEXT: vmv1r.v v10, v0
-; RV64-NEXT: vsetvli zero, a1, e64, m1, ta, mu
; RV64-NEXT: vmv1r.v v0, v9
+; RV64-NEXT: vsetvli zero, a1, e64, m1, ta, mu
; RV64-NEXT: vmslt.vx v10, v8, a0, v0.t
; RV64-NEXT: vmv.v.v v0, v10
; RV64-NEXT: ret
@@ -1759,8 +1777,8 @@ define <vscale x 2 x i1> @intrinsic_vmslt_mask_vx_nxv2i64_i64(<vscale x 2 x i1>
; RV64-LABEL: intrinsic_vmslt_mask_vx_nxv2i64_i64:
; RV64: # %bb.0: # %entry
; RV64-NEXT: vmv1r.v v11, v0
-; RV64-NEXT: vsetvli zero, a1, e64, m2, ta, mu
; RV64-NEXT: vmv1r.v v0, v10
+; RV64-NEXT: vsetvli zero, a1, e64, m2, ta, mu
; RV64-NEXT: vmslt.vx v11, v8, a0, v0.t
; RV64-NEXT: vmv1r.v v0, v11
; RV64-NEXT: ret
@@ -1833,8 +1851,8 @@ define <vscale x 4 x i1> @intrinsic_vmslt_mask_vx_nxv4i64_i64(<vscale x 4 x i1>
; RV64-LABEL: intrinsic_vmslt_mask_vx_nxv4i64_i64:
; RV64: # %bb.0: # %entry
; RV64-NEXT: vmv1r.v v13, v0
-; RV64-NEXT: vsetvli zero, a1, e64, m4, ta, mu
; RV64-NEXT: vmv1r.v v0, v12
+; RV64-NEXT: vsetvli zero, a1, e64, m4, ta, mu
; RV64-NEXT: vmslt.vx v13, v8, a0, v0.t
; RV64-NEXT: vmv1r.v v0, v13
; RV64-NEXT: ret
@@ -1868,8 +1886,8 @@ define <vscale x 1 x i1> @intrinsic_vmslt_mask_vi_nxv1i8_i8(<vscale x 1 x i1> %0
; CHECK-LABEL: intrinsic_vmslt_mask_vi_nxv1i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, mu
; CHECK-NEXT: vmsle.vi v10, v8, -15, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1903,8 +1921,8 @@ define <vscale x 2 x i1> @intrinsic_vmslt_mask_vi_nxv2i8_i8(<vscale x 2 x i1> %0
; CHECK-LABEL: intrinsic_vmslt_mask_vi_nxv2i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, mf4, ta, mu
; CHECK-NEXT: vmsle.vi v10, v8, -13, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1938,8 +1956,8 @@ define <vscale x 4 x i1> @intrinsic_vmslt_mask_vi_nxv4i8_i8(<vscale x 4 x i1> %0
; CHECK-LABEL: intrinsic_vmslt_mask_vi_nxv4i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, mu
; CHECK-NEXT: vmsle.vi v10, v8, -11, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1973,8 +1991,8 @@ define <vscale x 8 x i1> @intrinsic_vmslt_mask_vi_nxv8i8_i8(<vscale x 8 x i1> %0
; CHECK-LABEL: intrinsic_vmslt_mask_vi_nxv8i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, m1, ta, mu
; CHECK-NEXT: vmsle.vi v10, v8, -9, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -2008,8 +2026,8 @@ define <vscale x 16 x i1> @intrinsic_vmslt_mask_vi_nxv16i8_i8(<vscale x 16 x i1>
; CHECK-LABEL: intrinsic_vmslt_mask_vi_nxv16i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e8, m2, ta, mu
; CHECK-NEXT: vmsle.vi v11, v8, -7, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -2043,8 +2061,8 @@ define <vscale x 32 x i1> @intrinsic_vmslt_mask_vi_nxv32i8_i8(<vscale x 32 x i1>
; CHECK-LABEL: intrinsic_vmslt_mask_vi_nxv32i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, mu
; CHECK-NEXT: vmsle.vi v13, v8, -5, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -2078,8 +2096,8 @@ define <vscale x 1 x i1> @intrinsic_vmslt_mask_vi_nxv1i16_i16(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmslt_mask_vi_nxv1i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
; CHECK-NEXT: vmsle.vi v10, v8, -3, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -2113,8 +2131,8 @@ define <vscale x 2 x i1> @intrinsic_vmslt_mask_vi_nxv2i16_i16(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmslt_mask_vi_nxv2i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
; CHECK-NEXT: vmslt.vx v10, v8, zero, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -2148,8 +2166,8 @@ define <vscale x 4 x i1> @intrinsic_vmslt_mask_vi_nxv4i16_i16(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmslt_mask_vi_nxv4i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
; CHECK-NEXT: vmsle.vi v10, v8, 0, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -2183,8 +2201,8 @@ define <vscale x 8 x i1> @intrinsic_vmslt_mask_vi_nxv8i16_i16(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmslt_mask_vi_nxv8i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
; CHECK-NEXT: vmsle.vi v11, v8, 2, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -2218,8 +2236,8 @@ define <vscale x 16 x i1> @intrinsic_vmslt_mask_vi_nxv16i16_i16(<vscale x 16 x i
; CHECK-LABEL: intrinsic_vmslt_mask_vi_nxv16i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
; CHECK-NEXT: vmsle.vi v13, v8, 4, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -2253,8 +2271,8 @@ define <vscale x 1 x i1> @intrinsic_vmslt_mask_vi_nxv1i32_i32(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmslt_mask_vi_nxv1i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
; CHECK-NEXT: vmsle.vi v10, v8, 6, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -2288,8 +2306,8 @@ define <vscale x 2 x i1> @intrinsic_vmslt_mask_vi_nxv2i32_i32(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmslt_mask_vi_nxv2i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
; CHECK-NEXT: vmsle.vi v10, v8, 8, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -2323,8 +2341,8 @@ define <vscale x 4 x i1> @intrinsic_vmslt_mask_vi_nxv4i32_i32(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmslt_mask_vi_nxv4i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
; CHECK-NEXT: vmsle.vi v11, v8, 10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -2358,8 +2376,8 @@ define <vscale x 8 x i1> @intrinsic_vmslt_mask_vi_nxv8i32_i32(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmslt_mask_vi_nxv8i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
; CHECK-NEXT: vmsle.vi v13, v8, 12, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -2393,8 +2411,8 @@ define <vscale x 1 x i1> @intrinsic_vmslt_mask_vi_nxv1i64_i64(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmslt_mask_vi_nxv1i64_i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
; CHECK-NEXT: vmsle.vi v10, v8, 8, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -2428,8 +2446,8 @@ define <vscale x 2 x i1> @intrinsic_vmslt_mask_vi_nxv2i64_i64(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmslt_mask_vi_nxv2i64_i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
; CHECK-NEXT: vmsle.vi v11, v8, 8, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -2463,8 +2481,8 @@ define <vscale x 4 x i1> @intrinsic_vmslt_mask_vi_nxv4i64_i64(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmslt_mask_vi_nxv4i64_i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
; CHECK-NEXT: vmsle.vi v13, v8, 8, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vmsltu.ll b/llvm/test/CodeGen/RISCV/rvv/vmsltu.ll
index 76f3e449ab58f..7a8efa6c80fb6 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vmsltu.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vmsltu.ll
@@ -34,9 +34,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsltu.mask.nxv1i8(
define <vscale x 1 x i1> @intrinsic_vmsltu_mask_vv_nxv1i8_nxv1i8(<vscale x 1 x i1> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, <vscale x 1 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsltu_mask_vv_nxv1i8_nxv1i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, mu
-; CHECK-NEXT: vmsltu.vv v0, v8, v9
+; CHECK-NEXT: vmsltu.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmsltu.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -85,9 +86,10 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsltu.mask.nxv2i8(
define <vscale x 2 x i1> @intrinsic_vmsltu_mask_vv_nxv2i8_nxv2i8(<vscale x 2 x i1> %0, <vscale x 2 x i8> %1, <vscale x 2 x i8> %2, <vscale x 2 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsltu_mask_vv_nxv2i8_nxv2i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e8, mf4, ta, mu
-; CHECK-NEXT: vmsltu.vv v0, v8, v9
+; CHECK-NEXT: vmsltu.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmsltu.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -136,9 +138,10 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsltu.mask.nxv4i8(
define <vscale x 4 x i1> @intrinsic_vmsltu_mask_vv_nxv4i8_nxv4i8(<vscale x 4 x i1> %0, <vscale x 4 x i8> %1, <vscale x 4 x i8> %2, <vscale x 4 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsltu_mask_vv_nxv4i8_nxv4i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, mu
-; CHECK-NEXT: vmsltu.vv v0, v8, v9
+; CHECK-NEXT: vmsltu.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmsltu.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -187,9 +190,10 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsltu.mask.nxv8i8(
define <vscale x 8 x i1> @intrinsic_vmsltu_mask_vv_nxv8i8_nxv8i8(<vscale x 8 x i1> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, <vscale x 8 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsltu_mask_vv_nxv8i8_nxv8i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e8, m1, ta, mu
-; CHECK-NEXT: vmsltu.vv v0, v8, v9
+; CHECK-NEXT: vmsltu.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmsltu.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -238,11 +242,12 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsltu.mask.nxv16i8(
define <vscale x 16 x i1> @intrinsic_vmsltu_mask_vv_nxv16i8_nxv16i8(<vscale x 16 x i1> %0, <vscale x 16 x i8> %1, <vscale x 16 x i8> %2, <vscale x 16 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsltu_mask_vv_nxv16i8_nxv16i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e8, m2, ta, mu
-; CHECK-NEXT: vmsltu.vv v0, v8, v10
-; CHECK-NEXT: vmsltu.vv v14, v10, v12, v0.t
+; CHECK-NEXT: vmsltu.vv v14, v8, v10
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmsltu.vv v8, v10, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 16 x i1> @llvm.riscv.vmsltu.nxv16i8(
@@ -289,11 +294,12 @@ declare <vscale x 32 x i1> @llvm.riscv.vmsltu.mask.nxv32i8(
define <vscale x 32 x i1> @intrinsic_vmsltu_mask_vv_nxv32i8_nxv32i8(<vscale x 32 x i1> %0, <vscale x 32 x i8> %1, <vscale x 32 x i8> %2, <vscale x 32 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsltu_mask_vv_nxv32i8_nxv32i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, mu
-; CHECK-NEXT: vmsltu.vv v0, v8, v12
-; CHECK-NEXT: vmsltu.vv v20, v12, v16, v0.t
+; CHECK-NEXT: vmsltu.vv v20, v8, v12
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmsltu.vv v8, v12, v16, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 32 x i1> @llvm.riscv.vmsltu.nxv32i8(
@@ -340,9 +346,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsltu.mask.nxv1i16(
define <vscale x 1 x i1> @intrinsic_vmsltu_mask_vv_nxv1i16_nxv1i16(<vscale x 1 x i1> %0, <vscale x 1 x i16> %1, <vscale x 1 x i16> %2, <vscale x 1 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsltu_mask_vv_nxv1i16_nxv1i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
-; CHECK-NEXT: vmsltu.vv v0, v8, v9
+; CHECK-NEXT: vmsltu.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmsltu.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -391,9 +398,10 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsltu.mask.nxv2i16(
define <vscale x 2 x i1> @intrinsic_vmsltu_mask_vv_nxv2i16_nxv2i16(<vscale x 2 x i1> %0, <vscale x 2 x i16> %1, <vscale x 2 x i16> %2, <vscale x 2 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsltu_mask_vv_nxv2i16_nxv2i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
-; CHECK-NEXT: vmsltu.vv v0, v8, v9
+; CHECK-NEXT: vmsltu.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmsltu.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -442,9 +450,10 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsltu.mask.nxv4i16(
define <vscale x 4 x i1> @intrinsic_vmsltu_mask_vv_nxv4i16_nxv4i16(<vscale x 4 x i1> %0, <vscale x 4 x i16> %1, <vscale x 4 x i16> %2, <vscale x 4 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsltu_mask_vv_nxv4i16_nxv4i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
-; CHECK-NEXT: vmsltu.vv v0, v8, v9
+; CHECK-NEXT: vmsltu.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmsltu.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -493,11 +502,12 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsltu.mask.nxv8i16(
define <vscale x 8 x i1> @intrinsic_vmsltu_mask_vv_nxv8i16_nxv8i16(<vscale x 8 x i1> %0, <vscale x 8 x i16> %1, <vscale x 8 x i16> %2, <vscale x 8 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsltu_mask_vv_nxv8i16_nxv8i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
-; CHECK-NEXT: vmsltu.vv v0, v8, v10
-; CHECK-NEXT: vmsltu.vv v14, v10, v12, v0.t
+; CHECK-NEXT: vmsltu.vv v14, v8, v10
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmsltu.vv v8, v10, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 8 x i1> @llvm.riscv.vmsltu.nxv8i16(
@@ -544,11 +554,12 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsltu.mask.nxv16i16(
define <vscale x 16 x i1> @intrinsic_vmsltu_mask_vv_nxv16i16_nxv16i16(<vscale x 16 x i1> %0, <vscale x 16 x i16> %1, <vscale x 16 x i16> %2, <vscale x 16 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsltu_mask_vv_nxv16i16_nxv16i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
-; CHECK-NEXT: vmsltu.vv v0, v8, v12
-; CHECK-NEXT: vmsltu.vv v20, v12, v16, v0.t
+; CHECK-NEXT: vmsltu.vv v20, v8, v12
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmsltu.vv v8, v12, v16, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 16 x i1> @llvm.riscv.vmsltu.nxv16i16(
@@ -595,9 +606,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsltu.mask.nxv1i32(
define <vscale x 1 x i1> @intrinsic_vmsltu_mask_vv_nxv1i32_nxv1i32(<vscale x 1 x i1> %0, <vscale x 1 x i32> %1, <vscale x 1 x i32> %2, <vscale x 1 x i32> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsltu_mask_vv_nxv1i32_nxv1i32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
-; CHECK-NEXT: vmsltu.vv v0, v8, v9
+; CHECK-NEXT: vmsltu.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmsltu.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -646,9 +658,10 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsltu.mask.nxv2i32(
define <vscale x 2 x i1> @intrinsic_vmsltu_mask_vv_nxv2i32_nxv2i32(<vscale x 2 x i1> %0, <vscale x 2 x i32> %1, <vscale x 2 x i32> %2, <vscale x 2 x i32> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsltu_mask_vv_nxv2i32_nxv2i32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
-; CHECK-NEXT: vmsltu.vv v0, v8, v9
+; CHECK-NEXT: vmsltu.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmsltu.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -697,11 +710,12 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsltu.mask.nxv4i32(
define <vscale x 4 x i1> @intrinsic_vmsltu_mask_vv_nxv4i32_nxv4i32(<vscale x 4 x i1> %0, <vscale x 4 x i32> %1, <vscale x 4 x i32> %2, <vscale x 4 x i32> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsltu_mask_vv_nxv4i32_nxv4i32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
-; CHECK-NEXT: vmsltu.vv v0, v8, v10
-; CHECK-NEXT: vmsltu.vv v14, v10, v12, v0.t
+; CHECK-NEXT: vmsltu.vv v14, v8, v10
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmsltu.vv v8, v10, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 4 x i1> @llvm.riscv.vmsltu.nxv4i32(
@@ -748,11 +762,12 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsltu.mask.nxv8i32(
define <vscale x 8 x i1> @intrinsic_vmsltu_mask_vv_nxv8i32_nxv8i32(<vscale x 8 x i1> %0, <vscale x 8 x i32> %1, <vscale x 8 x i32> %2, <vscale x 8 x i32> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsltu_mask_vv_nxv8i32_nxv8i32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
-; CHECK-NEXT: vmsltu.vv v0, v8, v12
-; CHECK-NEXT: vmsltu.vv v20, v12, v16, v0.t
+; CHECK-NEXT: vmsltu.vv v20, v8, v12
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmsltu.vv v8, v12, v16, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 8 x i1> @llvm.riscv.vmsltu.nxv8i32(
@@ -799,9 +814,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsltu.mask.nxv1i64(
define <vscale x 1 x i1> @intrinsic_vmsltu_mask_vv_nxv1i64_nxv1i64(<vscale x 1 x i1> %0, <vscale x 1 x i64> %1, <vscale x 1 x i64> %2, <vscale x 1 x i64> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsltu_mask_vv_nxv1i64_nxv1i64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
-; CHECK-NEXT: vmsltu.vv v0, v8, v9
+; CHECK-NEXT: vmsltu.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmsltu.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -850,11 +866,12 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsltu.mask.nxv2i64(
define <vscale x 2 x i1> @intrinsic_vmsltu_mask_vv_nxv2i64_nxv2i64(<vscale x 2 x i1> %0, <vscale x 2 x i64> %1, <vscale x 2 x i64> %2, <vscale x 2 x i64> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsltu_mask_vv_nxv2i64_nxv2i64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
-; CHECK-NEXT: vmsltu.vv v0, v8, v10
-; CHECK-NEXT: vmsltu.vv v14, v10, v12, v0.t
+; CHECK-NEXT: vmsltu.vv v14, v8, v10
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmsltu.vv v8, v10, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 2 x i1> @llvm.riscv.vmsltu.nxv2i64(
@@ -901,11 +918,12 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsltu.mask.nxv4i64(
define <vscale x 4 x i1> @intrinsic_vmsltu_mask_vv_nxv4i64_nxv4i64(<vscale x 4 x i1> %0, <vscale x 4 x i64> %1, <vscale x 4 x i64> %2, <vscale x 4 x i64> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsltu_mask_vv_nxv4i64_nxv4i64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
-; CHECK-NEXT: vmsltu.vv v0, v8, v12
-; CHECK-NEXT: vmsltu.vv v20, v12, v16, v0.t
+; CHECK-NEXT: vmsltu.vv v20, v8, v12
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmsltu.vv v8, v12, v16, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 4 x i1> @llvm.riscv.vmsltu.nxv4i64(
@@ -953,8 +971,8 @@ define <vscale x 1 x i1> @intrinsic_vmsltu_mask_vx_nxv1i8_i8(<vscale x 1 x i1> %
; CHECK-LABEL: intrinsic_vmsltu_mask_vx_nxv1i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, mu
; CHECK-NEXT: vmsltu.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1000,8 +1018,8 @@ define <vscale x 2 x i1> @intrinsic_vmsltu_mask_vx_nxv2i8_i8(<vscale x 2 x i1> %
; CHECK-LABEL: intrinsic_vmsltu_mask_vx_nxv2i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e8, mf4, ta, mu
; CHECK-NEXT: vmsltu.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1047,8 +1065,8 @@ define <vscale x 4 x i1> @intrinsic_vmsltu_mask_vx_nxv4i8_i8(<vscale x 4 x i1> %
; CHECK-LABEL: intrinsic_vmsltu_mask_vx_nxv4i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e8, mf2, ta, mu
; CHECK-NEXT: vmsltu.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1094,8 +1112,8 @@ define <vscale x 8 x i1> @intrinsic_vmsltu_mask_vx_nxv8i8_i8(<vscale x 8 x i1> %
; CHECK-LABEL: intrinsic_vmsltu_mask_vx_nxv8i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e8, m1, ta, mu
; CHECK-NEXT: vmsltu.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -1141,8 +1159,8 @@ define <vscale x 16 x i1> @intrinsic_vmsltu_mask_vx_nxv16i8_i8(<vscale x 16 x i1
; CHECK-LABEL: intrinsic_vmsltu_mask_vx_nxv16i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a1, e8, m2, ta, mu
; CHECK-NEXT: vmsltu.vx v11, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -1188,8 +1206,8 @@ define <vscale x 32 x i1> @intrinsic_vmsltu_mask_vx_nxv32i8_i8(<vscale x 32 x i1
; CHECK-LABEL: intrinsic_vmsltu_mask_vx_nxv32i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a1, e8, m4, ta, mu
; CHECK-NEXT: vmsltu.vx v13, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -1235,8 +1253,8 @@ define <vscale x 1 x i1> @intrinsic_vmsltu_mask_vx_nxv1i16_i16(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmsltu_mask_vx_nxv1i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, mf4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e16, mf4, ta, mu
; CHECK-NEXT: vmsltu.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1282,8 +1300,8 @@ define <vscale x 2 x i1> @intrinsic_vmsltu_mask_vx_nxv2i16_i16(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmsltu_mask_vx_nxv2i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e16, mf2, ta, mu
; CHECK-NEXT: vmsltu.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1329,8 +1347,8 @@ define <vscale x 4 x i1> @intrinsic_vmsltu_mask_vx_nxv4i16_i16(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmsltu_mask_vx_nxv4i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e16, m1, ta, mu
; CHECK-NEXT: vmsltu.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -1376,8 +1394,8 @@ define <vscale x 8 x i1> @intrinsic_vmsltu_mask_vx_nxv8i16_i16(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmsltu_mask_vx_nxv8i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a1, e16, m2, ta, mu
; CHECK-NEXT: vmsltu.vx v11, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -1423,8 +1441,8 @@ define <vscale x 16 x i1> @intrinsic_vmsltu_mask_vx_nxv16i16_i16(<vscale x 16 x
; CHECK-LABEL: intrinsic_vmsltu_mask_vx_nxv16i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a1, e16, m4, ta, mu
; CHECK-NEXT: vmsltu.vx v13, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -1470,8 +1488,8 @@ define <vscale x 1 x i1> @intrinsic_vmsltu_mask_vx_nxv1i32_i32(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmsltu_mask_vx_nxv1i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e32, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e32, mf2, ta, mu
; CHECK-NEXT: vmsltu.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1517,8 +1535,8 @@ define <vscale x 2 x i1> @intrinsic_vmsltu_mask_vx_nxv2i32_i32(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmsltu_mask_vx_nxv2i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e32, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e32, m1, ta, mu
; CHECK-NEXT: vmsltu.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -1564,8 +1582,8 @@ define <vscale x 4 x i1> @intrinsic_vmsltu_mask_vx_nxv4i32_i32(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmsltu_mask_vx_nxv4i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a1, e32, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a1, e32, m2, ta, mu
; CHECK-NEXT: vmsltu.vx v11, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -1611,8 +1629,8 @@ define <vscale x 8 x i1> @intrinsic_vmsltu_mask_vx_nxv8i32_i32(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmsltu_mask_vx_nxv8i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a1, e32, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a1, e32, m4, ta, mu
; CHECK-NEXT: vmsltu.vx v13, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -1685,8 +1703,8 @@ define <vscale x 1 x i1> @intrinsic_vmsltu_mask_vx_nxv1i64_i64(<vscale x 1 x i1>
; RV64-LABEL: intrinsic_vmsltu_mask_vx_nxv1i64_i64:
; RV64: # %bb.0: # %entry
; RV64-NEXT: vmv1r.v v10, v0
-; RV64-NEXT: vsetvli zero, a1, e64, m1, ta, mu
; RV64-NEXT: vmv1r.v v0, v9
+; RV64-NEXT: vsetvli zero, a1, e64, m1, ta, mu
; RV64-NEXT: vmsltu.vx v10, v8, a0, v0.t
; RV64-NEXT: vmv.v.v v0, v10
; RV64-NEXT: ret
@@ -1759,8 +1777,8 @@ define <vscale x 2 x i1> @intrinsic_vmsltu_mask_vx_nxv2i64_i64(<vscale x 2 x i1>
; RV64-LABEL: intrinsic_vmsltu_mask_vx_nxv2i64_i64:
; RV64: # %bb.0: # %entry
; RV64-NEXT: vmv1r.v v11, v0
-; RV64-NEXT: vsetvli zero, a1, e64, m2, ta, mu
; RV64-NEXT: vmv1r.v v0, v10
+; RV64-NEXT: vsetvli zero, a1, e64, m2, ta, mu
; RV64-NEXT: vmsltu.vx v11, v8, a0, v0.t
; RV64-NEXT: vmv1r.v v0, v11
; RV64-NEXT: ret
@@ -1833,8 +1851,8 @@ define <vscale x 4 x i1> @intrinsic_vmsltu_mask_vx_nxv4i64_i64(<vscale x 4 x i1>
; RV64-LABEL: intrinsic_vmsltu_mask_vx_nxv4i64_i64:
; RV64: # %bb.0: # %entry
; RV64-NEXT: vmv1r.v v13, v0
-; RV64-NEXT: vsetvli zero, a1, e64, m4, ta, mu
; RV64-NEXT: vmv1r.v v0, v12
+; RV64-NEXT: vsetvli zero, a1, e64, m4, ta, mu
; RV64-NEXT: vmsltu.vx v13, v8, a0, v0.t
; RV64-NEXT: vmv1r.v v0, v13
; RV64-NEXT: ret
@@ -1868,8 +1886,8 @@ define <vscale x 1 x i1> @intrinsic_vmsltu_mask_vi_nxv1i8_i8(<vscale x 1 x i1> %
; CHECK-LABEL: intrinsic_vmsltu_mask_vi_nxv1i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, mu
; CHECK-NEXT: vmsleu.vi v10, v8, -15, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1903,8 +1921,8 @@ define <vscale x 2 x i1> @intrinsic_vmsltu_mask_vi_nxv2i8_i8(<vscale x 2 x i1> %
; CHECK-LABEL: intrinsic_vmsltu_mask_vi_nxv2i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, mf4, ta, mu
; CHECK-NEXT: vmsleu.vi v10, v8, -13, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1938,8 +1956,8 @@ define <vscale x 4 x i1> @intrinsic_vmsltu_mask_vi_nxv4i8_i8(<vscale x 4 x i1> %
; CHECK-LABEL: intrinsic_vmsltu_mask_vi_nxv4i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, mu
; CHECK-NEXT: vmsleu.vi v10, v8, -11, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1973,8 +1991,8 @@ define <vscale x 8 x i1> @intrinsic_vmsltu_mask_vi_nxv8i8_i8(<vscale x 8 x i1> %
; CHECK-LABEL: intrinsic_vmsltu_mask_vi_nxv8i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, m1, ta, mu
; CHECK-NEXT: vmsleu.vi v10, v8, -9, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -2008,8 +2026,8 @@ define <vscale x 16 x i1> @intrinsic_vmsltu_mask_vi_nxv16i8_i8(<vscale x 16 x i1
; CHECK-LABEL: intrinsic_vmsltu_mask_vi_nxv16i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e8, m2, ta, mu
; CHECK-NEXT: vmsleu.vi v11, v8, -7, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -2043,8 +2061,8 @@ define <vscale x 32 x i1> @intrinsic_vmsltu_mask_vi_nxv32i8_i8(<vscale x 32 x i1
; CHECK-LABEL: intrinsic_vmsltu_mask_vi_nxv32i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, mu
; CHECK-NEXT: vmsleu.vi v13, v8, -5, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -2078,8 +2096,8 @@ define <vscale x 1 x i1> @intrinsic_vmsltu_mask_vi_nxv1i16_i16(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmsltu_mask_vi_nxv1i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
; CHECK-NEXT: vmsleu.vi v10, v8, -3, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -2113,8 +2131,8 @@ define <vscale x 2 x i1> @intrinsic_vmsltu_mask_vi_nxv2i16_i16(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmsltu_mask_vi_nxv2i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
; CHECK-NEXT: vmsltu.vx v10, v8, zero, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -2148,8 +2166,8 @@ define <vscale x 4 x i1> @intrinsic_vmsltu_mask_vi_nxv4i16_i16(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmsltu_mask_vi_nxv4i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
; CHECK-NEXT: vmsleu.vi v10, v8, 0, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -2183,8 +2201,8 @@ define <vscale x 8 x i1> @intrinsic_vmsltu_mask_vi_nxv8i16_i16(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmsltu_mask_vi_nxv8i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
; CHECK-NEXT: vmsleu.vi v11, v8, 2, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -2218,8 +2236,8 @@ define <vscale x 16 x i1> @intrinsic_vmsltu_mask_vi_nxv16i16_i16(<vscale x 16 x
; CHECK-LABEL: intrinsic_vmsltu_mask_vi_nxv16i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
; CHECK-NEXT: vmsleu.vi v13, v8, 4, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -2253,8 +2271,8 @@ define <vscale x 1 x i1> @intrinsic_vmsltu_mask_vi_nxv1i32_i32(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmsltu_mask_vi_nxv1i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
; CHECK-NEXT: vmsleu.vi v10, v8, 6, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -2288,8 +2306,8 @@ define <vscale x 2 x i1> @intrinsic_vmsltu_mask_vi_nxv2i32_i32(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmsltu_mask_vi_nxv2i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
; CHECK-NEXT: vmsleu.vi v10, v8, 8, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -2323,8 +2341,8 @@ define <vscale x 4 x i1> @intrinsic_vmsltu_mask_vi_nxv4i32_i32(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmsltu_mask_vi_nxv4i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
; CHECK-NEXT: vmsleu.vi v11, v8, 10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -2358,8 +2376,8 @@ define <vscale x 8 x i1> @intrinsic_vmsltu_mask_vi_nxv8i32_i32(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmsltu_mask_vi_nxv8i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
; CHECK-NEXT: vmsleu.vi v13, v8, 12, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -2393,8 +2411,8 @@ define <vscale x 1 x i1> @intrinsic_vmsltu_mask_vi_nxv1i64_i64(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmsltu_mask_vi_nxv1i64_i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
; CHECK-NEXT: vmsleu.vi v10, v8, 14, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -2428,8 +2446,8 @@ define <vscale x 2 x i1> @intrinsic_vmsltu_mask_vi_nxv2i64_i64(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmsltu_mask_vi_nxv2i64_i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
; CHECK-NEXT: vmsleu.vi v11, v8, -16, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -2463,8 +2481,8 @@ define <vscale x 4 x i1> @intrinsic_vmsltu_mask_vi_nxv4i64_i64(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmsltu_mask_vi_nxv4i64_i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
; CHECK-NEXT: vmsleu.vi v13, v8, -14, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vmsne.ll b/llvm/test/CodeGen/RISCV/rvv/vmsne.ll
index 161c1bc4314fc..bd6bd8a804bcc 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vmsne.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vmsne.ll
@@ -34,9 +34,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsne.mask.nxv1i8(
define <vscale x 1 x i1> @intrinsic_vmsne_mask_vv_nxv1i8_nxv1i8(<vscale x 1 x i1> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, <vscale x 1 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsne_mask_vv_nxv1i8_nxv1i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, mu
-; CHECK-NEXT: vmsne.vv v0, v8, v9
+; CHECK-NEXT: vmsne.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmsne.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -85,9 +86,10 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsne.mask.nxv2i8(
define <vscale x 2 x i1> @intrinsic_vmsne_mask_vv_nxv2i8_nxv2i8(<vscale x 2 x i1> %0, <vscale x 2 x i8> %1, <vscale x 2 x i8> %2, <vscale x 2 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsne_mask_vv_nxv2i8_nxv2i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e8, mf4, ta, mu
-; CHECK-NEXT: vmsne.vv v0, v8, v9
+; CHECK-NEXT: vmsne.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmsne.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -136,9 +138,10 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsne.mask.nxv4i8(
define <vscale x 4 x i1> @intrinsic_vmsne_mask_vv_nxv4i8_nxv4i8(<vscale x 4 x i1> %0, <vscale x 4 x i8> %1, <vscale x 4 x i8> %2, <vscale x 4 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsne_mask_vv_nxv4i8_nxv4i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, mu
-; CHECK-NEXT: vmsne.vv v0, v8, v9
+; CHECK-NEXT: vmsne.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmsne.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -187,9 +190,10 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsne.mask.nxv8i8(
define <vscale x 8 x i1> @intrinsic_vmsne_mask_vv_nxv8i8_nxv8i8(<vscale x 8 x i1> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, <vscale x 8 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsne_mask_vv_nxv8i8_nxv8i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e8, m1, ta, mu
-; CHECK-NEXT: vmsne.vv v0, v8, v9
+; CHECK-NEXT: vmsne.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmsne.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -238,11 +242,12 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsne.mask.nxv16i8(
define <vscale x 16 x i1> @intrinsic_vmsne_mask_vv_nxv16i8_nxv16i8(<vscale x 16 x i1> %0, <vscale x 16 x i8> %1, <vscale x 16 x i8> %2, <vscale x 16 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsne_mask_vv_nxv16i8_nxv16i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e8, m2, ta, mu
-; CHECK-NEXT: vmsne.vv v0, v8, v10
-; CHECK-NEXT: vmsne.vv v14, v10, v12, v0.t
+; CHECK-NEXT: vmsne.vv v14, v8, v10
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmsne.vv v8, v10, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 16 x i1> @llvm.riscv.vmsne.nxv16i8(
@@ -289,11 +294,12 @@ declare <vscale x 32 x i1> @llvm.riscv.vmsne.mask.nxv32i8(
define <vscale x 32 x i1> @intrinsic_vmsne_mask_vv_nxv32i8_nxv32i8(<vscale x 32 x i1> %0, <vscale x 32 x i8> %1, <vscale x 32 x i8> %2, <vscale x 32 x i8> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsne_mask_vv_nxv32i8_nxv32i8:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, mu
-; CHECK-NEXT: vmsne.vv v0, v8, v12
-; CHECK-NEXT: vmsne.vv v20, v12, v16, v0.t
+; CHECK-NEXT: vmsne.vv v20, v8, v12
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmsne.vv v8, v12, v16, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 32 x i1> @llvm.riscv.vmsne.nxv32i8(
@@ -340,9 +346,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsne.mask.nxv1i16(
define <vscale x 1 x i1> @intrinsic_vmsne_mask_vv_nxv1i16_nxv1i16(<vscale x 1 x i1> %0, <vscale x 1 x i16> %1, <vscale x 1 x i16> %2, <vscale x 1 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsne_mask_vv_nxv1i16_nxv1i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
-; CHECK-NEXT: vmsne.vv v0, v8, v9
+; CHECK-NEXT: vmsne.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmsne.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -391,9 +398,10 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsne.mask.nxv2i16(
define <vscale x 2 x i1> @intrinsic_vmsne_mask_vv_nxv2i16_nxv2i16(<vscale x 2 x i1> %0, <vscale x 2 x i16> %1, <vscale x 2 x i16> %2, <vscale x 2 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsne_mask_vv_nxv2i16_nxv2i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
-; CHECK-NEXT: vmsne.vv v0, v8, v9
+; CHECK-NEXT: vmsne.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmsne.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -442,9 +450,10 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsne.mask.nxv4i16(
define <vscale x 4 x i1> @intrinsic_vmsne_mask_vv_nxv4i16_nxv4i16(<vscale x 4 x i1> %0, <vscale x 4 x i16> %1, <vscale x 4 x i16> %2, <vscale x 4 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsne_mask_vv_nxv4i16_nxv4i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
-; CHECK-NEXT: vmsne.vv v0, v8, v9
+; CHECK-NEXT: vmsne.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmsne.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -493,11 +502,12 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsne.mask.nxv8i16(
define <vscale x 8 x i1> @intrinsic_vmsne_mask_vv_nxv8i16_nxv8i16(<vscale x 8 x i1> %0, <vscale x 8 x i16> %1, <vscale x 8 x i16> %2, <vscale x 8 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsne_mask_vv_nxv8i16_nxv8i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
-; CHECK-NEXT: vmsne.vv v0, v8, v10
-; CHECK-NEXT: vmsne.vv v14, v10, v12, v0.t
+; CHECK-NEXT: vmsne.vv v14, v8, v10
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmsne.vv v8, v10, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 8 x i1> @llvm.riscv.vmsne.nxv8i16(
@@ -544,11 +554,12 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsne.mask.nxv16i16(
define <vscale x 16 x i1> @intrinsic_vmsne_mask_vv_nxv16i16_nxv16i16(<vscale x 16 x i1> %0, <vscale x 16 x i16> %1, <vscale x 16 x i16> %2, <vscale x 16 x i16> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsne_mask_vv_nxv16i16_nxv16i16:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
-; CHECK-NEXT: vmsne.vv v0, v8, v12
-; CHECK-NEXT: vmsne.vv v20, v12, v16, v0.t
+; CHECK-NEXT: vmsne.vv v20, v8, v12
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmsne.vv v8, v12, v16, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 16 x i1> @llvm.riscv.vmsne.nxv16i16(
@@ -595,9 +606,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsne.mask.nxv1i32(
define <vscale x 1 x i1> @intrinsic_vmsne_mask_vv_nxv1i32_nxv1i32(<vscale x 1 x i1> %0, <vscale x 1 x i32> %1, <vscale x 1 x i32> %2, <vscale x 1 x i32> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsne_mask_vv_nxv1i32_nxv1i32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
-; CHECK-NEXT: vmsne.vv v0, v8, v9
+; CHECK-NEXT: vmsne.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vmsne.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -646,9 +658,10 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsne.mask.nxv2i32(
define <vscale x 2 x i1> @intrinsic_vmsne_mask_vv_nxv2i32_nxv2i32(<vscale x 2 x i1> %0, <vscale x 2 x i32> %1, <vscale x 2 x i32> %2, <vscale x 2 x i32> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsne_mask_vv_nxv2i32_nxv2i32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
-; CHECK-NEXT: vmsne.vv v0, v8, v9
+; CHECK-NEXT: vmsne.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmsne.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -697,11 +710,12 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsne.mask.nxv4i32(
define <vscale x 4 x i1> @intrinsic_vmsne_mask_vv_nxv4i32_nxv4i32(<vscale x 4 x i1> %0, <vscale x 4 x i32> %1, <vscale x 4 x i32> %2, <vscale x 4 x i32> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsne_mask_vv_nxv4i32_nxv4i32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
-; CHECK-NEXT: vmsne.vv v0, v8, v10
-; CHECK-NEXT: vmsne.vv v14, v10, v12, v0.t
+; CHECK-NEXT: vmsne.vv v14, v8, v10
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmsne.vv v8, v10, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 4 x i1> @llvm.riscv.vmsne.nxv4i32(
@@ -748,11 +762,12 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsne.mask.nxv8i32(
define <vscale x 8 x i1> @intrinsic_vmsne_mask_vv_nxv8i32_nxv8i32(<vscale x 8 x i1> %0, <vscale x 8 x i32> %1, <vscale x 8 x i32> %2, <vscale x 8 x i32> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsne_mask_vv_nxv8i32_nxv8i32:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
-; CHECK-NEXT: vmsne.vv v0, v8, v12
-; CHECK-NEXT: vmsne.vv v20, v12, v16, v0.t
+; CHECK-NEXT: vmsne.vv v20, v8, v12
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmsne.vv v8, v12, v16, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 8 x i1> @llvm.riscv.vmsne.nxv8i32(
@@ -799,9 +814,10 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsne.mask.nxv1i64(
define <vscale x 1 x i1> @intrinsic_vmsne_mask_vv_nxv1i64_nxv1i64(<vscale x 1 x i1> %0, <vscale x 1 x i64> %1, <vscale x 1 x i64> %2, <vscale x 1 x i64> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsne_mask_vv_nxv1i64_nxv1i64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v11, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
-; CHECK-NEXT: vmsne.vv v0, v8, v9
+; CHECK-NEXT: vmsne.vv v8, v8, v9
+; CHECK-NEXT: vmv1r.v v11, v0
+; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: vmsne.vv v11, v9, v10, v0.t
; CHECK-NEXT: vmv.v.v v0, v11
; CHECK-NEXT: ret
@@ -850,11 +866,12 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsne.mask.nxv2i64(
define <vscale x 2 x i1> @intrinsic_vmsne_mask_vv_nxv2i64_nxv2i64(<vscale x 2 x i1> %0, <vscale x 2 x i64> %1, <vscale x 2 x i64> %2, <vscale x 2 x i64> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsne_mask_vv_nxv2i64_nxv2i64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v14, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
-; CHECK-NEXT: vmsne.vv v0, v8, v10
-; CHECK-NEXT: vmsne.vv v14, v10, v12, v0.t
+; CHECK-NEXT: vmsne.vv v14, v8, v10
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v14
+; CHECK-NEXT: vmsne.vv v8, v10, v12, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 2 x i1> @llvm.riscv.vmsne.nxv2i64(
@@ -901,11 +918,12 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsne.mask.nxv4i64(
define <vscale x 4 x i1> @intrinsic_vmsne_mask_vv_nxv4i64_nxv4i64(<vscale x 4 x i1> %0, <vscale x 4 x i64> %1, <vscale x 4 x i64> %2, <vscale x 4 x i64> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vmsne_mask_vv_nxv4i64_nxv4i64:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: vmv1r.v v20, v0
; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
-; CHECK-NEXT: vmsne.vv v0, v8, v12
-; CHECK-NEXT: vmsne.vv v20, v12, v16, v0.t
+; CHECK-NEXT: vmsne.vv v20, v8, v12
+; CHECK-NEXT: vmv1r.v v8, v0
; CHECK-NEXT: vmv1r.v v0, v20
+; CHECK-NEXT: vmsne.vv v8, v12, v16, v0.t
+; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: ret
entry:
%mask = call <vscale x 4 x i1> @llvm.riscv.vmsne.nxv4i64(
@@ -953,8 +971,8 @@ define <vscale x 1 x i1> @intrinsic_vmsne_mask_vx_nxv1i8_i8(<vscale x 1 x i1> %0
; CHECK-LABEL: intrinsic_vmsne_mask_vx_nxv1i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, mu
; CHECK-NEXT: vmsne.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1000,8 +1018,8 @@ define <vscale x 2 x i1> @intrinsic_vmsne_mask_vx_nxv2i8_i8(<vscale x 2 x i1> %0
; CHECK-LABEL: intrinsic_vmsne_mask_vx_nxv2i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e8, mf4, ta, mu
; CHECK-NEXT: vmsne.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1047,8 +1065,8 @@ define <vscale x 4 x i1> @intrinsic_vmsne_mask_vx_nxv4i8_i8(<vscale x 4 x i1> %0
; CHECK-LABEL: intrinsic_vmsne_mask_vx_nxv4i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e8, mf2, ta, mu
; CHECK-NEXT: vmsne.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1094,8 +1112,8 @@ define <vscale x 8 x i1> @intrinsic_vmsne_mask_vx_nxv8i8_i8(<vscale x 8 x i1> %0
; CHECK-LABEL: intrinsic_vmsne_mask_vx_nxv8i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e8, m1, ta, mu
; CHECK-NEXT: vmsne.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -1141,8 +1159,8 @@ define <vscale x 16 x i1> @intrinsic_vmsne_mask_vx_nxv16i8_i8(<vscale x 16 x i1>
; CHECK-LABEL: intrinsic_vmsne_mask_vx_nxv16i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a1, e8, m2, ta, mu
; CHECK-NEXT: vmsne.vx v11, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -1188,8 +1206,8 @@ define <vscale x 32 x i1> @intrinsic_vmsne_mask_vx_nxv32i8_i8(<vscale x 32 x i1>
; CHECK-LABEL: intrinsic_vmsne_mask_vx_nxv32i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a1, e8, m4, ta, mu
; CHECK-NEXT: vmsne.vx v13, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -1235,8 +1253,8 @@ define <vscale x 1 x i1> @intrinsic_vmsne_mask_vx_nxv1i16_i16(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmsne_mask_vx_nxv1i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, mf4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e16, mf4, ta, mu
; CHECK-NEXT: vmsne.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1282,8 +1300,8 @@ define <vscale x 2 x i1> @intrinsic_vmsne_mask_vx_nxv2i16_i16(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmsne_mask_vx_nxv2i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e16, mf2, ta, mu
; CHECK-NEXT: vmsne.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1329,8 +1347,8 @@ define <vscale x 4 x i1> @intrinsic_vmsne_mask_vx_nxv4i16_i16(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmsne_mask_vx_nxv4i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e16, m1, ta, mu
; CHECK-NEXT: vmsne.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -1376,8 +1394,8 @@ define <vscale x 8 x i1> @intrinsic_vmsne_mask_vx_nxv8i16_i16(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmsne_mask_vx_nxv8i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a1, e16, m2, ta, mu
; CHECK-NEXT: vmsne.vx v11, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -1423,8 +1441,8 @@ define <vscale x 16 x i1> @intrinsic_vmsne_mask_vx_nxv16i16_i16(<vscale x 16 x i
; CHECK-LABEL: intrinsic_vmsne_mask_vx_nxv16i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a1, e16, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a1, e16, m4, ta, mu
; CHECK-NEXT: vmsne.vx v13, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -1470,8 +1488,8 @@ define <vscale x 1 x i1> @intrinsic_vmsne_mask_vx_nxv1i32_i32(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmsne_mask_vx_nxv1i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e32, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e32, mf2, ta, mu
; CHECK-NEXT: vmsne.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1517,8 +1535,8 @@ define <vscale x 2 x i1> @intrinsic_vmsne_mask_vx_nxv2i32_i32(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmsne_mask_vx_nxv2i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a1, e32, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e32, m1, ta, mu
; CHECK-NEXT: vmsne.vx v10, v8, a0, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -1564,8 +1582,8 @@ define <vscale x 4 x i1> @intrinsic_vmsne_mask_vx_nxv4i32_i32(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmsne_mask_vx_nxv4i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a1, e32, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a1, e32, m2, ta, mu
; CHECK-NEXT: vmsne.vx v11, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -1611,8 +1629,8 @@ define <vscale x 8 x i1> @intrinsic_vmsne_mask_vx_nxv8i32_i32(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmsne_mask_vx_nxv8i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a1, e32, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a1, e32, m4, ta, mu
; CHECK-NEXT: vmsne.vx v13, v8, a0, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -1685,8 +1703,8 @@ define <vscale x 1 x i1> @intrinsic_vmsne_mask_vx_nxv1i64_i64(<vscale x 1 x i1>
; RV64-LABEL: intrinsic_vmsne_mask_vx_nxv1i64_i64:
; RV64: # %bb.0: # %entry
; RV64-NEXT: vmv1r.v v10, v0
-; RV64-NEXT: vsetvli zero, a1, e64, m1, ta, mu
; RV64-NEXT: vmv1r.v v0, v9
+; RV64-NEXT: vsetvli zero, a1, e64, m1, ta, mu
; RV64-NEXT: vmsne.vx v10, v8, a0, v0.t
; RV64-NEXT: vmv.v.v v0, v10
; RV64-NEXT: ret
@@ -1759,8 +1777,8 @@ define <vscale x 2 x i1> @intrinsic_vmsne_mask_vx_nxv2i64_i64(<vscale x 2 x i1>
; RV64-LABEL: intrinsic_vmsne_mask_vx_nxv2i64_i64:
; RV64: # %bb.0: # %entry
; RV64-NEXT: vmv1r.v v11, v0
-; RV64-NEXT: vsetvli zero, a1, e64, m2, ta, mu
; RV64-NEXT: vmv1r.v v0, v10
+; RV64-NEXT: vsetvli zero, a1, e64, m2, ta, mu
; RV64-NEXT: vmsne.vx v11, v8, a0, v0.t
; RV64-NEXT: vmv1r.v v0, v11
; RV64-NEXT: ret
@@ -1833,8 +1851,8 @@ define <vscale x 4 x i1> @intrinsic_vmsne_mask_vx_nxv4i64_i64(<vscale x 4 x i1>
; RV64-LABEL: intrinsic_vmsne_mask_vx_nxv4i64_i64:
; RV64: # %bb.0: # %entry
; RV64-NEXT: vmv1r.v v13, v0
-; RV64-NEXT: vsetvli zero, a1, e64, m4, ta, mu
; RV64-NEXT: vmv1r.v v0, v12
+; RV64-NEXT: vsetvli zero, a1, e64, m4, ta, mu
; RV64-NEXT: vmsne.vx v13, v8, a0, v0.t
; RV64-NEXT: vmv1r.v v0, v13
; RV64-NEXT: ret
@@ -1868,8 +1886,8 @@ define <vscale x 1 x i1> @intrinsic_vmsne_mask_vi_nxv1i8_i8(<vscale x 1 x i1> %0
; CHECK-LABEL: intrinsic_vmsne_mask_vi_nxv1i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, mu
; CHECK-NEXT: vmsne.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1903,8 +1921,8 @@ define <vscale x 2 x i1> @intrinsic_vmsne_mask_vi_nxv2i8_i8(<vscale x 2 x i1> %0
; CHECK-LABEL: intrinsic_vmsne_mask_vi_nxv2i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, mf4, ta, mu
; CHECK-NEXT: vmsne.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1938,8 +1956,8 @@ define <vscale x 4 x i1> @intrinsic_vmsne_mask_vi_nxv4i8_i8(<vscale x 4 x i1> %0
; CHECK-LABEL: intrinsic_vmsne_mask_vi_nxv4i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, mu
; CHECK-NEXT: vmsne.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -1973,8 +1991,8 @@ define <vscale x 8 x i1> @intrinsic_vmsne_mask_vi_nxv8i8_i8(<vscale x 8 x i1> %0
; CHECK-LABEL: intrinsic_vmsne_mask_vi_nxv8i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, m1, ta, mu
; CHECK-NEXT: vmsne.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -2008,8 +2026,8 @@ define <vscale x 16 x i1> @intrinsic_vmsne_mask_vi_nxv16i8_i8(<vscale x 16 x i1>
; CHECK-LABEL: intrinsic_vmsne_mask_vi_nxv16i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e8, m2, ta, mu
; CHECK-NEXT: vmsne.vi v11, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -2043,8 +2061,8 @@ define <vscale x 32 x i1> @intrinsic_vmsne_mask_vi_nxv32i8_i8(<vscale x 32 x i1>
; CHECK-LABEL: intrinsic_vmsne_mask_vi_nxv32i8_i8:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, mu
; CHECK-NEXT: vmsne.vi v13, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -2078,8 +2096,8 @@ define <vscale x 1 x i1> @intrinsic_vmsne_mask_vi_nxv1i16_i16(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmsne_mask_vi_nxv1i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
; CHECK-NEXT: vmsne.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -2113,8 +2131,8 @@ define <vscale x 2 x i1> @intrinsic_vmsne_mask_vi_nxv2i16_i16(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmsne_mask_vi_nxv2i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
; CHECK-NEXT: vmsne.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -2148,8 +2166,8 @@ define <vscale x 4 x i1> @intrinsic_vmsne_mask_vi_nxv4i16_i16(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmsne_mask_vi_nxv4i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
; CHECK-NEXT: vmsne.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -2183,8 +2201,8 @@ define <vscale x 8 x i1> @intrinsic_vmsne_mask_vi_nxv8i16_i16(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmsne_mask_vi_nxv8i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
; CHECK-NEXT: vmsne.vi v11, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -2218,8 +2236,8 @@ define <vscale x 16 x i1> @intrinsic_vmsne_mask_vi_nxv16i16_i16(<vscale x 16 x i
; CHECK-LABEL: intrinsic_vmsne_mask_vi_nxv16i16_i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
; CHECK-NEXT: vmsne.vi v13, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -2253,8 +2271,8 @@ define <vscale x 1 x i1> @intrinsic_vmsne_mask_vi_nxv1i32_i32(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmsne_mask_vi_nxv1i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
; CHECK-NEXT: vmsne.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -2288,8 +2306,8 @@ define <vscale x 2 x i1> @intrinsic_vmsne_mask_vi_nxv2i32_i32(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmsne_mask_vi_nxv2i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
; CHECK-NEXT: vmsne.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -2323,8 +2341,8 @@ define <vscale x 4 x i1> @intrinsic_vmsne_mask_vi_nxv4i32_i32(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmsne_mask_vi_nxv4i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
; CHECK-NEXT: vmsne.vi v11, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -2358,8 +2376,8 @@ define <vscale x 8 x i1> @intrinsic_vmsne_mask_vi_nxv8i32_i32(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmsne_mask_vi_nxv8i32_i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
; CHECK-NEXT: vmsne.vi v13, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
@@ -2393,8 +2411,8 @@ define <vscale x 1 x i1> @intrinsic_vmsne_mask_vi_nxv1i64_i64(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmsne_mask_vi_nxv1i64_i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, mu
; CHECK-NEXT: vmsne.vi v10, v8, 9, v0.t
; CHECK-NEXT: vmv.v.v v0, v10
; CHECK-NEXT: ret
@@ -2428,8 +2446,8 @@ define <vscale x 2 x i1> @intrinsic_vmsne_mask_vi_nxv2i64_i64(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmsne_mask_vi_nxv2i64_i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v11, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, mu
; CHECK-NEXT: vmsne.vi v11, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v11
; CHECK-NEXT: ret
@@ -2463,8 +2481,8 @@ define <vscale x 4 x i1> @intrinsic_vmsne_mask_vi_nxv4i64_i64(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmsne_mask_vi_nxv4i64_i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v13, v0
-; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, mu
; CHECK-NEXT: vmsne.vi v13, v8, 9, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vmsof.ll b/llvm/test/CodeGen/RISCV/rvv/vmsof.ll
index f6f90eddcd8c5..0c60681ea8de0 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vmsof.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vmsof.ll
@@ -32,8 +32,8 @@ define <vscale x 1 x i1> @intrinsic_vmsof_mask_m_nxv1i1_nxv1i1(<vscale x 1 x i1>
; CHECK-LABEL: intrinsic_vmsof_mask_m_nxv1i1_nxv1i1:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf8, tu, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, mf8, tu, mu
; CHECK-NEXT: vmsof.m v10, v8, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -74,8 +74,8 @@ define <vscale x 2 x i1> @intrinsic_vmsof_mask_m_nxv2i1_nxv2i1(<vscale x 2 x i1>
; CHECK-LABEL: intrinsic_vmsof_mask_m_nxv2i1_nxv2i1:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf4, tu, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, mf4, tu, mu
; CHECK-NEXT: vmsof.m v10, v8, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -116,8 +116,8 @@ define <vscale x 4 x i1> @intrinsic_vmsof_mask_m_nxv4i1_nxv4i1(<vscale x 4 x i1>
; CHECK-LABEL: intrinsic_vmsof_mask_m_nxv4i1_nxv4i1:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, mf2, tu, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, mf2, tu, mu
; CHECK-NEXT: vmsof.m v10, v8, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -158,8 +158,8 @@ define <vscale x 8 x i1> @intrinsic_vmsof_mask_m_nxv8i1_nxv8i1(<vscale x 8 x i1>
; CHECK-LABEL: intrinsic_vmsof_mask_m_nxv8i1_nxv8i1:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m1, tu, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, m1, tu, mu
; CHECK-NEXT: vmsof.m v10, v8, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -200,8 +200,8 @@ define <vscale x 16 x i1> @intrinsic_vmsof_mask_m_nxv16i1_nxv16i1(<vscale x 16 x
; CHECK-LABEL: intrinsic_vmsof_mask_m_nxv16i1_nxv16i1:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m2, tu, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, m2, tu, mu
; CHECK-NEXT: vmsof.m v10, v8, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -242,8 +242,8 @@ define <vscale x 32 x i1> @intrinsic_vmsof_mask_m_nxv32i1_nxv32i1(<vscale x 32 x
; CHECK-LABEL: intrinsic_vmsof_mask_m_nxv32i1_nxv32i1:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m4, tu, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, m4, tu, mu
; CHECK-NEXT: vmsof.m v10, v8, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
@@ -284,8 +284,8 @@ define <vscale x 64 x i1> @intrinsic_vmsof_mask_m_nxv64i1_nxv64i1(<vscale x 64 x
; CHECK-LABEL: intrinsic_vmsof_mask_m_nxv64i1_nxv64i1:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vmv1r.v v10, v0
-; CHECK-NEXT: vsetvli zero, a0, e8, m8, tu, mu
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, m8, tu, mu
; CHECK-NEXT: vmsof.m v10, v8, v0.t
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vp-cttz-elts.ll b/llvm/test/CodeGen/RISCV/rvv/vp-cttz-elts.ll
index 8b368bfaab08e..f7ca65801dc87 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vp-cttz-elts.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vp-cttz-elts.ll
@@ -6,8 +6,8 @@ define iXLen @bool_vec(<vscale x 2 x i1> %src, <vscale x 2 x i1> %m, i32 %evl) {
; RV32-LABEL: bool_vec:
; RV32: # %bb.0:
; RV32-NEXT: vmv1r.v v9, v0
-; RV32-NEXT: vsetvli zero, a0, e8, mf4, ta, ma
; RV32-NEXT: vmv1r.v v0, v8
+; RV32-NEXT: vsetvli zero, a0, e8, mf4, ta, ma
; RV32-NEXT: vfirst.m a1, v9, v0.t
; RV32-NEXT: bltz a1, .LBB0_2
; RV32-NEXT: # %bb.1:
@@ -20,8 +20,8 @@ define iXLen @bool_vec(<vscale x 2 x i1> %src, <vscale x 2 x i1> %m, i32 %evl) {
; RV64-NEXT: vmv1r.v v9, v0
; RV64-NEXT: slli a0, a0, 32
; RV64-NEXT: srli a0, a0, 32
-; RV64-NEXT: vsetvli zero, a0, e8, mf4, ta, ma
; RV64-NEXT: vmv1r.v v0, v8
+; RV64-NEXT: vsetvli zero, a0, e8, mf4, ta, ma
; RV64-NEXT: vfirst.m a1, v9, v0.t
; RV64-NEXT: bltz a1, .LBB0_2
; RV64-NEXT: # %bb.1:
@@ -36,8 +36,8 @@ define iXLen @bool_vec_zero_poison(<vscale x 2 x i1> %src, <vscale x 2 x i1> %m,
; RV32-LABEL: bool_vec_zero_poison:
; RV32: # %bb.0:
; RV32-NEXT: vmv1r.v v9, v0
-; RV32-NEXT: vsetvli zero, a0, e8, mf4, ta, ma
; RV32-NEXT: vmv1r.v v0, v8
+; RV32-NEXT: vsetvli zero, a0, e8, mf4, ta, ma
; RV32-NEXT: vfirst.m a0, v9, v0.t
; RV32-NEXT: ret
;
@@ -46,8 +46,8 @@ define iXLen @bool_vec_zero_poison(<vscale x 2 x i1> %src, <vscale x 2 x i1> %m,
; RV64-NEXT: vmv1r.v v9, v0
; RV64-NEXT: slli a0, a0, 32
; RV64-NEXT: srli a0, a0, 32
-; RV64-NEXT: vsetvli zero, a0, e8, mf4, ta, ma
; RV64-NEXT: vmv1r.v v0, v8
+; RV64-NEXT: vsetvli zero, a0, e8, mf4, ta, ma
; RV64-NEXT: vfirst.m a0, v9, v0.t
; RV64-NEXT: ret
%r = call iXLen @llvm.vp.cttz.elts.iXLen.nxv2i1(<vscale x 2 x i1> %src, i1 1, <vscale x 2 x i1> %m, i32 %evl)
diff --git a/llvm/test/CodeGen/RISCV/rvv/vp-reverse-int.ll b/llvm/test/CodeGen/RISCV/rvv/vp-reverse-int.ll
index 8b1660283cb7d..d0f2ce1ca8004 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vp-reverse-int.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vp-reverse-int.ll
@@ -450,14 +450,14 @@ define <vscale x 64 x i8> @test_vp_reverse_nxv64i8_masked(<vscale x 64 x i8> %sr
; CHECK-NEXT: addi a2, a2, -1
; CHECK-NEXT: vsetvli a3, zero, e16, m8, ta, ma
; CHECK-NEXT: vid.v v16
-; CHECK-NEXT: vrsub.vx v24, v16, a2
+; CHECK-NEXT: vrsub.vx v16, v16, a2
; CHECK-NEXT: vsetvli zero, zero, e8, m4, ta, ma
-; CHECK-NEXT: vrgatherei16.vv v20, v8, v24
-; CHECK-NEXT: vrgatherei16.vv v16, v12, v24
+; CHECK-NEXT: vrgatherei16.vv v28, v8, v16
+; CHECK-NEXT: vrgatherei16.vv v24, v12, v16
; CHECK-NEXT: slli a1, a1, 3
; CHECK-NEXT: sub a1, a1, a0
; CHECK-NEXT: vsetvli zero, a0, e8, m8, ta, ma
-; CHECK-NEXT: vslidedown.vx v8, v16, a1, v0.t
+; CHECK-NEXT: vslidedown.vx v8, v24, a1, v0.t
; CHECK-NEXT: ret
%dst = call <vscale x 64 x i8> @llvm.experimental.vp.reverse.nxv64i8(<vscale x 64 x i8> %src, <vscale x 64 x i1> %mask, i32 %evl)
ret <vscale x 64 x i8> %dst
diff --git a/llvm/test/CodeGen/RISCV/rvv/vp-reverse-mask-fixed-vectors.ll b/llvm/test/CodeGen/RISCV/rvv/vp-reverse-mask-fixed-vectors.ll
index a30ebf2d33b50..7f81b99eb0338 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vp-reverse-mask-fixed-vectors.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vp-reverse-mask-fixed-vectors.ll
@@ -8,8 +8,8 @@ define <2 x i1> @test_vp_reverse_v2i1_masked(<2 x i1> %src, <2 x i1> %mask, i32
; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, ma
; CHECK-NEXT: vmv.v.i v9, 0
; CHECK-NEXT: vmerge.vim v9, v9, 1, v0
-; CHECK-NEXT: vsetvli zero, zero, e16, mf4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, zero, e16, mf4, ta, ma
; CHECK-NEXT: vid.v v10, v0.t
; CHECK-NEXT: addi a0, a0, -1
; CHECK-NEXT: vrsub.vx v10, v10, a0, v0.t
@@ -45,8 +45,8 @@ define <4 x i1> @test_vp_reverse_v4i1_masked(<4 x i1> %src, <4 x i1> %mask, i32
; CHECK-NEXT: vsetvli zero, a0, e8, mf4, ta, ma
; CHECK-NEXT: vmv.v.i v9, 0
; CHECK-NEXT: vmerge.vim v9, v9, 1, v0
-; CHECK-NEXT: vsetvli zero, zero, e16, mf2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, zero, e16, mf2, ta, ma
; CHECK-NEXT: vid.v v10, v0.t
; CHECK-NEXT: addi a0, a0, -1
; CHECK-NEXT: vrsub.vx v10, v10, a0, v0.t
@@ -82,8 +82,8 @@ define <8 x i1> @test_vp_reverse_v8i1_masked(<8 x i1> %src, <8 x i1> %mask, i32
; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, ma
; CHECK-NEXT: vmv.v.i v9, 0
; CHECK-NEXT: vmerge.vim v9, v9, 1, v0
-; CHECK-NEXT: vsetvli zero, zero, e16, m1, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, zero, e16, m1, ta, ma
; CHECK-NEXT: vid.v v10, v0.t
; CHECK-NEXT: addi a0, a0, -1
; CHECK-NEXT: vrsub.vx v10, v10, a0, v0.t
@@ -119,8 +119,8 @@ define <16 x i1> @test_vp_reverse_v16i1_masked(<16 x i1> %src, <16 x i1> %mask,
; CHECK-NEXT: vsetvli zero, a0, e8, m1, ta, ma
; CHECK-NEXT: vmv.v.i v9, 0
; CHECK-NEXT: vmerge.vim v9, v9, 1, v0
-; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; CHECK-NEXT: vid.v v10, v0.t
; CHECK-NEXT: addi a0, a0, -1
; CHECK-NEXT: vrsub.vx v10, v10, a0, v0.t
diff --git a/llvm/test/CodeGen/RISCV/rvv/vp-reverse-mask.ll b/llvm/test/CodeGen/RISCV/rvv/vp-reverse-mask.ll
index ceb6a164e20df..acf7d16bda982 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vp-reverse-mask.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vp-reverse-mask.ll
@@ -7,8 +7,8 @@ define <vscale x 1 x i1> @test_vp_reverse_nxv1i1_masked(<vscale x 1 x i1> %src,
; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, ma
; CHECK-NEXT: vmv.v.i v9, 0
; CHECK-NEXT: vmerge.vim v9, v9, 1, v0
-; CHECK-NEXT: vsetvli zero, zero, e16, mf4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, zero, e16, mf4, ta, ma
; CHECK-NEXT: vid.v v10, v0.t
; CHECK-NEXT: addi a0, a0, -1
; CHECK-NEXT: vrsub.vx v10, v10, a0, v0.t
@@ -44,8 +44,8 @@ define <vscale x 2 x i1> @test_vp_reverse_nxv2i1_masked(<vscale x 2 x i1> %src,
; CHECK-NEXT: vsetvli zero, a0, e8, mf4, ta, ma
; CHECK-NEXT: vmv.v.i v9, 0
; CHECK-NEXT: vmerge.vim v9, v9, 1, v0
-; CHECK-NEXT: vsetvli zero, zero, e16, mf2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, zero, e16, mf2, ta, ma
; CHECK-NEXT: vid.v v10, v0.t
; CHECK-NEXT: addi a0, a0, -1
; CHECK-NEXT: vrsub.vx v10, v10, a0, v0.t
@@ -81,8 +81,8 @@ define <vscale x 4 x i1> @test_vp_reverse_nxv4i1_masked(<vscale x 4 x i1> %src,
; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, ma
; CHECK-NEXT: vmv.v.i v9, 0
; CHECK-NEXT: vmerge.vim v9, v9, 1, v0
-; CHECK-NEXT: vsetvli zero, zero, e16, m1, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, zero, e16, m1, ta, ma
; CHECK-NEXT: vid.v v10, v0.t
; CHECK-NEXT: addi a0, a0, -1
; CHECK-NEXT: vrsub.vx v10, v10, a0, v0.t
@@ -118,8 +118,8 @@ define <vscale x 8 x i1> @test_vp_reverse_nxv8i1_masked(<vscale x 8 x i1> %src,
; CHECK-NEXT: vsetvli zero, a0, e8, m1, ta, ma
; CHECK-NEXT: vmv.v.i v9, 0
; CHECK-NEXT: vmerge.vim v9, v9, 1, v0
-; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; CHECK-NEXT: vid.v v10, v0.t
; CHECK-NEXT: addi a0, a0, -1
; CHECK-NEXT: vrsub.vx v10, v10, a0, v0.t
@@ -155,8 +155,8 @@ define <vscale x 16 x i1> @test_vp_reverse_nxv16i1_masked(<vscale x 16 x i1> %sr
; CHECK-NEXT: vsetvli zero, a0, e8, m2, ta, ma
; CHECK-NEXT: vmv.v.i v10, 0
; CHECK-NEXT: vmerge.vim v10, v10, 1, v0
-; CHECK-NEXT: vsetvli zero, zero, e16, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, zero, e16, m4, ta, ma
; CHECK-NEXT: vid.v v12, v0.t
; CHECK-NEXT: addi a0, a0, -1
; CHECK-NEXT: vrsub.vx v12, v12, a0, v0.t
@@ -193,8 +193,8 @@ define <vscale x 32 x i1> @test_vp_reverse_nxv32i1_masked(<vscale x 32 x i1> %sr
; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, ma
; CHECK-NEXT: vmv.v.i v12, 0
; CHECK-NEXT: vmerge.vim v12, v12, 1, v0
-; CHECK-NEXT: vsetvli zero, zero, e16, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, zero, e16, m8, ta, ma
; CHECK-NEXT: vid.v v16, v0.t
; CHECK-NEXT: addi a0, a0, -1
; CHECK-NEXT: vrsub.vx v16, v16, a0, v0.t
@@ -242,8 +242,8 @@ define <vscale x 64 x i1> @test_vp_reverse_nxv64i1_masked(<vscale x 64 x i1> %sr
; CHECK-NEXT: vrgatherei16.vv v16, v28, v0
; CHECK-NEXT: slli a1, a1, 3
; CHECK-NEXT: sub a1, a1, a0
-; CHECK-NEXT: vsetvli zero, a0, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a0, e8, m8, ta, ma
; CHECK-NEXT: vslidedown.vx v16, v16, a1, v0.t
; CHECK-NEXT: vmsne.vi v8, v16, 0, v0.t
; CHECK-NEXT: vmv1r.v v0, v8
@@ -263,14 +263,14 @@ define <vscale x 64 x i1> @test_vp_reverse_nxv64i1(<vscale x 64 x i1> %src, i32
; CHECK-NEXT: addi a2, a2, -1
; CHECK-NEXT: vsetvli a3, zero, e16, m8, ta, ma
; CHECK-NEXT: vid.v v16
-; CHECK-NEXT: vrsub.vx v24, v16, a2
+; CHECK-NEXT: vrsub.vx v16, v16, a2
; CHECK-NEXT: vsetvli zero, zero, e8, m4, ta, ma
-; CHECK-NEXT: vrgatherei16.vv v20, v8, v24
-; CHECK-NEXT: vrgatherei16.vv v16, v12, v24
+; CHECK-NEXT: vrgatherei16.vv v28, v8, v16
+; CHECK-NEXT: vrgatherei16.vv v24, v12, v16
; CHECK-NEXT: slli a1, a1, 3
; CHECK-NEXT: sub a1, a1, a0
; CHECK-NEXT: vsetvli zero, a0, e8, m8, ta, ma
-; CHECK-NEXT: vslidedown.vx v8, v16, a1
+; CHECK-NEXT: vslidedown.vx v8, v24, a1
; CHECK-NEXT: vmsne.vi v0, v8, 0
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vp-splice-mask-fixed-vectors.ll b/llvm/test/CodeGen/RISCV/rvv/vp-splice-mask-fixed-vectors.ll
index ce0ae2022885a..9496cd82947d4 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vp-splice-mask-fixed-vectors.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vp-splice-mask-fixed-vectors.ll
@@ -68,8 +68,8 @@ define <2 x i1> @test_vp_splice_v2i1_masked(<2 x i1> %va, <2 x i1> %vb, <2 x i1>
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: vmerge.vim v10, v11, 1, v0
; CHECK-NEXT: addi a0, a0, -5
-; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, ma
; CHECK-NEXT: vslidedown.vi v10, v10, 5, v0.t
; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, mu
; CHECK-NEXT: vslideup.vx v10, v8, a0, v0.t
@@ -141,8 +141,8 @@ define <4 x i1> @test_vp_splice_v4i1_masked(<4 x i1> %va, <4 x i1> %vb, <4 x i1>
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: vmerge.vim v10, v11, 1, v0
; CHECK-NEXT: addi a0, a0, -5
-; CHECK-NEXT: vsetvli zero, a0, e8, mf4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, mf4, ta, ma
; CHECK-NEXT: vslidedown.vi v10, v10, 5, v0.t
; CHECK-NEXT: vsetvli zero, a1, e8, mf4, ta, mu
; CHECK-NEXT: vslideup.vx v10, v8, a0, v0.t
@@ -214,8 +214,8 @@ define <8 x i1> @test_vp_splice_v8i1_masked(<8 x i1> %va, <8 x i1> %vb, <8 x i1>
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: vmerge.vim v10, v11, 1, v0
; CHECK-NEXT: addi a0, a0, -5
-; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, ma
; CHECK-NEXT: vslidedown.vi v10, v10, 5, v0.t
; CHECK-NEXT: vsetvli zero, a1, e8, mf2, ta, mu
; CHECK-NEXT: vslideup.vx v10, v8, a0, v0.t
@@ -287,8 +287,8 @@ define <16 x i1> @test_vp_splice_v16i1_masked(<16 x i1> %va, <16 x i1> %vb, <16
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: vmerge.vim v10, v11, 1, v0
; CHECK-NEXT: addi a0, a0, -5
-; CHECK-NEXT: vsetvli zero, a0, e8, m1, ta, ma
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, m1, ta, ma
; CHECK-NEXT: vslidedown.vi v10, v10, 5, v0.t
; CHECK-NEXT: vsetvli zero, a1, e8, m1, ta, mu
; CHECK-NEXT: vslideup.vx v10, v8, a0, v0.t
diff --git a/llvm/test/CodeGen/RISCV/rvv/vp-splice-mask-vectors.ll b/llvm/test/CodeGen/RISCV/rvv/vp-splice-mask-vectors.ll
index 668cff2342936..9027630825227 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vp-splice-mask-vectors.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vp-splice-mask-vectors.ll
@@ -71,8 +71,8 @@ define <vscale x 1 x i1> @test_vp_splice_nxv1i1_masked(<vscale x 1 x i1> %va, <v
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: vmerge.vim v10, v11, 1, v0
; CHECK-NEXT: addi a0, a0, -5
-; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, ma
; CHECK-NEXT: vslidedown.vi v10, v10, 5, v0.t
; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, mu
; CHECK-NEXT: vslideup.vx v10, v8, a0, v0.t
@@ -144,8 +144,8 @@ define <vscale x 2 x i1> @test_vp_splice_nxv2i1_masked(<vscale x 2 x i1> %va, <v
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: vmerge.vim v10, v11, 1, v0
; CHECK-NEXT: addi a0, a0, -5
-; CHECK-NEXT: vsetvli zero, a0, e8, mf4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, mf4, ta, ma
; CHECK-NEXT: vslidedown.vi v10, v10, 5, v0.t
; CHECK-NEXT: vsetvli zero, a1, e8, mf4, ta, mu
; CHECK-NEXT: vslideup.vx v10, v8, a0, v0.t
@@ -217,8 +217,8 @@ define <vscale x 4 x i1> @test_vp_splice_nxv4i1_masked(<vscale x 4 x i1> %va, <v
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: vmerge.vim v10, v11, 1, v0
; CHECK-NEXT: addi a0, a0, -5
-; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, ma
; CHECK-NEXT: vslidedown.vi v10, v10, 5, v0.t
; CHECK-NEXT: vsetvli zero, a1, e8, mf2, ta, mu
; CHECK-NEXT: vslideup.vx v10, v8, a0, v0.t
@@ -290,8 +290,8 @@ define <vscale x 8 x i1> @test_vp_splice_nxv8i1_masked(<vscale x 8 x i1> %va, <v
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: vmerge.vim v10, v11, 1, v0
; CHECK-NEXT: addi a0, a0, -5
-; CHECK-NEXT: vsetvli zero, a0, e8, m1, ta, ma
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, m1, ta, ma
; CHECK-NEXT: vslidedown.vi v10, v10, 5, v0.t
; CHECK-NEXT: vsetvli zero, a1, e8, m1, ta, mu
; CHECK-NEXT: vslideup.vx v10, v8, a0, v0.t
@@ -363,8 +363,8 @@ define <vscale x 16 x i1> @test_vp_splice_nxv16i1_masked(<vscale x 16 x i1> %va,
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: vmerge.vim v10, v14, 1, v0
; CHECK-NEXT: addi a0, a0, -5
-; CHECK-NEXT: vsetvli zero, a0, e8, m2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, m2, ta, ma
; CHECK-NEXT: vslidedown.vi v10, v10, 5, v0.t
; CHECK-NEXT: vsetvli zero, a1, e8, m2, ta, mu
; CHECK-NEXT: vslideup.vx v10, v12, a0, v0.t
@@ -437,8 +437,8 @@ define <vscale x 32 x i1> @test_vp_splice_nxv32i1_masked(<vscale x 32 x i1> %va,
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: vmerge.vim v16, v16, 1, v0
; CHECK-NEXT: addi a0, a0, -5
-; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, ma
; CHECK-NEXT: vslidedown.vi v16, v16, 5, v0.t
; CHECK-NEXT: vsetvli zero, a1, e8, m4, ta, mu
; CHECK-NEXT: vslideup.vx v16, v12, a0, v0.t
@@ -511,8 +511,8 @@ define <vscale x 64 x i1> @test_vp_splice_nxv64i1_masked(<vscale x 64 x i1> %va,
; CHECK-NEXT: vmv1r.v v0, v10
; CHECK-NEXT: vmerge.vim v24, v24, 1, v0
; CHECK-NEXT: addi a0, a0, -5
-; CHECK-NEXT: vsetvli zero, a0, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e8, m8, ta, ma
; CHECK-NEXT: vslidedown.vi v24, v24, 5, v0.t
; CHECK-NEXT: vsetvli zero, a1, e8, m8, ta, mu
; CHECK-NEXT: vslideup.vx v24, v16, a0, v0.t
diff --git a/llvm/test/CodeGen/RISCV/rvv/vpgather-sdnode.ll b/llvm/test/CodeGen/RISCV/rvv/vpgather-sdnode.ll
index c86fee6305931..c0d7ecf74956b 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vpgather-sdnode.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vpgather-sdnode.ll
@@ -274,8 +274,8 @@ define <vscale x 32 x i8> @vpgather_baseidx_nxv32i8(ptr %base, <vscale x 32 x i8
; RV32-NEXT: .LBB12_2:
; RV32-NEXT: vsetvli a2, zero, e32, m8, ta, ma
; RV32-NEXT: vsext.vf4 v16, v8
-; RV32-NEXT: vsetvli zero, a1, e8, m2, ta, ma
; RV32-NEXT: vmv1r.v v0, v12
+; RV32-NEXT: vsetvli zero, a1, e8, m2, ta, ma
; RV32-NEXT: vluxei32.v v8, (a0), v16, v0.t
; RV32-NEXT: ret
;
@@ -308,8 +308,8 @@ define <vscale x 32 x i8> @vpgather_baseidx_nxv32i8(ptr %base, <vscale x 32 x i8
; RV64-NEXT: .LBB12_2:
; RV64-NEXT: vsetvli a6, zero, e64, m8, ta, ma
; RV64-NEXT: vsext.vf8 v16, v10
-; RV64-NEXT: vsetvli zero, a5, e8, m1, ta, ma
; RV64-NEXT: vmv1r.v v0, v13
+; RV64-NEXT: vsetvli zero, a5, e8, m1, ta, ma
; RV64-NEXT: vluxei64.v v10, (a0), v16, v0.t
; RV64-NEXT: bltu a1, a3, .LBB12_4
; RV64-NEXT: # %bb.3:
@@ -331,8 +331,8 @@ define <vscale x 32 x i8> @vpgather_baseidx_nxv32i8(ptr %base, <vscale x 32 x i8
; RV64-NEXT: .LBB12_6:
; RV64-NEXT: vsetvli a2, zero, e64, m8, ta, ma
; RV64-NEXT: vsext.vf8 v16, v8
-; RV64-NEXT: vsetvli zero, a1, e8, m1, ta, ma
; RV64-NEXT: vmv1r.v v0, v12
+; RV64-NEXT: vsetvli zero, a1, e8, m1, ta, ma
; RV64-NEXT: vluxei64.v v8, (a0), v16, v0.t
; RV64-NEXT: ret
%ptrs = getelementptr inbounds i8, ptr %base, <vscale x 32 x i8> %idxs
@@ -2269,18 +2269,18 @@ define <vscale x 16 x double> @vpgather_nxv16f64(<vscale x 16 x ptr> %ptrs, <vsc
; RV32-NEXT: sub a2, a0, a1
; RV32-NEXT: sltu a3, a0, a2
; RV32-NEXT: addi a3, a3, -1
+; RV32-NEXT: srli a4, a1, 3
+; RV32-NEXT: vsetvli a5, zero, e8, mf4, ta, ma
+; RV32-NEXT: vslidedown.vx v0, v0, a4
; RV32-NEXT: and a2, a3, a2
-; RV32-NEXT: srli a3, a1, 3
-; RV32-NEXT: vsetvli a4, zero, e8, mf4, ta, ma
-; RV32-NEXT: vslidedown.vx v0, v0, a3
; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
; RV32-NEXT: vluxei32.v v16, (zero), v12, v0.t
; RV32-NEXT: bltu a0, a1, .LBB102_2
; RV32-NEXT: # %bb.1:
; RV32-NEXT: mv a0, a1
; RV32-NEXT: .LBB102_2:
-; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vmv1r.v v0, v24
+; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV32-NEXT: vluxei32.v v24, (zero), v8, v0.t
; RV32-NEXT: vmv.v.v v8, v24
; RV32-NEXT: ret
@@ -2292,18 +2292,18 @@ define <vscale x 16 x double> @vpgather_nxv16f64(<vscale x 16 x ptr> %ptrs, <vsc
; RV64-NEXT: sub a2, a0, a1
; RV64-NEXT: sltu a3, a0, a2
; RV64-NEXT: addi a3, a3, -1
+; RV64-NEXT: srli a4, a1, 3
+; RV64-NEXT: vsetvli a5, zero, e8, mf4, ta, ma
+; RV64-NEXT: vslidedown.vx v0, v0, a4
; RV64-NEXT: and a2, a3, a2
-; RV64-NEXT: srli a3, a1, 3
-; RV64-NEXT: vsetvli a4, zero, e8, mf4, ta, ma
-; RV64-NEXT: vslidedown.vx v0, v0, a3
; RV64-NEXT: vsetvli zero, a2, e64, m8, ta, ma
; RV64-NEXT: vluxei64.v v16, (zero), v16, v0.t
; RV64-NEXT: bltu a0, a1, .LBB102_2
; RV64-NEXT: # %bb.1:
; RV64-NEXT: mv a0, a1
; RV64-NEXT: .LBB102_2:
-; RV64-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV64-NEXT: vmv1r.v v0, v24
+; RV64-NEXT: vsetvli zero, a0, e64, m8, ta, ma
; RV64-NEXT: vluxei64.v v8, (zero), v8, v0.t
; RV64-NEXT: ret
%v = call <vscale x 16 x double> @llvm.vp.gather.nxv16f64.nxv16p0(<vscale x 16 x ptr> %ptrs, <vscale x 16 x i1> %m, i32 %evl)
@@ -2319,20 +2319,20 @@ define <vscale x 16 x double> @vpgather_baseidx_nxv16i16_nxv16f64(ptr %base, <vs
; RV32-NEXT: vsll.vi v24, v16, 3
; RV32-NEXT: csrr a2, vlenb
; RV32-NEXT: sub a3, a1, a2
-; RV32-NEXT: sltu a4, a1, a3
-; RV32-NEXT: addi a4, a4, -1
-; RV32-NEXT: and a3, a4, a3
; RV32-NEXT: srli a4, a2, 3
; RV32-NEXT: vsetvli a5, zero, e8, mf4, ta, ma
; RV32-NEXT: vslidedown.vx v0, v0, a4
+; RV32-NEXT: sltu a4, a1, a3
+; RV32-NEXT: addi a4, a4, -1
+; RV32-NEXT: and a3, a4, a3
; RV32-NEXT: vsetvli zero, a3, e64, m8, ta, ma
; RV32-NEXT: vluxei32.v v16, (a0), v28, v0.t
; RV32-NEXT: bltu a1, a2, .LBB103_2
; RV32-NEXT: # %bb.1:
; RV32-NEXT: mv a1, a2
; RV32-NEXT: .LBB103_2:
-; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; RV32-NEXT: vmv1r.v v0, v12
+; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; RV32-NEXT: vluxei32.v v8, (a0), v24, v0.t
; RV32-NEXT: ret
;
@@ -2340,26 +2340,27 @@ define <vscale x 16 x double> @vpgather_baseidx_nxv16i16_nxv16f64(ptr %base, <vs
; RV64: # %bb.0:
; RV64-NEXT: vmv1r.v v12, v0
; RV64-NEXT: vsetvli a2, zero, e64, m8, ta, ma
-; RV64-NEXT: vsext.vf4 v16, v8
-; RV64-NEXT: vsll.vi v24, v16, 3
; RV64-NEXT: vsext.vf4 v16, v10
; RV64-NEXT: vsll.vi v16, v16, 3
; RV64-NEXT: csrr a2, vlenb
; RV64-NEXT: sub a3, a1, a2
-; RV64-NEXT: sltu a4, a1, a3
-; RV64-NEXT: addi a4, a4, -1
-; RV64-NEXT: and a3, a4, a3
; RV64-NEXT: srli a4, a2, 3
; RV64-NEXT: vsetvli a5, zero, e8, mf4, ta, ma
; RV64-NEXT: vslidedown.vx v0, v0, a4
+; RV64-NEXT: sltu a4, a1, a3
+; RV64-NEXT: addi a4, a4, -1
+; RV64-NEXT: and a3, a4, a3
; RV64-NEXT: vsetvli zero, a3, e64, m8, ta, ma
; RV64-NEXT: vluxei64.v v16, (a0), v16, v0.t
+; RV64-NEXT: vsetvli a3, zero, e64, m8, ta, ma
+; RV64-NEXT: vsext.vf4 v24, v8
+; RV64-NEXT: vsll.vi v24, v24, 3
; RV64-NEXT: bltu a1, a2, .LBB103_2
; RV64-NEXT: # %bb.1:
; RV64-NEXT: mv a1, a2
; RV64-NEXT: .LBB103_2:
-; RV64-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; RV64-NEXT: vmv1r.v v0, v12
+; RV64-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; RV64-NEXT: vluxei64.v v8, (a0), v24, v0.t
; RV64-NEXT: ret
%ptrs = getelementptr inbounds double, ptr %base, <vscale x 16 x i16> %idxs
@@ -2376,20 +2377,20 @@ define <vscale x 16 x double> @vpgather_baseidx_sext_nxv16i16_nxv16f64(ptr %base
; RV32-NEXT: vsll.vi v24, v16, 3
; RV32-NEXT: csrr a2, vlenb
; RV32-NEXT: sub a3, a1, a2
-; RV32-NEXT: sltu a4, a1, a3
-; RV32-NEXT: addi a4, a4, -1
-; RV32-NEXT: and a3, a4, a3
; RV32-NEXT: srli a4, a2, 3
; RV32-NEXT: vsetvli a5, zero, e8, mf4, ta, ma
; RV32-NEXT: vslidedown.vx v0, v0, a4
+; RV32-NEXT: sltu a4, a1, a3
+; RV32-NEXT: addi a4, a4, -1
+; RV32-NEXT: and a3, a4, a3
; RV32-NEXT: vsetvli zero, a3, e64, m8, ta, ma
; RV32-NEXT: vluxei32.v v16, (a0), v28, v0.t
; RV32-NEXT: bltu a1, a2, .LBB104_2
; RV32-NEXT: # %bb.1:
; RV32-NEXT: mv a1, a2
; RV32-NEXT: .LBB104_2:
-; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; RV32-NEXT: vmv1r.v v0, v12
+; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; RV32-NEXT: vluxei32.v v8, (a0), v24, v0.t
; RV32-NEXT: ret
;
@@ -2398,25 +2399,26 @@ define <vscale x 16 x double> @vpgather_baseidx_sext_nxv16i16_nxv16f64(ptr %base
; RV64-NEXT: vmv1r.v v12, v0
; RV64-NEXT: vsetvli a2, zero, e64, m8, ta, ma
; RV64-NEXT: vsext.vf4 v16, v10
-; RV64-NEXT: vsext.vf4 v24, v8
-; RV64-NEXT: vsll.vi v24, v24, 3
; RV64-NEXT: vsll.vi v16, v16, 3
; RV64-NEXT: csrr a2, vlenb
; RV64-NEXT: sub a3, a1, a2
-; RV64-NEXT: sltu a4, a1, a3
-; RV64-NEXT: addi a4, a4, -1
-; RV64-NEXT: and a3, a4, a3
; RV64-NEXT: srli a4, a2, 3
; RV64-NEXT: vsetvli a5, zero, e8, mf4, ta, ma
; RV64-NEXT: vslidedown.vx v0, v0, a4
+; RV64-NEXT: sltu a4, a1, a3
+; RV64-NEXT: addi a4, a4, -1
+; RV64-NEXT: and a3, a4, a3
; RV64-NEXT: vsetvli zero, a3, e64, m8, ta, ma
; RV64-NEXT: vluxei64.v v16, (a0), v16, v0.t
+; RV64-NEXT: vsetvli a3, zero, e64, m8, ta, ma
+; RV64-NEXT: vsext.vf4 v24, v8
+; RV64-NEXT: vsll.vi v24, v24, 3
; RV64-NEXT: bltu a1, a2, .LBB104_2
; RV64-NEXT: # %bb.1:
; RV64-NEXT: mv a1, a2
; RV64-NEXT: .LBB104_2:
-; RV64-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; RV64-NEXT: vmv1r.v v0, v12
+; RV64-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; RV64-NEXT: vluxei64.v v8, (a0), v24, v0.t
; RV64-NEXT: ret
%eidxs = sext <vscale x 16 x i16> %idxs to <vscale x 16 x i64>
@@ -2434,20 +2436,20 @@ define <vscale x 16 x double> @vpgather_baseidx_zext_nxv16i16_nxv16f64(ptr %base
; RV32-NEXT: vsll.vi v24, v16, 3
; RV32-NEXT: csrr a2, vlenb
; RV32-NEXT: sub a3, a1, a2
-; RV32-NEXT: sltu a4, a1, a3
-; RV32-NEXT: addi a4, a4, -1
-; RV32-NEXT: and a3, a4, a3
; RV32-NEXT: srli a4, a2, 3
; RV32-NEXT: vsetvli a5, zero, e8, mf4, ta, ma
; RV32-NEXT: vslidedown.vx v0, v0, a4
+; RV32-NEXT: sltu a4, a1, a3
+; RV32-NEXT: addi a4, a4, -1
+; RV32-NEXT: and a3, a4, a3
; RV32-NEXT: vsetvli zero, a3, e64, m8, ta, ma
; RV32-NEXT: vluxei32.v v16, (a0), v28, v0.t
; RV32-NEXT: bltu a1, a2, .LBB105_2
; RV32-NEXT: # %bb.1:
; RV32-NEXT: mv a1, a2
; RV32-NEXT: .LBB105_2:
-; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; RV32-NEXT: vmv1r.v v0, v12
+; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; RV32-NEXT: vluxei32.v v8, (a0), v24, v0.t
; RV32-NEXT: ret
;
@@ -2459,20 +2461,20 @@ define <vscale x 16 x double> @vpgather_baseidx_zext_nxv16i16_nxv16f64(ptr %base
; RV64-NEXT: vsll.vi v24, v16, 3
; RV64-NEXT: csrr a2, vlenb
; RV64-NEXT: sub a3, a1, a2
-; RV64-NEXT: sltu a4, a1, a3
-; RV64-NEXT: addi a4, a4, -1
-; RV64-NEXT: and a3, a4, a3
; RV64-NEXT: srli a4, a2, 3
; RV64-NEXT: vsetvli a5, zero, e8, mf4, ta, ma
; RV64-NEXT: vslidedown.vx v0, v0, a4
+; RV64-NEXT: sltu a4, a1, a3
+; RV64-NEXT: addi a4, a4, -1
+; RV64-NEXT: and a3, a4, a3
; RV64-NEXT: vsetvli zero, a3, e64, m8, ta, ma
; RV64-NEXT: vluxei32.v v16, (a0), v28, v0.t
; RV64-NEXT: bltu a1, a2, .LBB105_2
; RV64-NEXT: # %bb.1:
; RV64-NEXT: mv a1, a2
; RV64-NEXT: .LBB105_2:
-; RV64-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; RV64-NEXT: vmv1r.v v0, v12
+; RV64-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; RV64-NEXT: vluxei32.v v8, (a0), v24, v0.t
; RV64-NEXT: ret
%eidxs = zext <vscale x 16 x i16> %idxs to <vscale x 16 x i64>
diff --git a/llvm/test/CodeGen/RISCV/rvv/vpload.ll b/llvm/test/CodeGen/RISCV/rvv/vpload.ll
index f07c16476c56a..1b1e9153a2fd5 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vpload.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vpload.ll
@@ -444,18 +444,18 @@ define <vscale x 16 x double> @vpload_nxv16f64(ptr %ptr, <vscale x 16 x i1> %m,
; CHECK-NEXT: addi a4, a4, -1
; CHECK-NEXT: and a3, a4, a3
; CHECK-NEXT: slli a4, a2, 3
-; CHECK-NEXT: add a4, a0, a4
; CHECK-NEXT: srli a5, a2, 3
; CHECK-NEXT: vsetvli a6, zero, e8, mf4, ta, ma
; CHECK-NEXT: vslidedown.vx v0, v0, a5
+; CHECK-NEXT: add a4, a0, a4
; CHECK-NEXT: vsetvli zero, a3, e64, m8, ta, ma
; CHECK-NEXT: vle64.v v16, (a4), v0.t
; CHECK-NEXT: bltu a1, a2, .LBB37_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a1, a2
; CHECK-NEXT: .LBB37_2:
-; CHECK-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; CHECK-NEXT: vle64.v v8, (a0), v0.t
; CHECK-NEXT: ret
%load = call <vscale x 16 x double> @llvm.vp.load.nxv16f64.p0(ptr %ptr, <vscale x 16 x i1> %m, i32 %evl)
@@ -489,10 +489,10 @@ define <vscale x 16 x double> @vpload_nxv17f64(ptr %ptr, ptr %out, <vscale x 17
; CHECK-NEXT: addi a7, a7, -1
; CHECK-NEXT: and a6, a7, a6
; CHECK-NEXT: slli a7, a3, 3
-; CHECK-NEXT: add a7, a0, a7
; CHECK-NEXT: srli t0, a3, 3
; CHECK-NEXT: vsetvli t1, zero, e8, mf4, ta, ma
; CHECK-NEXT: vslidedown.vx v0, v8, t0
+; CHECK-NEXT: add a7, a0, a7
; CHECK-NEXT: vsetvli zero, a6, e64, m8, ta, ma
; CHECK-NEXT: vle64.v v16, (a7), v0.t
; CHECK-NEXT: sub a5, a2, a5
@@ -504,18 +504,18 @@ define <vscale x 16 x double> @vpload_nxv17f64(ptr %ptr, ptr %out, <vscale x 17
; CHECK-NEXT: mv a2, a3
; CHECK-NEXT: .LBB38_4:
; CHECK-NEXT: slli a5, a3, 4
-; CHECK-NEXT: add a5, a0, a5
; CHECK-NEXT: srli a6, a3, 2
; CHECK-NEXT: vsetvli a7, zero, e8, mf2, ta, ma
; CHECK-NEXT: vslidedown.vx v0, v8, a6
+; CHECK-NEXT: add a5, a0, a5
; CHECK-NEXT: vsetvli zero, a2, e64, m8, ta, ma
; CHECK-NEXT: vle64.v v24, (a5), v0.t
; CHECK-NEXT: bltu a4, a3, .LBB38_6
; CHECK-NEXT: # %bb.5:
; CHECK-NEXT: mv a4, a3
; CHECK-NEXT: .LBB38_6:
-; CHECK-NEXT: vsetvli zero, a4, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a4, e64, m8, ta, ma
; CHECK-NEXT: vle64.v v8, (a0), v0.t
; CHECK-NEXT: vs1r.v v24, (a1)
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vpmerge-sdnode.ll b/llvm/test/CodeGen/RISCV/rvv/vpmerge-sdnode.ll
index 76efdda15bf77..094e6c9cc754f 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vpmerge-sdnode.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vpmerge-sdnode.ll
@@ -369,23 +369,23 @@ define <vscale x 128 x i8> @vpmerge_vv_nxv128i8(<vscale x 128 x i8> %va, <vscale
; CHECK-NEXT: slli a1, a1, 3
; CHECK-NEXT: add a4, a0, a1
; CHECK-NEXT: vl8r.v v16, (a4)
-; CHECK-NEXT: vl8r.v v8, (a0)
-; CHECK-NEXT: vsetvli a0, zero, e8, m8, ta, ma
-; CHECK-NEXT: sub a0, a3, a1
+; CHECK-NEXT: vsetvli a4, zero, e8, m8, ta, ma
; CHECK-NEXT: vlm.v v0, (a2)
-; CHECK-NEXT: sltu a2, a3, a0
-; CHECK-NEXT: addi a2, a2, -1
-; CHECK-NEXT: and a0, a2, a0
-; CHECK-NEXT: vsetvli zero, a0, e8, m8, tu, ma
+; CHECK-NEXT: sub a2, a3, a1
+; CHECK-NEXT: sltu a4, a3, a2
+; CHECK-NEXT: vl8r.v v8, (a0)
+; CHECK-NEXT: addi a4, a4, -1
+; CHECK-NEXT: and a2, a4, a2
+; CHECK-NEXT: vsetvli zero, a2, e8, m8, tu, ma
; CHECK-NEXT: vmerge.vvm v16, v16, v24, v0
; CHECK-NEXT: bltu a3, a1, .LBB28_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a3, a1
; CHECK-NEXT: .LBB28_2:
-; CHECK-NEXT: vsetvli zero, a3, e8, m8, tu, ma
; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, a3, e8, m8, tu, ma
; CHECK-NEXT: vmerge.vvm v8, v8, v24, v0
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
@@ -414,8 +414,8 @@ define <vscale x 128 x i8> @vpmerge_vx_nxv128i8(i8 %a, <vscale x 128 x i8> %vb,
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a2, a1
; CHECK-NEXT: .LBB29_2:
-; CHECK-NEXT: vsetvli zero, a2, e8, m8, tu, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a2, e8, m8, tu, ma
; CHECK-NEXT: vmerge.vxm v8, v8, a0, v0
; CHECK-NEXT: ret
%elt.head = insertelement <vscale x 128 x i8> poison, i8 %a, i32 0
@@ -442,8 +442,8 @@ define <vscale x 128 x i8> @vpmerge_vi_nxv128i8(<vscale x 128 x i8> %vb, <vscale
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a1, a0
; CHECK-NEXT: .LBB30_2:
-; CHECK-NEXT: vsetvli zero, a1, e8, m8, tu, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a1, e8, m8, tu, ma
; CHECK-NEXT: vmerge.vim v8, v8, 2, v0
; CHECK-NEXT: ret
%v = call <vscale x 128 x i8> @llvm.vp.merge.nxv128i8(<vscale x 128 x i1> %m, <vscale x 128 x i8> splat (i8 2), <vscale x 128 x i8> %vb, i32 %evl)
diff --git a/llvm/test/CodeGen/RISCV/rvv/vpscatter-sdnode.ll b/llvm/test/CodeGen/RISCV/rvv/vpscatter-sdnode.ll
index 351fc500145ea..59662db42898f 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vpscatter-sdnode.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vpscatter-sdnode.ll
@@ -2124,10 +2124,10 @@ define void @vpscatter_nxv16f64(<vscale x 16 x double> %val, <vscale x 16 x ptr>
; RV32-NEXT: sub a2, a1, a0
; RV32-NEXT: sltu a1, a1, a2
; RV32-NEXT: addi a1, a1, -1
-; RV32-NEXT: and a1, a1, a2
; RV32-NEXT: srli a0, a0, 3
-; RV32-NEXT: vsetvli a2, zero, e8, mf4, ta, ma
+; RV32-NEXT: vsetvli a3, zero, e8, mf4, ta, ma
; RV32-NEXT: vslidedown.vx v0, v0, a0
+; RV32-NEXT: and a1, a1, a2
; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; RV32-NEXT: vsoxei32.v v16, (zero), v28, v0.t
; RV32-NEXT: ret
@@ -2157,13 +2157,13 @@ define void @vpscatter_nxv16f64(<vscale x 16 x double> %val, <vscale x 16 x ptr>
; RV64-NEXT: sub a0, a2, a1
; RV64-NEXT: sltu a2, a2, a0
; RV64-NEXT: addi a2, a2, -1
-; RV64-NEXT: and a0, a2, a0
; RV64-NEXT: srli a1, a1, 3
-; RV64-NEXT: vsetvli a2, zero, e8, mf4, ta, ma
+; RV64-NEXT: vsetvli a3, zero, e8, mf4, ta, ma
; RV64-NEXT: vslidedown.vx v0, v0, a1
+; RV64-NEXT: and a0, a2, a0
+; RV64-NEXT: addi a1, sp, 16
+; RV64-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
; RV64-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV64-NEXT: addi a0, sp, 16
-; RV64-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; RV64-NEXT: vsoxei64.v v16, (zero), v8, v0.t
; RV64-NEXT: csrr a0, vlenb
; RV64-NEXT: slli a0, a0, 3
@@ -2192,10 +2192,10 @@ define void @vpscatter_baseidx_nxv16i16_nxv16f64(<vscale x 16 x double> %val, pt
; RV32-NEXT: sub a3, a2, a1
; RV32-NEXT: sltu a2, a2, a3
; RV32-NEXT: addi a2, a2, -1
-; RV32-NEXT: and a2, a2, a3
; RV32-NEXT: srli a1, a1, 3
-; RV32-NEXT: vsetvli a3, zero, e8, mf4, ta, ma
+; RV32-NEXT: vsetvli a4, zero, e8, mf4, ta, ma
; RV32-NEXT: vslidedown.vx v0, v0, a1
+; RV32-NEXT: and a2, a2, a3
; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
; RV32-NEXT: vsoxei32.v v16, (a0), v28, v0.t
; RV32-NEXT: ret
@@ -2232,11 +2232,10 @@ define void @vpscatter_baseidx_nxv16i16_nxv16f64(<vscale x 16 x double> %val, pt
; RV64-NEXT: sub a3, a2, a1
; RV64-NEXT: sltu a2, a2, a3
; RV64-NEXT: addi a2, a2, -1
-; RV64-NEXT: and a2, a2, a3
; RV64-NEXT: srli a1, a1, 3
-; RV64-NEXT: vsetvli a3, zero, e8, mf4, ta, ma
+; RV64-NEXT: vsetvli a4, zero, e8, mf4, ta, ma
; RV64-NEXT: vslidedown.vx v0, v0, a1
-; RV64-NEXT: vsetvli zero, a2, e64, m8, ta, ma
+; RV64-NEXT: and a2, a2, a3
; RV64-NEXT: csrr a1, vlenb
; RV64-NEXT: slli a1, a1, 3
; RV64-NEXT: add a1, sp, a1
@@ -2244,6 +2243,7 @@ define void @vpscatter_baseidx_nxv16i16_nxv16f64(<vscale x 16 x double> %val, pt
; RV64-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
; RV64-NEXT: addi a1, sp, 16
; RV64-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
+; RV64-NEXT: vsetvli zero, a2, e64, m8, ta, ma
; RV64-NEXT: vsoxei64.v v8, (a0), v16, v0.t
; RV64-NEXT: csrr a0, vlenb
; RV64-NEXT: slli a0, a0, 4
@@ -2273,10 +2273,10 @@ define void @vpscatter_baseidx_sext_nxv16i16_nxv16f64(<vscale x 16 x double> %va
; RV32-NEXT: sub a3, a2, a1
; RV32-NEXT: sltu a2, a2, a3
; RV32-NEXT: addi a2, a2, -1
-; RV32-NEXT: and a2, a2, a3
; RV32-NEXT: srli a1, a1, 3
-; RV32-NEXT: vsetvli a3, zero, e8, mf4, ta, ma
+; RV32-NEXT: vsetvli a4, zero, e8, mf4, ta, ma
; RV32-NEXT: vslidedown.vx v0, v0, a1
+; RV32-NEXT: and a2, a2, a3
; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
; RV32-NEXT: vsoxei32.v v16, (a0), v28, v0.t
; RV32-NEXT: ret
@@ -2308,22 +2308,22 @@ define void @vpscatter_baseidx_sext_nxv16i16_nxv16f64(<vscale x 16 x double> %va
; RV64-NEXT: # %bb.1:
; RV64-NEXT: mv a3, a1
; RV64-NEXT: .LBB101_2:
+; RV64-NEXT: addi a4, sp, 16
+; RV64-NEXT: vl1r.v v0, (a4) # Unknown-size Folded Reload
; RV64-NEXT: vsetvli zero, a3, e64, m8, ta, ma
-; RV64-NEXT: addi a3, sp, 16
-; RV64-NEXT: vl1r.v v0, (a3) # Unknown-size Folded Reload
; RV64-NEXT: vsoxei64.v v8, (a0), v24, v0.t
; RV64-NEXT: sub a3, a2, a1
; RV64-NEXT: sltu a2, a2, a3
; RV64-NEXT: addi a2, a2, -1
-; RV64-NEXT: and a2, a2, a3
; RV64-NEXT: srli a1, a1, 3
-; RV64-NEXT: vsetvli a3, zero, e8, mf4, ta, ma
+; RV64-NEXT: vsetvli a4, zero, e8, mf4, ta, ma
; RV64-NEXT: vslidedown.vx v0, v0, a1
-; RV64-NEXT: vsetvli zero, a2, e64, m8, ta, ma
+; RV64-NEXT: and a2, a2, a3
; RV64-NEXT: csrr a1, vlenb
; RV64-NEXT: add a1, sp, a1
; RV64-NEXT: addi a1, a1, 16
; RV64-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
+; RV64-NEXT: vsetvli zero, a2, e64, m8, ta, ma
; RV64-NEXT: vsoxei64.v v8, (a0), v16, v0.t
; RV64-NEXT: csrr a0, vlenb
; RV64-NEXT: li a1, 10
@@ -2355,10 +2355,10 @@ define void @vpscatter_baseidx_zext_nxv16i16_nxv16f64(<vscale x 16 x double> %va
; RV32-NEXT: sub a3, a2, a1
; RV32-NEXT: sltu a2, a2, a3
; RV32-NEXT: addi a2, a2, -1
-; RV32-NEXT: and a2, a2, a3
; RV32-NEXT: srli a1, a1, 3
-; RV32-NEXT: vsetvli a3, zero, e8, mf4, ta, ma
+; RV32-NEXT: vsetvli a4, zero, e8, mf4, ta, ma
; RV32-NEXT: vslidedown.vx v0, v0, a1
+; RV32-NEXT: and a2, a2, a3
; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
; RV32-NEXT: vsoxei32.v v16, (a0), v28, v0.t
; RV32-NEXT: ret
@@ -2380,10 +2380,10 @@ define void @vpscatter_baseidx_zext_nxv16i16_nxv16f64(<vscale x 16 x double> %va
; RV64-NEXT: sub a3, a2, a1
; RV64-NEXT: sltu a2, a2, a3
; RV64-NEXT: addi a2, a2, -1
-; RV64-NEXT: and a2, a2, a3
; RV64-NEXT: srli a1, a1, 3
-; RV64-NEXT: vsetvli a3, zero, e8, mf4, ta, ma
+; RV64-NEXT: vsetvli a4, zero, e8, mf4, ta, ma
; RV64-NEXT: vslidedown.vx v0, v0, a1
+; RV64-NEXT: and a2, a2, a3
; RV64-NEXT: vsetvli zero, a2, e64, m8, ta, ma
; RV64-NEXT: vsoxei32.v v16, (a0), v28, v0.t
; RV64-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vpstore.ll b/llvm/test/CodeGen/RISCV/rvv/vpstore.ll
index c12fc0497742a..ce0ee38bc7047 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vpstore.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vpstore.ll
@@ -380,10 +380,10 @@ define void @vpstore_nxv16f64(<vscale x 16 x double> %val, ptr %ptr, <vscale x 1
; CHECK-NEXT: addi a1, a1, -1
; CHECK-NEXT: and a1, a1, a3
; CHECK-NEXT: slli a3, a2, 3
-; CHECK-NEXT: add a0, a0, a3
; CHECK-NEXT: srli a2, a2, 3
-; CHECK-NEXT: vsetvli a3, zero, e8, mf4, ta, ma
+; CHECK-NEXT: vsetvli a4, zero, e8, mf4, ta, ma
; CHECK-NEXT: vslidedown.vx v0, v0, a2
+; CHECK-NEXT: add a0, a0, a3
; CHECK-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; CHECK-NEXT: vse64.v v16, (a0), v0.t
; CHECK-NEXT: ret
@@ -420,36 +420,36 @@ define void @vpstore_nxv17f64(<vscale x 17 x double> %val, ptr %ptr, <vscale x 1
; CHECK-NEXT: vl8re64.v v0, (a0)
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vs8r.v v0, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: vsetvli zero, a6, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a6, e64, m8, ta, ma
; CHECK-NEXT: vse64.v v8, (a1), v0.t
; CHECK-NEXT: sub a0, a5, a3
; CHECK-NEXT: sltu a5, a5, a0
; CHECK-NEXT: addi a5, a5, -1
-; CHECK-NEXT: and a0, a5, a0
-; CHECK-NEXT: slli a5, a3, 3
-; CHECK-NEXT: add a5, a1, a5
-; CHECK-NEXT: srli a6, a3, 3
+; CHECK-NEXT: and a5, a5, a0
+; CHECK-NEXT: slli a0, a3, 3
+; CHECK-NEXT: add a6, a1, a0
+; CHECK-NEXT: srli a0, a3, 3
; CHECK-NEXT: vsetvli a7, zero, e8, mf4, ta, ma
-; CHECK-NEXT: vslidedown.vx v0, v24, a6
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT: vslidedown.vx v0, v24, a0
; CHECK-NEXT: sub a0, a2, a4
; CHECK-NEXT: sltu a2, a2, a0
; CHECK-NEXT: addi a2, a2, -1
; CHECK-NEXT: and a0, a2, a0
-; CHECK-NEXT: vse64.v v16, (a5), v0.t
+; CHECK-NEXT: vsetvli zero, a5, e64, m8, ta, ma
+; CHECK-NEXT: vse64.v v16, (a6), v0.t
; CHECK-NEXT: bltu a0, a3, .LBB31_6
; CHECK-NEXT: # %bb.5:
; CHECK-NEXT: mv a0, a3
; CHECK-NEXT: .LBB31_6:
; CHECK-NEXT: slli a2, a3, 4
-; CHECK-NEXT: add a1, a1, a2
; CHECK-NEXT: srli a3, a3, 2
-; CHECK-NEXT: vsetvli a2, zero, e8, mf2, ta, ma
+; CHECK-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
; CHECK-NEXT: vslidedown.vx v0, v24, a3
+; CHECK-NEXT: add a1, a1, a2
+; CHECK-NEXT: addi a2, sp, 16
+; CHECK-NEXT: vl8r.v v8, (a2) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vse64.v v8, (a1), v0.t
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
diff --git a/llvm/test/CodeGen/RISCV/rvv/vreductions-fp-sdnode.ll b/llvm/test/CodeGen/RISCV/rvv/vreductions-fp-sdnode.ll
index 6d42b15273cf8..fa78301338346 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vreductions-fp-sdnode.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vreductions-fp-sdnode.ll
@@ -1018,11 +1018,11 @@ declare half @llvm.vector.reduce.fmin.nxv10f16(<vscale x 10 x half>)
define half @vreduce_fmin_nxv10f16(<vscale x 10 x half> %v) {
; CHECK-LABEL: vreduce_fmin_nxv10f16:
; CHECK: # %bb.0:
+; CHECK-NEXT: lui a0, %hi(.LCPI73_0)
+; CHECK-NEXT: addi a0, a0, %lo(.LCPI73_0)
+; CHECK-NEXT: vsetvli a1, zero, e16, m1, ta, ma
+; CHECK-NEXT: vlse16.v v12, (a0), zero
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: lui a1, %hi(.LCPI73_0)
-; CHECK-NEXT: addi a1, a1, %lo(.LCPI73_0)
-; CHECK-NEXT: vsetvli a2, zero, e16, m1, ta, ma
-; CHECK-NEXT: vlse16.v v12, (a1), zero
; CHECK-NEXT: srli a0, a0, 2
; CHECK-NEXT: add a1, a0, a0
; CHECK-NEXT: vsetvli zero, a1, e16, m1, ta, ma
diff --git a/llvm/test/CodeGen/RISCV/rvv/vreductions-fp-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vreductions-fp-vp.ll
index 46560fc501c6f..f21b42e9519b6 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vreductions-fp-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vreductions-fp-vp.ll
@@ -192,8 +192,8 @@ define half @vpreduce_fadd_nxv64f16(half %s, <vscale x 64 x half> %v, <vscale x
; ZVFH-NEXT: vfmv.s.f v25, fa0
; ZVFH-NEXT: vsetvli zero, a0, e16, m8, ta, ma
; ZVFH-NEXT: vfredusum.vs v25, v8, v25, v0.t
-; ZVFH-NEXT: vsetvli zero, a1, e16, m8, ta, ma
; ZVFH-NEXT: vmv1r.v v0, v24
+; ZVFH-NEXT: vsetvli zero, a1, e16, m8, ta, ma
; ZVFH-NEXT: vfredusum.vs v25, v16, v25, v0.t
; ZVFH-NEXT: vfmv.f.s fa0, v25
; ZVFH-NEXT: ret
@@ -247,8 +247,8 @@ define half @vpreduce_fadd_nxv64f16(half %s, <vscale x 64 x half> %v, <vscale x
; ZVFHMIN-NEXT: vfmv.s.f v8, fa5
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
-; ZVFHMIN-NEXT: vsetvli zero, a5, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v6
+; ZVFHMIN-NEXT: vsetvli zero, a5, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfredusum.vs v8, v24, v8, v0.t
; ZVFHMIN-NEXT: vfmv.f.s fa5, v8
; ZVFHMIN-NEXT: fcvt.h.s fa5, fa5
@@ -257,8 +257,8 @@ define half @vpreduce_fadd_nxv64f16(half %s, <vscale x 64 x half> %v, <vscale x
; ZVFHMIN-NEXT: vfmv.s.f v8, fa5
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v16
-; ZVFHMIN-NEXT: vsetvli zero, a1, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v7
+; ZVFHMIN-NEXT: vsetvli zero, a1, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfredusum.vs v8, v24, v8, v0.t
; ZVFHMIN-NEXT: vfmv.f.s fa5, v8
; ZVFHMIN-NEXT: fcvt.h.s fa5, fa5
@@ -298,8 +298,8 @@ define half @vpreduce_ord_fadd_nxv64f16(half %s, <vscale x 64 x half> %v, <vscal
; ZVFH-NEXT: vfmv.s.f v25, fa0
; ZVFH-NEXT: vsetvli zero, a0, e16, m8, ta, ma
; ZVFH-NEXT: vfredosum.vs v25, v8, v25, v0.t
-; ZVFH-NEXT: vsetvli zero, a1, e16, m8, ta, ma
; ZVFH-NEXT: vmv1r.v v0, v24
+; ZVFH-NEXT: vsetvli zero, a1, e16, m8, ta, ma
; ZVFH-NEXT: vfredosum.vs v25, v16, v25, v0.t
; ZVFH-NEXT: vfmv.f.s fa0, v25
; ZVFH-NEXT: ret
@@ -353,8 +353,8 @@ define half @vpreduce_ord_fadd_nxv64f16(half %s, <vscale x 64 x half> %v, <vscal
; ZVFHMIN-NEXT: vfmv.s.f v8, fa5
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
-; ZVFHMIN-NEXT: vsetvli zero, a5, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v6
+; ZVFHMIN-NEXT: vsetvli zero, a5, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfredosum.vs v8, v24, v8, v0.t
; ZVFHMIN-NEXT: vfmv.f.s fa5, v8
; ZVFHMIN-NEXT: fcvt.h.s fa5, fa5
@@ -363,8 +363,8 @@ define half @vpreduce_ord_fadd_nxv64f16(half %s, <vscale x 64 x half> %v, <vscal
; ZVFHMIN-NEXT: vfmv.s.f v8, fa5
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v16
-; ZVFHMIN-NEXT: vsetvli zero, a1, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v7
+; ZVFHMIN-NEXT: vsetvli zero, a1, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfredosum.vs v8, v24, v8, v0.t
; ZVFHMIN-NEXT: vfmv.f.s fa5, v8
; ZVFHMIN-NEXT: fcvt.h.s fa5, fa5
diff --git a/llvm/test/CodeGen/RISCV/rvv/vreductions-int-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vreductions-int-vp.ll
index 7bcf37b1af3c8..c41ddaee75a86 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vreductions-int-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vreductions-int-vp.ll
@@ -1115,8 +1115,8 @@ define signext i32 @vpreduce_umax_nxv32i32(i32 signext %s, <vscale x 32 x i32> %
; CHECK-NEXT: vmv.s.x v25, a0
; CHECK-NEXT: vsetvli zero, a1, e32, m8, ta, ma
; CHECK-NEXT: vredmaxu.vs v25, v8, v25, v0.t
-; CHECK-NEXT: vsetvli zero, a2, e32, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a2, e32, m8, ta, ma
; CHECK-NEXT: vredmaxu.vs v25, v16, v25, v0.t
; CHECK-NEXT: vmv.x.s a0, v25
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vreductions-mask-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vreductions-mask-vp.ll
index 94ed7e568a01e..39666bb6119a0 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vreductions-mask-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vreductions-mask-vp.ll
@@ -24,8 +24,8 @@ define zeroext i1 @vpreduce_or_nxv1i1(i1 zeroext %s, <vscale x 1 x i1> %v, <vsca
; CHECK-LABEL: vpreduce_or_nxv1i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: snez a1, a1
; CHECK-NEXT: or a0, a1, a0
@@ -40,8 +40,8 @@ define zeroext i1 @vpreduce_xor_nxv1i1(i1 zeroext %s, <vscale x 1 x i1> %v, <vsc
; CHECK-LABEL: vpreduce_xor_nxv1i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: andi a1, a1, 1
; CHECK-NEXT: xor a0, a1, a0
@@ -72,8 +72,8 @@ define zeroext i1 @vpreduce_or_nxv2i1(i1 zeroext %s, <vscale x 2 x i1> %v, <vsca
; CHECK-LABEL: vpreduce_or_nxv2i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, mf4, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: snez a1, a1
; CHECK-NEXT: or a0, a1, a0
@@ -88,8 +88,8 @@ define zeroext i1 @vpreduce_xor_nxv2i1(i1 zeroext %s, <vscale x 2 x i1> %v, <vsc
; CHECK-LABEL: vpreduce_xor_nxv2i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, mf4, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: andi a1, a1, 1
; CHECK-NEXT: xor a0, a1, a0
@@ -120,8 +120,8 @@ define zeroext i1 @vpreduce_or_nxv4i1(i1 zeroext %s, <vscale x 4 x i1> %v, <vsca
; CHECK-LABEL: vpreduce_or_nxv4i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, mf2, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: snez a1, a1
; CHECK-NEXT: or a0, a1, a0
@@ -136,8 +136,8 @@ define zeroext i1 @vpreduce_xor_nxv4i1(i1 zeroext %s, <vscale x 4 x i1> %v, <vsc
; CHECK-LABEL: vpreduce_xor_nxv4i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, mf2, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: andi a1, a1, 1
; CHECK-NEXT: xor a0, a1, a0
@@ -168,8 +168,8 @@ define zeroext i1 @vpreduce_or_nxv8i1(i1 zeroext %s, <vscale x 8 x i1> %v, <vsca
; CHECK-LABEL: vpreduce_or_nxv8i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m1, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, m1, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: snez a1, a1
; CHECK-NEXT: or a0, a1, a0
@@ -184,8 +184,8 @@ define zeroext i1 @vpreduce_xor_nxv8i1(i1 zeroext %s, <vscale x 8 x i1> %v, <vsc
; CHECK-LABEL: vpreduce_xor_nxv8i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m1, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, m1, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: andi a1, a1, 1
; CHECK-NEXT: xor a0, a1, a0
@@ -216,8 +216,8 @@ define zeroext i1 @vpreduce_or_nxv16i1(i1 zeroext %s, <vscale x 16 x i1> %v, <vs
; CHECK-LABEL: vpreduce_or_nxv16i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, m2, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: snez a1, a1
; CHECK-NEXT: or a0, a1, a0
@@ -232,8 +232,8 @@ define zeroext i1 @vpreduce_xor_nxv16i1(i1 zeroext %s, <vscale x 16 x i1> %v, <v
; CHECK-LABEL: vpreduce_xor_nxv16i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, m2, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: andi a1, a1, 1
; CHECK-NEXT: xor a0, a1, a0
@@ -264,8 +264,8 @@ define zeroext i1 @vpreduce_or_nxv32i1(i1 zeroext %s, <vscale x 32 x i1> %v, <vs
; CHECK-LABEL: vpreduce_or_nxv32i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, m4, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: snez a1, a1
; CHECK-NEXT: or a0, a1, a0
@@ -280,8 +280,8 @@ define zeroext i1 @vpreduce_xor_nxv32i1(i1 zeroext %s, <vscale x 32 x i1> %v, <v
; CHECK-LABEL: vpreduce_xor_nxv32i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, m4, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: andi a1, a1, 1
; CHECK-NEXT: xor a0, a1, a0
@@ -296,8 +296,8 @@ define zeroext i1 @vpreduce_or_nxv40i1(i1 zeroext %s, <vscale x 40 x i1> %v, <vs
; CHECK-LABEL: vpreduce_or_nxv40i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, m8, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: snez a1, a1
; CHECK-NEXT: or a0, a1, a0
@@ -328,8 +328,8 @@ define zeroext i1 @vpreduce_or_nxv64i1(i1 zeroext %s, <vscale x 64 x i1> %v, <vs
; CHECK-LABEL: vpreduce_or_nxv64i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, m8, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: snez a1, a1
; CHECK-NEXT: or a0, a1, a0
@@ -344,8 +344,8 @@ define zeroext i1 @vpreduce_xor_nxv64i1(i1 zeroext %s, <vscale x 64 x i1> %v, <v
; CHECK-LABEL: vpreduce_xor_nxv64i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, m8, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: andi a1, a1, 1
; CHECK-NEXT: xor a0, a1, a0
@@ -366,16 +366,16 @@ define zeroext i1 @vpreduce_or_nxv128i1(i1 zeroext %s, <vscale x 128 x i1> %v, <
; CHECK-NEXT: sltu a4, a1, a3
; CHECK-NEXT: addi a4, a4, -1
; CHECK-NEXT: and a3, a4, a3
-; CHECK-NEXT: vsetvli zero, a3, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v10
+; CHECK-NEXT: vsetvli zero, a3, e8, m8, ta, ma
; CHECK-NEXT: vcpop.m a3, v8, v0.t
; CHECK-NEXT: snez a3, a3
; CHECK-NEXT: bltu a1, a2, .LBB22_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a1, a2
; CHECK-NEXT: .LBB22_2:
-; CHECK-NEXT: vsetvli zero, a1, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a1, e8, m8, ta, ma
; CHECK-NEXT: vcpop.m a1, v11, v0.t
; CHECK-NEXT: snez a1, a1
; CHECK-NEXT: or a0, a1, a0
@@ -391,8 +391,8 @@ define zeroext i1 @vpreduce_add_nxv1i1(i1 zeroext %s, <vscale x 1 x i1> %v, <vsc
; CHECK-LABEL: vpreduce_add_nxv1i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: andi a1, a1, 1
; CHECK-NEXT: xor a0, a1, a0
@@ -407,8 +407,8 @@ define zeroext i1 @vpreduce_add_nxv2i1(i1 zeroext %s, <vscale x 2 x i1> %v, <vsc
; CHECK-LABEL: vpreduce_add_nxv2i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, mf4, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: andi a1, a1, 1
; CHECK-NEXT: xor a0, a1, a0
@@ -423,8 +423,8 @@ define zeroext i1 @vpreduce_add_nxv4i1(i1 zeroext %s, <vscale x 4 x i1> %v, <vsc
; CHECK-LABEL: vpreduce_add_nxv4i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, mf2, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: andi a1, a1, 1
; CHECK-NEXT: xor a0, a1, a0
@@ -439,8 +439,8 @@ define zeroext i1 @vpreduce_add_nxv8i1(i1 zeroext %s, <vscale x 8 x i1> %v, <vsc
; CHECK-LABEL: vpreduce_add_nxv8i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m1, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, m1, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: andi a1, a1, 1
; CHECK-NEXT: xor a0, a1, a0
@@ -455,8 +455,8 @@ define zeroext i1 @vpreduce_add_nxv16i1(i1 zeroext %s, <vscale x 16 x i1> %v, <v
; CHECK-LABEL: vpreduce_add_nxv16i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, m2, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: andi a1, a1, 1
; CHECK-NEXT: xor a0, a1, a0
@@ -471,8 +471,8 @@ define zeroext i1 @vpreduce_add_nxv32i1(i1 zeroext %s, <vscale x 32 x i1> %v, <v
; CHECK-LABEL: vpreduce_add_nxv32i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, m4, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: andi a1, a1, 1
; CHECK-NEXT: xor a0, a1, a0
@@ -487,8 +487,8 @@ define zeroext i1 @vpreduce_add_nxv64i1(i1 zeroext %s, <vscale x 64 x i1> %v, <v
; CHECK-LABEL: vpreduce_add_nxv64i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, m8, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: andi a1, a1, 1
; CHECK-NEXT: xor a0, a1, a0
@@ -616,8 +616,8 @@ define zeroext i1 @vpreduce_smin_nxv1i1(i1 zeroext %s, <vscale x 1 x i1> %v, <vs
; CHECK-LABEL: vpreduce_smin_nxv1i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: snez a1, a1
; CHECK-NEXT: or a0, a1, a0
@@ -632,8 +632,8 @@ define zeroext i1 @vpreduce_smin_nxv2i1(i1 zeroext %s, <vscale x 2 x i1> %v, <vs
; CHECK-LABEL: vpreduce_smin_nxv2i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, mf4, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: snez a1, a1
; CHECK-NEXT: or a0, a1, a0
@@ -648,8 +648,8 @@ define zeroext i1 @vpreduce_smin_nxv4i1(i1 zeroext %s, <vscale x 4 x i1> %v, <vs
; CHECK-LABEL: vpreduce_smin_nxv4i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, mf2, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: snez a1, a1
; CHECK-NEXT: or a0, a1, a0
@@ -664,8 +664,8 @@ define zeroext i1 @vpreduce_smin_nxv8i1(i1 zeroext %s, <vscale x 8 x i1> %v, <vs
; CHECK-LABEL: vpreduce_smin_nxv8i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m1, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, m1, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: snez a1, a1
; CHECK-NEXT: or a0, a1, a0
@@ -680,8 +680,8 @@ define zeroext i1 @vpreduce_smin_nxv16i1(i1 zeroext %s, <vscale x 16 x i1> %v, <
; CHECK-LABEL: vpreduce_smin_nxv16i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, m2, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: snez a1, a1
; CHECK-NEXT: or a0, a1, a0
@@ -696,8 +696,8 @@ define zeroext i1 @vpreduce_smin_nxv32i1(i1 zeroext %s, <vscale x 32 x i1> %v, <
; CHECK-LABEL: vpreduce_smin_nxv32i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, m4, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: snez a1, a1
; CHECK-NEXT: or a0, a1, a0
@@ -712,8 +712,8 @@ define zeroext i1 @vpreduce_smin_nxv64i1(i1 zeroext %s, <vscale x 64 x i1> %v, <
; CHECK-LABEL: vpreduce_smin_nxv64i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, m8, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: snez a1, a1
; CHECK-NEXT: or a0, a1, a0
@@ -728,8 +728,8 @@ define zeroext i1 @vpreduce_umax_nxv1i1(i1 zeroext %s, <vscale x 1 x i1> %v, <vs
; CHECK-LABEL: vpreduce_umax_nxv1i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: snez a1, a1
; CHECK-NEXT: or a0, a1, a0
@@ -744,8 +744,8 @@ define zeroext i1 @vpreduce_umax_nxv2i1(i1 zeroext %s, <vscale x 2 x i1> %v, <vs
; CHECK-LABEL: vpreduce_umax_nxv2i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, mf4, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: snez a1, a1
; CHECK-NEXT: or a0, a1, a0
@@ -760,8 +760,8 @@ define zeroext i1 @vpreduce_umax_nxv4i1(i1 zeroext %s, <vscale x 4 x i1> %v, <vs
; CHECK-LABEL: vpreduce_umax_nxv4i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, mf2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, mf2, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: snez a1, a1
; CHECK-NEXT: or a0, a1, a0
@@ -776,8 +776,8 @@ define zeroext i1 @vpreduce_umax_nxv8i1(i1 zeroext %s, <vscale x 8 x i1> %v, <vs
; CHECK-LABEL: vpreduce_umax_nxv8i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m1, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, m1, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: snez a1, a1
; CHECK-NEXT: or a0, a1, a0
@@ -792,8 +792,8 @@ define zeroext i1 @vpreduce_umax_nxv16i1(i1 zeroext %s, <vscale x 16 x i1> %v, <
; CHECK-LABEL: vpreduce_umax_nxv16i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m2, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, m2, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: snez a1, a1
; CHECK-NEXT: or a0, a1, a0
@@ -808,8 +808,8 @@ define zeroext i1 @vpreduce_umax_nxv32i1(i1 zeroext %s, <vscale x 32 x i1> %v, <
; CHECK-LABEL: vpreduce_umax_nxv32i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, m4, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: snez a1, a1
; CHECK-NEXT: or a0, a1, a0
@@ -824,8 +824,8 @@ define zeroext i1 @vpreduce_umax_nxv64i1(i1 zeroext %s, <vscale x 64 x i1> %v, <
; CHECK-LABEL: vpreduce_umax_nxv64i1:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vsetvli zero, a1, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: vsetvli zero, a1, e8, m8, ta, ma
; CHECK-NEXT: vcpop.m a1, v9, v0.t
; CHECK-NEXT: snez a1, a1
; CHECK-NEXT: or a0, a1, a0
diff --git a/llvm/test/CodeGen/RISCV/rvv/vrgatherei16-subreg-liveness.ll b/llvm/test/CodeGen/RISCV/rvv/vrgatherei16-subreg-liveness.ll
index 462d49991ae4f..e95e9fabe9342 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vrgatherei16-subreg-liveness.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vrgatherei16-subreg-liveness.ll
@@ -12,11 +12,11 @@
define internal void @foo(<vscale x 1 x i16> %v15, <vscale x 1 x i16> %0, <vscale x 1 x i16> %vs12.i.i.i, <vscale x 1 x i16> %1, <vscale x 8 x i8> %v37) {
; NOSUBREG-LABEL: foo:
; NOSUBREG: # %bb.0: # %loopIR.preheader.i.i
-; NOSUBREG-NEXT: vsetvli a0, zero, e16, m2, ta, ma
-; NOSUBREG-NEXT: vmv.v.i v14, 0
-; NOSUBREG-NEXT: vsetvli zero, zero, e8, m1, ta, ma
+; NOSUBREG-NEXT: vsetvli a0, zero, e8, m1, ta, ma
; NOSUBREG-NEXT: vmv.v.i v9, 0
-; NOSUBREG-NEXT: vmv.v.i v8, 0
+; NOSUBREG-NEXT: vsetvli zero, zero, e16, m2, ta, ma
+; NOSUBREG-NEXT: vmv.v.i v14, 0
+; NOSUBREG-NEXT: vmv1r.v v8, v9
; NOSUBREG-NEXT: vsetivli zero, 4, e8, m1, tu, ma
; NOSUBREG-NEXT: vrgatherei16.vv v8, v9, v14
; NOSUBREG-NEXT: .LBB0_1: # %loopIR3.i.i
@@ -32,11 +32,11 @@ define internal void @foo(<vscale x 1 x i16> %v15, <vscale x 1 x i16> %0, <vscal
;
; SUBREG-LABEL: foo:
; SUBREG: # %bb.0: # %loopIR.preheader.i.i
-; SUBREG-NEXT: vsetvli a0, zero, e16, m2, ta, ma
-; SUBREG-NEXT: vmv.v.i v14, 0
-; SUBREG-NEXT: vsetvli zero, zero, e8, m1, ta, ma
+; SUBREG-NEXT: vsetvli a0, zero, e8, m1, ta, ma
; SUBREG-NEXT: vmv.v.i v9, 0
-; SUBREG-NEXT: vmv.v.i v8, 0
+; SUBREG-NEXT: vsetvli zero, zero, e16, m2, ta, ma
+; SUBREG-NEXT: vmv.v.i v14, 0
+; SUBREG-NEXT: vmv1r.v v8, v9
; SUBREG-NEXT: vsetivli zero, 4, e8, m1, tu, ma
; SUBREG-NEXT: vrgatherei16.vv v8, v9, v14
; SUBREG-NEXT: .LBB0_1: # %loopIR3.i.i
diff --git a/llvm/test/CodeGen/RISCV/rvv/vsadd-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vsadd-vp.ll
index f9ea5143cfcb7..71b91f56e89a5 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vsadd-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vsadd-vp.ll
@@ -587,8 +587,8 @@ define <vscale x 128 x i8> @vsadd_vi_nxv128i8(<vscale x 128 x i8> %va, <vscale x
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a1, a0
; CHECK-NEXT: .LBB50_2:
-; CHECK-NEXT: vsetvli zero, a1, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a1, e8, m8, ta, ma
; CHECK-NEXT: vsadd.vi v8, v8, -1, v0.t
; CHECK-NEXT: ret
%v = call <vscale x 128 x i8> @llvm.vp.sadd.sat.nxv128i8(<vscale x 128 x i8> %va, <vscale x 128 x i8> splat (i8 -1), <vscale x 128 x i1> %m, i32 %evl)
@@ -1366,8 +1366,8 @@ define <vscale x 32 x i32> @vsadd_vi_nxv32i32(<vscale x 32 x i32> %va, <vscale x
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB118_2:
-; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; CHECK-NEXT: vsadd.vi v8, v8, -1, v0.t
; CHECK-NEXT: ret
%v = call <vscale x 32 x i32> @llvm.vp.sadd.sat.nxv32i32(<vscale x 32 x i32> %va, <vscale x 32 x i32> splat (i32 -1), <vscale x 32 x i1> %m, i32 %evl)
diff --git a/llvm/test/CodeGen/RISCV/rvv/vsaddu-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vsaddu-vp.ll
index 745b93b257085..454a4ebab04a2 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vsaddu-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vsaddu-vp.ll
@@ -586,8 +586,8 @@ define <vscale x 128 x i8> @vsaddu_vi_nxv128i8(<vscale x 128 x i8> %va, <vscale
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a1, a0
; CHECK-NEXT: .LBB50_2:
-; CHECK-NEXT: vsetvli zero, a1, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a1, e8, m8, ta, ma
; CHECK-NEXT: vsaddu.vi v8, v8, -1, v0.t
; CHECK-NEXT: ret
%v = call <vscale x 128 x i8> @llvm.vp.uadd.sat.nxv128i8(<vscale x 128 x i8> %va, <vscale x 128 x i8> splat (i8 -1), <vscale x 128 x i1> %m, i32 %evl)
@@ -1365,8 +1365,8 @@ define <vscale x 32 x i32> @vsaddu_vi_nxv32i32(<vscale x 32 x i32> %va, <vscale
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB118_2:
-; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; CHECK-NEXT: vsaddu.vi v8, v8, -1, v0.t
; CHECK-NEXT: ret
%v = call <vscale x 32 x i32> @llvm.vp.uadd.sat.nxv32i32(<vscale x 32 x i32> %va, <vscale x 32 x i32> splat (i32 -1), <vscale x 32 x i1> %m, i32 %evl)
diff --git a/llvm/test/CodeGen/RISCV/rvv/vselect-fp.ll b/llvm/test/CodeGen/RISCV/rvv/vselect-fp.ll
index 4457c1002acc7..53b8e4a78b756 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vselect-fp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vselect-fp.ll
@@ -211,12 +211,12 @@ define <vscale x 32 x half> @vfmerge_fv_nxv32f16(<vscale x 32 x half> %va, half
; CHECK-ZVFHMIN: # %bb.0:
; CHECK-ZVFHMIN-NEXT: fcvt.s.h fa5, fa0
; CHECK-ZVFHMIN-NEXT: vsetvli a0, zero, e32, m8, ta, ma
-; CHECK-ZVFHMIN-NEXT: vfmv.v.f v24, fa5
+; CHECK-ZVFHMIN-NEXT: vfmv.v.f v16, fa5
; CHECK-ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; CHECK-ZVFHMIN-NEXT: vfncvt.f.f.w v16, v24
-; CHECK-ZVFHMIN-NEXT: vmv.v.v v20, v16
+; CHECK-ZVFHMIN-NEXT: vfncvt.f.f.w v24, v16
+; CHECK-ZVFHMIN-NEXT: vmv.v.v v28, v24
; CHECK-ZVFHMIN-NEXT: vsetvli a0, zero, e16, m8, ta, ma
-; CHECK-ZVFHMIN-NEXT: vmerge.vvm v8, v8, v16, v0
+; CHECK-ZVFHMIN-NEXT: vmerge.vvm v8, v8, v24, v0
; CHECK-ZVFHMIN-NEXT: ret
%head = insertelement <vscale x 32 x half> poison, half %b, i32 0
%splat = shufflevector <vscale x 32 x half> %head, <vscale x 32 x half> poison, <vscale x 32 x i32> zeroinitializer
diff --git a/llvm/test/CodeGen/RISCV/rvv/vselect-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vselect-vp.ll
index 0a5e50160fbc9..ee0617c931480 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vselect-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vselect-vp.ll
@@ -354,11 +354,17 @@ define <vscale x 32 x i32> @select_nxv32i32(<vscale x 32 x i1> %a, <vscale x 32
; CHECK-NEXT: addi sp, sp, -16
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a1, a1, 4
+; CHECK-NEXT: slli a1, a1, 5
; CHECK-NEXT: sub sp, sp, a1
-; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
+; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x20, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 32 * vlenb
; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a1, a1, 3
+; CHECK-NEXT: slli a1, a1, 4
+; CHECK-NEXT: add a1, sp, a1
+; CHECK-NEXT: addi a1, a1, 16
+; CHECK-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
+; CHECK-NEXT: csrr a1, vlenb
+; CHECK-NEXT: li a3, 24
+; CHECK-NEXT: mul a1, a1, a3
; CHECK-NEXT: add a1, sp, a1
; CHECK-NEXT: addi a1, a1, 16
; CHECK-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
@@ -367,35 +373,51 @@ define <vscale x 32 x i32> @select_nxv32i32(<vscale x 32 x i1> %a, <vscale x 32
; CHECK-NEXT: slli a1, a3, 3
; CHECK-NEXT: add a1, a0, a1
; CHECK-NEXT: vl8re32.v v8, (a1)
+; CHECK-NEXT: csrr a1, vlenb
+; CHECK-NEXT: slli a1, a1, 3
+; CHECK-NEXT: add a1, sp, a1
+; CHECK-NEXT: addi a1, a1, 16
+; CHECK-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
; CHECK-NEXT: slli a1, a3, 1
; CHECK-NEXT: sub a4, a2, a1
; CHECK-NEXT: sltu a5, a2, a4
; CHECK-NEXT: addi a5, a5, -1
-; CHECK-NEXT: and a4, a5, a4
; CHECK-NEXT: srli a3, a3, 2
-; CHECK-NEXT: vl8re32.v v0, (a0)
+; CHECK-NEXT: vl8re32.v v8, (a0)
; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vs8r.v v0, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: vsetvli a0, zero, e8, mf2, ta, ma
-; CHECK-NEXT: vslidedown.vx v0, v24, a3
+; CHECK-NEXT: vslidedown.vx v0, v0, a3
+; CHECK-NEXT: and a4, a5, a4
+; CHECK-NEXT: csrr a0, vlenb
+; CHECK-NEXT: slli a0, a0, 4
+; CHECK-NEXT: add a0, sp, a0
+; CHECK-NEXT: addi a0, a0, 16
+; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: csrr a0, vlenb
+; CHECK-NEXT: slli a0, a0, 3
+; CHECK-NEXT: add a0, sp, a0
+; CHECK-NEXT: addi a0, a0, 16
+; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, a4, e32, m8, ta, ma
; CHECK-NEXT: vmerge.vvm v16, v8, v16, v0
; CHECK-NEXT: bltu a2, a1, .LBB27_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a2, a1
; CHECK-NEXT: .LBB27_2:
-; CHECK-NEXT: vsetvli zero, a2, e32, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
+; CHECK-NEXT: li a1, 24
+; CHECK-NEXT: mul a0, a0, a1
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vmerge.vvm v8, v24, v8, v0
+; CHECK-NEXT: addi a0, sp, 16
+; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, a2, e32, m8, ta, ma
+; CHECK-NEXT: vmerge.vvm v8, v8, v24, v0
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 4
+; CHECK-NEXT: slli a0, a0, 5
; CHECK-NEXT: add sp, sp, a0
; CHECK-NEXT: addi sp, sp, 16
; CHECK-NEXT: ret
@@ -411,11 +433,17 @@ define <vscale x 32 x i32> @select_evl_nxv32i32(<vscale x 32 x i1> %a, <vscale x
; CHECK-NEXT: addi sp, sp, -16
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a1, a1, 4
+; CHECK-NEXT: slli a1, a1, 5
; CHECK-NEXT: sub sp, sp, a1
-; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
+; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x20, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 32 * vlenb
; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a1, a1, 3
+; CHECK-NEXT: slli a1, a1, 4
+; CHECK-NEXT: add a1, sp, a1
+; CHECK-NEXT: addi a1, a1, 16
+; CHECK-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
+; CHECK-NEXT: csrr a1, vlenb
+; CHECK-NEXT: li a2, 24
+; CHECK-NEXT: mul a1, a1, a2
; CHECK-NEXT: add a1, sp, a1
; CHECK-NEXT: addi a1, a1, 16
; CHECK-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
@@ -424,35 +452,51 @@ define <vscale x 32 x i32> @select_evl_nxv32i32(<vscale x 32 x i1> %a, <vscale x
; CHECK-NEXT: slli a2, a1, 3
; CHECK-NEXT: add a2, a0, a2
; CHECK-NEXT: vl8re32.v v8, (a2)
+; CHECK-NEXT: csrr a2, vlenb
+; CHECK-NEXT: slli a2, a2, 3
+; CHECK-NEXT: add a2, sp, a2
+; CHECK-NEXT: addi a2, a2, 16
+; CHECK-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
; CHECK-NEXT: slli a2, a1, 1
; CHECK-NEXT: sub a3, a1, a2
; CHECK-NEXT: sltu a4, a1, a3
; CHECK-NEXT: addi a4, a4, -1
-; CHECK-NEXT: and a3, a4, a3
-; CHECK-NEXT: srli a4, a1, 2
-; CHECK-NEXT: vl8re32.v v0, (a0)
+; CHECK-NEXT: srli a5, a1, 2
+; CHECK-NEXT: vl8re32.v v8, (a0)
; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vs8r.v v0, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: vsetvli a0, zero, e8, mf2, ta, ma
-; CHECK-NEXT: vslidedown.vx v0, v24, a4
+; CHECK-NEXT: vslidedown.vx v0, v0, a5
+; CHECK-NEXT: and a3, a4, a3
+; CHECK-NEXT: csrr a0, vlenb
+; CHECK-NEXT: slli a0, a0, 4
+; CHECK-NEXT: add a0, sp, a0
+; CHECK-NEXT: addi a0, a0, 16
+; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: csrr a0, vlenb
+; CHECK-NEXT: slli a0, a0, 3
+; CHECK-NEXT: add a0, sp, a0
+; CHECK-NEXT: addi a0, a0, 16
+; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; CHECK-NEXT: vmerge.vvm v16, v8, v16, v0
; CHECK-NEXT: bltu a1, a2, .LBB28_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a1, a2
; CHECK-NEXT: .LBB28_2:
-; CHECK-NEXT: vsetvli zero, a1, e32, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
+; CHECK-NEXT: li a2, 24
+; CHECK-NEXT: mul a0, a0, a2
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vmerge.vvm v8, v24, v8, v0
+; CHECK-NEXT: addi a0, sp, 16
+; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, a1, e32, m8, ta, ma
+; CHECK-NEXT: vmerge.vvm v8, v8, v24, v0
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 4
+; CHECK-NEXT: slli a0, a0, 5
; CHECK-NEXT: add sp, sp, a0
; CHECK-NEXT: addi sp, sp, 16
; CHECK-NEXT: ret
@@ -714,20 +758,19 @@ define <vscale x 16 x double> @select_nxv16f64(<vscale x 16 x i1> %a, <vscale x
; CHECK-NEXT: sub a3, a2, a1
; CHECK-NEXT: sltu a4, a2, a3
; CHECK-NEXT: addi a4, a4, -1
-; CHECK-NEXT: and a3, a4, a3
-; CHECK-NEXT: srli a4, a1, 3
+; CHECK-NEXT: srli a5, a1, 3
; CHECK-NEXT: vl8re64.v v0, (a0)
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vs8r.v v0, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: vsetvli a0, zero, e8, mf4, ta, ma
-; CHECK-NEXT: vslidedown.vx v0, v24, a4
+; CHECK-NEXT: vslidedown.vx v0, v24, a5
+; CHECK-NEXT: and a3, a4, a3
; CHECK-NEXT: vsetvli zero, a3, e64, m8, ta, ma
; CHECK-NEXT: vmerge.vvm v16, v8, v16, v0
; CHECK-NEXT: bltu a2, a1, .LBB48_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a2, a1
; CHECK-NEXT: .LBB48_2:
-; CHECK-NEXT: vsetvli zero, a2, e64, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
@@ -736,6 +779,7 @@ define <vscale x 16 x double> @select_nxv16f64(<vscale x 16 x i1> %a, <vscale x
; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, a2, e64, m8, ta, ma
; CHECK-NEXT: vmerge.vvm v8, v24, v8, v0
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 4
diff --git a/llvm/test/CodeGen/RISCV/rvv/vsetvli-insert-crossbb.ll b/llvm/test/CodeGen/RISCV/rvv/vsetvli-insert-crossbb.ll
index 25aa3a7081a16..beebf0b9c5686 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vsetvli-insert-crossbb.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vsetvli-insert-crossbb.ll
@@ -342,13 +342,13 @@ define <vscale x 1 x double> @test8(i64 %avl, i8 zeroext %cond, <vscale x 1 x do
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vs1r.v v8, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: call foo
-; CHECK-NEXT: vsetvli zero, s0, e64, m1, ta, ma
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: add a0, a0, sp
; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vl1r.v v8, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl1r.v v9, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, s0, e64, m1, ta, ma
; CHECK-NEXT: vfsub.vv v8, v9, v8
; CHECK-NEXT: .LBB6_3: # %if.then
; CHECK-NEXT: csrr a0, vlenb
diff --git a/llvm/test/CodeGen/RISCV/rvv/vsetvli-regression.ll b/llvm/test/CodeGen/RISCV/rvv/vsetvli-regression.ll
index f658a2c6b24a6..c3b19b59ec3d6 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vsetvli-regression.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vsetvli-regression.ll
@@ -11,9 +11,10 @@ define i32 @illegal_preserve_vl(<vscale x 2 x i32> %a, <vscale x 4 x i64> %x, pt
; CHECK: # %bb.0:
; CHECK-NEXT: vsetvli a1, zero, e64, m4, ta, ma
; CHECK-NEXT: vadd.vv v12, v12, v12
-; CHECK-NEXT: vs4r.v v12, (a0)
; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, ma
-; CHECK-NEXT: vmv.x.s a0, v8
+; CHECK-NEXT: vmv.x.s a1, v8
+; CHECK-NEXT: vs4r.v v12, (a0)
+; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: ret
%index = add <vscale x 4 x i64> %x, %x
store <vscale x 4 x i64> %index, ptr %y
diff --git a/llvm/test/CodeGen/RISCV/rvv/vsext-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vsext-vp.ll
index 834e7dd85aea0..9b5a1a54ad5df 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vsext-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vsext-vp.ll
@@ -167,8 +167,8 @@ define <vscale x 32 x i32> @vsext_nxv32i8_nxv32i32(<vscale x 32 x i8> %a, <vscal
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB12_2:
-; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; CHECK-NEXT: vsext.vf4 v24, v8, v0.t
; CHECK-NEXT: vmv.v.v v8, v24
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vsitofp-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vsitofp-vp.ll
index 706876dc38547..ca44a9a64de4c 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vsitofp-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vsitofp-vp.ll
@@ -386,7 +386,15 @@ declare <vscale x 32 x half> @llvm.vp.sitofp.nxv32f16.nxv32i32(<vscale x 32 x i3
define <vscale x 32 x half> @vsitofp_nxv32f16_nxv32i32(<vscale x 32 x i32> %va, <vscale x 32 x i1> %m, i32 zeroext %evl) {
; ZVFH-LABEL: vsitofp_nxv32f16_nxv32i32:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vmv1r.v v24, v0
+; ZVFH-NEXT: addi sp, sp, -16
+; ZVFH-NEXT: .cfi_def_cfa_offset 16
+; ZVFH-NEXT: csrr a1, vlenb
+; ZVFH-NEXT: slli a1, a1, 3
+; ZVFH-NEXT: sub sp, sp, a1
+; ZVFH-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; ZVFH-NEXT: vmv1r.v v7, v0
+; ZVFH-NEXT: addi a1, sp, 16
+; ZVFH-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
; ZVFH-NEXT: csrr a1, vlenb
; ZVFH-NEXT: srli a2, a1, 2
; ZVFH-NEXT: vsetvli a3, zero, e8, mf2, ta, ma
@@ -396,16 +404,22 @@ define <vscale x 32 x half> @vsitofp_nxv32f16_nxv32i32(<vscale x 32 x i32> %va,
; ZVFH-NEXT: sltu a3, a0, a2
; ZVFH-NEXT: addi a3, a3, -1
; ZVFH-NEXT: and a2, a3, a2
+; ZVFH-NEXT: addi a3, sp, 16
+; ZVFH-NEXT: vl8r.v v24, (a3) # Unknown-size Folded Reload
; ZVFH-NEXT: vsetvli zero, a2, e16, m4, ta, ma
-; ZVFH-NEXT: vfncvt.f.x.w v28, v16, v0.t
+; ZVFH-NEXT: vfncvt.f.x.w v20, v24, v0.t
; ZVFH-NEXT: bltu a0, a1, .LBB25_2
; ZVFH-NEXT: # %bb.1:
; ZVFH-NEXT: mv a0, a1
; ZVFH-NEXT: .LBB25_2:
+; ZVFH-NEXT: vmv1r.v v0, v7
; ZVFH-NEXT: vsetvli zero, a0, e16, m4, ta, ma
-; ZVFH-NEXT: vmv1r.v v0, v24
-; ZVFH-NEXT: vfncvt.f.x.w v24, v8, v0.t
-; ZVFH-NEXT: vmv8r.v v8, v24
+; ZVFH-NEXT: vfncvt.f.x.w v16, v8, v0.t
+; ZVFH-NEXT: vmv8r.v v8, v16
+; ZVFH-NEXT: csrr a0, vlenb
+; ZVFH-NEXT: slli a0, a0, 3
+; ZVFH-NEXT: add sp, sp, a0
+; ZVFH-NEXT: addi sp, sp, 16
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vsitofp_nxv32f16_nxv32i32:
@@ -428,8 +442,8 @@ define <vscale x 32 x half> @vsitofp_nxv32f16_nxv32i32(<vscale x 32 x i32> %va,
; ZVFHMIN-NEXT: # %bb.1:
; ZVFHMIN-NEXT: mv a0, a1
; ZVFHMIN-NEXT: .LBB25_2:
-; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v7
+; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfcvt.f.x.v v8, v8, v0.t
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v16, v8
@@ -460,8 +474,8 @@ define <vscale x 32 x float> @vsitofp_nxv32f32_nxv32i32(<vscale x 32 x i32> %va,
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB26_2:
-; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; CHECK-NEXT: vfcvt.f.x.v v8, v8, v0.t
; CHECK-NEXT: ret
%v = call <vscale x 32 x float> @llvm.vp.sitofp.nxv32f32.nxv32i32(<vscale x 32 x i32> %va, <vscale x 32 x i1> %m, i32 %evl)
diff --git a/llvm/test/CodeGen/RISCV/rvv/vssub-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vssub-vp.ll
index b56a0f40176cf..613b58b0f1b88 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vssub-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vssub-vp.ll
@@ -593,22 +593,22 @@ define <vscale x 128 x i8> @vssub_vi_nxv128i8(<vscale x 128 x i8> %va, <vscale x
; CHECK-NEXT: vmv1r.v v24, v0
; CHECK-NEXT: vsetvli a2, zero, e8, m8, ta, ma
; CHECK-NEXT: vlm.v v0, (a0)
-; CHECK-NEXT: csrr a2, vlenb
-; CHECK-NEXT: slli a2, a2, 3
-; CHECK-NEXT: sub a0, a1, a2
-; CHECK-NEXT: sltu a3, a1, a0
+; CHECK-NEXT: csrr a0, vlenb
+; CHECK-NEXT: slli a0, a0, 3
+; CHECK-NEXT: sub a2, a1, a0
+; CHECK-NEXT: sltu a3, a1, a2
; CHECK-NEXT: addi a3, a3, -1
-; CHECK-NEXT: and a3, a3, a0
-; CHECK-NEXT: li a0, -1
+; CHECK-NEXT: and a3, a3, a2
+; CHECK-NEXT: li a2, -1
; CHECK-NEXT: vsetvli zero, a3, e8, m8, ta, ma
-; CHECK-NEXT: vssub.vx v16, v16, a0, v0.t
-; CHECK-NEXT: bltu a1, a2, .LBB50_2
+; CHECK-NEXT: vssub.vx v16, v16, a2, v0.t
+; CHECK-NEXT: bltu a1, a0, .LBB50_2
; CHECK-NEXT: # %bb.1:
-; CHECK-NEXT: mv a1, a2
+; CHECK-NEXT: mv a1, a0
; CHECK-NEXT: .LBB50_2:
-; CHECK-NEXT: vsetvli zero, a1, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
-; CHECK-NEXT: vssub.vx v8, v8, a0, v0.t
+; CHECK-NEXT: vsetvli zero, a1, e8, m8, ta, ma
+; CHECK-NEXT: vssub.vx v8, v8, a2, v0.t
; CHECK-NEXT: ret
%v = call <vscale x 128 x i8> @llvm.vp.ssub.sat.nxv128i8(<vscale x 128 x i8> %va, <vscale x 128 x i8> splat (i8 -1), <vscale x 128 x i1> %m, i32 %evl)
ret <vscale x 128 x i8> %v
@@ -1393,25 +1393,25 @@ define <vscale x 32 x i32> @vssub_vi_nxv32i32(<vscale x 32 x i32> %va, <vscale x
; CHECK-LABEL: vssub_vi_nxv32i32:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v24, v0
-; CHECK-NEXT: csrr a2, vlenb
-; CHECK-NEXT: srli a1, a2, 2
+; CHECK-NEXT: csrr a1, vlenb
+; CHECK-NEXT: srli a2, a1, 2
; CHECK-NEXT: vsetvli a3, zero, e8, mf2, ta, ma
-; CHECK-NEXT: vslidedown.vx v0, v0, a1
-; CHECK-NEXT: slli a2, a2, 1
-; CHECK-NEXT: sub a1, a0, a2
-; CHECK-NEXT: sltu a3, a0, a1
+; CHECK-NEXT: vslidedown.vx v0, v0, a2
+; CHECK-NEXT: slli a1, a1, 1
+; CHECK-NEXT: sub a2, a0, a1
+; CHECK-NEXT: sltu a3, a0, a2
; CHECK-NEXT: addi a3, a3, -1
-; CHECK-NEXT: and a3, a3, a1
-; CHECK-NEXT: li a1, -1
+; CHECK-NEXT: and a3, a3, a2
+; CHECK-NEXT: li a2, -1
; CHECK-NEXT: vsetvli zero, a3, e32, m8, ta, ma
-; CHECK-NEXT: vssub.vx v16, v16, a1, v0.t
-; CHECK-NEXT: bltu a0, a2, .LBB118_2
+; CHECK-NEXT: vssub.vx v16, v16, a2, v0.t
+; CHECK-NEXT: bltu a0, a1, .LBB118_2
; CHECK-NEXT: # %bb.1:
-; CHECK-NEXT: mv a0, a2
+; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB118_2:
-; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
-; CHECK-NEXT: vssub.vx v8, v8, a1, v0.t
+; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
+; CHECK-NEXT: vssub.vx v8, v8, a2, v0.t
; CHECK-NEXT: ret
%v = call <vscale x 32 x i32> @llvm.vp.ssub.sat.nxv32i32(<vscale x 32 x i32> %va, <vscale x 32 x i32> splat (i32 -1), <vscale x 32 x i1> %m, i32 %evl)
ret <vscale x 32 x i32> %v
diff --git a/llvm/test/CodeGen/RISCV/rvv/vssubu-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vssubu-vp.ll
index 8275c3081c7c1..8c729d7d9bfb6 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vssubu-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vssubu-vp.ll
@@ -591,22 +591,22 @@ define <vscale x 128 x i8> @vssubu_vi_nxv128i8(<vscale x 128 x i8> %va, <vscale
; CHECK-NEXT: vmv1r.v v24, v0
; CHECK-NEXT: vsetvli a2, zero, e8, m8, ta, ma
; CHECK-NEXT: vlm.v v0, (a0)
-; CHECK-NEXT: csrr a2, vlenb
-; CHECK-NEXT: slli a2, a2, 3
-; CHECK-NEXT: sub a0, a1, a2
-; CHECK-NEXT: sltu a3, a1, a0
+; CHECK-NEXT: csrr a0, vlenb
+; CHECK-NEXT: slli a0, a0, 3
+; CHECK-NEXT: sub a2, a1, a0
+; CHECK-NEXT: sltu a3, a1, a2
; CHECK-NEXT: addi a3, a3, -1
-; CHECK-NEXT: and a3, a3, a0
-; CHECK-NEXT: li a0, -1
+; CHECK-NEXT: and a3, a3, a2
+; CHECK-NEXT: li a2, -1
; CHECK-NEXT: vsetvli zero, a3, e8, m8, ta, ma
-; CHECK-NEXT: vssubu.vx v16, v16, a0, v0.t
-; CHECK-NEXT: bltu a1, a2, .LBB50_2
+; CHECK-NEXT: vssubu.vx v16, v16, a2, v0.t
+; CHECK-NEXT: bltu a1, a0, .LBB50_2
; CHECK-NEXT: # %bb.1:
-; CHECK-NEXT: mv a1, a2
+; CHECK-NEXT: mv a1, a0
; CHECK-NEXT: .LBB50_2:
-; CHECK-NEXT: vsetvli zero, a1, e8, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
-; CHECK-NEXT: vssubu.vx v8, v8, a0, v0.t
+; CHECK-NEXT: vsetvli zero, a1, e8, m8, ta, ma
+; CHECK-NEXT: vssubu.vx v8, v8, a2, v0.t
; CHECK-NEXT: ret
%v = call <vscale x 128 x i8> @llvm.vp.usub.sat.nxv128i8(<vscale x 128 x i8> %va, <vscale x 128 x i8> splat (i8 -1), <vscale x 128 x i1> %m, i32 %evl)
ret <vscale x 128 x i8> %v
@@ -1391,25 +1391,25 @@ define <vscale x 32 x i32> @vssubu_vi_nxv32i32(<vscale x 32 x i32> %va, <vscale
; CHECK-LABEL: vssubu_vi_nxv32i32:
; CHECK: # %bb.0:
; CHECK-NEXT: vmv1r.v v24, v0
-; CHECK-NEXT: csrr a2, vlenb
-; CHECK-NEXT: srli a1, a2, 2
+; CHECK-NEXT: csrr a1, vlenb
+; CHECK-NEXT: srli a2, a1, 2
; CHECK-NEXT: vsetvli a3, zero, e8, mf2, ta, ma
-; CHECK-NEXT: vslidedown.vx v0, v0, a1
-; CHECK-NEXT: slli a2, a2, 1
-; CHECK-NEXT: sub a1, a0, a2
-; CHECK-NEXT: sltu a3, a0, a1
+; CHECK-NEXT: vslidedown.vx v0, v0, a2
+; CHECK-NEXT: slli a1, a1, 1
+; CHECK-NEXT: sub a2, a0, a1
+; CHECK-NEXT: sltu a3, a0, a2
; CHECK-NEXT: addi a3, a3, -1
-; CHECK-NEXT: and a3, a3, a1
-; CHECK-NEXT: li a1, -1
+; CHECK-NEXT: and a3, a3, a2
+; CHECK-NEXT: li a2, -1
; CHECK-NEXT: vsetvli zero, a3, e32, m8, ta, ma
-; CHECK-NEXT: vssubu.vx v16, v16, a1, v0.t
-; CHECK-NEXT: bltu a0, a2, .LBB118_2
+; CHECK-NEXT: vssubu.vx v16, v16, a2, v0.t
+; CHECK-NEXT: bltu a0, a1, .LBB118_2
; CHECK-NEXT: # %bb.1:
-; CHECK-NEXT: mv a0, a2
+; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB118_2:
-; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
-; CHECK-NEXT: vssubu.vx v8, v8, a1, v0.t
+; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
+; CHECK-NEXT: vssubu.vx v8, v8, a2, v0.t
; CHECK-NEXT: ret
%v = call <vscale x 32 x i32> @llvm.vp.usub.sat.nxv32i32(<vscale x 32 x i32> %va, <vscale x 32 x i32> splat (i32 -1), <vscale x 32 x i1> %m, i32 %evl)
ret <vscale x 32 x i32> %v
diff --git a/llvm/test/CodeGen/RISCV/rvv/vtrunc-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vtrunc-vp.ll
index 4857810e7a170..27755c166cc52 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vtrunc-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vtrunc-vp.ll
@@ -174,8 +174,8 @@ define <vscale x 15 x i16> @vtrunc_nxv15i16_nxv15i64(<vscale x 15 x i64> %a, <vs
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB12_2:
-; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, ma
; CHECK-NEXT: vnsrl.wi v20, v8, 0, v0.t
; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; CHECK-NEXT: vnsrl.wi v16, v20, 0, v0.t
@@ -232,8 +232,8 @@ define <vscale x 32 x i7> @vtrunc_nxv32i7_nxv32i32(<vscale x 32 x i32> %a, <vsca
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB15_2:
-; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, ma
; CHECK-NEXT: vnsrl.wi v20, v8, 0, v0.t
; CHECK-NEXT: vsetvli zero, zero, e8, m2, ta, ma
; CHECK-NEXT: vnsrl.wi v16, v20, 0, v0.t
@@ -266,8 +266,8 @@ define <vscale x 32 x i8> @vtrunc_nxv32i8_nxv32i32(<vscale x 32 x i32> %a, <vsca
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB16_2:
-; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, ma
; CHECK-NEXT: vnsrl.wi v20, v8, 0, v0.t
; CHECK-NEXT: vsetvli zero, zero, e8, m2, ta, ma
; CHECK-NEXT: vnsrl.wi v16, v20, 0, v0.t
@@ -312,11 +312,11 @@ define <vscale x 32 x i32> @vtrunc_nxv32i64_nxv32i32(<vscale x 32 x i64> %a, <vs
; CHECK-NEXT: sub a6, a5, a1
; CHECK-NEXT: sltu a7, a5, a6
; CHECK-NEXT: addi a7, a7, -1
-; CHECK-NEXT: and a6, a7, a6
-; CHECK-NEXT: vsetvli a7, zero, e8, mf4, ta, ma
; CHECK-NEXT: vl8re64.v v24, (a0)
+; CHECK-NEXT: vsetvli a0, zero, e8, mf4, ta, ma
; CHECK-NEXT: vslidedown.vx v0, v16, a3
-; CHECK-NEXT: vsetvli zero, a6, e32, m4, ta, ma
+; CHECK-NEXT: and a0, a7, a6
+; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, ma
; CHECK-NEXT: vnsrl.wi v20, v8, 0, v0.t
; CHECK-NEXT: bltu a5, a1, .LBB17_2
; CHECK-NEXT: # %bb.1:
@@ -324,8 +324,8 @@ define <vscale x 32 x i32> @vtrunc_nxv32i64_nxv32i32(<vscale x 32 x i64> %a, <vs
; CHECK-NEXT: .LBB17_2:
; CHECK-NEXT: vsetvli a0, zero, e8, mf4, ta, ma
; CHECK-NEXT: vslidedown.vx v6, v7, a3
-; CHECK-NEXT: vsetvli zero, a5, e32, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v16
+; CHECK-NEXT: vsetvli zero, a5, e32, m4, ta, ma
; CHECK-NEXT: vnsrl.wi v16, v24, 0, v0.t
; CHECK-NEXT: bltu a2, a4, .LBB17_4
; CHECK-NEXT: # %bb.3:
@@ -335,22 +335,22 @@ define <vscale x 32 x i32> @vtrunc_nxv32i64_nxv32i32(<vscale x 32 x i64> %a, <vs
; CHECK-NEXT: sltu a3, a2, a0
; CHECK-NEXT: addi a3, a3, -1
; CHECK-NEXT: and a0, a3, a0
-; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v6
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: addi a3, sp, 16
+; CHECK-NEXT: vl8r.v v8, (a3) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, ma
; CHECK-NEXT: vnsrl.wi v28, v8, 0, v0.t
; CHECK-NEXT: bltu a2, a1, .LBB17_6
; CHECK-NEXT: # %bb.5:
; CHECK-NEXT: mv a2, a1
; CHECK-NEXT: .LBB17_6:
-; CHECK-NEXT: vsetvli zero, a2, e32, m4, ta, ma
; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vsetvli zero, a2, e32, m4, ta, ma
; CHECK-NEXT: vnsrl.wi v24, v8, 0, v0.t
; CHECK-NEXT: vmv8r.v v8, v24
; CHECK-NEXT: csrr a0, vlenb
diff --git a/llvm/test/CodeGen/RISCV/rvv/vuitofp-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vuitofp-vp.ll
index e083d594db25d..e5941dc7b5ab2 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vuitofp-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vuitofp-vp.ll
@@ -386,7 +386,15 @@ declare <vscale x 32 x half> @llvm.vp.uitofp.nxv32f16.nxv32i32(<vscale x 32 x i3
define <vscale x 32 x half> @vuitofp_nxv32f16_nxv32i32(<vscale x 32 x i32> %va, <vscale x 32 x i1> %m, i32 zeroext %evl) {
; ZVFH-LABEL: vuitofp_nxv32f16_nxv32i32:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vmv1r.v v24, v0
+; ZVFH-NEXT: addi sp, sp, -16
+; ZVFH-NEXT: .cfi_def_cfa_offset 16
+; ZVFH-NEXT: csrr a1, vlenb
+; ZVFH-NEXT: slli a1, a1, 3
+; ZVFH-NEXT: sub sp, sp, a1
+; ZVFH-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; ZVFH-NEXT: vmv1r.v v7, v0
+; ZVFH-NEXT: addi a1, sp, 16
+; ZVFH-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
; ZVFH-NEXT: csrr a1, vlenb
; ZVFH-NEXT: srli a2, a1, 2
; ZVFH-NEXT: vsetvli a3, zero, e8, mf2, ta, ma
@@ -396,16 +404,22 @@ define <vscale x 32 x half> @vuitofp_nxv32f16_nxv32i32(<vscale x 32 x i32> %va,
; ZVFH-NEXT: sltu a3, a0, a2
; ZVFH-NEXT: addi a3, a3, -1
; ZVFH-NEXT: and a2, a3, a2
+; ZVFH-NEXT: addi a3, sp, 16
+; ZVFH-NEXT: vl8r.v v24, (a3) # Unknown-size Folded Reload
; ZVFH-NEXT: vsetvli zero, a2, e16, m4, ta, ma
-; ZVFH-NEXT: vfncvt.f.xu.w v28, v16, v0.t
+; ZVFH-NEXT: vfncvt.f.xu.w v20, v24, v0.t
; ZVFH-NEXT: bltu a0, a1, .LBB25_2
; ZVFH-NEXT: # %bb.1:
; ZVFH-NEXT: mv a0, a1
; ZVFH-NEXT: .LBB25_2:
+; ZVFH-NEXT: vmv1r.v v0, v7
; ZVFH-NEXT: vsetvli zero, a0, e16, m4, ta, ma
-; ZVFH-NEXT: vmv1r.v v0, v24
-; ZVFH-NEXT: vfncvt.f.xu.w v24, v8, v0.t
-; ZVFH-NEXT: vmv8r.v v8, v24
+; ZVFH-NEXT: vfncvt.f.xu.w v16, v8, v0.t
+; ZVFH-NEXT: vmv8r.v v8, v16
+; ZVFH-NEXT: csrr a0, vlenb
+; ZVFH-NEXT: slli a0, a0, 3
+; ZVFH-NEXT: add sp, sp, a0
+; ZVFH-NEXT: addi sp, sp, 16
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vuitofp_nxv32f16_nxv32i32:
@@ -428,8 +442,8 @@ define <vscale x 32 x half> @vuitofp_nxv32f16_nxv32i32(<vscale x 32 x i32> %va,
; ZVFHMIN-NEXT: # %bb.1:
; ZVFHMIN-NEXT: mv a0, a1
; ZVFHMIN-NEXT: .LBB25_2:
-; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmv1r.v v0, v7
+; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfcvt.f.xu.v v8, v8, v0.t
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v16, v8
@@ -460,8 +474,8 @@ define <vscale x 32 x float> @vuitofp_nxv32f32_nxv32i32(<vscale x 32 x i32> %va,
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB26_2:
-; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; CHECK-NEXT: vfcvt.f.xu.v v8, v8, v0.t
; CHECK-NEXT: ret
%v = call <vscale x 32 x float> @llvm.vp.uitofp.nxv32f32.nxv32i32(<vscale x 32 x i32> %va, <vscale x 32 x i1> %m, i32 %evl)
diff --git a/llvm/test/CodeGen/RISCV/rvv/vxrm-insert.ll b/llvm/test/CodeGen/RISCV/rvv/vxrm-insert.ll
index c5f34eee31189..a869b433a4952 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vxrm-insert.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vxrm-insert.ll
@@ -85,9 +85,9 @@ define <vscale x 1 x i8> @test3(<vscale x 1 x i8> %0, <vscale x 1 x i8> %1, <vsc
; RV32-NEXT: vaadd.vv v8, v8, v9
; RV32-NEXT: call foo
; RV32-NEXT: csrwi vxrm, 0
-; RV32-NEXT: vsetvli zero, s0, e8, mf8, ta, ma
; RV32-NEXT: addi a0, sp, 16
; RV32-NEXT: vl1r.v v9, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vsetvli zero, s0, e8, mf8, ta, ma
; RV32-NEXT: vaadd.vv v8, v8, v9
; RV32-NEXT: csrr a0, vlenb
; RV32-NEXT: slli a0, a0, 1
@@ -113,9 +113,9 @@ define <vscale x 1 x i8> @test3(<vscale x 1 x i8> %0, <vscale x 1 x i8> %1, <vsc
; RV64-NEXT: vaadd.vv v8, v8, v9
; RV64-NEXT: call foo
; RV64-NEXT: csrwi vxrm, 0
-; RV64-NEXT: vsetvli zero, s0, e8, mf8, ta, ma
; RV64-NEXT: addi a0, sp, 16
; RV64-NEXT: vl1r.v v9, (a0) # Unknown-size Folded Reload
+; RV64-NEXT: vsetvli zero, s0, e8, mf8, ta, ma
; RV64-NEXT: vaadd.vv v8, v8, v9
; RV64-NEXT: csrr a0, vlenb
; RV64-NEXT: slli a0, a0, 1
diff --git a/llvm/test/CodeGen/RISCV/rvv/vzext-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vzext-vp.ll
index 400f89b1ef77d..3b5541c1a2440 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vzext-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vzext-vp.ll
@@ -167,8 +167,8 @@ define <vscale x 32 x i32> @vzext_nxv32i8_nxv32i32(<vscale x 32 x i8> %a, <vscal
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB12_2:
-; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; CHECK-NEXT: vmv1r.v v0, v12
+; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; CHECK-NEXT: vzext.vf4 v24, v8, v0.t
; CHECK-NEXT: vmv.v.v v8, v24
; CHECK-NEXT: ret
diff --git a/llvm/test/CodeGen/RISCV/srem-seteq-illegal-types.ll b/llvm/test/CodeGen/RISCV/srem-seteq-illegal-types.ll
index 9ecfa50178316..3ed98ac6d5b1d 100644
--- a/llvm/test/CodeGen/RISCV/srem-seteq-illegal-types.ll
+++ b/llvm/test/CodeGen/RISCV/srem-seteq-illegal-types.ll
@@ -643,9 +643,9 @@ define void @test_srem_vec(ptr %X) nounwind {
; RV32MV-NEXT: mv a1, s4
; RV32MV-NEXT: li a3, 0
; RV32MV-NEXT: call __moddi3
-; RV32MV-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32MV-NEXT: addi a2, sp, 16
; RV32MV-NEXT: vl2r.v v8, (a2) # Unknown-size Folded Reload
+; RV32MV-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32MV-NEXT: vslide1down.vx v8, v8, a0
; RV32MV-NEXT: vslide1down.vx v8, v8, a1
; RV32MV-NEXT: addi a0, sp, 16
@@ -655,9 +655,9 @@ define void @test_srem_vec(ptr %X) nounwind {
; RV32MV-NEXT: mv a0, s1
; RV32MV-NEXT: mv a1, s3
; RV32MV-NEXT: call __moddi3
-; RV32MV-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32MV-NEXT: addi a2, sp, 16
; RV32MV-NEXT: vl2r.v v8, (a2) # Unknown-size Folded Reload
+; RV32MV-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; RV32MV-NEXT: vslide1down.vx v8, v8, a0
; RV32MV-NEXT: vslide1down.vx v8, v8, a1
; RV32MV-NEXT: vslidedown.vi v8, v8, 2
More information about the llvm-commits
mailing list