[llvm] [RegAlloc][RISCV] Increase the spill weight by target factor (PR #113675)
Pengcheng Wang via llvm-commits
llvm-commits at lists.llvm.org
Thu Dec 5 05:13:08 PST 2024
https://github.com/wangpc-pp updated https://github.com/llvm/llvm-project/pull/113675
>From 92eac5d2cbd39a2a87d38d984a4ac4c9c082f544 Mon Sep 17 00:00:00 2001
From: Wang Pengcheng <wangpengcheng.pp at bytedance.com>
Date: Fri, 25 Oct 2024 18:23:03 +0800
Subject: [PATCH] [RegAlloc][RISCV] Increase the spill weight by target factor
Currently, the spill weight is only determined by isDef/isUse and
block frequency. However, for registers with different register
classes, the costs of spilling them are different.
For example, for `LMUL>1` registers (in which, several physical regsiter
compound a bigger logical register), the costs are larger than
`LMUL=1` case (in which, there is only one physical register).
To sovle this problem, a new target hook `getSpillWeightFactor` is
added. Targets can override the default factor (which is 1) according
to the register classes.
For RISC-V, the factors are set to the `RegClassWeight` which is
used to track regsiter pressure. The values of `RegClassWeight`
are the number of register units.
I believe all the targets can benefit from this change, but I will
shrink the range of tests to RISC-V only.
Partially fixes #113489.
---
llvm/include/llvm/CodeGen/LiveIntervals.h | 4 +-
.../include/llvm/CodeGen/TargetRegisterInfo.h | 4 +
llvm/lib/CodeGen/CalcSpillWeights.cpp | 12 +-
llvm/lib/CodeGen/LiveIntervals.cpp | 8 +-
llvm/lib/CodeGen/TargetRegisterInfo.cpp | 5 +
llvm/lib/Target/RISCV/RISCVRegisterInfo.cpp | 5 +
llvm/lib/Target/RISCV/RISCVRegisterInfo.h | 3 +
llvm/test/CodeGen/RISCV/rvv/abs-vp.ll | 37 +-
llvm/test/CodeGen/RISCV/rvv/bitreverse-vp.ll | 136 +-
llvm/test/CodeGen/RISCV/rvv/bswap-vp.ll | 94 +-
llvm/test/CodeGen/RISCV/rvv/ceil-vp.ll | 61 +-
llvm/test/CodeGen/RISCV/rvv/compressstore.ll | 38 +-
llvm/test/CodeGen/RISCV/rvv/ctpop-vp.ll | 207 +-
llvm/test/CodeGen/RISCV/rvv/cttz-vp.ll | 311 +--
llvm/test/CodeGen/RISCV/rvv/expandload.ll | 97 +-
.../RISCV/rvv/fixed-vectors-bitreverse-vp.ll | 226 +--
.../RISCV/rvv/fixed-vectors-bswap-vp.ll | 96 +-
.../RISCV/rvv/fixed-vectors-ceil-vp.ll | 58 +-
.../RISCV/rvv/fixed-vectors-ctlz-vp.ll | 628 ++----
.../RISCV/rvv/fixed-vectors-ctpop-vp.ll | 182 +-
.../RISCV/rvv/fixed-vectors-cttz-vp.ll | 404 ++--
.../RISCV/rvv/fixed-vectors-floor-vp.ll | 58 +-
.../RISCV/rvv/fixed-vectors-fmaximum-vp.ll | 59 +-
.../RISCV/rvv/fixed-vectors-fminimum-vp.ll | 59 +-
.../rvv/fixed-vectors-int-explodevector.ll | 26 +-
.../rvv/fixed-vectors-interleaved-access.ll | 632 +++---
.../RISCV/rvv/fixed-vectors-nearbyint-vp.ll | 16 -
.../RISCV/rvv/fixed-vectors-reduction-fp.ll | 296 ++-
.../RISCV/rvv/fixed-vectors-rint-vp.ll | 16 -
.../RISCV/rvv/fixed-vectors-round-vp.ll | 58 +-
.../RISCV/rvv/fixed-vectors-roundeven-vp.ll | 58 +-
.../RISCV/rvv/fixed-vectors-roundtozero-vp.ll | 58 +-
.../RISCV/rvv/fixed-vectors-setcc-fp-vp.ll | 624 +++---
.../RISCV/rvv/fixed-vectors-setcc-int-vp.ll | 58 +-
.../RISCV/rvv/fixed-vectors-trunc-vp.ll | 148 +-
.../RISCV/rvv/fixed-vectors-vcopysign-vp.ll | 28 +-
.../RISCV/rvv/fixed-vectors-vfmax-vp.ll | 28 +-
.../RISCV/rvv/fixed-vectors-vfmin-vp.ll | 28 +-
.../RISCV/rvv/fixed-vectors-vselect-vp.ll | 84 +-
llvm/test/CodeGen/RISCV/rvv/floor-vp.ll | 61 +-
.../test/CodeGen/RISCV/rvv/fmaximum-sdnode.ll | 38 +-
llvm/test/CodeGen/RISCV/rvv/fmaximum-vp.ll | 243 +--
.../test/CodeGen/RISCV/rvv/fminimum-sdnode.ll | 38 +-
llvm/test/CodeGen/RISCV/rvv/fminimum-vp.ll | 243 +--
llvm/test/CodeGen/RISCV/rvv/fshr-fshl-vp.ll | 286 +--
.../test/CodeGen/RISCV/rvv/mscatter-sdnode.ll | 82 +-
llvm/test/CodeGen/RISCV/rvv/rint-vp.ll | 45 -
llvm/test/CodeGen/RISCV/rvv/round-vp.ll | 61 +-
llvm/test/CodeGen/RISCV/rvv/roundeven-vp.ll | 61 +-
llvm/test/CodeGen/RISCV/rvv/roundtozero-vp.ll | 61 +-
llvm/test/CodeGen/RISCV/rvv/setcc-fp-vp.ll | 286 ++-
llvm/test/CodeGen/RISCV/rvv/setcc-int-vp.ll | 60 +-
.../test/CodeGen/RISCV/rvv/strided-vpstore.ll | 30 +-
.../RISCV/rvv/vector-deinterleave-load.ll | 51 +-
.../CodeGen/RISCV/rvv/vector-deinterleave.ll | 90 +-
llvm/test/CodeGen/RISCV/rvv/vfadd-vp.ll | 184 +-
llvm/test/CodeGen/RISCV/rvv/vfdiv-vp.ll | 184 +-
llvm/test/CodeGen/RISCV/rvv/vfma-vp.ll | 1754 ++++++++---------
.../RISCV/rvv/vfmadd-constrained-sdnode.ll | 110 +-
llvm/test/CodeGen/RISCV/rvv/vfmadd-sdnode.ll | 306 +--
.../RISCV/rvv/vfmsub-constrained-sdnode.ll | 17 +-
llvm/test/CodeGen/RISCV/rvv/vfmul-vp.ll | 92 +-
llvm/test/CodeGen/RISCV/rvv/vfptrunc-vp.ll | 26 +-
llvm/test/CodeGen/RISCV/rvv/vfsub-vp.ll | 184 +-
llvm/test/CodeGen/RISCV/rvv/vfwmacc-vp.ll | 24 +-
llvm/test/CodeGen/RISCV/rvv/vfwnmacc-vp.ll | 24 +-
llvm/test/CodeGen/RISCV/rvv/vfwnmsac-vp.ll | 24 +-
llvm/test/CodeGen/RISCV/rvv/vpstore.ll | 30 +-
llvm/test/CodeGen/RISCV/rvv/vselect-fp.ll | 9 +-
llvm/test/CodeGen/RISCV/rvv/vselect-vp.ll | 96 +-
llvm/test/CodeGen/RISCV/rvv/vtrunc-vp.ll | 9 +-
71 files changed, 3637 insertions(+), 6122 deletions(-)
diff --git a/llvm/include/llvm/CodeGen/LiveIntervals.h b/llvm/include/llvm/CodeGen/LiveIntervals.h
index 161bb247a0e968..a58ba178ac8484 100644
--- a/llvm/include/llvm/CodeGen/LiveIntervals.h
+++ b/llvm/include/llvm/CodeGen/LiveIntervals.h
@@ -117,14 +117,14 @@ class LiveIntervals {
/// If \p PSI is provided the calculation is altered for optsize functions.
static float getSpillWeight(bool isDef, bool isUse,
const MachineBlockFrequencyInfo *MBFI,
- const MachineInstr &MI,
+ const MachineInstr &MI, unsigned Factor = 1,
ProfileSummaryInfo *PSI = nullptr);
/// Calculate the spill weight to assign to a single instruction.
/// If \p PSI is provided the calculation is altered for optsize functions.
static float getSpillWeight(bool isDef, bool isUse,
const MachineBlockFrequencyInfo *MBFI,
- const MachineBasicBlock *MBB,
+ const MachineBasicBlock *MBB, unsigned Factor = 1,
ProfileSummaryInfo *PSI = nullptr);
LiveInterval &getInterval(Register Reg) {
diff --git a/llvm/include/llvm/CodeGen/TargetRegisterInfo.h b/llvm/include/llvm/CodeGen/TargetRegisterInfo.h
index 292fa3c94969be..7dd272fed996d1 100644
--- a/llvm/include/llvm/CodeGen/TargetRegisterInfo.h
+++ b/llvm/include/llvm/CodeGen/TargetRegisterInfo.h
@@ -926,6 +926,10 @@ class TargetRegisterInfo : public MCRegisterInfo {
/// Returns a -1 terminated array of pressure set IDs.
virtual const int *getRegUnitPressureSets(unsigned RegUnit) const = 0;
+ /// Get the factor of spill weight for this register class.
+ virtual unsigned
+ getSpillWeightScaleFactor(const TargetRegisterClass *RC) const;
+
/// Get a list of 'hint' registers that the register allocator should try
/// first when allocating a physical register for the virtual register
/// VirtReg. These registers are effectively moved to the front of the
diff --git a/llvm/lib/CodeGen/CalcSpillWeights.cpp b/llvm/lib/CodeGen/CalcSpillWeights.cpp
index 6a9dc60584a332..3bc0159fefd02b 100644
--- a/llvm/lib/CodeGen/CalcSpillWeights.cpp
+++ b/llvm/lib/CodeGen/CalcSpillWeights.cpp
@@ -188,6 +188,7 @@ float VirtRegAuxInfo::weightCalcHelper(LiveInterval &LI, SlotIndex *Start,
// Do not update future local split artifacts.
bool ShouldUpdateLI = !IsLocalSplitArtifact;
+ unsigned Factor = TRI.getSpillWeightScaleFactor(MRI.getRegClass(LI.reg()));
if (IsLocalSplitArtifact) {
MachineBasicBlock *LocalMBB = LIS.getMBBFromIndex(*End);
assert(LocalMBB == LIS.getMBBFromIndex(*Start) &&
@@ -198,10 +199,10 @@ float VirtRegAuxInfo::weightCalcHelper(LiveInterval &LI, SlotIndex *Start,
// localLI = COPY other
// ...
// other = COPY localLI
- TotalWeight +=
- LiveIntervals::getSpillWeight(true, false, &MBFI, LocalMBB, PSI);
- TotalWeight +=
- LiveIntervals::getSpillWeight(false, true, &MBFI, LocalMBB, PSI);
+ TotalWeight += LiveIntervals::getSpillWeight(true, false, &MBFI, LocalMBB,
+ Factor, PSI);
+ TotalWeight += LiveIntervals::getSpillWeight(false, true, &MBFI, LocalMBB,
+ Factor, PSI);
NumInstr += 2;
}
@@ -271,7 +272,8 @@ float VirtRegAuxInfo::weightCalcHelper(LiveInterval &LI, SlotIndex *Start,
// Calculate instr weight.
bool Reads, Writes;
std::tie(Reads, Writes) = MI->readsWritesVirtualRegister(LI.reg());
- Weight = LiveIntervals::getSpillWeight(Writes, Reads, &MBFI, *MI, PSI);
+ Weight =
+ LiveIntervals::getSpillWeight(Writes, Reads, &MBFI, *MI, Factor, PSI);
// Give extra weight to what looks like a loop induction variable update.
if (Writes && IsExiting && LIS.isLiveOutOfMBB(LI, MBB))
diff --git a/llvm/lib/CodeGen/LiveIntervals.cpp b/llvm/lib/CodeGen/LiveIntervals.cpp
index f9ee6e4563f8d6..e3ecc46b10366a 100644
--- a/llvm/lib/CodeGen/LiveIntervals.cpp
+++ b/llvm/lib/CodeGen/LiveIntervals.cpp
@@ -886,22 +886,22 @@ LiveIntervals::hasPHIKill(const LiveInterval &LI, const VNInfo *VNI) const {
float LiveIntervals::getSpillWeight(bool isDef, bool isUse,
const MachineBlockFrequencyInfo *MBFI,
- const MachineInstr &MI,
+ const MachineInstr &MI, unsigned Factor,
ProfileSummaryInfo *PSI) {
- return getSpillWeight(isDef, isUse, MBFI, MI.getParent(), PSI);
+ return getSpillWeight(isDef, isUse, MBFI, MI.getParent(), Factor, PSI);
}
float LiveIntervals::getSpillWeight(bool isDef, bool isUse,
const MachineBlockFrequencyInfo *MBFI,
const MachineBasicBlock *MBB,
- ProfileSummaryInfo *PSI) {
+ unsigned Factor, ProfileSummaryInfo *PSI) {
float Weight = isDef + isUse;
const auto *MF = MBB->getParent();
// When optimizing for size we only consider the codesize impact of spilling
// the register, not the runtime impact.
if (PSI && llvm::shouldOptimizeForSize(MF, PSI, MBFI))
return Weight;
- return Weight * MBFI->getBlockFreqRelativeToEntryBlock(MBB);
+ return Weight * MBFI->getBlockFreqRelativeToEntryBlock(MBB) * Factor;
}
LiveRange::Segment
diff --git a/llvm/lib/CodeGen/TargetRegisterInfo.cpp b/llvm/lib/CodeGen/TargetRegisterInfo.cpp
index 032f1a33e75c43..93c59cb134d8c0 100644
--- a/llvm/lib/CodeGen/TargetRegisterInfo.cpp
+++ b/llvm/lib/CodeGen/TargetRegisterInfo.cpp
@@ -414,6 +414,11 @@ bool TargetRegisterInfo::shouldRewriteCopySrc(const TargetRegisterClass *DefRC,
return shareSameRegisterFile(*this, DefRC, DefSubReg, SrcRC, SrcSubReg);
}
+unsigned TargetRegisterInfo::getSpillWeightScaleFactor(
+ const TargetRegisterClass *RC) const {
+ return 1;
+}
+
// Compute target-independent register allocator hints to help eliminate copies.
bool TargetRegisterInfo::getRegAllocationHints(
Register VirtReg, ArrayRef<MCPhysReg> Order,
diff --git a/llvm/lib/Target/RISCV/RISCVRegisterInfo.cpp b/llvm/lib/Target/RISCV/RISCVRegisterInfo.cpp
index cfcc3119257f65..9e02f1ecc60cde 100644
--- a/llvm/lib/Target/RISCV/RISCVRegisterInfo.cpp
+++ b/llvm/lib/Target/RISCV/RISCVRegisterInfo.cpp
@@ -805,6 +805,11 @@ RISCVRegisterInfo::getRegisterCostTableIndex(const MachineFunction &MF) const {
: 0;
}
+unsigned RISCVRegisterInfo::getSpillWeightScaleFactor(
+ const TargetRegisterClass *RC) const {
+ return getRegClassWeight(RC).RegWeight;
+}
+
// Add two address hints to improve chances of being able to use a compressed
// instruction.
bool RISCVRegisterInfo::getRegAllocationHints(
diff --git a/llvm/lib/Target/RISCV/RISCVRegisterInfo.h b/llvm/lib/Target/RISCV/RISCVRegisterInfo.h
index 3ab79694e175c8..9b4317873fec61 100644
--- a/llvm/lib/Target/RISCV/RISCVRegisterInfo.h
+++ b/llvm/lib/Target/RISCV/RISCVRegisterInfo.h
@@ -127,6 +127,9 @@ struct RISCVRegisterInfo : public RISCVGenRegisterInfo {
unsigned getRegisterCostTableIndex(const MachineFunction &MF) const override;
+ unsigned
+ getSpillWeightScaleFactor(const TargetRegisterClass *RC) const override;
+
bool getRegAllocationHints(Register VirtReg, ArrayRef<MCPhysReg> Order,
SmallVectorImpl<MCPhysReg> &Hints,
const MachineFunction &MF, const VirtRegMap *VRM,
diff --git a/llvm/test/CodeGen/RISCV/rvv/abs-vp.ll b/llvm/test/CodeGen/RISCV/rvv/abs-vp.ll
index 163d9145bc3623..d52217d8fe2474 100644
--- a/llvm/test/CodeGen/RISCV/rvv/abs-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/abs-vp.ll
@@ -561,18 +561,7 @@ declare <vscale x 16 x i64> @llvm.vp.abs.nxv16i64(<vscale x 16 x i64>, i1 immarg
define <vscale x 16 x i64> @vp_abs_nxv16i64(<vscale x 16 x i64> %va, <vscale x 16 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_abs_nxv16i64:
; 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: 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: vmv1r.v v7, v0
; CHECK-NEXT: csrr a1, vlenb
; CHECK-NEXT: srli a2, a1, 3
; CHECK-NEXT: sub a3, a0, a1
@@ -582,30 +571,16 @@ define <vscale x 16 x i64> @vp_abs_nxv16i64(<vscale x 16 x i64> %va, <vscale x 1
; CHECK-NEXT: addi a2, a2, -1
; CHECK-NEXT: and a2, a2, a3
; CHECK-NEXT: vsetvli zero, a2, e64, m8, ta, ma
-; CHECK-NEXT: vrsub.vi v8, v16, 0, v0.t
-; CHECK-NEXT: vmax.vv v8, v16, v8, v0.t
-; CHECK-NEXT: addi a2, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
+; CHECK-NEXT: vrsub.vi v24, v16, 0, v0.t
+; CHECK-NEXT: vmax.vv v16, v16, v24, v0.t
; CHECK-NEXT: bltu a0, a1, .LBB46_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB46_2:
-; CHECK-NEXT: vmv1r.v v0, v24
-; 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: vmv1r.v v0, v7
; 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
-; 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: .cfi_def_cfa sp, 16
-; CHECK-NEXT: addi sp, sp, 16
-; CHECK-NEXT: .cfi_def_cfa_offset 0
+; CHECK-NEXT: vrsub.vi v24, v8, 0, v0.t
+; CHECK-NEXT: vmax.vv v8, v8, v24, v0.t
; CHECK-NEXT: ret
%v = call <vscale x 16 x i64> @llvm.vp.abs.nxv16i64(<vscale x 16 x i64> %va, i1 false, <vscale x 16 x i1> %m, i32 %evl)
ret <vscale x 16 x i64> %v
diff --git a/llvm/test/CodeGen/RISCV/rvv/bitreverse-vp.ll b/llvm/test/CodeGen/RISCV/rvv/bitreverse-vp.ll
index 66a1178cddb66c..6b5f5d786b60c2 100644
--- a/llvm/test/CodeGen/RISCV/rvv/bitreverse-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/bitreverse-vp.ll
@@ -2311,35 +2311,37 @@ define <vscale x 7 x i64> @vp_bitreverse_nxv7i64(<vscale x 7 x i64> %va, <vscale
; RV32-NEXT: vsll.vx v24, v24, a4, v0.t
; RV32-NEXT: vor.vv v16, v16, v24, 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, 16
; RV32-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
; RV32-NEXT: vsetvli a3, zero, e64, m8, ta, ma
; RV32-NEXT: vlse64.v v16, (a5), zero
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: slli a3, a3, 3
+; 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: lui a3, 4080
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vx v24, v8, a3, 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: vand.vx v16, v8, a3, v0.t
+; RV32-NEXT: vsll.vi v24, v16, 24, v0.t
; RV32-NEXT: csrr a5, vlenb
; RV32-NEXT: slli a5, a5, 4
; RV32-NEXT: add a5, sp, a5
; RV32-NEXT: addi a5, a5, 16
+; RV32-NEXT: vl8r.v v16, (a5) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v16, v8, v16, v0.t
+; RV32-NEXT: vsll.vi v16, v16, 8, v0.t
+; RV32-NEXT: vor.vv v16, v24, v16, v0.t
+; 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: 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
+; 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
@@ -2353,7 +2355,7 @@ define <vscale x 7 x i64> @vp_bitreverse_nxv7i64(<vscale x 7 x i64> %va, <vscale
; RV32-NEXT: vand.vx v24, v24, a3, v0.t
; RV32-NEXT: vsrl.vi v8, v8, 8, v0.t
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: slli a1, a1, 3
+; RV32-NEXT: slli a1, a1, 4
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
; RV32-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
@@ -2371,7 +2373,7 @@ define <vscale x 7 x i64> @vp_bitreverse_nxv7i64(<vscale x 7 x i64> %va, <vscale
; RV32-NEXT: vsetvli a4, zero, e32, m8, ta, ma
; RV32-NEXT: vmv.v.x v24, a1
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: slli a1, a1, 4
+; RV32-NEXT: slli a1, a1, 3
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
; RV32-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
@@ -2700,35 +2702,37 @@ define <vscale x 8 x i64> @vp_bitreverse_nxv8i64(<vscale x 8 x i64> %va, <vscale
; RV32-NEXT: vsll.vx v24, v24, a4, v0.t
; RV32-NEXT: vor.vv v16, v16, v24, 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, 16
; RV32-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
; RV32-NEXT: vsetvli a3, zero, e64, m8, ta, ma
; RV32-NEXT: vlse64.v v16, (a5), zero
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: slli a3, a3, 3
+; 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: lui a3, 4080
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vx v24, v8, a3, 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: vand.vx v16, v8, a3, v0.t
+; RV32-NEXT: vsll.vi v24, v16, 24, v0.t
; RV32-NEXT: csrr a5, vlenb
; RV32-NEXT: slli a5, a5, 4
; RV32-NEXT: add a5, sp, a5
; RV32-NEXT: addi a5, a5, 16
+; RV32-NEXT: vl8r.v v16, (a5) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v16, v8, v16, v0.t
+; RV32-NEXT: vsll.vi v16, v16, 8, v0.t
+; RV32-NEXT: vor.vv v16, v24, v16, v0.t
+; 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: 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
+; 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
@@ -2742,7 +2746,7 @@ define <vscale x 8 x i64> @vp_bitreverse_nxv8i64(<vscale x 8 x i64> %va, <vscale
; RV32-NEXT: vand.vx v24, v24, a3, v0.t
; RV32-NEXT: vsrl.vi v8, v8, 8, v0.t
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: slli a1, a1, 3
+; RV32-NEXT: slli a1, a1, 4
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
; RV32-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
@@ -2760,7 +2764,7 @@ define <vscale x 8 x i64> @vp_bitreverse_nxv8i64(<vscale x 8 x i64> %va, <vscale
; RV32-NEXT: vsetvli a4, zero, e32, m8, ta, ma
; RV32-NEXT: vmv.v.x v24, a1
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: slli a1, a1, 4
+; RV32-NEXT: slli a1, a1, 3
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
; RV32-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
@@ -3069,18 +3073,7 @@ declare <vscale x 64 x i16> @llvm.vp.bitreverse.nxv64i16(<vscale x 64 x i16>, <v
define <vscale x 64 x i16> @vp_bitreverse_nxv64i16(<vscale x 64 x i16> %va, <vscale x 64 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_bitreverse_nxv64i16:
; 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: 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: vmv1r.v v7, v0
; CHECK-NEXT: csrr a3, vlenb
; CHECK-NEXT: lui a1, 1
; CHECK-NEXT: lui a2, 3
@@ -3097,63 +3090,48 @@ define <vscale x 64 x i16> @vp_bitreverse_nxv64i16(<vscale x 64 x i16> %va, <vsc
; CHECK-NEXT: addi a2, a2, 819
; CHECK-NEXT: addi a1, a6, 1365
; CHECK-NEXT: vsetvli zero, a5, e16, m8, ta, ma
-; CHECK-NEXT: vsrl.vi v8, v16, 8, v0.t
+; CHECK-NEXT: vsrl.vi v24, v16, 8, v0.t
; CHECK-NEXT: vsll.vi v16, v16, 8, v0.t
-; CHECK-NEXT: vor.vv v8, v16, v8, v0.t
-; CHECK-NEXT: vsrl.vi v16, v8, 4, v0.t
+; CHECK-NEXT: vor.vv v16, v16, v24, v0.t
+; CHECK-NEXT: vsrl.vi v24, v16, 4, v0.t
+; CHECK-NEXT: vand.vx v24, v24, a4, v0.t
; CHECK-NEXT: vand.vx v16, v16, a4, v0.t
-; CHECK-NEXT: vand.vx v8, v8, a4, v0.t
-; CHECK-NEXT: vsll.vi v8, v8, 4, v0.t
-; CHECK-NEXT: vor.vv v8, v16, v8, v0.t
-; CHECK-NEXT: vsrl.vi v16, v8, 2, v0.t
+; CHECK-NEXT: vsll.vi v16, v16, 4, v0.t
+; CHECK-NEXT: vor.vv v16, v24, v16, v0.t
+; CHECK-NEXT: vsrl.vi v24, v16, 2, v0.t
+; CHECK-NEXT: vand.vx v24, v24, a2, v0.t
; CHECK-NEXT: vand.vx v16, v16, a2, v0.t
-; CHECK-NEXT: vand.vx v8, v8, a2, v0.t
-; CHECK-NEXT: vsll.vi v8, v8, 2, v0.t
-; CHECK-NEXT: vor.vv v8, v16, v8, v0.t
-; CHECK-NEXT: vsrl.vi v16, v8, 1, v0.t
+; CHECK-NEXT: vsll.vi v16, v16, 2, v0.t
+; CHECK-NEXT: vor.vv v16, v24, v16, v0.t
+; CHECK-NEXT: vsrl.vi v24, v16, 1, v0.t
+; CHECK-NEXT: vand.vx v24, v24, a1, v0.t
; CHECK-NEXT: vand.vx v16, v16, a1, v0.t
-; CHECK-NEXT: vand.vx v8, v8, a1, v0.t
-; CHECK-NEXT: vsll.vi v8, v8, 1, v0.t
-; CHECK-NEXT: vor.vv v8, v16, v8, v0.t
-; CHECK-NEXT: addi a5, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a5) # Unknown-size Folded Spill
+; CHECK-NEXT: vsll.vi v16, v16, 1, v0.t
+; CHECK-NEXT: vor.vv v16, v24, v16, v0.t
; CHECK-NEXT: bltu a0, a3, .LBB46_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a3
; CHECK-NEXT: .LBB46_2:
-; CHECK-NEXT: vmv1r.v v0, v24
-; CHECK-NEXT: csrr a3, vlenb
-; CHECK-NEXT: slli a3, a3, 3
-; CHECK-NEXT: add a3, sp, a3
-; CHECK-NEXT: addi a3, a3, 16
-; CHECK-NEXT: vl8r.v v8, (a3) # Unknown-size Folded Reload
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, a0, e16, m8, ta, ma
-; CHECK-NEXT: vsrl.vi v16, v8, 8, v0.t
+; CHECK-NEXT: vsrl.vi v24, v8, 8, v0.t
; CHECK-NEXT: vsll.vi v8, v8, 8, v0.t
-; CHECK-NEXT: vor.vv v8, v8, v16, v0.t
-; CHECK-NEXT: vsrl.vi v16, v8, 4, v0.t
-; CHECK-NEXT: vand.vx v16, v16, a4, v0.t
+; CHECK-NEXT: vor.vv v8, v8, v24, v0.t
+; CHECK-NEXT: vsrl.vi v24, v8, 4, v0.t
+; CHECK-NEXT: vand.vx v24, v24, a4, v0.t
; CHECK-NEXT: vand.vx v8, v8, a4, v0.t
; CHECK-NEXT: vsll.vi v8, v8, 4, v0.t
-; CHECK-NEXT: vor.vv v8, v16, v8, v0.t
-; CHECK-NEXT: vsrl.vi v16, v8, 2, v0.t
-; CHECK-NEXT: vand.vx v16, v16, a2, v0.t
+; CHECK-NEXT: vor.vv v8, v24, v8, v0.t
+; CHECK-NEXT: vsrl.vi v24, v8, 2, v0.t
+; CHECK-NEXT: vand.vx v24, v24, a2, v0.t
; CHECK-NEXT: vand.vx v8, v8, a2, v0.t
; CHECK-NEXT: vsll.vi v8, v8, 2, v0.t
-; CHECK-NEXT: vor.vv v8, v16, v8, v0.t
-; CHECK-NEXT: vsrl.vi v16, v8, 1, v0.t
-; CHECK-NEXT: vand.vx v16, v16, a1, v0.t
+; CHECK-NEXT: vor.vv v8, v24, v8, v0.t
+; CHECK-NEXT: vsrl.vi v24, v8, 1, v0.t
+; CHECK-NEXT: vand.vx v24, v24, a1, v0.t
; CHECK-NEXT: vand.vx v8, v8, a1, v0.t
; CHECK-NEXT: vsll.vi v8, v8, 1, v0.t
-; CHECK-NEXT: vor.vv v8, v16, v8, v0.t
-; CHECK-NEXT: addi a0, sp, 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: .cfi_def_cfa sp, 16
-; CHECK-NEXT: addi sp, sp, 16
-; CHECK-NEXT: .cfi_def_cfa_offset 0
+; CHECK-NEXT: vor.vv v8, v24, v8, v0.t
; CHECK-NEXT: ret
;
; CHECK-ZVBB-LABEL: vp_bitreverse_nxv64i16:
diff --git a/llvm/test/CodeGen/RISCV/rvv/bswap-vp.ll b/llvm/test/CodeGen/RISCV/rvv/bswap-vp.ll
index 1c95ec8fafd4f1..ca783060cfe641 100644
--- a/llvm/test/CodeGen/RISCV/rvv/bswap-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/bswap-vp.ll
@@ -1048,35 +1048,37 @@ define <vscale x 7 x i64> @vp_bswap_nxv7i64(<vscale x 7 x i64> %va, <vscale x 7
; RV32-NEXT: vsll.vx v24, v24, a4, v0.t
; RV32-NEXT: vor.vv v16, v16, v24, 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, 16
; RV32-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
; RV32-NEXT: vsetvli a3, zero, e64, m8, ta, ma
; RV32-NEXT: vlse64.v v16, (a5), zero
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: slli a3, a3, 3
+; 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: lui a3, 4080
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vx v24, v8, a3, 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: vand.vx v16, v8, a3, v0.t
+; RV32-NEXT: vsll.vi v24, v16, 24, v0.t
; 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 v16, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v16, v8, v16, v0.t
+; RV32-NEXT: vsll.vi v16, v16, 8, v0.t
+; RV32-NEXT: vor.vv v16, v24, 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, 16
; 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
+; RV32-NEXT: slli a0, a0, 3
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 16
; RV32-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
@@ -1090,7 +1092,7 @@ define <vscale x 7 x i64> @vp_bswap_nxv7i64(<vscale x 7 x i64> %va, <vscale x 7
; RV32-NEXT: vand.vx v24, v24, a3, v0.t
; RV32-NEXT: vsrl.vi v8, v8, 8, v0.t
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 3
+; RV32-NEXT: slli a0, a0, 4
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 16
; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
@@ -1100,7 +1102,7 @@ define <vscale x 7 x i64> @vp_bswap_nxv7i64(<vscale x 7 x i64> %va, <vscale x 7
; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; RV32-NEXT: vor.vv v8, v8, v16, 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, 16
; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
@@ -1323,35 +1325,37 @@ define <vscale x 8 x i64> @vp_bswap_nxv8i64(<vscale x 8 x i64> %va, <vscale x 8
; RV32-NEXT: vsll.vx v24, v24, a4, v0.t
; RV32-NEXT: vor.vv v16, v16, v24, 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, 16
; RV32-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
; RV32-NEXT: vsetvli a3, zero, e64, m8, ta, ma
; RV32-NEXT: vlse64.v v16, (a5), zero
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: slli a3, a3, 3
+; 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: lui a3, 4080
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vx v24, v8, a3, 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: vand.vx v16, v8, a3, v0.t
+; RV32-NEXT: vsll.vi v24, v16, 24, v0.t
; 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 v16, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v16, v8, v16, v0.t
+; RV32-NEXT: vsll.vi v16, v16, 8, v0.t
+; RV32-NEXT: vor.vv v16, v24, 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, 16
; 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
+; RV32-NEXT: slli a0, a0, 3
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 16
; RV32-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
@@ -1365,7 +1369,7 @@ define <vscale x 8 x i64> @vp_bswap_nxv8i64(<vscale x 8 x i64> %va, <vscale x 8
; RV32-NEXT: vand.vx v24, v24, a3, v0.t
; RV32-NEXT: vsrl.vi v8, v8, 8, v0.t
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 3
+; RV32-NEXT: slli a0, a0, 4
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 16
; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
@@ -1375,7 +1379,7 @@ define <vscale x 8 x i64> @vp_bswap_nxv8i64(<vscale x 8 x i64> %va, <vscale x 8
; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; RV32-NEXT: vor.vv v8, v8, v16, 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, 16
; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
@@ -1578,18 +1582,7 @@ declare <vscale x 64 x i16> @llvm.vp.bswap.nxv64i16(<vscale x 64 x i16>, <vscale
define <vscale x 64 x i16> @vp_bswap_nxv64i16(<vscale x 64 x i16> %va, <vscale x 64 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_bswap_nxv64i16:
; 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: 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: vmv1r.v v7, v0
; CHECK-NEXT: csrr a1, vlenb
; CHECK-NEXT: srli a2, a1, 1
; CHECK-NEXT: slli a1, a1, 2
@@ -1600,33 +1593,18 @@ define <vscale x 64 x i16> @vp_bswap_nxv64i16(<vscale x 64 x i16> %va, <vscale x
; CHECK-NEXT: addi a3, a3, -1
; CHECK-NEXT: and a2, a3, a2
; CHECK-NEXT: vsetvli zero, a2, e16, m8, ta, ma
-; CHECK-NEXT: vsrl.vi v8, v16, 8, v0.t
+; CHECK-NEXT: vsrl.vi v24, v16, 8, v0.t
; CHECK-NEXT: vsll.vi v16, v16, 8, v0.t
-; CHECK-NEXT: vor.vv v8, v16, v8, v0.t
-; CHECK-NEXT: addi a2, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
+; CHECK-NEXT: vor.vv v16, v16, v24, 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, v24
-; 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: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, a0, e16, m8, ta, ma
-; CHECK-NEXT: vsrl.vi v16, v8, 8, v0.t
+; CHECK-NEXT: vsrl.vi v24, v8, 8, v0.t
; CHECK-NEXT: vsll.vi v8, v8, 8, v0.t
-; CHECK-NEXT: vor.vv v8, v8, v16, v0.t
-; CHECK-NEXT: addi a0, sp, 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: .cfi_def_cfa sp, 16
-; CHECK-NEXT: addi sp, sp, 16
-; CHECK-NEXT: .cfi_def_cfa_offset 0
+; CHECK-NEXT: vor.vv v8, v8, v24, v0.t
; CHECK-NEXT: ret
;
; CHECK-ZVKB-LABEL: vp_bswap_nxv64i16:
diff --git a/llvm/test/CodeGen/RISCV/rvv/ceil-vp.ll b/llvm/test/CodeGen/RISCV/rvv/ceil-vp.ll
index 7d0b0118a72725..2ad48c3a70100c 100644
--- a/llvm/test/CodeGen/RISCV/rvv/ceil-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/ceil-vp.ll
@@ -273,12 +273,6 @@ declare <vscale x 32 x bfloat> @llvm.vp.ceil.nxv32bf16(<vscale x 32 x bfloat>, <
define <vscale x 32 x bfloat> @vp_ceil_vv_nxv32bf16(<vscale x 32 x bfloat> %va, <vscale x 32 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_ceil_vv_nxv32bf16:
; 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 v7, v0
; CHECK-NEXT: csrr a2, vlenb
; CHECK-NEXT: vsetvli a1, zero, e16, m4, ta, ma
@@ -302,11 +296,7 @@ define <vscale x 32 x bfloat> @vp_ceil_vv_nxv32bf16(<vscale x 32 x bfloat> %va,
; CHECK-NEXT: vmv1r.v v0, v12
; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v16, v24, v0.t
-; CHECK-NEXT: addi a3, sp, 16
-; CHECK-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
; CHECK-NEXT: fsrm a2
-; CHECK-NEXT: addi a2, sp, 16
-; CHECK-NEXT: vl8r.v v16, (a2) # Unknown-size Folded Reload
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, mu
; CHECK-NEXT: vfsgnj.vv v24, v16, v24, v0.t
@@ -332,12 +322,6 @@ define <vscale x 32 x bfloat> @vp_ceil_vv_nxv32bf16(<vscale x 32 x bfloat> %va,
; CHECK-NEXT: vfsgnj.vv v24, v16, v24, v0.t
; CHECK-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; CHECK-NEXT: vfncvtbf16.f.f.w v8, v24
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add sp, sp, a0
-; CHECK-NEXT: .cfi_def_cfa sp, 16
-; CHECK-NEXT: addi sp, sp, 16
-; CHECK-NEXT: .cfi_def_cfa_offset 0
; CHECK-NEXT: ret
%v = call <vscale x 32 x bfloat> @llvm.vp.ceil.nxv32bf16(<vscale x 32 x bfloat> %va, <vscale x 32 x i1> %m, i32 %evl)
ret <vscale x 32 x bfloat> %v
@@ -840,12 +824,6 @@ define <vscale x 32 x half> @vp_ceil_vv_nxv32f16(<vscale x 32 x half> %va, <vsca
;
; ZVFHMIN-LABEL: vp_ceil_vv_nxv32f16:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: addi sp, sp, -16
-; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
-; ZVFHMIN-NEXT: csrr a1, vlenb
-; 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 v7, v0
; ZVFHMIN-NEXT: csrr a2, vlenb
; ZVFHMIN-NEXT: vsetvli a1, zero, e16, m4, ta, ma
@@ -869,11 +847,7 @@ define <vscale x 32 x half> @vp_ceil_vv_nxv32f16(<vscale x 32 x half> %va, <vsca
; ZVFHMIN-NEXT: vmv1r.v v0, v12
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfcvt.x.f.v v16, v24, v0.t
-; ZVFHMIN-NEXT: addi a3, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: fsrm a2
-; ZVFHMIN-NEXT: addi a2, sp, 16
-; ZVFHMIN-NEXT: vl8r.v v16, (a2) # 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, v16, v24, v0.t
@@ -899,12 +873,6 @@ define <vscale x 32 x half> @vp_ceil_vv_nxv32f16(<vscale x 32 x half> %va, <vsca
; ZVFHMIN-NEXT: vfsgnj.vv v24, v16, v24, v0.t
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; 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
-; ZVFHMIN-NEXT: .cfi_def_cfa sp, 16
-; ZVFHMIN-NEXT: addi sp, sp, 16
-; ZVFHMIN-NEXT: .cfi_def_cfa_offset 0
; ZVFHMIN-NEXT: ret
%v = call <vscale x 32 x half> @llvm.vp.ceil.nxv32f16(<vscale x 32 x half> %va, <vscale x 32 x i1> %m, i32 %evl)
ret <vscale x 32 x half> %v
@@ -1419,12 +1387,6 @@ declare <vscale x 16 x double> @llvm.vp.ceil.nxv16f64(<vscale x 16 x double>, <v
define <vscale x 16 x double> @vp_ceil_vv_nxv16f64(<vscale x 16 x double> %va, <vscale x 16 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_ceil_vv_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, 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: csrr a1, vlenb
; CHECK-NEXT: lui a2, %hi(.LCPI44_0)
@@ -1445,40 +1407,27 @@ define <vscale x 16 x double> @vp_ceil_vv_nxv16f64(<vscale x 16 x double> %va, <
; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; 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: 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, v24, v16, v0.t
-; CHECK-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
; CHECK-NEXT: bltu a0, a1, .LBB44_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB44_2:
; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; 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 v7, 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: 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: addi a0, sp, 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 sp, sp, a0
-; CHECK-NEXT: .cfi_def_cfa sp, 16
-; CHECK-NEXT: addi sp, sp, 16
-; CHECK-NEXT: .cfi_def_cfa_offset 0
+; CHECK-NEXT: vfsgnj.vv v8, v24, v8, v0.t
; CHECK-NEXT: ret
%v = call <vscale x 16 x double> @llvm.vp.ceil.nxv16f64(<vscale x 16 x double> %va, <vscale x 16 x i1> %m, i32 %evl)
ret <vscale x 16 x double> %v
diff --git a/llvm/test/CodeGen/RISCV/rvv/compressstore.ll b/llvm/test/CodeGen/RISCV/rvv/compressstore.ll
index a407cd048ffe3f..49e54032c1de61 100644
--- a/llvm/test/CodeGen/RISCV/rvv/compressstore.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/compressstore.ll
@@ -224,55 +224,39 @@ define void @test_compresstore_v256i8(ptr %p, <256 x i1> %mask, <256 x i8> %data
;
; RV32-LABEL: test_compresstore_v256i8:
; RV32: # %bb.0: # %entry
-; RV32-NEXT: addi sp, sp, -16
-; RV32-NEXT: .cfi_def_cfa_offset 16
-; RV32-NEXT: csrr a2, vlenb
-; RV32-NEXT: slli a2, a2, 3
-; RV32-NEXT: sub sp, sp, a2
-; RV32-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
-; RV32-NEXT: vmv8r.v v24, v16
+; RV32-NEXT: vmv1r.v v7, v8
; RV32-NEXT: li a2, 128
; RV32-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; RV32-NEXT: vslidedown.vi v9, v0, 1
; RV32-NEXT: li a3, 32
; RV32-NEXT: vmv.x.s a4, v0
; RV32-NEXT: vsetvli zero, a2, e8, m8, ta, ma
-; RV32-NEXT: vle8.v v16, (a1)
-; RV32-NEXT: addi a1, sp, 16
-; RV32-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: vle8.v v24, (a1)
; RV32-NEXT: vsetivli zero, 1, e64, m1, ta, ma
-; RV32-NEXT: vsrl.vx v10, v9, a3
+; RV32-NEXT: vsrl.vx v6, v9, a3
; RV32-NEXT: vmv.x.s a1, v9
-; RV32-NEXT: vsrl.vx v9, v0, a3
+; RV32-NEXT: vsrl.vx v5, v0, a3
; RV32-NEXT: vsetvli zero, a2, e8, m8, ta, ma
-; RV32-NEXT: vcompress.vm v16, v24, v0
+; RV32-NEXT: vcompress.vm v8, v16, v0
; RV32-NEXT: vcpop.m a3, v0
; RV32-NEXT: cpop a4, a4
; RV32-NEXT: vsetvli zero, a2, e64, m1, ta, ma
-; RV32-NEXT: vmv.x.s a5, v10
-; RV32-NEXT: vmv.x.s a6, v9
+; RV32-NEXT: vmv.x.s a5, v6
+; RV32-NEXT: vmv.x.s a6, v5
; RV32-NEXT: vsetvli zero, a3, e8, m8, ta, ma
-; RV32-NEXT: vse8.v v16, (a0)
+; RV32-NEXT: vse8.v v8, (a0)
; RV32-NEXT: cpop a1, a1
; RV32-NEXT: cpop a3, a6
; RV32-NEXT: cpop a5, a5
; RV32-NEXT: add a3, a4, a3
; RV32-NEXT: add a1, a1, a5
; RV32-NEXT: add a1, a3, a1
-; RV32-NEXT: addi a3, sp, 16
-; RV32-NEXT: vl8r.v v24, (a3) # Unknown-size Folded Reload
; RV32-NEXT: vsetvli zero, a2, e8, m8, ta, ma
-; RV32-NEXT: vcompress.vm v16, v24, v8
+; RV32-NEXT: vcompress.vm v8, v24, v7
; RV32-NEXT: add a0, a0, a1
-; RV32-NEXT: vcpop.m a1, v8
+; RV32-NEXT: vcpop.m a1, v7
; RV32-NEXT: vsetvli zero, a1, e8, m8, ta, ma
-; RV32-NEXT: vse8.v v16, (a0)
-; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 3
-; RV32-NEXT: add sp, sp, a0
-; RV32-NEXT: .cfi_def_cfa sp, 16
-; RV32-NEXT: addi sp, sp, 16
-; RV32-NEXT: .cfi_def_cfa_offset 0
+; RV32-NEXT: vse8.v v8, (a0)
; RV32-NEXT: ret
entry:
tail call void @llvm.masked.compressstore.v256i8(<256 x i8> %data, ptr align 1 %p, <256 x i1> %mask)
diff --git a/llvm/test/CodeGen/RISCV/rvv/ctpop-vp.ll b/llvm/test/CodeGen/RISCV/rvv/ctpop-vp.ll
index 9e75dc9dccffde..d3929e936d1542 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, 0x30, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 48 * vlenb
; RV32-NEXT: vmv1r.v v7, v0
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a2, 24
-; 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 v8, (a1) # Unknown-size Folded Spill
@@ -2038,88 +2037,60 @@ define <vscale x 16 x i64> @vp_ctpop_nxv16i64(<vscale x 16 x i64> %va, <vscale x
; RV32-NEXT: addi a2, a2, 1365
; RV32-NEXT: vsetvli a4, zero, e32, m8, ta, ma
; RV32-NEXT: vmv.v.x v8, a2
-; RV32-NEXT: csrr a2, vlenb
-; RV32-NEXT: slli a2, a2, 5
-; RV32-NEXT: add a2, sp, a2
-; RV32-NEXT: addi a2, a2, 16
-; RV32-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
; RV32-NEXT: sltu a2, a0, a3
; RV32-NEXT: addi a2, a2, -1
; RV32-NEXT: and a2, a2, a3
; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
-; RV32-NEXT: vsrl.vi v8, v16, 1, v0.t
+; RV32-NEXT: vsrl.vi v24, 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: 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 v8, (a3) # Unknown-size Folded Reload
-; 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 v24, (a3) # Unknown-size Folded Reload
-; RV32-NEXT: vand.vv v8, v24, v8, v0.t
-; RV32-NEXT: vsub.vv v16, v16, v8, v0.t
+; RV32-NEXT: vand.vv v24, v24, v8, v0.t
+; RV32-NEXT: vsub.vv v16, v16, v24, v0.t
; 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: 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: 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, v16, v8, v0.t
+; RV32-NEXT: vand.vv v24, v16, v8, v0.t
; RV32-NEXT: vsrl.vi v16, v16, 2, 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, 16
-; RV32-NEXT: vl8r.v v24, (a3) # Unknown-size Folded Reload
-; RV32-NEXT: vand.vv v16, v16, v24, v0.t
-; 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: vs8r.v v8, (a3) # Unknown-size Folded Spill
+; RV32-NEXT: vand.vv v16, v16, v8, v0.t
+; RV32-NEXT: vadd.vv v16, v24, v16, v0.t
+; RV32-NEXT: vsrl.vi v24, v16, 4, v0.t
+; RV32-NEXT: vadd.vv v16, v16, v24, 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 v16, a3
+; 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 v16, (a3) # Unknown-size Folded Spill
+; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v8, v8, v16, v0.t
+; RV32-NEXT: vand.vv v16, v16, v8, 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 v16, a3
+; RV32-NEXT: vmv.v.x v8, a3
; 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 v16, (a3) # Unknown-size Folded Spill
+; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
-; RV32-NEXT: vmul.vv v8, v8, v16, 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: addi a3, sp, 16
; RV32-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
; RV32-NEXT: bltu a0, a1, .LBB46_2
@@ -2127,60 +2098,32 @@ define <vscale x 16 x i64> @vp_ctpop_nxv16i64(<vscale x 16 x i64> %va, <vscale x
; RV32-NEXT: mv a0, a1
; RV32-NEXT: .LBB46_2:
; RV32-NEXT: vmv1r.v v0, v7
-; RV32-NEXT: li a3, 24
-; RV32-NEXT: mul a1, a1, a3
+; RV32-NEXT: slli a1, a1, 5
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
-; RV32-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vl8r.v v8, (a1) # 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
-; 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 v8, v24, v8, v0.t
-; RV32-NEXT: vsub.vv v8, v16, v8, v0.t
-; 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: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; RV32-NEXT: vsrl.vi v16, v8, 1, 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 v8, (a0) # Unknown-size Folded Reload
-; 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: 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: vs8r.v v16, (a0) # Unknown-size Folded Spill
-; RV32-NEXT: vmv8r.v v16, v8
-; 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 v8, (a0) # Unknown-size Folded Reload
-; RV32-NEXT: vsrl.vi v8, v8, 2, v0.t
-; RV32-NEXT: vand.vv v8, v8, v16, v0.t
+; RV32-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v24, v16, v24, v0.t
+; RV32-NEXT: vsub.vv v24, v8, v24, 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: vsrl.vi v16, v8, 4, v0.t
-; RV32-NEXT: vadd.vv v8, v8, v16, v0.t
+; RV32-NEXT: vand.vv v8, v24, v16, v0.t
+; RV32-NEXT: vsrl.vi v24, v24, 2, v0.t
+; RV32-NEXT: vand.vv v24, v24, v16, v0.t
+; RV32-NEXT: vadd.vv v8, v8, v24, v0.t
+; RV32-NEXT: vsrl.vi v24, v8, 4, v0.t
+; RV32-NEXT: vadd.vv v8, v8, v24, v0.t
; RV32-NEXT: csrr a0, vlenb
; RV32-NEXT: slli a0, a0, 4
; RV32-NEXT: add a0, sp, a0
@@ -2207,28 +2150,17 @@ define <vscale x 16 x i64> @vp_ctpop_nxv16i64(<vscale x 16 x i64> %va, <vscale x
;
; RV64-LABEL: vp_ctpop_nxv16i64:
; RV64: # %bb.0:
-; RV64-NEXT: addi sp, sp, -16
-; RV64-NEXT: .cfi_def_cfa_offset 16
-; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: slli a1, a1, 4
-; RV64-NEXT: sub sp, sp, a1
-; 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 a1, vlenb
-; RV64-NEXT: slli a1, a1, 3
-; RV64-NEXT: add a1, sp, a1
-; RV64-NEXT: addi a1, a1, 16
-; RV64-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
; RV64-NEXT: csrr a1, vlenb
; RV64-NEXT: srli a2, a1, 3
; RV64-NEXT: vsetvli a3, zero, e8, mf4, ta, ma
-; RV64-NEXT: vslidedown.vx v24, v0, a2
+; RV64-NEXT: vslidedown.vx v7, v0, a2
; RV64-NEXT: mv a2, a0
; RV64-NEXT: bltu a0, a1, .LBB46_2
; RV64-NEXT: # %bb.1:
; RV64-NEXT: mv a2, a1
; RV64-NEXT: .LBB46_2:
; RV64-NEXT: vsetvli zero, a2, e64, m8, ta, ma
-; RV64-NEXT: vsrl.vi v16, v8, 1, v0.t
+; RV64-NEXT: vsrl.vi v24, v8, 1, v0.t
; RV64-NEXT: lui a2, 349525
; RV64-NEXT: lui a3, 209715
; RV64-NEXT: lui a4, 61681
@@ -2238,58 +2170,43 @@ define <vscale x 16 x i64> @vp_ctpop_nxv16i64(<vscale x 16 x i64> %va, <vscale x
; RV64-NEXT: addiw a4, a4, -241
; RV64-NEXT: addiw a5, a5, 257
; RV64-NEXT: slli a6, a2, 32
-; RV64-NEXT: add a6, a2, a6
-; RV64-NEXT: slli a2, a3, 32
-; RV64-NEXT: add a7, a3, a2
-; RV64-NEXT: slli a2, a4, 32
-; RV64-NEXT: add a2, a4, a2
-; RV64-NEXT: slli a3, a5, 32
-; RV64-NEXT: add a3, a5, a3
-; RV64-NEXT: li a4, 56
+; RV64-NEXT: add a2, a2, a6
+; RV64-NEXT: slli a6, a3, 32
+; RV64-NEXT: add a3, a3, a6
+; RV64-NEXT: slli a6, a4, 32
+; RV64-NEXT: add a4, a4, a6
+; RV64-NEXT: slli a6, a5, 32
+; RV64-NEXT: add a5, a5, a6
+; RV64-NEXT: li a6, 56
; RV64-NEXT: sub a1, a0, a1
; RV64-NEXT: sltu a0, a0, a1
; RV64-NEXT: addi a0, a0, -1
; RV64-NEXT: and a0, a0, a1
-; RV64-NEXT: vand.vx v16, v16, a6, v0.t
-; RV64-NEXT: vsub.vv v8, v8, v16, v0.t
-; RV64-NEXT: vand.vx v16, v8, a7, v0.t
+; RV64-NEXT: vand.vx v24, v24, a2, v0.t
+; RV64-NEXT: vsub.vv v8, v8, v24, v0.t
+; RV64-NEXT: vand.vx v24, v8, a3, v0.t
; RV64-NEXT: vsrl.vi v8, v8, 2, v0.t
-; RV64-NEXT: vand.vx v8, v8, a7, v0.t
-; RV64-NEXT: vadd.vv v8, v16, v8, v0.t
-; RV64-NEXT: vsrl.vi v16, v8, 4, v0.t
-; RV64-NEXT: vadd.vv v8, v8, v16, v0.t
-; RV64-NEXT: vand.vx v8, v8, a2, v0.t
-; RV64-NEXT: vmul.vx v8, v8, a3, v0.t
-; RV64-NEXT: vsrl.vx v8, v8, a4, v0.t
-; RV64-NEXT: addi a1, sp, 16
-; RV64-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
-; RV64-NEXT: vmv1r.v v0, v24
-; 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: vand.vx v8, v8, a3, v0.t
+; RV64-NEXT: vadd.vv v8, v24, v8, v0.t
+; RV64-NEXT: vsrl.vi v24, v8, 4, v0.t
+; RV64-NEXT: vadd.vv v8, v8, v24, v0.t
+; RV64-NEXT: vand.vx v8, v8, a4, v0.t
+; RV64-NEXT: vmul.vx v8, v8, a5, v0.t
+; RV64-NEXT: vsrl.vx v8, v8, a6, v0.t
+; RV64-NEXT: vmv1r.v v0, v7
; RV64-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV64-NEXT: vsrl.vi v16, v8, 1, v0.t
-; RV64-NEXT: vand.vx v16, v16, a6, v0.t
-; RV64-NEXT: vsub.vv v16, v8, v16, v0.t
-; RV64-NEXT: vand.vx v8, v16, a7, v0.t
+; RV64-NEXT: vsrl.vi v24, v16, 1, v0.t
+; RV64-NEXT: vand.vx v24, v24, a2, v0.t
+; RV64-NEXT: vsub.vv v16, v16, v24, v0.t
+; RV64-NEXT: vand.vx v24, v16, a3, v0.t
; RV64-NEXT: vsrl.vi v16, v16, 2, v0.t
-; RV64-NEXT: vand.vx v16, v16, a7, v0.t
-; RV64-NEXT: vadd.vv v8, v8, v16, v0.t
-; RV64-NEXT: vsrl.vi v16, v8, 4, v0.t
-; RV64-NEXT: vadd.vv v8, v8, v16, v0.t
-; RV64-NEXT: vand.vx v8, v8, a2, v0.t
-; RV64-NEXT: vmul.vx v8, v8, a3, v0.t
-; RV64-NEXT: vsrl.vx v16, v8, a4, v0.t
-; RV64-NEXT: addi a0, sp, 16
-; RV64-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; RV64-NEXT: csrr a0, vlenb
-; RV64-NEXT: slli a0, a0, 4
-; RV64-NEXT: add sp, sp, a0
-; RV64-NEXT: .cfi_def_cfa sp, 16
-; RV64-NEXT: addi sp, sp, 16
-; RV64-NEXT: .cfi_def_cfa_offset 0
+; RV64-NEXT: vand.vx v16, v16, a3, v0.t
+; RV64-NEXT: vadd.vv v16, v24, v16, v0.t
+; RV64-NEXT: vsrl.vi v24, v16, 4, v0.t
+; RV64-NEXT: vadd.vv v16, v16, v24, v0.t
+; RV64-NEXT: vand.vx v16, v16, a4, v0.t
+; RV64-NEXT: vmul.vx v16, v16, a5, v0.t
+; RV64-NEXT: vsrl.vx v16, v16, a6, v0.t
; RV64-NEXT: ret
;
; CHECK-ZVBB-LABEL: vp_ctpop_nxv16i64:
diff --git a/llvm/test/CodeGen/RISCV/rvv/cttz-vp.ll b/llvm/test/CodeGen/RISCV/rvv/cttz-vp.ll
index 9e6295b6644171..5d2327f60534c6 100644
--- a/llvm/test/CodeGen/RISCV/rvv/cttz-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/cttz-vp.ll
@@ -2242,11 +2242,11 @@ define <vscale x 16 x i64> @vp_cttz_nxv16i64(<vscale x 16 x i64> %va, <vscale x
; RV32-NEXT: addi sp, sp, -16
; RV32-NEXT: .cfi_def_cfa_offset 16
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a2, 56
+; RV32-NEXT: li a2, 48
; RV32-NEXT: mul a1, a1, a2
; RV32-NEXT: sub sp, sp, a1
-; 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: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x30, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 48 * vlenb
+; RV32-NEXT: vmv1r.v v7, v0
; RV32-NEXT: csrr a1, vlenb
; RV32-NEXT: slli a1, a1, 5
; RV32-NEXT: add a1, sp, a1
@@ -2264,226 +2264,115 @@ define <vscale x 16 x i64> @vp_cttz_nxv16i64(<vscale x 16 x i64> %va, <vscale x
; RV32-NEXT: lui a4, 349525
; RV32-NEXT: addi a4, a4, 1365
; RV32-NEXT: vsetvli zero, a3, e64, m8, ta, ma
-; RV32-NEXT: vsub.vx v8, v16, a2, v0.t
+; RV32-NEXT: vsub.vx v24, v16, a2, v0.t
; RV32-NEXT: vnot.v v16, v16, v0.t
-; RV32-NEXT: vand.vv v8, v16, v8, v0.t
-; RV32-NEXT: csrr a5, vlenb
-; RV32-NEXT: li a6, 48
-; RV32-NEXT: mul a5, a5, a6
-; RV32-NEXT: add a5, sp, a5
-; RV32-NEXT: addi a5, a5, 16
-; RV32-NEXT: vs8r.v v8, (a5) # Unknown-size Folded Spill
+; RV32-NEXT: vand.vv v16, v16, v24, v0.t
; RV32-NEXT: vsetvli a5, zero, e32, 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: vs8r.v v16, (a4) # Unknown-size Folded Spill
-; 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: vl8r.v v8, (a4) # Unknown-size Folded Reload
+; RV32-NEXT: vmv.v.x v8, a4
; RV32-NEXT: vsetvli zero, a3, e64, m8, ta, ma
-; RV32-NEXT: vsrl.vi v8, v8, 1, v0.t
-; 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 v8, (a4) # Unknown-size Folded Spill
+; RV32-NEXT: vsrl.vi v24, v16, 1, v0.t
; 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: 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: vl8r.v v8, (a4) # Unknown-size Folded Reload
-; RV32-NEXT: vand.vv v16, v8, v16, 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: vl8r.v v8, (a4) # Unknown-size Folded Reload
-; RV32-NEXT: vsub.vv v16, v8, v16, v0.t
+; RV32-NEXT: vs8r.v v8, (a4) # Unknown-size Folded Spill
+; RV32-NEXT: vand.vv v24, v24, v8, v0.t
+; RV32-NEXT: vsub.vv v16, v16, v24, v0.t
; 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: 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: 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: vl8r.v v8, (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: vand.vv v24, v16, v8, v0.t
+; RV32-NEXT: vsrl.vi v16, v16, 2, v0.t
; 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 v8, (a4) # Unknown-size Folded Spill
-; RV32-NEXT: vsrl.vi v16, v16, 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: vl8r.v v8, (a4) # Unknown-size Folded Reload
; RV32-NEXT: vand.vv v16, v16, v8, v0.t
-; 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: vl8r.v v8, (a4) # 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: vadd.vv v16, v24, v16, v0.t
+; RV32-NEXT: vsrl.vi v24, v16, 4, v0.t
+; RV32-NEXT: vadd.vv v16, v16, v24, 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 v16, a4
+; 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: 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: vsetvli zero, a3, e64, m8, ta, ma
-; RV32-NEXT: vand.vv v8, v8, v16, v0.t
+; RV32-NEXT: vand.vv v16, v16, v8, 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 v16, a4
+; RV32-NEXT: vmv.v.x v8, a4
; RV32-NEXT: csrr a4, vlenb
-; RV32-NEXT: slli a4, a4, 4
+; 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: vs8r.v v8, (a4) # Unknown-size Folded Spill
; RV32-NEXT: vsetvli zero, a3, e64, m8, ta, ma
-; RV32-NEXT: vmul.vv v8, v8, v16, 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: csrr a4, vlenb
-; RV32-NEXT: slli a4, a4, 3
-; RV32-NEXT: add a4, sp, a4
-; RV32-NEXT: addi a4, a4, 16
+; 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: vmv1r.v v0, v24
+; RV32-NEXT: vmv1r.v v0, v7
; 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: vsub.vx v24, v8, a2, v0.t
; RV32-NEXT: vnot.v v8, v8, v0.t
-; RV32-NEXT: vand.vv v8, v8, v16, v0.t
-; RV32-NEXT: addi a0, sp, 16
-; 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: add a0, sp, a0
-; RV32-NEXT: addi a0, a0, 16
-; RV32-NEXT: vs8r.v v8, (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, 16
-; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; 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 v8, (a0) # Unknown-size Folded Reload
-; RV32-NEXT: vand.vv v16, v8, v16, v0.t
-; RV32-NEXT: addi a0, sp, 16
-; RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; RV32-NEXT: vsub.vv v8, v8, v16, v0.t
+; RV32-NEXT: vand.vv v8, v8, v24, v0.t
+; RV32-NEXT: vsrl.vi v16, v8, 1, 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: 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: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v24, v16, v24, v0.t
+; RV32-NEXT: vsub.vv v8, v8, v24, 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, 16
; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; RV32-NEXT: vand.vv v16, v16, v8, v0.t
-; 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: vs8r.v v16, (a0) # Unknown-size Folded Spill
-; RV32-NEXT: vmv8r.v v16, v8
-; 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 v8, (a0) # 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 v24, v8, 4, v0.t
+; RV32-NEXT: vadd.vv v8, v8, v24, v0.t
; 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: 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: csrr a0, vlenb
-; RV32-NEXT: li a1, 24
-; RV32-NEXT: mul a0, a0, a1
+; RV32-NEXT: slli a0, a0, 4
; 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: 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, 16
; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; RV32-NEXT: vmul.vv v8, v8, v16, v0.t
; RV32-NEXT: vsrl.vx v8, v8, a3, 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: addi a0, sp, 16
; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: li a1, 56
+; RV32-NEXT: li a1, 48
; RV32-NEXT: mul a0, a0, a1
; RV32-NEXT: add sp, sp, a0
; RV32-NEXT: .cfi_def_cfa sp, 16
@@ -2493,18 +2382,7 @@ define <vscale x 16 x i64> @vp_cttz_nxv16i64(<vscale x 16 x i64> %va, <vscale x
;
; RV64-LABEL: vp_cttz_nxv16i64:
; RV64: # %bb.0:
-; RV64-NEXT: addi sp, sp, -16
-; RV64-NEXT: .cfi_def_cfa_offset 16
-; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: slli a1, a1, 4
-; RV64-NEXT: sub sp, sp, a1
-; 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: vmv1r.v v24, v0
-; 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: vs8r.v v8, (a1) # Unknown-size Folded Spill
+; RV64-NEXT: vmv1r.v v7, v0
; RV64-NEXT: csrr a1, vlenb
; RV64-NEXT: li a2, 1
; RV64-NEXT: lui a3, 349525
@@ -2532,56 +2410,42 @@ define <vscale x 16 x i64> @vp_cttz_nxv16i64(<vscale x 16 x i64> %va, <vscale x
; RV64-NEXT: and t0, a5, t0
; RV64-NEXT: li a5, 56
; RV64-NEXT: vsetvli zero, t0, e64, m8, ta, ma
-; RV64-NEXT: vsub.vx v8, v16, a2, v0.t
+; RV64-NEXT: vsub.vx v24, v16, a2, v0.t
; RV64-NEXT: vnot.v v16, v16, v0.t
-; RV64-NEXT: vand.vv v8, v16, v8, v0.t
-; RV64-NEXT: vsrl.vi v16, v8, 1, v0.t
-; RV64-NEXT: vand.vx v16, v16, a7, v0.t
-; RV64-NEXT: vsub.vv v8, v8, v16, v0.t
-; RV64-NEXT: vand.vx v16, v8, a6, v0.t
-; RV64-NEXT: vsrl.vi v8, v8, 2, v0.t
-; RV64-NEXT: vand.vx v8, v8, a6, v0.t
-; RV64-NEXT: vadd.vv v8, v16, v8, v0.t
-; RV64-NEXT: vsrl.vi v16, v8, 4, v0.t
-; RV64-NEXT: vadd.vv v8, v8, v16, v0.t
-; RV64-NEXT: vand.vx v8, v8, a3, v0.t
-; RV64-NEXT: vmul.vx v8, v8, a4, v0.t
-; RV64-NEXT: vsrl.vx v8, v8, a5, v0.t
-; RV64-NEXT: addi t0, sp, 16
-; RV64-NEXT: vs8r.v v8, (t0) # Unknown-size Folded Spill
+; RV64-NEXT: vand.vv v16, v16, v24, v0.t
+; RV64-NEXT: vsrl.vi v24, v16, 1, v0.t
+; RV64-NEXT: vand.vx v24, v24, a7, v0.t
+; RV64-NEXT: vsub.vv v16, v16, v24, v0.t
+; RV64-NEXT: vand.vx v24, v16, a6, v0.t
+; RV64-NEXT: vsrl.vi v16, v16, 2, v0.t
+; RV64-NEXT: vand.vx v16, v16, a6, v0.t
+; RV64-NEXT: vadd.vv v16, v24, v16, v0.t
+; RV64-NEXT: vsrl.vi v24, v16, 4, v0.t
+; RV64-NEXT: vadd.vv v16, v16, v24, v0.t
+; RV64-NEXT: vand.vx v16, v16, a3, v0.t
+; RV64-NEXT: vmul.vx v16, v16, a4, v0.t
+; RV64-NEXT: vsrl.vx v16, v16, a5, v0.t
; RV64-NEXT: bltu a0, a1, .LBB46_2
; RV64-NEXT: # %bb.1:
; RV64-NEXT: mv a0, a1
; RV64-NEXT: .LBB46_2:
-; RV64-NEXT: vmv1r.v v0, v24
-; 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: vmv1r.v v0, v7
; RV64-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV64-NEXT: vsub.vx v16, v8, a2, v0.t
+; RV64-NEXT: vsub.vx v24, v8, a2, v0.t
; RV64-NEXT: vnot.v v8, v8, v0.t
-; RV64-NEXT: vand.vv v8, v8, v16, v0.t
-; RV64-NEXT: vsrl.vi v16, v8, 1, v0.t
-; RV64-NEXT: vand.vx v16, v16, a7, v0.t
-; RV64-NEXT: vsub.vv v8, v8, v16, v0.t
-; RV64-NEXT: vand.vx v16, v8, a6, v0.t
+; RV64-NEXT: vand.vv v8, v8, v24, v0.t
+; RV64-NEXT: vsrl.vi v24, v8, 1, v0.t
+; RV64-NEXT: vand.vx v24, v24, a7, v0.t
+; RV64-NEXT: vsub.vv v8, v8, v24, v0.t
+; RV64-NEXT: vand.vx v24, v8, a6, v0.t
; RV64-NEXT: vsrl.vi v8, v8, 2, v0.t
; RV64-NEXT: vand.vx v8, v8, a6, v0.t
-; RV64-NEXT: vadd.vv v8, v16, v8, v0.t
-; RV64-NEXT: vsrl.vi v16, v8, 4, v0.t
-; RV64-NEXT: vadd.vv v8, v8, v16, v0.t
+; RV64-NEXT: vadd.vv v8, v24, v8, v0.t
+; RV64-NEXT: vsrl.vi v24, v8, 4, v0.t
+; RV64-NEXT: vadd.vv v8, v8, v24, v0.t
; RV64-NEXT: vand.vx v8, v8, a3, v0.t
; RV64-NEXT: vmul.vx v8, v8, a4, v0.t
; RV64-NEXT: vsrl.vx v8, v8, a5, v0.t
-; RV64-NEXT: addi a0, sp, 16
-; RV64-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; RV64-NEXT: csrr a0, vlenb
-; RV64-NEXT: slli a0, a0, 4
-; RV64-NEXT: add sp, sp, a0
-; RV64-NEXT: .cfi_def_cfa sp, 16
-; RV64-NEXT: addi sp, sp, 16
-; RV64-NEXT: .cfi_def_cfa_offset 0
; RV64-NEXT: ret
;
; CHECK-ZVBB-LABEL: vp_cttz_nxv16i64:
@@ -3996,18 +3860,7 @@ define <vscale x 8 x i64> @vp_cttz_zero_undef_nxv8i64_unmasked(<vscale x 8 x i64
define <vscale x 16 x i64> @vp_cttz_zero_undef_nxv16i64(<vscale x 16 x i64> %va, <vscale x 16 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_cttz_zero_undef_nxv16i64:
; 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: 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: vmv1r.v v7, v0
; CHECK-NEXT: csrr a1, vlenb
; CHECK-NEXT: fsrmi a3, 1
; CHECK-NEXT: srli a2, a1, 3
@@ -4019,40 +3872,26 @@ define <vscale x 16 x i64> @vp_cttz_zero_undef_nxv16i64(<vscale x 16 x i64> %va,
; CHECK-NEXT: and a4, a2, a4
; CHECK-NEXT: li a2, 52
; CHECK-NEXT: vsetvli zero, a4, e64, m8, ta, ma
-; CHECK-NEXT: vrsub.vi v8, v16, 0, v0.t
-; CHECK-NEXT: vand.vv v8, v16, v8, v0.t
-; CHECK-NEXT: vfcvt.f.xu.v v8, v8, v0.t
+; CHECK-NEXT: vrsub.vi v24, v16, 0, v0.t
+; CHECK-NEXT: vand.vv v16, v16, v24, v0.t
+; CHECK-NEXT: vfcvt.f.xu.v v16, v16, v0.t
; CHECK-NEXT: fsrm a3
-; CHECK-NEXT: vsrl.vx v8, v8, a2, v0.t
+; CHECK-NEXT: vsrl.vx v16, v16, a2, v0.t
; CHECK-NEXT: li a3, 1023
-; CHECK-NEXT: vsub.vx v8, v8, a3, v0.t
-; CHECK-NEXT: addi a4, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a4) # Unknown-size Folded Spill
+; CHECK-NEXT: vsub.vx v16, v16, a3, v0.t
; CHECK-NEXT: bltu a0, a1, .LBB94_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB94_2:
-; CHECK-NEXT: vmv1r.v v0, v24
-; 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: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: vrsub.vi v16, v8, 0, v0.t
+; CHECK-NEXT: vrsub.vi v24, v8, 0, v0.t
; CHECK-NEXT: fsrmi a0, 1
-; CHECK-NEXT: vand.vv v8, v8, v16, v0.t
+; CHECK-NEXT: vand.vv v8, v8, v24, v0.t
; CHECK-NEXT: vfcvt.f.xu.v v8, v8, v0.t
; CHECK-NEXT: vsrl.vx v8, v8, a2, v0.t
; CHECK-NEXT: vsub.vx v8, v8, a3, v0.t
; CHECK-NEXT: fsrm a0
-; CHECK-NEXT: addi a0, sp, 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: .cfi_def_cfa sp, 16
-; CHECK-NEXT: addi sp, sp, 16
-; CHECK-NEXT: .cfi_def_cfa_offset 0
; CHECK-NEXT: ret
;
; CHECK-ZVBB-LABEL: vp_cttz_zero_undef_nxv16i64:
diff --git a/llvm/test/CodeGen/RISCV/rvv/expandload.ll b/llvm/test/CodeGen/RISCV/rvv/expandload.ll
index b32d85bb1943a5..7b970f7152237c 100644
--- a/llvm/test/CodeGen/RISCV/rvv/expandload.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/expandload.ll
@@ -230,7 +230,7 @@ define <256 x i8> @test_expandload_v256i8(ptr %base, <256 x i1> %mask, <256 x i8
; CHECK-RV32-NEXT: vmv1r.v v7, v8
; CHECK-RV32-NEXT: li a2, 128
; CHECK-RV32-NEXT: vsetivli zero, 1, e64, m1, ta, ma
-; CHECK-RV32-NEXT: vslidedown.vi v9, v0, 1
+; CHECK-RV32-NEXT: vslidedown.vi v8, v0, 1
; CHECK-RV32-NEXT: li a3, 32
; CHECK-RV32-NEXT: vmv.x.s a4, v0
; CHECK-RV32-NEXT: vsetvli zero, a2, e8, m8, ta, ma
@@ -241,15 +241,15 @@ define <256 x i8> @test_expandload_v256i8(ptr %base, <256 x i1> %mask, <256 x i8
; CHECK-RV32-NEXT: addi a1, a1, 16
; CHECK-RV32-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
; CHECK-RV32-NEXT: vsetivli zero, 1, e64, m1, ta, ma
-; CHECK-RV32-NEXT: vsrl.vx v10, v9, a3
-; CHECK-RV32-NEXT: vsrl.vx v11, v0, a3
-; CHECK-RV32-NEXT: vmv.x.s a1, v9
+; CHECK-RV32-NEXT: vsrl.vx v9, v8, a3
+; CHECK-RV32-NEXT: vsrl.vx v10, v0, a3
+; CHECK-RV32-NEXT: vmv.x.s a1, v8
; CHECK-RV32-NEXT: vsetvli zero, a2, e8, m8, ta, ma
; CHECK-RV32-NEXT: vcpop.m a3, v0
; CHECK-RV32-NEXT: cpop a4, a4
; CHECK-RV32-NEXT: vsetvli zero, a2, e64, m1, ta, ma
-; CHECK-RV32-NEXT: vmv.x.s a5, v10
-; CHECK-RV32-NEXT: vmv.x.s a6, v11
+; CHECK-RV32-NEXT: vmv.x.s a5, v9
+; CHECK-RV32-NEXT: vmv.x.s a6, v10
; CHECK-RV32-NEXT: vsetvli zero, a3, e8, m8, ta, ma
; CHECK-RV32-NEXT: vle8.v v8, (a0)
; CHECK-RV32-NEXT: csrr a3, vlenb
@@ -290,12 +290,7 @@ define <256 x i8> @test_expandload_v256i8(ptr %base, <256 x i1> %mask, <256 x i8
; CHECK-RV32-NEXT: add a0, sp, a0
; CHECK-RV32-NEXT: addi a0, a0, 16
; CHECK-RV32-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
-; CHECK-RV32-NEXT: viota.m v16, v7
-; CHECK-RV32-NEXT: csrr a0, vlenb
-; CHECK-RV32-NEXT: slli a0, a0, 4
-; CHECK-RV32-NEXT: add a0, sp, a0
-; CHECK-RV32-NEXT: addi a0, a0, 16
-; CHECK-RV32-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
+; CHECK-RV32-NEXT: viota.m v8, v7
; CHECK-RV32-NEXT: vmv1r.v v0, v7
; CHECK-RV32-NEXT: csrr a0, vlenb
; CHECK-RV32-NEXT: slli a0, a0, 3
@@ -304,11 +299,6 @@ define <256 x i8> @test_expandload_v256i8(ptr %base, <256 x i1> %mask, <256 x i8
; CHECK-RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-RV32-NEXT: addi a0, sp, 16
; CHECK-RV32-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
-; CHECK-RV32-NEXT: csrr a0, vlenb
-; CHECK-RV32-NEXT: slli a0, a0, 4
-; CHECK-RV32-NEXT: add a0, sp, a0
-; CHECK-RV32-NEXT: addi a0, a0, 16
-; CHECK-RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; CHECK-RV32-NEXT: vrgather.vv v16, v24, v8, v0.t
; CHECK-RV32-NEXT: csrr a0, vlenb
; CHECK-RV32-NEXT: li a1, 24
@@ -341,7 +331,7 @@ define <256 x i8> @test_expandload_v256i8(ptr %base, <256 x i1> %mask, <256 x i8
; CHECK-RV64-NEXT: vmv1r.v v7, v8
; CHECK-RV64-NEXT: li a2, 128
; CHECK-RV64-NEXT: vsetivli zero, 1, e64, m1, ta, ma
-; CHECK-RV64-NEXT: vslidedown.vi v9, v0, 1
+; CHECK-RV64-NEXT: vslidedown.vi v8, v0, 1
; CHECK-RV64-NEXT: vmv.x.s a3, v0
; CHECK-RV64-NEXT: vsetvli zero, a2, e8, m8, ta, ma
; CHECK-RV64-NEXT: vle8.v v16, (a1)
@@ -351,7 +341,7 @@ define <256 x i8> @test_expandload_v256i8(ptr %base, <256 x i1> %mask, <256 x i8
; CHECK-RV64-NEXT: addi a1, a1, 16
; CHECK-RV64-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
; CHECK-RV64-NEXT: vsetvli zero, a2, e64, m1, ta, ma
-; CHECK-RV64-NEXT: vmv.x.s a1, v9
+; CHECK-RV64-NEXT: vmv.x.s a1, v8
; CHECK-RV64-NEXT: vsetvli zero, a2, e8, m8, ta, ma
; CHECK-RV64-NEXT: vcpop.m a4, v0
; CHECK-RV64-NEXT: vsetvli zero, a4, e8, m8, ta, ma
@@ -372,31 +362,26 @@ define <256 x i8> @test_expandload_v256i8(ptr %base, <256 x i1> %mask, <256 x i8
; CHECK-RV64-NEXT: addi a0, sp, 16
; CHECK-RV64-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; CHECK-RV64-NEXT: vsetvli zero, a2, e8, m8, ta, mu
-; CHECK-RV64-NEXT: viota.m v24, v0
-; CHECK-RV64-NEXT: csrr a0, vlenb
-; CHECK-RV64-NEXT: li a1, 24
-; CHECK-RV64-NEXT: mul a0, a0, a1
-; CHECK-RV64-NEXT: add a0, sp, a0
-; CHECK-RV64-NEXT: addi a0, a0, 16
-; CHECK-RV64-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-RV64-NEXT: viota.m v16, v0
; CHECK-RV64-NEXT: csrr a0, vlenb
; CHECK-RV64-NEXT: slli a0, a0, 4
; CHECK-RV64-NEXT: add a0, sp, a0
; CHECK-RV64-NEXT: addi a0, a0, 16
-; CHECK-RV64-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; CHECK-RV64-NEXT: vrgather.vv v8, v16, v24, v0.t
+; CHECK-RV64-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
; CHECK-RV64-NEXT: csrr a0, vlenb
; CHECK-RV64-NEXT: li a1, 24
; CHECK-RV64-NEXT: mul a0, a0, a1
; CHECK-RV64-NEXT: add a0, sp, a0
; CHECK-RV64-NEXT: addi a0, a0, 16
-; CHECK-RV64-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
-; CHECK-RV64-NEXT: viota.m v16, v7
+; CHECK-RV64-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-RV64-NEXT: vrgather.vv v8, v24, v16, v0.t
; CHECK-RV64-NEXT: csrr a0, vlenb
-; CHECK-RV64-NEXT: slli a0, a0, 4
+; CHECK-RV64-NEXT: li a1, 24
+; CHECK-RV64-NEXT: mul a0, a0, a1
; CHECK-RV64-NEXT: add a0, sp, a0
; CHECK-RV64-NEXT: addi a0, a0, 16
-; CHECK-RV64-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
+; CHECK-RV64-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; CHECK-RV64-NEXT: viota.m v8, v7
; CHECK-RV64-NEXT: vmv1r.v v0, v7
; CHECK-RV64-NEXT: csrr a0, vlenb
; CHECK-RV64-NEXT: slli a0, a0, 3
@@ -405,11 +390,6 @@ define <256 x i8> @test_expandload_v256i8(ptr %base, <256 x i1> %mask, <256 x i8
; CHECK-RV64-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-RV64-NEXT: addi a0, sp, 16
; CHECK-RV64-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
-; CHECK-RV64-NEXT: csrr a0, vlenb
-; CHECK-RV64-NEXT: slli a0, a0, 4
-; CHECK-RV64-NEXT: add a0, sp, a0
-; CHECK-RV64-NEXT: addi a0, a0, 16
-; CHECK-RV64-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; CHECK-RV64-NEXT: vrgather.vv v16, v24, v8, v0.t
; CHECK-RV64-NEXT: csrr a0, vlenb
; CHECK-RV64-NEXT: li a1, 24
@@ -664,12 +644,12 @@ define <128 x i16> @test_expandload_v128i16(ptr %base, <128 x i1> %mask, <128 x
; CHECK-RV32-NEXT: addi sp, sp, -16
; CHECK-RV32-NEXT: .cfi_def_cfa_offset 16
; CHECK-RV32-NEXT: csrr a1, vlenb
-; CHECK-RV32-NEXT: slli a1, a1, 5
-; CHECK-RV32-NEXT: sub sp, sp, a1
-; CHECK-RV32-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x20, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 32 * vlenb
-; CHECK-RV32-NEXT: csrr a1, vlenb
; CHECK-RV32-NEXT: li a2, 24
; CHECK-RV32-NEXT: mul a1, a1, a2
+; CHECK-RV32-NEXT: sub sp, sp, a1
+; CHECK-RV32-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x18, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 24 * vlenb
+; CHECK-RV32-NEXT: csrr a1, vlenb
+; CHECK-RV32-NEXT: slli a1, a1, 4
; CHECK-RV32-NEXT: add a1, sp, a1
; CHECK-RV32-NEXT: addi a1, a1, 16
; CHECK-RV32-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
@@ -682,58 +662,49 @@ define <128 x i16> @test_expandload_v128i16(ptr %base, <128 x i1> %mask, <128 x
; CHECK-RV32-NEXT: vsetvli zero, a1, e8, m4, ta, ma
; CHECK-RV32-NEXT: vcpop.m a4, v0
; CHECK-RV32-NEXT: vsetivli zero, 1, e64, m1, ta, ma
-; CHECK-RV32-NEXT: vsrl.vx v25, v0, a2
+; CHECK-RV32-NEXT: vsrl.vx v6, v0, a2
; CHECK-RV32-NEXT: vsetvli zero, a1, e8, m4, ta, ma
; CHECK-RV32-NEXT: vcpop.m a2, v7
; CHECK-RV32-NEXT: vsetvli zero, a4, e16, m8, ta, ma
; CHECK-RV32-NEXT: vle16.v v16, (a0)
-; CHECK-RV32-NEXT: csrr a5, vlenb
-; CHECK-RV32-NEXT: slli a5, a5, 4
-; CHECK-RV32-NEXT: add a5, sp, a5
-; CHECK-RV32-NEXT: addi a5, a5, 16
-; CHECK-RV32-NEXT: vs8r.v v16, (a5) # Unknown-size Folded Spill
; CHECK-RV32-NEXT: vsetvli zero, a4, e64, m1, ta, ma
-; CHECK-RV32-NEXT: vmv.x.s a4, v25
+; CHECK-RV32-NEXT: vmv.x.s a4, v6
; CHECK-RV32-NEXT: cpop a4, a4
; CHECK-RV32-NEXT: cpop a3, a3
; CHECK-RV32-NEXT: add a3, a3, a4
; CHECK-RV32-NEXT: slli a3, a3, 1
; CHECK-RV32-NEXT: add a0, a0, a3
; CHECK-RV32-NEXT: vsetvli zero, a2, e16, m8, ta, ma
-; CHECK-RV32-NEXT: vle16.v v16, (a0)
+; CHECK-RV32-NEXT: vle16.v v24, (a0)
; CHECK-RV32-NEXT: csrr a0, vlenb
; CHECK-RV32-NEXT: slli a0, a0, 3
; CHECK-RV32-NEXT: add a0, sp, a0
; CHECK-RV32-NEXT: addi a0, a0, 16
-; CHECK-RV32-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
+; CHECK-RV32-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; CHECK-RV32-NEXT: vsetvli zero, a1, e16, m8, ta, mu
-; CHECK-RV32-NEXT: viota.m v16, v0
-; CHECK-RV32-NEXT: csrr a0, vlenb
-; CHECK-RV32-NEXT: slli a0, a0, 4
-; CHECK-RV32-NEXT: add a0, sp, a0
-; CHECK-RV32-NEXT: addi a0, a0, 16
-; CHECK-RV32-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
-; CHECK-RV32-NEXT: vrgather.vv v8, v24, v16, v0.t
+; CHECK-RV32-NEXT: viota.m v24, v0
+; CHECK-RV32-NEXT: vrgather.vv v8, v16, v24, v0.t
; CHECK-RV32-NEXT: addi a0, sp, 16
; CHECK-RV32-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; CHECK-RV32-NEXT: viota.m v8, v7
; CHECK-RV32-NEXT: vmv1r.v v0, v7
; CHECK-RV32-NEXT: csrr a0, vlenb
-; CHECK-RV32-NEXT: li a1, 24
-; CHECK-RV32-NEXT: mul a0, a0, a1
+; CHECK-RV32-NEXT: slli a0, a0, 4
; CHECK-RV32-NEXT: add a0, sp, a0
; CHECK-RV32-NEXT: addi a0, a0, 16
-; CHECK-RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; CHECK-RV32-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
; CHECK-RV32-NEXT: csrr a0, vlenb
; CHECK-RV32-NEXT: slli a0, a0, 3
; CHECK-RV32-NEXT: add a0, sp, a0
; CHECK-RV32-NEXT: addi a0, a0, 16
-; CHECK-RV32-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
-; CHECK-RV32-NEXT: vrgather.vv v16, v24, v8, v0.t
+; CHECK-RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; CHECK-RV32-NEXT: vrgather.vv v24, v16, v8, v0.t
; CHECK-RV32-NEXT: addi a0, sp, 16
; CHECK-RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-RV32-NEXT: vmv.v.v v16, v24
; CHECK-RV32-NEXT: csrr a0, vlenb
-; CHECK-RV32-NEXT: slli a0, a0, 5
+; CHECK-RV32-NEXT: li a1, 24
+; CHECK-RV32-NEXT: mul a0, a0, a1
; CHECK-RV32-NEXT: add sp, sp, a0
; CHECK-RV32-NEXT: .cfi_def_cfa sp, 16
; CHECK-RV32-NEXT: addi sp, sp, 16
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 3eb5d36b4896a7..9305879147d320 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-bitreverse-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-bitreverse-vp.ll
@@ -1659,7 +1659,6 @@ define <15 x i64> @vp_bitreverse_v15i64(<15 x i64> %va, <15 x i1> %m, i32 zeroex
; 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: vmv8r.v v24, v8
; RV32-NEXT: lui a2, 1044480
; RV32-NEXT: lui a3, 61681
; RV32-NEXT: lui a4, 209715
@@ -1682,58 +1681,60 @@ define <15 x i64> @vp_bitreverse_v15i64(<15 x i64> %va, <15 x i1> %m, i32 zeroex
; RV32-NEXT: addi a5, a6, -256
; RV32-NEXT: sw a4, 24(sp)
; RV32-NEXT: sw a4, 28(sp)
-; RV32-NEXT: vand.vx v8, v8, a5, v0.t
-; RV32-NEXT: vsll.vx v8, v8, a2, v0.t
-; RV32-NEXT: vor.vv v8, v16, v8, v0.t
+; RV32-NEXT: vand.vx v24, v8, a5, v0.t
+; RV32-NEXT: vsll.vx v24, v24, a2, v0.t
+; RV32-NEXT: vor.vv v16, v16, v24, v0.t
; RV32-NEXT: csrr a4, vlenb
-; RV32-NEXT: slli a4, a4, 4
+; RV32-NEXT: slli a4, a4, 3
; RV32-NEXT: add a4, sp, a4
; RV32-NEXT: addi a4, a4, 48
-; RV32-NEXT: vs8r.v v8, (a4) # Unknown-size Folded Spill
+; RV32-NEXT: vs8r.v v16, (a4) # Unknown-size Folded Spill
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v8, (a3), zero
+; RV32-NEXT: vlse64.v v16, (a3), zero
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: slli a3, a3, 3
+; RV32-NEXT: slli a3, a3, 4
; 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 v16, (a3) # Unknown-size Folded Spill
; RV32-NEXT: lui a3, 4080
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vx v16, v24, a3, v0.t
-; RV32-NEXT: vsll.vi v16, v16, 24, v0.t
-; RV32-NEXT: addi a4, sp, 48
-; RV32-NEXT: vs8r.v v16, (a4) # Unknown-size Folded Spill
-; RV32-NEXT: vand.vv v16, v24, v8, v0.t
-; RV32-NEXT: vsll.vi v16, v16, 8, v0.t
-; RV32-NEXT: vl8r.v v8, (a4) # Unknown-size Folded Reload
-; RV32-NEXT: vor.vv v16, v8, v16, v0.t
+; RV32-NEXT: vand.vx v16, v8, a3, v0.t
+; RV32-NEXT: vsll.vi v24, v16, 24, v0.t
; RV32-NEXT: csrr a4, vlenb
; RV32-NEXT: slli a4, a4, 4
; RV32-NEXT: add a4, sp, a4
; RV32-NEXT: addi a4, a4, 48
-; RV32-NEXT: vl8r.v v8, (a4) # Unknown-size Folded Reload
-; RV32-NEXT: vor.vv v16, v8, v16, v0.t
+; RV32-NEXT: vl8r.v v16, (a4) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v16, v8, v16, v0.t
+; RV32-NEXT: vsll.vi v16, v16, 8, v0.t
+; RV32-NEXT: vor.vv v16, v24, v16, v0.t
; RV32-NEXT: csrr a4, vlenb
-; RV32-NEXT: slli a4, a4, 4
+; RV32-NEXT: slli a4, a4, 3
+; RV32-NEXT: add a4, sp, a4
+; RV32-NEXT: addi a4, a4, 48
+; RV32-NEXT: vl8r.v v24, (a4) # Unknown-size Folded Reload
+; RV32-NEXT: vor.vv v16, v24, v16, v0.t
+; 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: vsrl.vx v16, v24, a1, v0.t
-; RV32-NEXT: vsrl.vx v8, v24, a2, v0.t
-; RV32-NEXT: vand.vx v8, v8, a5, v0.t
-; RV32-NEXT: vor.vv v8, v8, v16, v0.t
+; RV32-NEXT: vsrl.vx v16, v8, a1, v0.t
+; RV32-NEXT: vsrl.vx v24, v8, a2, v0.t
+; RV32-NEXT: vand.vx v24, v24, a5, v0.t
+; RV32-NEXT: vor.vv v16, v24, v16, v0.t
; RV32-NEXT: addi a1, sp, 48
-; RV32-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
-; RV32-NEXT: vsrl.vi v8, v24, 24, v0.t
-; RV32-NEXT: vand.vx v16, v8, a3, v0.t
-; RV32-NEXT: vsrl.vi v8, v24, 8, v0.t
+; RV32-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: vsrl.vi v24, v8, 24, v0.t
+; RV32-NEXT: vand.vx v24, v24, a3, v0.t
+; RV32-NEXT: vsrl.vi v8, v8, 8, v0.t
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: slli a1, a1, 3
+; 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: vand.vv v8, v8, v24, v0.t
-; RV32-NEXT: vor.vv v8, v8, v16, v0.t
+; 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, 48
; RV32-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
; RV32-NEXT: vor.vv v8, v8, v16, v0.t
@@ -1743,7 +1744,7 @@ define <15 x i64> @vp_bitreverse_v15i64(<15 x i64> %va, <15 x i1> %m, i32 zeroex
; RV32-NEXT: vlse64.v v24, (a1), zero
; RV32-NEXT: addi a1, sp, 24
; 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: vl8r.v v16, (a3) # Unknown-size Folded Reload
@@ -2055,7 +2056,6 @@ define <16 x i64> @vp_bitreverse_v16i64(<16 x i64> %va, <16 x i1> %m, i32 zeroex
; 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: vmv8r.v v24, v8
; RV32-NEXT: lui a2, 1044480
; RV32-NEXT: lui a3, 61681
; RV32-NEXT: lui a4, 209715
@@ -2078,58 +2078,60 @@ define <16 x i64> @vp_bitreverse_v16i64(<16 x i64> %va, <16 x i1> %m, i32 zeroex
; RV32-NEXT: addi a5, a6, -256
; RV32-NEXT: sw a4, 24(sp)
; RV32-NEXT: sw a4, 28(sp)
-; RV32-NEXT: vand.vx v8, v8, a5, v0.t
-; RV32-NEXT: vsll.vx v8, v8, a2, v0.t
-; RV32-NEXT: vor.vv v8, v16, v8, v0.t
+; RV32-NEXT: vand.vx v24, v8, a5, v0.t
+; RV32-NEXT: vsll.vx v24, v24, a2, v0.t
+; RV32-NEXT: vor.vv v16, v16, v24, v0.t
; RV32-NEXT: csrr a4, vlenb
-; RV32-NEXT: slli a4, a4, 4
+; RV32-NEXT: slli a4, a4, 3
; RV32-NEXT: add a4, sp, a4
; RV32-NEXT: addi a4, a4, 48
-; RV32-NEXT: vs8r.v v8, (a4) # Unknown-size Folded Spill
+; RV32-NEXT: vs8r.v v16, (a4) # Unknown-size Folded Spill
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v8, (a3), zero
+; RV32-NEXT: vlse64.v v16, (a3), zero
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: slli a3, a3, 3
+; RV32-NEXT: slli a3, a3, 4
; 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 v16, (a3) # Unknown-size Folded Spill
; RV32-NEXT: lui a3, 4080
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vx v16, v24, a3, v0.t
-; RV32-NEXT: vsll.vi v16, v16, 24, v0.t
-; RV32-NEXT: addi a4, sp, 48
-; RV32-NEXT: vs8r.v v16, (a4) # Unknown-size Folded Spill
-; RV32-NEXT: vand.vv v16, v24, v8, v0.t
-; RV32-NEXT: vsll.vi v16, v16, 8, v0.t
-; RV32-NEXT: vl8r.v v8, (a4) # Unknown-size Folded Reload
-; RV32-NEXT: vor.vv v16, v8, v16, v0.t
+; RV32-NEXT: vand.vx v16, v8, a3, v0.t
+; RV32-NEXT: vsll.vi v24, v16, 24, v0.t
; RV32-NEXT: csrr a4, vlenb
; RV32-NEXT: slli a4, a4, 4
; RV32-NEXT: add a4, sp, a4
; RV32-NEXT: addi a4, a4, 48
-; RV32-NEXT: vl8r.v v8, (a4) # Unknown-size Folded Reload
-; RV32-NEXT: vor.vv v16, v8, v16, v0.t
+; RV32-NEXT: vl8r.v v16, (a4) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v16, v8, v16, v0.t
+; RV32-NEXT: vsll.vi v16, v16, 8, v0.t
+; RV32-NEXT: vor.vv v16, v24, v16, v0.t
; RV32-NEXT: csrr a4, vlenb
-; RV32-NEXT: slli a4, a4, 4
+; RV32-NEXT: slli a4, a4, 3
+; RV32-NEXT: add a4, sp, a4
+; RV32-NEXT: addi a4, a4, 48
+; RV32-NEXT: vl8r.v v24, (a4) # Unknown-size Folded Reload
+; RV32-NEXT: vor.vv v16, v24, v16, v0.t
+; 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: vsrl.vx v16, v24, a1, v0.t
-; RV32-NEXT: vsrl.vx v8, v24, a2, v0.t
-; RV32-NEXT: vand.vx v8, v8, a5, v0.t
-; RV32-NEXT: vor.vv v8, v8, v16, v0.t
+; RV32-NEXT: vsrl.vx v16, v8, a1, v0.t
+; RV32-NEXT: vsrl.vx v24, v8, a2, v0.t
+; RV32-NEXT: vand.vx v24, v24, a5, v0.t
+; RV32-NEXT: vor.vv v16, v24, v16, v0.t
; RV32-NEXT: addi a1, sp, 48
-; RV32-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
-; RV32-NEXT: vsrl.vi v8, v24, 24, v0.t
-; RV32-NEXT: vand.vx v16, v8, a3, v0.t
-; RV32-NEXT: vsrl.vi v8, v24, 8, v0.t
+; RV32-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: vsrl.vi v24, v8, 24, v0.t
+; RV32-NEXT: vand.vx v24, v24, a3, v0.t
+; RV32-NEXT: vsrl.vi v8, v8, 8, v0.t
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: slli a1, a1, 3
+; 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: vand.vv v8, v8, v24, v0.t
-; RV32-NEXT: vor.vv v8, v8, v16, v0.t
+; 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, 48
; RV32-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
; RV32-NEXT: vor.vv v8, v8, v16, v0.t
@@ -2139,7 +2141,7 @@ define <16 x i64> @vp_bitreverse_v16i64(<16 x i64> %va, <16 x i1> %m, i32 zeroex
; RV32-NEXT: vlse64.v v24, (a1), zero
; RV32-NEXT: addi a1, sp, 24
; 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: vl8r.v v16, (a3) # Unknown-size Folded Reload
@@ -2444,27 +2446,16 @@ declare <128 x i16> @llvm.vp.bitreverse.v128i16(<128 x i16>, <128 x i1>, i32)
define <128 x i16> @vp_bitreverse_v128i16(<128 x i16> %va, <128 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_bitreverse_v128i16:
; 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: 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: li a2, 64
; CHECK-NEXT: vsetivli zero, 8, e8, m1, ta, ma
-; CHECK-NEXT: vslidedown.vi v24, v0, 8
+; CHECK-NEXT: vslidedown.vi v7, v0, 8
; CHECK-NEXT: mv a1, a0
; CHECK-NEXT: bltu a0, a2, .LBB34_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: li a1, 64
; CHECK-NEXT: .LBB34_2:
; CHECK-NEXT: vsetvli zero, a1, e16, m8, ta, ma
-; CHECK-NEXT: vsrl.vi v16, v8, 8, v0.t
+; CHECK-NEXT: vsrl.vi v24, v8, 8, v0.t
; CHECK-NEXT: lui a1, 1
; CHECK-NEXT: lui a2, 3
; CHECK-NEXT: addi a3, a0, -64
@@ -2473,60 +2464,45 @@ define <128 x i16> @vp_bitreverse_v128i16(<128 x i16> %va, <128 x i1> %m, i32 ze
; CHECK-NEXT: and a3, a0, a3
; CHECK-NEXT: lui a0, 5
; CHECK-NEXT: vsll.vi v8, v8, 8, v0.t
-; CHECK-NEXT: addi a4, a1, -241
-; CHECK-NEXT: addi a1, a2, 819
+; CHECK-NEXT: addi a1, a1, -241
+; CHECK-NEXT: addi a2, a2, 819
; CHECK-NEXT: addi a0, a0, 1365
-; CHECK-NEXT: vor.vv v8, v8, v16, v0.t
-; CHECK-NEXT: vsrl.vi v16, v8, 4, v0.t
-; CHECK-NEXT: vand.vx v16, v16, a4, v0.t
-; CHECK-NEXT: vand.vx v8, v8, a4, v0.t
-; CHECK-NEXT: vsll.vi v8, v8, 4, v0.t
-; CHECK-NEXT: vor.vv v8, v16, v8, v0.t
-; CHECK-NEXT: vsrl.vi v16, v8, 2, v0.t
-; CHECK-NEXT: vand.vx v16, v16, a1, v0.t
+; CHECK-NEXT: vor.vv v8, v8, v24, v0.t
+; CHECK-NEXT: vsrl.vi v24, v8, 4, v0.t
+; CHECK-NEXT: vand.vx v24, v24, a1, v0.t
; CHECK-NEXT: vand.vx v8, v8, a1, v0.t
+; CHECK-NEXT: vsll.vi v8, v8, 4, v0.t
+; CHECK-NEXT: vor.vv v8, v24, v8, v0.t
+; CHECK-NEXT: vsrl.vi v24, v8, 2, v0.t
+; CHECK-NEXT: vand.vx v24, v24, a2, v0.t
+; CHECK-NEXT: vand.vx v8, v8, a2, v0.t
; CHECK-NEXT: vsll.vi v8, v8, 2, v0.t
-; CHECK-NEXT: vor.vv v8, v16, v8, v0.t
-; CHECK-NEXT: vsrl.vi v16, v8, 1, v0.t
-; CHECK-NEXT: vand.vx v16, v16, a0, v0.t
+; CHECK-NEXT: vor.vv v8, v24, v8, v0.t
+; CHECK-NEXT: vsrl.vi v24, v8, 1, v0.t
+; CHECK-NEXT: vand.vx v24, v24, a0, v0.t
; CHECK-NEXT: vand.vx v8, v8, a0, v0.t
; CHECK-NEXT: vsll.vi v8, v8, 1, v0.t
-; CHECK-NEXT: vor.vv v8, v16, v8, v0.t
-; CHECK-NEXT: addi a2, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
-; CHECK-NEXT: vmv1r.v v0, v24
-; 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: vor.vv v8, v24, v8, v0.t
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, a3, 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
-; CHECK-NEXT: vsrl.vi v16, v8, 4, v0.t
-; CHECK-NEXT: vand.vx v16, v16, a4, v0.t
-; CHECK-NEXT: vand.vx v8, v8, a4, v0.t
-; CHECK-NEXT: vsll.vi v8, v8, 4, v0.t
-; CHECK-NEXT: vor.vv v8, v16, v8, v0.t
-; CHECK-NEXT: vsrl.vi v16, v8, 2, v0.t
+; CHECK-NEXT: vsrl.vi v24, v16, 8, v0.t
+; CHECK-NEXT: vsll.vi v16, v16, 8, v0.t
+; CHECK-NEXT: vor.vv v16, v16, v24, v0.t
+; CHECK-NEXT: vsrl.vi v24, v16, 4, v0.t
+; CHECK-NEXT: vand.vx v24, v24, a1, v0.t
; CHECK-NEXT: vand.vx v16, v16, a1, v0.t
-; CHECK-NEXT: vand.vx v8, v8, a1, v0.t
-; CHECK-NEXT: vsll.vi v8, v8, 2, v0.t
-; CHECK-NEXT: vor.vv v8, v16, v8, v0.t
-; CHECK-NEXT: vsrl.vi v16, v8, 1, v0.t
+; CHECK-NEXT: vsll.vi v16, v16, 4, v0.t
+; CHECK-NEXT: vor.vv v16, v24, v16, v0.t
+; CHECK-NEXT: vsrl.vi v24, v16, 2, v0.t
+; CHECK-NEXT: vand.vx v24, v24, a2, v0.t
+; CHECK-NEXT: vand.vx v16, v16, a2, v0.t
+; CHECK-NEXT: vsll.vi v16, v16, 2, v0.t
+; CHECK-NEXT: vor.vv v16, v24, v16, v0.t
+; CHECK-NEXT: vsrl.vi v24, v16, 1, v0.t
+; CHECK-NEXT: vand.vx v24, v24, a0, v0.t
; CHECK-NEXT: vand.vx v16, v16, a0, v0.t
-; CHECK-NEXT: vand.vx v8, v8, a0, v0.t
-; CHECK-NEXT: vsll.vi v8, v8, 1, v0.t
-; CHECK-NEXT: vor.vv v16, v16, v8, v0.t
-; CHECK-NEXT: addi a0, sp, 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 sp, sp, a0
-; CHECK-NEXT: .cfi_def_cfa sp, 16
-; CHECK-NEXT: addi sp, sp, 16
-; CHECK-NEXT: .cfi_def_cfa_offset 0
+; CHECK-NEXT: vsll.vi v16, v16, 1, v0.t
+; CHECK-NEXT: vor.vv v16, v24, v16, v0.t
; CHECK-NEXT: ret
%v = call <128 x i16> @llvm.vp.bitreverse.v128i16(<128 x i16> %va, <128 x i1> %m, i32 %evl)
ret <128 x i16> %v
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 d765e4c0b8f6a9..5f382c3fdc834e 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-bswap-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-bswap-vp.ll
@@ -782,35 +782,37 @@ define <15 x i64> @vp_bswap_v15i64(<15 x i64> %va, <15 x i1> %m, i32 zeroext %ev
; RV32-NEXT: vsll.vx v24, v24, a4, v0.t
; RV32-NEXT: vor.vv v16, v16, v24, 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, 16
; RV32-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV32-NEXT: vlse64.v v16, (a5), zero
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: slli a3, a3, 3
+; 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: lui a3, 4080
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vx v24, v8, a3, 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: vand.vx v16, v8, a3, v0.t
+; RV32-NEXT: vsll.vi v24, v16, 24, v0.t
; 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 v16, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v16, v8, v16, v0.t
+; RV32-NEXT: vsll.vi v16, v16, 8, v0.t
+; RV32-NEXT: vor.vv v16, v24, 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, 16
; 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
+; RV32-NEXT: slli a0, a0, 3
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 16
; RV32-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
@@ -824,7 +826,7 @@ define <15 x i64> @vp_bswap_v15i64(<15 x i64> %va, <15 x i1> %m, i32 zeroext %ev
; RV32-NEXT: vand.vx v24, v24, a3, v0.t
; RV32-NEXT: vsrl.vi v8, v8, 8, v0.t
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 3
+; RV32-NEXT: slli a0, a0, 4
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 16
; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
@@ -834,7 +836,7 @@ define <15 x i64> @vp_bswap_v15i64(<15 x i64> %va, <15 x i1> %m, i32 zeroext %ev
; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; RV32-NEXT: vor.vv v8, v8, v16, 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, 16
; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
@@ -1045,35 +1047,37 @@ define <16 x i64> @vp_bswap_v16i64(<16 x i64> %va, <16 x i1> %m, i32 zeroext %ev
; RV32-NEXT: vsll.vx v24, v24, a4, v0.t
; RV32-NEXT: vor.vv v16, v16, v24, 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, 16
; RV32-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV32-NEXT: vlse64.v v16, (a5), zero
; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: slli a3, a3, 3
+; 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: lui a3, 4080
; RV32-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV32-NEXT: vand.vx v24, v8, a3, 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: vand.vx v16, v8, a3, v0.t
+; RV32-NEXT: vsll.vi v24, v16, 24, v0.t
; 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 v16, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v16, v8, v16, v0.t
+; RV32-NEXT: vsll.vi v16, v16, 8, v0.t
+; RV32-NEXT: vor.vv v16, v24, 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, 16
; 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
+; RV32-NEXT: slli a0, a0, 3
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 16
; RV32-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
@@ -1087,7 +1091,7 @@ define <16 x i64> @vp_bswap_v16i64(<16 x i64> %va, <16 x i1> %m, i32 zeroext %ev
; RV32-NEXT: vand.vx v24, v24, a3, v0.t
; RV32-NEXT: vsrl.vi v8, v8, 8, v0.t
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 3
+; RV32-NEXT: slli a0, a0, 4
; RV32-NEXT: add a0, sp, a0
; RV32-NEXT: addi a0, a0, 16
; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
@@ -1097,7 +1101,7 @@ define <16 x i64> @vp_bswap_v16i64(<16 x i64> %va, <16 x i1> %m, i32 zeroext %ev
; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; RV32-NEXT: vor.vv v8, v8, v16, 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, 16
; RV32-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
@@ -1287,53 +1291,27 @@ declare <128 x i16> @llvm.vp.bswap.v128i16(<128 x i16>, <128 x i1>, i32)
define <128 x i16> @vp_bswap_v128i16(<128 x i16> %va, <128 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_bswap_v128i16:
; 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: 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: li a2, 64
; CHECK-NEXT: vsetivli zero, 8, e8, m1, ta, ma
-; CHECK-NEXT: vslidedown.vi v24, v0, 8
+; CHECK-NEXT: vslidedown.vi v7, v0, 8
; CHECK-NEXT: mv a1, a0
; CHECK-NEXT: bltu a0, a2, .LBB26_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: li a1, 64
; CHECK-NEXT: .LBB26_2:
; CHECK-NEXT: vsetvli zero, a1, e16, m8, ta, ma
-; CHECK-NEXT: vsrl.vi v16, v8, 8, v0.t
+; CHECK-NEXT: vsrl.vi v24, v8, 8, v0.t
; CHECK-NEXT: vsll.vi v8, v8, 8, v0.t
-; CHECK-NEXT: vor.vv v8, v8, v16, v0.t
-; CHECK-NEXT: addi a1, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
+; CHECK-NEXT: vor.vv v8, v8, v24, v0.t
; CHECK-NEXT: addi a1, a0, -64
; CHECK-NEXT: sltu a0, a0, a1
; CHECK-NEXT: addi a0, a0, -1
; CHECK-NEXT: and a0, a0, a1
-; CHECK-NEXT: vmv1r.v v0, v24
-; 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: vmv1r.v v0, v7
; 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
-; CHECK-NEXT: addi a0, sp, 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 sp, sp, a0
-; CHECK-NEXT: .cfi_def_cfa sp, 16
-; CHECK-NEXT: addi sp, sp, 16
-; CHECK-NEXT: .cfi_def_cfa_offset 0
+; CHECK-NEXT: vsrl.vi v24, v16, 8, v0.t
+; CHECK-NEXT: vsll.vi v16, v16, 8, v0.t
+; CHECK-NEXT: vor.vv v16, v16, v24, v0.t
; CHECK-NEXT: ret
%v = call <128 x i16> @llvm.vp.bswap.v128i16(<128 x i16> %va, <128 x i1> %m, i32 %evl)
ret <128 x i16> %v
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 511242aa677c2a..4b85a95f633ba4 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-ceil-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-ceil-vp.ll
@@ -737,29 +737,18 @@ 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: vmv1r.v v6, v0
; CHECK-NEXT: li a2, 16
; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
-; CHECK-NEXT: vslidedown.vi v24, v0, 2
+; 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: vmv1r.v v0, v25
+; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; CHECK-NEXT: vfabs.v v16, v8, v0.t
+; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: lui a1, %hi(.LCPI26_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI26_0)(a1)
; CHECK-NEXT: addi a1, a0, -16
@@ -767,43 +756,28 @@ define <32 x double> @vp_ceil_v32f64(<32 x double> %va, <32 x i1> %m, i32 zeroex
; CHECK-NEXT: addi a0, a0, -1
; CHECK-NEXT: and a0, a0, a1
; 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, v25
+; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
-; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
+; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: fsrm a1
-; 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: addi a1, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
-; CHECK-NEXT: vmv1r.v v0, v24
-; 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: vfsgnj.vv v8, v24, v8, v0.t
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; 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, v24
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
-; 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: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 4
-; CHECK-NEXT: add sp, sp, a0
-; CHECK-NEXT: .cfi_def_cfa sp, 16
-; CHECK-NEXT: addi sp, sp, 16
-; CHECK-NEXT: .cfi_def_cfa_offset 0
+; CHECK-NEXT: vfsgnj.vv v16, v24, v16, v0.t
; CHECK-NEXT: ret
%v = call <32 x double> @llvm.vp.ceil.v32f64(<32 x double> %va, <32 x i1> %m, i32 %evl)
ret <32 x double> %v
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 9d0d42cf754c5e..4f7f5b83ae3902 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-ctlz-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-ctlz-vp.ll
@@ -2050,17 +2050,17 @@ define <32 x i64> @vp_ctlz_v32i64(<32 x i64> %va, <32 x i1> %m, 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, 56
+; RV32-NEXT: li a2, 48
; RV32-NEXT: mul a1, a1, a2
; 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: .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: vsetivli zero, 2, e8, mf4, ta, ma
-; RV32-NEXT: vslidedown.vi v24, v0, 2
+; RV32-NEXT: vslidedown.vi v7, v0, 2
; RV32-NEXT: lui a1, 349525
; RV32-NEXT: lui a2, 209715
; RV32-NEXT: addi a1, a1, 1365
@@ -2108,129 +2108,57 @@ define <32 x i64> @vp_ctlz_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %evl
; RV32-NEXT: addi a3, a3, 48
; RV32-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
-; RV32-NEXT: vnot.v v16, 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 v16, (a3) # Unknown-size Folded Spill
+; RV32-NEXT: vnot.v v8, v8, v0.t
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v8, (a4), 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: vsetvli zero, a2, e64, m8, ta, ma
-; RV32-NEXT: vsrl.vi v8, v16, 1, v0.t
+; RV32-NEXT: vlse64.v v16, (a4), 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 v8, (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, 48
-; RV32-NEXT: vl8r.v v16, (a3) # Unknown-size Folded Reload
-; 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: vl8r.v v8, (a3) # Unknown-size Folded Reload
-; 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: vl8r.v v16, (a3) # Unknown-size Folded Reload
-; RV32-NEXT: vsub.vv v16, v16, 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 v16, (a3) # Unknown-size Folded Spill
+; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vi v16, v8, 1, 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: 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: li a4, 24
+; 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 v16, (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, 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: vl8r.v v24, (a3) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v16, v16, v24, v0.t
+; RV32-NEXT: vsub.vv v8, v8, v16, 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: vl8r.v v8, (a3) # Unknown-size Folded Reload
-; RV32-NEXT: vadd.vv v8, 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: vl8r.v v24, (a3) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v16, v8, v24, v0.t
+; RV32-NEXT: vsrl.vi v8, v8, 2, v0.t
+; RV32-NEXT: vand.vv v8, v8, v24, v0.t
+; RV32-NEXT: vadd.vv v8, v16, v8, v0.t
; 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 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: vlse64.v v16, (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: vs8r.v v8, (a3) # Unknown-size Folded Spill
-; 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 v8, (a3) # Unknown-size Folded Reload
+; RV32-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
+; RV32-NEXT: vlse64.v v24, (a4), zero
+; RV32-NEXT: addi a3, sp, 48
+; RV32-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
; 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: vadd.vv v8, v8, v16, 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: vl8r.v v16, (a2) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v8, v8, v16, v0.t
+; RV32-NEXT: vmul.vv v8, v8, v24, v0.t
; RV32-NEXT: li a2, 56
; RV32-NEXT: vsrl.vx v8, v8, a2, v0.t
; RV32-NEXT: csrr a3, vlenb
@@ -2242,7 +2170,7 @@ 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: vmv1r.v v0, v24
+; RV32-NEXT: vmv1r.v v0, v7
; RV32-NEXT: csrr a3, vlenb
; RV32-NEXT: slli a3, a3, 4
; RV32-NEXT: add a3, sp, a3
@@ -2262,86 +2190,34 @@ define <32 x i64> @vp_ctlz_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %evl
; 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: addi a0, sp, 48
-; 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, 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: vsrl.vi v16, v8, 1, 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: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; 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 v8, (a0) # Unknown-size Folded Reload
-; RV32-NEXT: vand.vv v16, v8, v16, v0.t
-; RV32-NEXT: addi a0, sp, 48
-; RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v16, v16, v24, v0.t
; 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, v16, v8, 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: vs8r.v v16, (a0) # Unknown-size Folded Spill
-; 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 v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v16, v8, v24, v0.t
; RV32-NEXT: vsrl.vi v8, v8, 2, 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, 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, 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, v24, v0.t
; 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: csrr a0, vlenb
-; RV32-NEXT: li a1, 24
-; RV32-NEXT: mul a0, a0, a1
+; 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: 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: addi a0, sp, 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
@@ -2351,7 +2227,7 @@ define <32 x i64> @vp_ctlz_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %evl
; 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, 56
+; RV32-NEXT: li a1, 48
; RV32-NEXT: mul a0, a0, a1
; RV32-NEXT: add sp, sp, a0
; RV32-NEXT: .cfi_def_cfa sp, 48
@@ -2361,27 +2237,16 @@ define <32 x i64> @vp_ctlz_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %evl
;
; RV64-LABEL: vp_ctlz_v32i64:
; RV64: # %bb.0:
-; RV64-NEXT: addi sp, sp, -16
-; RV64-NEXT: .cfi_def_cfa_offset 16
-; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: slli a1, a1, 4
-; RV64-NEXT: sub sp, sp, a1
-; 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 a1, vlenb
-; RV64-NEXT: slli a1, a1, 3
-; RV64-NEXT: add a1, sp, a1
-; RV64-NEXT: addi a1, a1, 16
-; RV64-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
; RV64-NEXT: li a2, 16
; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
-; RV64-NEXT: vslidedown.vi v24, v0, 2
+; RV64-NEXT: vslidedown.vi v7, v0, 2
; RV64-NEXT: mv a1, a0
; RV64-NEXT: bltu a0, a2, .LBB34_2
; RV64-NEXT: # %bb.1:
; RV64-NEXT: li a1, 16
; RV64-NEXT: .LBB34_2:
; RV64-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; RV64-NEXT: vsrl.vi v16, v8, 1, v0.t
+; RV64-NEXT: vsrl.vi v24, v8, 1, v0.t
; RV64-NEXT: li a1, 32
; RV64-NEXT: lui a2, 349525
; RV64-NEXT: lui a3, 209715
@@ -2404,72 +2269,57 @@ define <32 x i64> @vp_ctlz_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %evl
; RV64-NEXT: addi a0, a0, -1
; RV64-NEXT: and a6, a0, a6
; RV64-NEXT: li a0, 56
-; RV64-NEXT: vor.vv v8, v8, v16, v0.t
-; RV64-NEXT: vsrl.vi v16, v8, 2, v0.t
-; RV64-NEXT: vor.vv v8, v8, v16, v0.t
-; RV64-NEXT: vsrl.vi v16, v8, 4, v0.t
-; RV64-NEXT: vor.vv v8, v8, v16, v0.t
-; RV64-NEXT: vsrl.vi v16, v8, 8, v0.t
-; RV64-NEXT: vor.vv v8, v8, v16, v0.t
-; RV64-NEXT: vsrl.vi v16, v8, 16, v0.t
-; RV64-NEXT: vor.vv v8, v8, v16, v0.t
-; RV64-NEXT: vsrl.vx v16, v8, a1, v0.t
-; RV64-NEXT: vor.vv v8, v8, v16, v0.t
+; RV64-NEXT: vor.vv v8, v8, v24, v0.t
+; RV64-NEXT: vsrl.vi v24, v8, 2, v0.t
+; RV64-NEXT: vor.vv v8, v8, v24, v0.t
+; RV64-NEXT: vsrl.vi v24, v8, 4, v0.t
+; RV64-NEXT: vor.vv v8, v8, v24, v0.t
+; RV64-NEXT: vsrl.vi v24, v8, 8, v0.t
+; RV64-NEXT: vor.vv v8, v8, v24, v0.t
+; RV64-NEXT: vsrl.vi v24, v8, 16, v0.t
+; RV64-NEXT: vor.vv v8, v8, v24, v0.t
+; RV64-NEXT: vsrl.vx v24, v8, a1, v0.t
+; RV64-NEXT: vor.vv v8, v8, v24, v0.t
; RV64-NEXT: vnot.v v8, v8, v0.t
-; RV64-NEXT: vsrl.vi v16, v8, 1, v0.t
-; RV64-NEXT: vand.vx v16, v16, a5, v0.t
-; RV64-NEXT: vsub.vv v8, v8, v16, v0.t
-; RV64-NEXT: vand.vx v16, v8, a4, v0.t
+; RV64-NEXT: vsrl.vi v24, v8, 1, v0.t
+; RV64-NEXT: vand.vx v24, v24, a5, v0.t
+; RV64-NEXT: vsub.vv v8, v8, v24, v0.t
+; RV64-NEXT: vand.vx v24, v8, a4, v0.t
; RV64-NEXT: vsrl.vi v8, v8, 2, v0.t
; RV64-NEXT: vand.vx v8, v8, a4, v0.t
-; RV64-NEXT: vadd.vv v8, v16, v8, v0.t
-; RV64-NEXT: vsrl.vi v16, v8, 4, v0.t
-; RV64-NEXT: vadd.vv v8, v8, v16, v0.t
+; RV64-NEXT: vadd.vv v8, v24, v8, v0.t
+; RV64-NEXT: vsrl.vi v24, v8, 4, v0.t
+; RV64-NEXT: vadd.vv v8, v8, v24, v0.t
; RV64-NEXT: vand.vx v8, v8, a2, v0.t
; RV64-NEXT: vmul.vx v8, v8, a3, v0.t
; RV64-NEXT: vsrl.vx v8, v8, a0, v0.t
-; RV64-NEXT: addi a7, sp, 16
-; RV64-NEXT: vs8r.v v8, (a7) # Unknown-size Folded Spill
-; RV64-NEXT: vmv1r.v v0, v24
-; 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: vmv1r.v v0, v7
; RV64-NEXT: vsetvli zero, a6, 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
-; RV64-NEXT: vor.vv v8, v16, v8, v0.t
-; RV64-NEXT: vsrl.vi v16, v8, 4, v0.t
-; RV64-NEXT: vor.vv v8, v8, v16, v0.t
-; RV64-NEXT: vsrl.vi v16, v8, 8, v0.t
-; RV64-NEXT: vor.vv v8, v8, v16, v0.t
-; RV64-NEXT: vsrl.vi v16, v8, 16, v0.t
-; RV64-NEXT: vor.vv v8, v8, v16, v0.t
-; RV64-NEXT: vsrl.vx v16, v8, a1, v0.t
-; RV64-NEXT: vor.vv v8, v8, v16, v0.t
-; RV64-NEXT: vnot.v v8, v8, v0.t
-; RV64-NEXT: vsrl.vi v16, v8, 1, v0.t
-; RV64-NEXT: vand.vx v16, v16, a5, v0.t
-; RV64-NEXT: vsub.vv v8, v8, v16, v0.t
-; RV64-NEXT: vand.vx v16, v8, a4, v0.t
-; RV64-NEXT: vsrl.vi v8, v8, 2, v0.t
-; RV64-NEXT: vand.vx v8, v8, a4, v0.t
-; RV64-NEXT: vadd.vv v8, v16, v8, v0.t
-; RV64-NEXT: vsrl.vi v16, v8, 4, v0.t
-; RV64-NEXT: vadd.vv v8, v8, v16, v0.t
-; RV64-NEXT: vand.vx v8, v8, a2, v0.t
-; RV64-NEXT: vmul.vx v8, v8, a3, v0.t
-; RV64-NEXT: vsrl.vx v16, v8, a0, v0.t
-; RV64-NEXT: addi a0, sp, 16
-; RV64-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; RV64-NEXT: csrr a0, vlenb
-; RV64-NEXT: slli a0, a0, 4
-; RV64-NEXT: add sp, sp, a0
-; RV64-NEXT: .cfi_def_cfa sp, 16
-; RV64-NEXT: addi sp, sp, 16
-; RV64-NEXT: .cfi_def_cfa_offset 0
+; RV64-NEXT: vsrl.vi v24, v16, 1, v0.t
+; RV64-NEXT: vor.vv v16, v16, v24, v0.t
+; RV64-NEXT: vsrl.vi v24, v16, 2, v0.t
+; RV64-NEXT: vor.vv v16, v16, v24, v0.t
+; RV64-NEXT: vsrl.vi v24, v16, 4, v0.t
+; RV64-NEXT: vor.vv v16, v16, v24, v0.t
+; RV64-NEXT: vsrl.vi v24, v16, 8, v0.t
+; RV64-NEXT: vor.vv v16, v16, v24, v0.t
+; RV64-NEXT: vsrl.vi v24, v16, 16, v0.t
+; RV64-NEXT: vor.vv v16, v16, v24, v0.t
+; RV64-NEXT: vsrl.vx v24, v16, a1, v0.t
+; RV64-NEXT: vor.vv v16, v16, v24, v0.t
+; RV64-NEXT: vnot.v v16, v16, v0.t
+; RV64-NEXT: vsrl.vi v24, v16, 1, v0.t
+; RV64-NEXT: vand.vx v24, v24, a5, v0.t
+; RV64-NEXT: vsub.vv v16, v16, v24, v0.t
+; RV64-NEXT: vand.vx v24, v16, a4, v0.t
+; RV64-NEXT: vsrl.vi v16, v16, 2, v0.t
+; RV64-NEXT: vand.vx v16, v16, a4, v0.t
+; RV64-NEXT: vadd.vv v16, v24, v16, v0.t
+; RV64-NEXT: vsrl.vi v24, v16, 4, v0.t
+; RV64-NEXT: vadd.vv v16, v16, v24, v0.t
+; RV64-NEXT: vand.vx v16, v16, a2, v0.t
+; RV64-NEXT: vmul.vx v16, v16, a3, v0.t
+; RV64-NEXT: vsrl.vx v16, v16, a0, v0.t
; RV64-NEXT: ret
%v = call <32 x i64> @llvm.vp.ctlz.v32i64(<32 x i64> %va, i1 false, <32 x i1> %m, i32 %evl)
ret <32 x i64> %v
@@ -4756,17 +4606,17 @@ define <32 x i64> @vp_ctlz_zero_undef_v32i64(<32 x i64> %va, <32 x i1> %m, i32 z
; RV32-NEXT: addi sp, sp, -48
; RV32-NEXT: .cfi_def_cfa_offset 48
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a2, 56
+; RV32-NEXT: li a2, 48
; RV32-NEXT: mul a1, a1, a2
; 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: .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: vsetivli zero, 2, e8, mf4, ta, ma
-; RV32-NEXT: vslidedown.vi v24, v0, 2
+; RV32-NEXT: vslidedown.vi v7, v0, 2
; RV32-NEXT: lui a1, 349525
; RV32-NEXT: lui a2, 209715
; RV32-NEXT: addi a1, a1, 1365
@@ -4814,129 +4664,57 @@ define <32 x i64> @vp_ctlz_zero_undef_v32i64(<32 x i64> %va, <32 x i1> %m, i32 z
; RV32-NEXT: addi a3, a3, 48
; RV32-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
-; RV32-NEXT: vnot.v v16, 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 v16, (a3) # Unknown-size Folded Spill
+; RV32-NEXT: vnot.v v8, v8, v0.t
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v8, (a4), 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: vsetvli zero, a2, e64, m8, ta, ma
-; RV32-NEXT: vsrl.vi v8, v16, 1, v0.t
+; RV32-NEXT: vlse64.v v16, (a4), 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 v8, (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, 48
-; RV32-NEXT: vl8r.v v16, (a3) # Unknown-size Folded Reload
-; 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: vl8r.v v8, (a3) # Unknown-size Folded Reload
-; 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: vl8r.v v16, (a3) # Unknown-size Folded Reload
-; RV32-NEXT: vsub.vv v16, v16, 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 v16, (a3) # Unknown-size Folded Spill
+; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
+; RV32-NEXT: vsrl.vi v16, v8, 1, 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: 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: li a4, 24
+; 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 v16, (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, 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: vl8r.v v24, (a3) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v16, v16, v24, v0.t
+; RV32-NEXT: vsub.vv v8, v8, v16, 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: vl8r.v v8, (a3) # Unknown-size Folded Reload
-; RV32-NEXT: vadd.vv v8, 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: vl8r.v v24, (a3) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v16, v8, v24, v0.t
+; RV32-NEXT: vsrl.vi v8, v8, 2, v0.t
+; RV32-NEXT: vand.vv v8, v8, v24, v0.t
+; RV32-NEXT: vadd.vv v8, v16, v8, v0.t
; 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 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: vlse64.v v16, (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: vs8r.v v8, (a3) # Unknown-size Folded Spill
-; 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 v8, (a3) # Unknown-size Folded Reload
+; RV32-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
+; RV32-NEXT: vlse64.v v24, (a4), zero
+; RV32-NEXT: addi a3, sp, 48
+; RV32-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
; 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: vadd.vv v8, v8, v16, 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: vl8r.v v16, (a2) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v8, v8, v16, v0.t
+; RV32-NEXT: vmul.vv v8, v8, v24, v0.t
; RV32-NEXT: li a2, 56
; RV32-NEXT: vsrl.vx v8, v8, a2, v0.t
; RV32-NEXT: csrr a3, vlenb
@@ -4948,7 +4726,7 @@ 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: vmv1r.v v0, v24
+; RV32-NEXT: vmv1r.v v0, v7
; RV32-NEXT: csrr a3, vlenb
; RV32-NEXT: slli a3, a3, 4
; RV32-NEXT: add a3, sp, a3
@@ -4968,86 +4746,34 @@ define <32 x i64> @vp_ctlz_zero_undef_v32i64(<32 x i64> %va, <32 x i1> %m, i32 z
; 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: addi a0, sp, 48
-; 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, 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: vsrl.vi v16, v8, 1, 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: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; 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 v8, (a0) # Unknown-size Folded Reload
-; RV32-NEXT: vand.vv v16, v8, v16, v0.t
-; RV32-NEXT: addi a0, sp, 48
-; RV32-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v16, v16, v24, v0.t
; 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, v16, v8, 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: vs8r.v v16, (a0) # Unknown-size Folded Spill
-; 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 v8, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v16, v8, v24, v0.t
; RV32-NEXT: vsrl.vi v8, v8, 2, 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, 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, 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, v24, v0.t
; 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: csrr a0, vlenb
-; RV32-NEXT: li a1, 24
-; RV32-NEXT: mul a0, a0, a1
+; 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: 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: addi a0, sp, 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
@@ -5057,7 +4783,7 @@ define <32 x i64> @vp_ctlz_zero_undef_v32i64(<32 x i64> %va, <32 x i1> %m, i32 z
; 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, 56
+; RV32-NEXT: li a1, 48
; RV32-NEXT: mul a0, a0, a1
; RV32-NEXT: add sp, sp, a0
; RV32-NEXT: .cfi_def_cfa sp, 48
@@ -5067,27 +4793,16 @@ define <32 x i64> @vp_ctlz_zero_undef_v32i64(<32 x i64> %va, <32 x i1> %m, i32 z
;
; RV64-LABEL: vp_ctlz_zero_undef_v32i64:
; RV64: # %bb.0:
-; RV64-NEXT: addi sp, sp, -16
-; RV64-NEXT: .cfi_def_cfa_offset 16
-; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: slli a1, a1, 4
-; RV64-NEXT: sub sp, sp, a1
-; 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 a1, vlenb
-; RV64-NEXT: slli a1, a1, 3
-; RV64-NEXT: add a1, sp, a1
-; RV64-NEXT: addi a1, a1, 16
-; RV64-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
; RV64-NEXT: li a2, 16
; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
-; RV64-NEXT: vslidedown.vi v24, v0, 2
+; RV64-NEXT: vslidedown.vi v7, v0, 2
; RV64-NEXT: mv a1, a0
; RV64-NEXT: bltu a0, a2, .LBB70_2
; RV64-NEXT: # %bb.1:
; RV64-NEXT: li a1, 16
; RV64-NEXT: .LBB70_2:
; RV64-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; RV64-NEXT: vsrl.vi v16, v8, 1, v0.t
+; RV64-NEXT: vsrl.vi v24, v8, 1, v0.t
; RV64-NEXT: li a1, 32
; RV64-NEXT: lui a2, 349525
; RV64-NEXT: lui a3, 209715
@@ -5110,72 +4825,57 @@ define <32 x i64> @vp_ctlz_zero_undef_v32i64(<32 x i64> %va, <32 x i1> %m, i32 z
; RV64-NEXT: addi a0, a0, -1
; RV64-NEXT: and a6, a0, a6
; RV64-NEXT: li a0, 56
-; RV64-NEXT: vor.vv v8, v8, v16, v0.t
-; RV64-NEXT: vsrl.vi v16, v8, 2, v0.t
-; RV64-NEXT: vor.vv v8, v8, v16, v0.t
-; RV64-NEXT: vsrl.vi v16, v8, 4, v0.t
-; RV64-NEXT: vor.vv v8, v8, v16, v0.t
-; RV64-NEXT: vsrl.vi v16, v8, 8, v0.t
-; RV64-NEXT: vor.vv v8, v8, v16, v0.t
-; RV64-NEXT: vsrl.vi v16, v8, 16, v0.t
-; RV64-NEXT: vor.vv v8, v8, v16, v0.t
-; RV64-NEXT: vsrl.vx v16, v8, a1, v0.t
-; RV64-NEXT: vor.vv v8, v8, v16, v0.t
+; RV64-NEXT: vor.vv v8, v8, v24, v0.t
+; RV64-NEXT: vsrl.vi v24, v8, 2, v0.t
+; RV64-NEXT: vor.vv v8, v8, v24, v0.t
+; RV64-NEXT: vsrl.vi v24, v8, 4, v0.t
+; RV64-NEXT: vor.vv v8, v8, v24, v0.t
+; RV64-NEXT: vsrl.vi v24, v8, 8, v0.t
+; RV64-NEXT: vor.vv v8, v8, v24, v0.t
+; RV64-NEXT: vsrl.vi v24, v8, 16, v0.t
+; RV64-NEXT: vor.vv v8, v8, v24, v0.t
+; RV64-NEXT: vsrl.vx v24, v8, a1, v0.t
+; RV64-NEXT: vor.vv v8, v8, v24, v0.t
; RV64-NEXT: vnot.v v8, v8, v0.t
-; RV64-NEXT: vsrl.vi v16, v8, 1, v0.t
-; RV64-NEXT: vand.vx v16, v16, a5, v0.t
-; RV64-NEXT: vsub.vv v8, v8, v16, v0.t
-; RV64-NEXT: vand.vx v16, v8, a4, v0.t
+; RV64-NEXT: vsrl.vi v24, v8, 1, v0.t
+; RV64-NEXT: vand.vx v24, v24, a5, v0.t
+; RV64-NEXT: vsub.vv v8, v8, v24, v0.t
+; RV64-NEXT: vand.vx v24, v8, a4, v0.t
; RV64-NEXT: vsrl.vi v8, v8, 2, v0.t
; RV64-NEXT: vand.vx v8, v8, a4, v0.t
-; RV64-NEXT: vadd.vv v8, v16, v8, v0.t
-; RV64-NEXT: vsrl.vi v16, v8, 4, v0.t
-; RV64-NEXT: vadd.vv v8, v8, v16, v0.t
+; RV64-NEXT: vadd.vv v8, v24, v8, v0.t
+; RV64-NEXT: vsrl.vi v24, v8, 4, v0.t
+; RV64-NEXT: vadd.vv v8, v8, v24, v0.t
; RV64-NEXT: vand.vx v8, v8, a2, v0.t
; RV64-NEXT: vmul.vx v8, v8, a3, v0.t
; RV64-NEXT: vsrl.vx v8, v8, a0, v0.t
-; RV64-NEXT: addi a7, sp, 16
-; RV64-NEXT: vs8r.v v8, (a7) # Unknown-size Folded Spill
-; RV64-NEXT: vmv1r.v v0, v24
-; 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: vmv1r.v v0, v7
; RV64-NEXT: vsetvli zero, a6, 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
-; RV64-NEXT: vor.vv v8, v16, v8, v0.t
-; RV64-NEXT: vsrl.vi v16, v8, 4, v0.t
-; RV64-NEXT: vor.vv v8, v8, v16, v0.t
-; RV64-NEXT: vsrl.vi v16, v8, 8, v0.t
-; RV64-NEXT: vor.vv v8, v8, v16, v0.t
-; RV64-NEXT: vsrl.vi v16, v8, 16, v0.t
-; RV64-NEXT: vor.vv v8, v8, v16, v0.t
-; RV64-NEXT: vsrl.vx v16, v8, a1, v0.t
-; RV64-NEXT: vor.vv v8, v8, v16, v0.t
-; RV64-NEXT: vnot.v v8, v8, v0.t
-; RV64-NEXT: vsrl.vi v16, v8, 1, v0.t
-; RV64-NEXT: vand.vx v16, v16, a5, v0.t
-; RV64-NEXT: vsub.vv v8, v8, v16, v0.t
-; RV64-NEXT: vand.vx v16, v8, a4, v0.t
-; RV64-NEXT: vsrl.vi v8, v8, 2, v0.t
-; RV64-NEXT: vand.vx v8, v8, a4, v0.t
-; RV64-NEXT: vadd.vv v8, v16, v8, v0.t
-; RV64-NEXT: vsrl.vi v16, v8, 4, v0.t
-; RV64-NEXT: vadd.vv v8, v8, v16, v0.t
-; RV64-NEXT: vand.vx v8, v8, a2, v0.t
-; RV64-NEXT: vmul.vx v8, v8, a3, v0.t
-; RV64-NEXT: vsrl.vx v16, v8, a0, v0.t
-; RV64-NEXT: addi a0, sp, 16
-; RV64-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; RV64-NEXT: csrr a0, vlenb
-; RV64-NEXT: slli a0, a0, 4
-; RV64-NEXT: add sp, sp, a0
-; RV64-NEXT: .cfi_def_cfa sp, 16
-; RV64-NEXT: addi sp, sp, 16
-; RV64-NEXT: .cfi_def_cfa_offset 0
+; RV64-NEXT: vsrl.vi v24, v16, 1, v0.t
+; RV64-NEXT: vor.vv v16, v16, v24, v0.t
+; RV64-NEXT: vsrl.vi v24, v16, 2, v0.t
+; RV64-NEXT: vor.vv v16, v16, v24, v0.t
+; RV64-NEXT: vsrl.vi v24, v16, 4, v0.t
+; RV64-NEXT: vor.vv v16, v16, v24, v0.t
+; RV64-NEXT: vsrl.vi v24, v16, 8, v0.t
+; RV64-NEXT: vor.vv v16, v16, v24, v0.t
+; RV64-NEXT: vsrl.vi v24, v16, 16, v0.t
+; RV64-NEXT: vor.vv v16, v16, v24, v0.t
+; RV64-NEXT: vsrl.vx v24, v16, a1, v0.t
+; RV64-NEXT: vor.vv v16, v16, v24, v0.t
+; RV64-NEXT: vnot.v v16, v16, v0.t
+; RV64-NEXT: vsrl.vi v24, v16, 1, v0.t
+; RV64-NEXT: vand.vx v24, v24, a5, v0.t
+; RV64-NEXT: vsub.vv v16, v16, v24, v0.t
+; RV64-NEXT: vand.vx v24, v16, a4, v0.t
+; RV64-NEXT: vsrl.vi v16, v16, 2, v0.t
+; RV64-NEXT: vand.vx v16, v16, a4, v0.t
+; RV64-NEXT: vadd.vv v16, v24, v16, v0.t
+; RV64-NEXT: vsrl.vi v24, v16, 4, v0.t
+; RV64-NEXT: vadd.vv v16, v16, v24, v0.t
+; RV64-NEXT: vand.vx v16, v16, a2, v0.t
+; RV64-NEXT: vmul.vx v16, v16, a3, v0.t
+; RV64-NEXT: vsrl.vx v16, v16, a0, v0.t
; RV64-NEXT: ret
%v = call <32 x i64> @llvm.vp.ctlz.v32i64(<32 x i64> %va, i1 true, <32 x i1> %m, i32 %evl)
ret <32 x i64> %v
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 5e73e6df9170c2..82f6deb85c44d6 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-ctpop-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-ctpop-vp.ll
@@ -1565,75 +1565,53 @@ define <32 x i64> @vp_ctpop_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %ev
; RV32-NEXT: addi a2, sp, 32
; RV32-NEXT: vlse64.v v16, (a2), zero
; RV32-NEXT: csrr a2, vlenb
-; RV32-NEXT: slli a2, a2, 5
+; 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 v16, (a2) # Unknown-size Folded Spill
; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; RV32-NEXT: vsrl.vi v24, v8, 1, v0.t
+; RV32-NEXT: vsrl.vi v16, v8, 1, 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: vand.vv v16, v24, v16, v0.t
+; RV32-NEXT: vl8r.v v24, (a2) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v16, v16, v24, v0.t
; RV32-NEXT: vsub.vv v8, v8, v16, 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 v16, (a2) # Unknown-size Folded Reload
-; RV32-NEXT: vand.vv v24, 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: vs8r.v v24, (a2) # Unknown-size Folded Spill
+; RV32-NEXT: vl8r.v v24, (a2) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v16, v8, v24, v0.t
; RV32-NEXT: vsrl.vi v8, v8, 2, v0.t
-; RV32-NEXT: vand.vv v8, 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 v16, (a2) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v8, v8, v24, v0.t
; 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, 24
-; 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: 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: vs8r.v v16, (a2) # Unknown-size Folded Spill
+; RV32-NEXT: vlse64.v v24, (a3), zero
+; RV32-NEXT: addi a2, sp, 48
+; RV32-NEXT: vs8r.v v24, (a2) # Unknown-size Folded Spill
; RV32-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; 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: vsrl.vi v16, v8, 4, v0.t
+; RV32-NEXT: vadd.vv v8, v8, v16, v0.t
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a2, 24
-; RV32-NEXT: mul a1, a1, a2
+; RV32-NEXT: slli a1, a1, 5
; 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: vl8r.v v16, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v8, v8, v16, v0.t
+; RV32-NEXT: vmul.vv v8, v8, v24, v0.t
; RV32-NEXT: li a1, 56
; RV32-NEXT: vsrl.vx v8, v8, a1, v0.t
; RV32-NEXT: csrr a2, vlenb
@@ -1652,35 +1630,35 @@ define <32 x i64> @vp_ctpop_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %ev
; 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: vsrl.vi v8, v16, 1, 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 v8, (a0) # Unknown-size Folded Reload
-; RV32-NEXT: vand.vv v8, v24, v8, v0.t
+; RV32-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v8, v8, v24, v0.t
; RV32-NEXT: vsub.vv v8, v16, v8, v0.t
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 5
+; 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, v8, v16, v0.t
+; RV32-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v16, v8, v24, 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: vand.vv v8, v8, v24, v0.t
+; 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, 24
-; RV32-NEXT: mul a0, a0, a2
+; 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: vand.vv v8, v8, v16, v0.t
+; RV32-NEXT: addi a0, sp, 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
@@ -1699,27 +1677,16 @@ define <32 x i64> @vp_ctpop_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %ev
;
; RV64-LABEL: vp_ctpop_v32i64:
; RV64: # %bb.0:
-; RV64-NEXT: addi sp, sp, -16
-; RV64-NEXT: .cfi_def_cfa_offset 16
-; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: slli a1, a1, 4
-; RV64-NEXT: sub sp, sp, a1
-; 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 a1, vlenb
-; RV64-NEXT: slli a1, a1, 3
-; RV64-NEXT: add a1, sp, a1
-; RV64-NEXT: addi a1, a1, 16
-; RV64-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
; RV64-NEXT: li a2, 16
; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
-; RV64-NEXT: vslidedown.vi v24, v0, 2
+; RV64-NEXT: vslidedown.vi v7, v0, 2
; RV64-NEXT: mv a1, a0
; RV64-NEXT: bltu a0, a2, .LBB34_2
; RV64-NEXT: # %bb.1:
; RV64-NEXT: li a1, 16
; RV64-NEXT: .LBB34_2:
; RV64-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; RV64-NEXT: vsrl.vi v16, v8, 1, v0.t
+; RV64-NEXT: vsrl.vi v24, v8, 1, v0.t
; RV64-NEXT: lui a1, 349525
; RV64-NEXT: lui a2, 209715
; RV64-NEXT: lui a3, 61681
@@ -1729,58 +1696,43 @@ define <32 x i64> @vp_ctpop_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %ev
; RV64-NEXT: addiw a3, a3, -241
; RV64-NEXT: addiw a4, a4, 257
; RV64-NEXT: slli a5, a1, 32
-; RV64-NEXT: add a5, a1, a5
-; RV64-NEXT: slli a1, a2, 32
-; RV64-NEXT: add a6, a2, a1
-; RV64-NEXT: slli a1, a3, 32
-; RV64-NEXT: add a1, a3, a1
-; RV64-NEXT: slli a2, a4, 32
-; RV64-NEXT: add a2, a4, a2
-; RV64-NEXT: addi a3, a0, -16
-; RV64-NEXT: sltu a0, a0, a3
+; RV64-NEXT: add a1, a1, a5
+; RV64-NEXT: slli a5, a2, 32
+; RV64-NEXT: add a2, a2, a5
+; RV64-NEXT: slli a5, a3, 32
+; RV64-NEXT: add a3, a3, a5
+; RV64-NEXT: slli a5, a4, 32
+; RV64-NEXT: add a4, a4, a5
+; RV64-NEXT: addi a5, a0, -16
+; RV64-NEXT: sltu a0, a0, a5
; RV64-NEXT: addi a0, a0, -1
-; RV64-NEXT: and a0, a0, a3
-; RV64-NEXT: li a3, 56
-; RV64-NEXT: vand.vx v16, v16, a5, v0.t
-; RV64-NEXT: vsub.vv v8, v8, v16, v0.t
-; RV64-NEXT: vand.vx v16, v8, a6, v0.t
+; RV64-NEXT: and a0, a0, a5
+; RV64-NEXT: li a5, 56
+; RV64-NEXT: vand.vx v24, v24, a1, v0.t
+; RV64-NEXT: vsub.vv v8, v8, v24, v0.t
+; RV64-NEXT: vand.vx v24, v8, a2, v0.t
; RV64-NEXT: vsrl.vi v8, v8, 2, v0.t
-; RV64-NEXT: vand.vx v8, v8, a6, v0.t
-; RV64-NEXT: vadd.vv v8, v16, v8, v0.t
-; RV64-NEXT: vsrl.vi v16, v8, 4, v0.t
-; RV64-NEXT: vadd.vv v8, v8, v16, v0.t
-; RV64-NEXT: vand.vx v8, v8, a1, v0.t
-; RV64-NEXT: vmul.vx v8, v8, a2, v0.t
-; RV64-NEXT: vsrl.vx v8, v8, a3, v0.t
-; RV64-NEXT: addi a4, sp, 16
-; RV64-NEXT: vs8r.v v8, (a4) # Unknown-size Folded Spill
-; RV64-NEXT: vmv1r.v v0, v24
-; 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 v8, (a4) # Unknown-size Folded Reload
+; RV64-NEXT: vand.vx v8, v8, a2, v0.t
+; RV64-NEXT: vadd.vv v8, v24, v8, v0.t
+; RV64-NEXT: vsrl.vi v24, v8, 4, v0.t
+; RV64-NEXT: vadd.vv v8, v8, v24, v0.t
+; RV64-NEXT: vand.vx v8, v8, a3, v0.t
+; RV64-NEXT: vmul.vx v8, v8, a4, v0.t
+; RV64-NEXT: vsrl.vx v8, v8, a5, v0.t
+; RV64-NEXT: vmv1r.v v0, v7
; RV64-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; RV64-NEXT: vsrl.vi v16, v8, 1, v0.t
-; RV64-NEXT: vand.vx v16, v16, a5, v0.t
-; RV64-NEXT: vsub.vv v16, v8, v16, v0.t
-; RV64-NEXT: vand.vx v8, v16, a6, v0.t
+; RV64-NEXT: vsrl.vi v24, v16, 1, v0.t
+; RV64-NEXT: vand.vx v24, v24, a1, v0.t
+; RV64-NEXT: vsub.vv v16, v16, v24, v0.t
+; RV64-NEXT: vand.vx v24, v16, a2, v0.t
; RV64-NEXT: vsrl.vi v16, v16, 2, v0.t
-; RV64-NEXT: vand.vx v16, v16, a6, v0.t
-; RV64-NEXT: vadd.vv v8, v8, v16, v0.t
-; RV64-NEXT: vsrl.vi v16, v8, 4, v0.t
-; RV64-NEXT: vadd.vv v8, v8, v16, v0.t
-; RV64-NEXT: vand.vx v8, v8, a1, v0.t
-; RV64-NEXT: vmul.vx v8, v8, a2, v0.t
-; RV64-NEXT: vsrl.vx v16, v8, a3, v0.t
-; RV64-NEXT: addi a0, sp, 16
-; RV64-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; RV64-NEXT: csrr a0, vlenb
-; RV64-NEXT: slli a0, a0, 4
-; RV64-NEXT: add sp, sp, a0
-; RV64-NEXT: .cfi_def_cfa sp, 16
-; RV64-NEXT: addi sp, sp, 16
-; RV64-NEXT: .cfi_def_cfa_offset 0
+; RV64-NEXT: vand.vx v16, v16, a2, v0.t
+; RV64-NEXT: vadd.vv v16, v24, v16, v0.t
+; RV64-NEXT: vsrl.vi v24, v16, 4, v0.t
+; RV64-NEXT: vadd.vv v16, v16, v24, v0.t
+; RV64-NEXT: vand.vx v16, v16, a3, v0.t
+; RV64-NEXT: vmul.vx v16, v16, a4, v0.t
+; RV64-NEXT: vsrl.vx v16, v16, a5, v0.t
; RV64-NEXT: ret
%v = call <32 x i64> @llvm.vp.ctpop.v32i64(<32 x i64> %va, <32 x i1> %m, i32 %evl)
ret <32 x i64> %v
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 cd4b19f11d1602..13fcaa3654f0c4 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-cttz-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-cttz-vp.ll
@@ -1769,99 +1769,64 @@ define <32 x i64> @vp_cttz_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %evl
; 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: 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, 48
-; RV32-NEXT: vs8r.v v8, (a4) # Unknown-size Folded Spill
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v8, (a3), zero
+; RV32-NEXT: vlse64.v v16, (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: vs8r.v v16, (a3) # Unknown-size Folded Spill
; RV32-NEXT: addi a3, sp, 32
-; 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: vs8r.v v8, (a3) # Unknown-size Folded Spill
+; RV32-NEXT: vlse64.v v16, (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: vl8r.v v16, (a3) # Unknown-size Folded Reload
+; RV32-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
-; RV32-NEXT: vsrl.vi v24, v16, 1, v0.t
+; RV32-NEXT: vsrl.vi v16, v8, 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, 48
-; RV32-NEXT: vl8r.v v16, (a3) # Unknown-size Folded Reload
-; RV32-NEXT: vand.vv v24, v24, v16, 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: vl8r.v v16, (a3) # Unknown-size Folded Reload
-; RV32-NEXT: vsub.vv v24, v16, v24, v0.t
-; RV32-NEXT: vand.vv v16, v24, v8, 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 v16, (a3) # 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 v24, (a3) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v16, v16, v24, v0.t
+; RV32-NEXT: vsub.vv v8, v8, v16, 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: vl8r.v v8, (a3) # Unknown-size Folded Reload
-; RV32-NEXT: vadd.vv v8, 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: vl8r.v v24, (a3) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v16, v8, v24, v0.t
+; RV32-NEXT: vsrl.vi v8, v8, 2, v0.t
+; RV32-NEXT: vand.vv v8, v8, v24, v0.t
+; RV32-NEXT: vadd.vv v8, v16, v8, v0.t
; 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 v16, (a3), zero
-; RV32-NEXT: addi a3, sp, 48
-; RV32-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: vlse64.v v8, (a4), 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 v8, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: slli a3, a3, 3
+; RV32-NEXT: slli a3, a3, 5
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
-; RV32-NEXT: vl8r.v v8, (a3) # Unknown-size Folded Reload
+; RV32-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
+; RV32-NEXT: vlse64.v v24, (a4), zero
+; RV32-NEXT: addi a3, sp, 48
+; RV32-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
-; 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: vsrl.vi v16, v8, 4, v0.t
+; RV32-NEXT: vadd.vv v8, v8, v16, v0.t
; RV32-NEXT: csrr a2, vlenb
-; RV32-NEXT: li a3, 24
-; RV32-NEXT: mul a2, a2, a3
+; 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: vl8r.v v16, (a2) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v8, v8, v16, v0.t
+; RV32-NEXT: vmul.vv v8, v8, v24, v0.t
; RV32-NEXT: li a2, 56
; RV32-NEXT: vsrl.vx v8, v8, a2, v0.t
; RV32-NEXT: csrr a3, vlenb
@@ -1883,52 +1848,35 @@ define <32 x i64> @vp_cttz_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %evl
; 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
-; RV32-NEXT: vsrl.vi v24, v8, 1, v0.t
+; RV32-NEXT: vsrl.vi v16, v8, 1, 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: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; RV32-NEXT: vand.vv v16, v24, v16, v0.t
-; RV32-NEXT: vsub.vv v24, v8, v16, v0.t
+; RV32-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v16, v16, v24, v0.t
+; RV32-NEXT: vsub.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 v8, (a0) # Unknown-size Folded Reload
-; RV32-NEXT: vand.vv v16, v24, v8, 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: vs8r.v v16, (a0) # Unknown-size Folded Spill
-; RV32-NEXT: vsrl.vi v8, v24, 2, v0.t
+; RV32-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v16, v8, v24, v0.t
+; RV32-NEXT: vsrl.vi v8, v8, 2, v0.t
+; RV32-NEXT: vand.vv v8, v8, v24, v0.t
+; 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: 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: vand.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: 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 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: vmul.vv v8, v8, v16, v0.t
; RV32-NEXT: vsrl.vx v16, v8, a2, v0.t
; RV32-NEXT: csrr a0, vlenb
@@ -1947,20 +1895,9 @@ define <32 x i64> @vp_cttz_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %evl
;
; RV64-LABEL: vp_cttz_v32i64:
; RV64: # %bb.0:
-; RV64-NEXT: addi sp, sp, -16
-; RV64-NEXT: .cfi_def_cfa_offset 16
-; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: slli a1, a1, 4
-; RV64-NEXT: sub sp, sp, a1
-; 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 a1, vlenb
-; RV64-NEXT: slli a1, a1, 3
-; RV64-NEXT: add a1, sp, a1
-; RV64-NEXT: addi a1, a1, 16
-; RV64-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
; RV64-NEXT: li a1, 16
; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
-; RV64-NEXT: vslidedown.vi v24, v0, 2
+; RV64-NEXT: vslidedown.vi v7, v0, 2
; RV64-NEXT: mv a4, a0
; RV64-NEXT: bltu a0, a1, .LBB34_2
; RV64-NEXT: # %bb.1:
@@ -1989,53 +1926,38 @@ define <32 x i64> @vp_cttz_v32i64(<32 x i64> %va, <32 x i1> %m, i32 zeroext %evl
; RV64-NEXT: and a7, a0, a7
; RV64-NEXT: li a0, 56
; RV64-NEXT: vsetvli zero, a4, e64, m8, ta, ma
-; RV64-NEXT: vsub.vx v16, v8, a1, v0.t
+; RV64-NEXT: vsub.vx v24, v8, a1, v0.t
; RV64-NEXT: vnot.v v8, v8, v0.t
-; RV64-NEXT: vand.vv v8, v8, v16, v0.t
-; RV64-NEXT: vsrl.vi v16, v8, 1, v0.t
-; RV64-NEXT: vand.vx v16, v16, a6, v0.t
-; RV64-NEXT: vsub.vv v8, v8, v16, v0.t
-; RV64-NEXT: vand.vx v16, v8, a5, v0.t
+; RV64-NEXT: vand.vv v8, v8, v24, v0.t
+; RV64-NEXT: vsrl.vi v24, v8, 1, v0.t
+; RV64-NEXT: vand.vx v24, v24, a6, v0.t
+; RV64-NEXT: vsub.vv v8, v8, v24, v0.t
+; RV64-NEXT: vand.vx v24, v8, a5, v0.t
; RV64-NEXT: vsrl.vi v8, v8, 2, v0.t
; RV64-NEXT: vand.vx v8, v8, a5, v0.t
-; RV64-NEXT: vadd.vv v8, v16, v8, v0.t
-; RV64-NEXT: vsrl.vi v16, v8, 4, v0.t
-; RV64-NEXT: vadd.vv v8, v8, v16, v0.t
+; RV64-NEXT: vadd.vv v8, v24, v8, v0.t
+; RV64-NEXT: vsrl.vi v24, v8, 4, v0.t
+; RV64-NEXT: vadd.vv v8, v8, v24, v0.t
; RV64-NEXT: vand.vx v8, v8, a2, v0.t
; RV64-NEXT: vmul.vx v8, v8, a3, v0.t
; RV64-NEXT: vsrl.vx v8, v8, a0, v0.t
-; RV64-NEXT: addi a4, sp, 16
-; RV64-NEXT: vs8r.v v8, (a4) # Unknown-size Folded Spill
-; RV64-NEXT: vmv1r.v v0, v24
-; 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 v8, (a4) # Unknown-size Folded Reload
+; RV64-NEXT: vmv1r.v v0, v7
; RV64-NEXT: vsetvli zero, a7, 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
-; RV64-NEXT: vsrl.vi v16, v8, 1, v0.t
-; RV64-NEXT: vand.vx v16, v16, a6, v0.t
-; RV64-NEXT: vsub.vv v8, v8, v16, v0.t
-; RV64-NEXT: vand.vx v16, v8, a5, v0.t
-; RV64-NEXT: vsrl.vi v8, v8, 2, v0.t
-; RV64-NEXT: vand.vx v8, v8, a5, v0.t
-; RV64-NEXT: vadd.vv v8, v16, v8, v0.t
-; RV64-NEXT: vsrl.vi v16, v8, 4, v0.t
-; RV64-NEXT: vadd.vv v8, v8, v16, v0.t
-; RV64-NEXT: vand.vx v8, v8, a2, v0.t
-; RV64-NEXT: vmul.vx v8, v8, a3, v0.t
-; RV64-NEXT: vsrl.vx v16, v8, a0, v0.t
-; RV64-NEXT: addi a0, sp, 16
-; RV64-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; RV64-NEXT: csrr a0, vlenb
-; RV64-NEXT: slli a0, a0, 4
-; RV64-NEXT: add sp, sp, a0
-; RV64-NEXT: .cfi_def_cfa sp, 16
-; RV64-NEXT: addi sp, sp, 16
-; RV64-NEXT: .cfi_def_cfa_offset 0
+; RV64-NEXT: vsub.vx v24, v16, a1, v0.t
+; RV64-NEXT: vnot.v v16, v16, v0.t
+; RV64-NEXT: vand.vv v16, v16, v24, v0.t
+; RV64-NEXT: vsrl.vi v24, v16, 1, v0.t
+; RV64-NEXT: vand.vx v24, v24, a6, v0.t
+; RV64-NEXT: vsub.vv v16, v16, v24, v0.t
+; RV64-NEXT: vand.vx v24, v16, a5, v0.t
+; RV64-NEXT: vsrl.vi v16, v16, 2, v0.t
+; RV64-NEXT: vand.vx v16, v16, a5, v0.t
+; RV64-NEXT: vadd.vv v16, v24, v16, v0.t
+; RV64-NEXT: vsrl.vi v24, v16, 4, v0.t
+; RV64-NEXT: vadd.vv v16, v16, v24, v0.t
+; RV64-NEXT: vand.vx v16, v16, a2, v0.t
+; RV64-NEXT: vmul.vx v16, v16, a3, v0.t
+; RV64-NEXT: vsrl.vx v16, v16, a0, v0.t
; RV64-NEXT: ret
%v = call <32 x i64> @llvm.vp.cttz.v32i64(<32 x i64> %va, i1 false, <32 x i1> %m, i32 %evl)
ret <32 x i64> %v
@@ -3955,99 +3877,64 @@ define <32 x i64> @vp_cttz_zero_undef_v32i64(<32 x i64> %va, <32 x i1> %m, i32 z
; 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: 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, 48
-; RV32-NEXT: vs8r.v v8, (a4) # Unknown-size Folded Spill
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV32-NEXT: vlse64.v v8, (a3), zero
+; RV32-NEXT: vlse64.v v16, (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: vs8r.v v16, (a3) # Unknown-size Folded Spill
; RV32-NEXT: addi a3, sp, 32
-; 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: vs8r.v v8, (a3) # Unknown-size Folded Spill
+; RV32-NEXT: vlse64.v v16, (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: vl8r.v v16, (a3) # Unknown-size Folded Reload
+; RV32-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
-; RV32-NEXT: vsrl.vi v24, v16, 1, v0.t
+; RV32-NEXT: vsrl.vi v16, v8, 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, 48
-; RV32-NEXT: vl8r.v v16, (a3) # Unknown-size Folded Reload
-; RV32-NEXT: vand.vv v24, v24, v16, 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: vl8r.v v16, (a3) # Unknown-size Folded Reload
-; RV32-NEXT: vsub.vv v24, v16, v24, v0.t
-; RV32-NEXT: vand.vv v16, v24, v8, 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 v16, (a3) # 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 v24, (a3) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v16, v16, v24, v0.t
+; RV32-NEXT: vsub.vv v8, v8, v16, 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: vl8r.v v8, (a3) # Unknown-size Folded Reload
-; RV32-NEXT: vadd.vv v8, 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: vl8r.v v24, (a3) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v16, v8, v24, v0.t
+; RV32-NEXT: vsrl.vi v8, v8, 2, v0.t
+; RV32-NEXT: vand.vv v8, v8, v24, v0.t
+; RV32-NEXT: vadd.vv v8, v16, v8, v0.t
; 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 v16, (a3), zero
-; RV32-NEXT: addi a3, sp, 48
-; RV32-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: vlse64.v v8, (a4), 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 v8, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: slli a3, a3, 3
+; RV32-NEXT: slli a3, a3, 5
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 48
-; RV32-NEXT: vl8r.v v8, (a3) # Unknown-size Folded Reload
+; RV32-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
+; RV32-NEXT: vlse64.v v24, (a4), zero
+; RV32-NEXT: addi a3, sp, 48
+; RV32-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
-; 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: vsrl.vi v16, v8, 4, v0.t
+; RV32-NEXT: vadd.vv v8, v8, v16, v0.t
; RV32-NEXT: csrr a2, vlenb
-; RV32-NEXT: li a3, 24
-; RV32-NEXT: mul a2, a2, a3
+; 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: vl8r.v v16, (a2) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v8, v8, v16, v0.t
+; RV32-NEXT: vmul.vv v8, v8, v24, v0.t
; RV32-NEXT: li a2, 56
; RV32-NEXT: vsrl.vx v8, v8, a2, v0.t
; RV32-NEXT: csrr a3, vlenb
@@ -4069,52 +3956,35 @@ define <32 x i64> @vp_cttz_zero_undef_v32i64(<32 x i64> %va, <32 x i1> %m, i32 z
; 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
-; RV32-NEXT: vsrl.vi v24, v8, 1, v0.t
+; RV32-NEXT: vsrl.vi v16, v8, 1, 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: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; RV32-NEXT: vand.vv v16, v24, v16, v0.t
-; RV32-NEXT: vsub.vv v24, v8, v16, v0.t
+; RV32-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v16, v16, v24, v0.t
+; RV32-NEXT: vsub.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 v8, (a0) # Unknown-size Folded Reload
-; RV32-NEXT: vand.vv v16, v24, v8, 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: vs8r.v v16, (a0) # Unknown-size Folded Spill
-; RV32-NEXT: vsrl.vi v8, v24, 2, v0.t
+; RV32-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; RV32-NEXT: vand.vv v16, v8, v24, v0.t
+; RV32-NEXT: vsrl.vi v8, v8, 2, v0.t
+; RV32-NEXT: vand.vv v8, v8, v24, v0.t
+; 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: 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: vand.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: 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 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: vmul.vv v8, v8, v16, v0.t
; RV32-NEXT: vsrl.vx v16, v8, a2, v0.t
; RV32-NEXT: csrr a0, vlenb
@@ -4133,20 +4003,9 @@ define <32 x i64> @vp_cttz_zero_undef_v32i64(<32 x i64> %va, <32 x i1> %m, i32 z
;
; RV64-LABEL: vp_cttz_zero_undef_v32i64:
; RV64: # %bb.0:
-; RV64-NEXT: addi sp, sp, -16
-; RV64-NEXT: .cfi_def_cfa_offset 16
-; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: slli a1, a1, 4
-; RV64-NEXT: sub sp, sp, a1
-; 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 a1, vlenb
-; RV64-NEXT: slli a1, a1, 3
-; RV64-NEXT: add a1, sp, a1
-; RV64-NEXT: addi a1, a1, 16
-; RV64-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
; RV64-NEXT: li a1, 16
; RV64-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
-; RV64-NEXT: vslidedown.vi v24, v0, 2
+; RV64-NEXT: vslidedown.vi v7, v0, 2
; RV64-NEXT: mv a4, a0
; RV64-NEXT: bltu a0, a1, .LBB70_2
; RV64-NEXT: # %bb.1:
@@ -4175,53 +4034,38 @@ define <32 x i64> @vp_cttz_zero_undef_v32i64(<32 x i64> %va, <32 x i1> %m, i32 z
; RV64-NEXT: and a7, a0, a7
; RV64-NEXT: li a0, 56
; RV64-NEXT: vsetvli zero, a4, e64, m8, ta, ma
-; RV64-NEXT: vsub.vx v16, v8, a1, v0.t
+; RV64-NEXT: vsub.vx v24, v8, a1, v0.t
; RV64-NEXT: vnot.v v8, v8, v0.t
-; RV64-NEXT: vand.vv v8, v8, v16, v0.t
-; RV64-NEXT: vsrl.vi v16, v8, 1, v0.t
-; RV64-NEXT: vand.vx v16, v16, a6, v0.t
-; RV64-NEXT: vsub.vv v8, v8, v16, v0.t
-; RV64-NEXT: vand.vx v16, v8, a5, v0.t
+; RV64-NEXT: vand.vv v8, v8, v24, v0.t
+; RV64-NEXT: vsrl.vi v24, v8, 1, v0.t
+; RV64-NEXT: vand.vx v24, v24, a6, v0.t
+; RV64-NEXT: vsub.vv v8, v8, v24, v0.t
+; RV64-NEXT: vand.vx v24, v8, a5, v0.t
; RV64-NEXT: vsrl.vi v8, v8, 2, v0.t
; RV64-NEXT: vand.vx v8, v8, a5, v0.t
-; RV64-NEXT: vadd.vv v8, v16, v8, v0.t
-; RV64-NEXT: vsrl.vi v16, v8, 4, v0.t
-; RV64-NEXT: vadd.vv v8, v8, v16, v0.t
+; RV64-NEXT: vadd.vv v8, v24, v8, v0.t
+; RV64-NEXT: vsrl.vi v24, v8, 4, v0.t
+; RV64-NEXT: vadd.vv v8, v8, v24, v0.t
; RV64-NEXT: vand.vx v8, v8, a2, v0.t
; RV64-NEXT: vmul.vx v8, v8, a3, v0.t
; RV64-NEXT: vsrl.vx v8, v8, a0, v0.t
-; RV64-NEXT: addi a4, sp, 16
-; RV64-NEXT: vs8r.v v8, (a4) # Unknown-size Folded Spill
-; RV64-NEXT: vmv1r.v v0, v24
-; 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 v8, (a4) # Unknown-size Folded Reload
+; RV64-NEXT: vmv1r.v v0, v7
; RV64-NEXT: vsetvli zero, a7, 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
-; RV64-NEXT: vsrl.vi v16, v8, 1, v0.t
-; RV64-NEXT: vand.vx v16, v16, a6, v0.t
-; RV64-NEXT: vsub.vv v8, v8, v16, v0.t
-; RV64-NEXT: vand.vx v16, v8, a5, v0.t
-; RV64-NEXT: vsrl.vi v8, v8, 2, v0.t
-; RV64-NEXT: vand.vx v8, v8, a5, v0.t
-; RV64-NEXT: vadd.vv v8, v16, v8, v0.t
-; RV64-NEXT: vsrl.vi v16, v8, 4, v0.t
-; RV64-NEXT: vadd.vv v8, v8, v16, v0.t
-; RV64-NEXT: vand.vx v8, v8, a2, v0.t
-; RV64-NEXT: vmul.vx v8, v8, a3, v0.t
-; RV64-NEXT: vsrl.vx v16, v8, a0, v0.t
-; RV64-NEXT: addi a0, sp, 16
-; RV64-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; RV64-NEXT: csrr a0, vlenb
-; RV64-NEXT: slli a0, a0, 4
-; RV64-NEXT: add sp, sp, a0
-; RV64-NEXT: .cfi_def_cfa sp, 16
-; RV64-NEXT: addi sp, sp, 16
-; RV64-NEXT: .cfi_def_cfa_offset 0
+; RV64-NEXT: vsub.vx v24, v16, a1, v0.t
+; RV64-NEXT: vnot.v v16, v16, v0.t
+; RV64-NEXT: vand.vv v16, v16, v24, v0.t
+; RV64-NEXT: vsrl.vi v24, v16, 1, v0.t
+; RV64-NEXT: vand.vx v24, v24, a6, v0.t
+; RV64-NEXT: vsub.vv v16, v16, v24, v0.t
+; RV64-NEXT: vand.vx v24, v16, a5, v0.t
+; RV64-NEXT: vsrl.vi v16, v16, 2, v0.t
+; RV64-NEXT: vand.vx v16, v16, a5, v0.t
+; RV64-NEXT: vadd.vv v16, v24, v16, v0.t
+; RV64-NEXT: vsrl.vi v24, v16, 4, v0.t
+; RV64-NEXT: vadd.vv v16, v16, v24, v0.t
+; RV64-NEXT: vand.vx v16, v16, a2, v0.t
+; RV64-NEXT: vmul.vx v16, v16, a3, v0.t
+; RV64-NEXT: vsrl.vx v16, v16, a0, v0.t
; RV64-NEXT: ret
%v = call <32 x i64> @llvm.vp.cttz.v32i64(<32 x i64> %va, i1 true, <32 x i1> %m, i32 %evl)
ret <32 x i64> %v
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 02e99ea513e69b..97d544980bc9eb 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-floor-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-floor-vp.ll
@@ -737,29 +737,18 @@ 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: vmv1r.v v6, v0
; CHECK-NEXT: li a2, 16
; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
-; CHECK-NEXT: vslidedown.vi v24, v0, 2
+; 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: vmv1r.v v0, v25
+; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; CHECK-NEXT: vfabs.v v16, v8, v0.t
+; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: lui a1, %hi(.LCPI26_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI26_0)(a1)
; CHECK-NEXT: addi a1, a0, -16
@@ -767,43 +756,28 @@ define <32 x double> @vp_floor_v32f64(<32 x double> %va, <32 x i1> %m, i32 zeroe
; CHECK-NEXT: addi a0, a0, -1
; CHECK-NEXT: and a0, a0, a1
; 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, v25
+; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
-; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
+; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: fsrm a1
-; 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: addi a1, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
-; CHECK-NEXT: vmv1r.v v0, v24
-; 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: vfsgnj.vv v8, v24, v8, v0.t
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; 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, v24
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
-; 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: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 4
-; CHECK-NEXT: add sp, sp, a0
-; CHECK-NEXT: .cfi_def_cfa sp, 16
-; CHECK-NEXT: addi sp, sp, 16
-; CHECK-NEXT: .cfi_def_cfa_offset 0
+; CHECK-NEXT: vfsgnj.vv v16, v24, v16, v0.t
; CHECK-NEXT: ret
%v = call <32 x double> @llvm.vp.floor.v32f64(<32 x double> %va, <32 x i1> %m, i32 %evl)
ret <32 x double> %v
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 f43934afc370df..617e226df92986 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fmaximum-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fmaximum-vp.ll
@@ -595,23 +595,18 @@ 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 v25, v0
-; 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 v16, (a1) # Unknown-size Folded Spill
+; CHECK-NEXT: vmv1r.v v6, v0
; 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
+; CHECK-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
; CHECK-NEXT: addi a1, a0, 128
; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; CHECK-NEXT: vle64.v v16, (a1)
; 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
@@ -624,43 +619,29 @@ define <32 x double> @vfmax_vv_v32f64(<32 x double> %va, <32 x double> %vb, <32
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: li a0, 16
; CHECK-NEXT: .LBB24_2:
-; CHECK-NEXT: vmv1r.v v0, v25
-; 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: vl8r.v v8, (a1) # Unknown-size Folded Reload
+; 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: vmv8r.v v8, v16
+; CHECK-NEXT: vmerge.vvm v24, v8, v16, v0
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a1, 24
-; 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: vmerge.vvm v16, v16, v8, v0
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: vmv1r.v v0, v25
-; CHECK-NEXT: vmfeq.vv v26, v8, v8, v0.t
+; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vmv1r.v v0, v6
+; CHECK-NEXT: vmfeq.vv v26, v16, v16, v0.t
; CHECK-NEXT: vmv1r.v v0, v26
+; CHECK-NEXT: vmerge.vvm v8, v16, v8, v0
+; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a1, 24
-; 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: vmerge.vvm v8, v8, v16, v0
-; CHECK-NEXT: vmv1r.v v0, v25
-; CHECK-NEXT: addi a0, sp, 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: li a1, 24
-; 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: vs8r.v v8, (a0) # Unknown-size Folded Spill
@@ -670,7 +651,8 @@ define <32 x double> @vfmax_vv_v32f64(<32 x double> %va, <32 x double> %vb, <32
; CHECK-NEXT: and a0, a1, a0
; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a1, a1, 4
+; 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 v16, (a1) # Unknown-size Folded Reload
@@ -678,7 +660,7 @@ define <32 x double> @vfmax_vv_v32f64(<32 x double> %va, <32 x double> %vb, <32
; 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, 3
+; 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
@@ -688,13 +670,12 @@ define <32 x double> @vfmax_vv_v32f64(<32 x double> %va, <32 x double> %vb, <32
; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vmfeq.vv v25, v8, v8, v0.t
; CHECK-NEXT: vmv1r.v v0, v25
-; CHECK-NEXT: vmerge.vvm v8, v8, v16, v0
+; CHECK-NEXT: vmerge.vvm v24, v8, v16, v0
; CHECK-NEXT: vmv1r.v v0, v7
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vfmax.vv v16, v8, v16, v0.t
+; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vfmax.vv v16, v24, v8, v0.t
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a1, 24
-; 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 v8, (a0) # Unknown-size Folded Reload
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 7067cc21ab56d5..cef45258c61df8 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fminimum-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fminimum-vp.ll
@@ -595,23 +595,18 @@ 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 v25, v0
-; 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 v16, (a1) # Unknown-size Folded Spill
+; CHECK-NEXT: vmv1r.v v6, v0
; 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
+; CHECK-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
; CHECK-NEXT: addi a1, a0, 128
; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; CHECK-NEXT: vle64.v v16, (a1)
; 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
@@ -624,43 +619,29 @@ define <32 x double> @vfmin_vv_v32f64(<32 x double> %va, <32 x double> %vb, <32
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: li a0, 16
; CHECK-NEXT: .LBB24_2:
-; CHECK-NEXT: vmv1r.v v0, v25
-; 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: vl8r.v v8, (a1) # Unknown-size Folded Reload
+; 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: vmv8r.v v8, v16
+; CHECK-NEXT: vmerge.vvm v24, v8, v16, v0
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a1, 24
-; 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: vmerge.vvm v16, v16, v8, v0
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: vmv1r.v v0, v25
-; CHECK-NEXT: vmfeq.vv v26, v8, v8, v0.t
+; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vmv1r.v v0, v6
+; CHECK-NEXT: vmfeq.vv v26, v16, v16, v0.t
; CHECK-NEXT: vmv1r.v v0, v26
+; CHECK-NEXT: vmerge.vvm v8, v16, v8, v0
+; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a1, 24
-; 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: vmerge.vvm v8, v8, v16, v0
-; CHECK-NEXT: vmv1r.v v0, v25
-; CHECK-NEXT: addi a0, sp, 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: li a1, 24
-; 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: vs8r.v v8, (a0) # Unknown-size Folded Spill
@@ -670,7 +651,8 @@ define <32 x double> @vfmin_vv_v32f64(<32 x double> %va, <32 x double> %vb, <32
; CHECK-NEXT: and a0, a1, a0
; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a1, a1, 4
+; 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 v16, (a1) # Unknown-size Folded Reload
@@ -678,7 +660,7 @@ define <32 x double> @vfmin_vv_v32f64(<32 x double> %va, <32 x double> %vb, <32
; 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, 3
+; 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
@@ -688,13 +670,12 @@ define <32 x double> @vfmin_vv_v32f64(<32 x double> %va, <32 x double> %vb, <32
; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vmfeq.vv v25, v8, v8, v0.t
; CHECK-NEXT: vmv1r.v v0, v25
-; CHECK-NEXT: vmerge.vvm v8, v8, v16, v0
+; CHECK-NEXT: vmerge.vvm v24, v8, v16, v0
; CHECK-NEXT: vmv1r.v v0, v7
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vfmin.vv v16, v8, v16, v0.t
+; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vfmin.vv v16, v24, v8, v0.t
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a1, 24
-; 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 v8, (a0) # Unknown-size Folded Reload
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 a25014295f9e88..ea2809aa95dade 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-int-explodevector.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-int-explodevector.ll
@@ -869,9 +869,8 @@ define i64 @explode_16xi64(<16 x i64> %v) {
; RV32-NEXT: .cfi_offset s10, -44
; RV32-NEXT: .cfi_offset s11, -48
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 3
; RV32-NEXT: sub sp, sp, a0
-; RV32-NEXT: .cfi_escape 0x0f, 0x0e, 0x72, 0x00, 0x11, 0xc0, 0x00, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 64 + 8 * vlenb
+; RV32-NEXT: .cfi_escape 0x0f, 0x0e, 0x72, 0x00, 0x11, 0xc0, 0x00, 0x22, 0x11, 0x01, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 64 + 1 * vlenb
; RV32-NEXT: vsetivli zero, 1, e64, m8, ta, ma
; RV32-NEXT: vslidedown.vi v24, v8, 2
; RV32-NEXT: li a0, 32
@@ -913,8 +912,6 @@ define i64 @explode_16xi64(<16 x i64> %v) {
; RV32-NEXT: vsrl.vx v16, v16, a0
; RV32-NEXT: vmv.x.s s3, v16
; RV32-NEXT: vslidedown.vi v16, v8, 13
-; RV32-NEXT: addi s4, sp, 16
-; RV32-NEXT: vs8r.v v16, (s4) # Unknown-size Folded Spill
; RV32-NEXT: vmv.x.s s4, v24
; RV32-NEXT: vsrl.vx v24, v24, a0
; RV32-NEXT: vmv.x.s s5, v24
@@ -924,19 +921,21 @@ define i64 @explode_16xi64(<16 x i64> %v) {
; RV32-NEXT: vmv.x.s s7, v0
; RV32-NEXT: vmv.s.x v7, zero
; RV32-NEXT: vsetivli zero, 2, e64, m1, ta, ma
-; RV32-NEXT: vredxor.vs v16, v8, v7
+; RV32-NEXT: vredxor.vs v7, v8, v7
+; RV32-NEXT: addi s8, sp, 16
+; RV32-NEXT: vs1r.v v7, (s8) # Unknown-size Folded Spill
; RV32-NEXT: vsetivli zero, 1, e64, m8, ta, ma
; RV32-NEXT: vslidedown.vi v8, v8, 15
-; RV32-NEXT: addi s8, sp, 16
-; RV32-NEXT: vl8r.v v0, (s8) # Unknown-size Folded Reload
-; RV32-NEXT: vmv.x.s s8, v0
-; RV32-NEXT: vsrl.vx v0, v0, a0
-; RV32-NEXT: vmv.x.s s9, v0
+; RV32-NEXT: vmv.x.s s8, v16
+; RV32-NEXT: vsrl.vx v16, v16, a0
+; RV32-NEXT: vmv.x.s s9, v16
; RV32-NEXT: vsrl.vx v0, v24, a0
+; RV32-NEXT: addi s10, sp, 16
+; RV32-NEXT: vl1r.v v17, (s10) # Unknown-size Folded Reload
; RV32-NEXT: vsetivli zero, 1, e64, m1, ta, ma
-; RV32-NEXT: vsrl.vx v17, v16, a0
-; RV32-NEXT: vmv.x.s s10, v16
-; RV32-NEXT: vmv.x.s s11, v17
+; RV32-NEXT: vsrl.vx v16, v17, a0
+; RV32-NEXT: vmv.x.s s10, v17
+; RV32-NEXT: vmv.x.s s11, v16
; RV32-NEXT: vsetivli zero, 1, e64, m8, ta, ma
; RV32-NEXT: vsrl.vx v16, v8, a0
; RV32-NEXT: add a2, s11, a2
@@ -1000,7 +999,6 @@ define i64 @explode_16xi64(<16 x i64> %v) {
; RV32-NEXT: sltu a2, a0, a2
; RV32-NEXT: add a1, a1, a2
; RV32-NEXT: csrr a2, vlenb
-; RV32-NEXT: slli a2, a2, 3
; RV32-NEXT: add sp, sp, a2
; RV32-NEXT: .cfi_def_cfa sp, 64
; RV32-NEXT: lw s0, 60(sp) # 4-byte Folded Reload
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 651674ee9a5022..a49da09426d11a 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-interleaved-access.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-interleaved-access.ll
@@ -183,59 +183,72 @@ 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: slli a3, a2, 6
-; RV32-NEXT: add a2, a3, a2
+; RV32-NEXT: li a3, 61
+; RV32-NEXT: mul a2, a2, a3
; RV32-NEXT: sub sp, sp, a2
-; RV32-NEXT: .cfi_escape 0x0f, 0x0e, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0xc1, 0x00, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 65 * vlenb
+; RV32-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x3d, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 61 * vlenb
; RV32-NEXT: addi a3, a1, 256
; RV32-NEXT: addi a4, a1, 128
; RV32-NEXT: li a2, 32
; RV32-NEXT: lui a5, 12291
; RV32-NEXT: vsetvli zero, a2, e32, m8, ta, mu
-; RV32-NEXT: vle32.v v24, (a1)
+; RV32-NEXT: vle32.v v0, (a1)
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a6, 41
+; RV32-NEXT: li a6, 45
; RV32-NEXT: mul a1, a1, a6
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
-; RV32-NEXT: vs8r.v v24, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: vs8r.v v0, (a1) # Unknown-size Folded Spill
; RV32-NEXT: lui a1, %hi(.LCPI8_0)
; RV32-NEXT: addi a1, a1, %lo(.LCPI8_0)
-; RV32-NEXT: vle16.v v4, (a1)
+; RV32-NEXT: vle16.v v8, (a1)
+; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: li a6, 37
+; RV32-NEXT: mul a1, a1, a6
+; 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, 1
; RV32-NEXT: addi a5, a5, 3
; RV32-NEXT: vle32.v v8, (a4)
; RV32-NEXT: csrr a4, vlenb
-; RV32-NEXT: li a6, 57
+; RV32-NEXT: li a6, 53
; RV32-NEXT: mul a4, a4, a6
; RV32-NEXT: add a4, sp, a4
; RV32-NEXT: addi a4, a4, 16
; RV32-NEXT: vs8r.v v8, (a4) # Unknown-size Folded Spill
; RV32-NEXT: addi a1, a1, -64
; RV32-NEXT: vle32.v v16, (a3)
-; RV32-NEXT: vmv.s.x v3, a5
-; RV32-NEXT: vmv.s.x v0, a1
+; RV32-NEXT: csrr a3, vlenb
+; RV32-NEXT: li a4, 29
+; RV32-NEXT: mul a3, a3, a4
+; RV32-NEXT: add a3, sp, a3
+; RV32-NEXT: addi a3, a3, 16
+; RV32-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
+; RV32-NEXT: vmv.s.x v28, a5
+; RV32-NEXT: vmv.s.x v24, a1
; 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: vs1r.v v0, (a1) # Unknown-size Folded Spill
-; RV32-NEXT: vcompress.vm v8, v24, v3
+; RV32-NEXT: vs1r.v v24, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: vcompress.vm v8, v0, v28
+; RV32-NEXT: vmv1r.v v0, v24
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a3, 57
+; RV32-NEXT: li a3, 53
; 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: csrr a1, vlenb
+; 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, v0.t
; RV32-NEXT: lui a1, 12
-; RV32-NEXT: csrr a3, vlenb
-; RV32-NEXT: li a4, 49
-; RV32-NEXT: mul a3, a3, a4
-; RV32-NEXT: add a3, sp, a3
-; RV32-NEXT: addi a3, a3, 16
-; RV32-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
; RV32-NEXT: vsetivli zero, 16, e32, m4, ta, ma
; RV32-NEXT: vslideup.vi v12, v16, 4
; RV32-NEXT: csrr a3, vlenb
@@ -245,22 +258,22 @@ define {<8 x i64>, <8 x i64>, <8 x i64>, <8 x i64>, <8 x i64>, <8 x i64>} @load_
; RV32-NEXT: addi a3, a3, 16
; RV32-NEXT: vs4r.v v12, (a3) # Unknown-size Folded Spill
; RV32-NEXT: vmv.s.x v0, a1
-; RV32-NEXT: vsetivli zero, 16, e32, m8, ta, ma
-; RV32-NEXT: vslidedown.vi v24, v16, 16
; 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: vs1r.v v0, (a1) # Unknown-size Folded Spill
-; RV32-NEXT: vsetivli zero, 16, e32, m4, ta, mu
-; RV32-NEXT: vslideup.vi v12, v24, 10, v0.t
+; RV32-NEXT: vsetivli zero, 16, e32, m8, ta, ma
+; RV32-NEXT: vslidedown.vi v24, v16, 16
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: slli a3, a1, 5
-; 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: vs8r.v v24, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: vsetivli zero, 16, e32, m4, ta, mu
+; RV32-NEXT: vslideup.vi v12, v24, 10, v0.t
; RV32-NEXT: vsetivli zero, 12, e32, m4, tu, ma
; RV32-NEXT: vmv.v.v v12, v8
; RV32-NEXT: csrr a1, vlenb
@@ -273,16 +286,16 @@ define {<8 x i64>, <8 x i64>, <8 x i64>, <8 x i64>, <8 x i64>, <8 x i64>} @load_
; RV32-NEXT: lui a3, %hi(.LCPI8_1)
; RV32-NEXT: addi a3, a3, %lo(.LCPI8_1)
; RV32-NEXT: vsetvli zero, a2, e32, m8, ta, mu
-; RV32-NEXT: vle16.v v28, (a3)
+; RV32-NEXT: vle16.v v20, (a3)
; RV32-NEXT: addi a1, a1, 12
-; RV32-NEXT: vmv.s.x v20, a1
+; RV32-NEXT: vmv.s.x v24, a1
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a3, 41
+; 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: vl8r.v v0, (a1) # Unknown-size Folded Reload
-; RV32-NEXT: vcompress.vm v8, v0, v20
+; RV32-NEXT: vcompress.vm v8, v0, v24
; RV32-NEXT: csrr a1, vlenb
; RV32-NEXT: li a3, 13
; RV32-NEXT: mul a1, a1, a3
@@ -290,14 +303,14 @@ define {<8 x i64>, <8 x i64>, <8 x i64>, <8 x i64>, <8 x i64>, <8 x i64>} @load_
; RV32-NEXT: addi a1, a1, 16
; RV32-NEXT: vl1r.v v0, (a1) # Unknown-size Folded Reload
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a3, 57
+; RV32-NEXT: li a3, 53
; 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, v28, v0.t
+; RV32-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vrgatherei16.vv v8, v24, v20, v0.t
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a3, 49
+; RV32-NEXT: li a3, 29
; RV32-NEXT: mul a1, a1, a3
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
@@ -310,6 +323,12 @@ define {<8 x i64>, <8 x i64>, <8 x i64>, <8 x i64>, <8 x i64>, <8 x i64>} @load_
; 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 a3, 37
+; 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: vslideup.vi v12, v24, 8, v0.t
; RV32-NEXT: vsetivli zero, 12, e32, m4, tu, ma
; RV32-NEXT: vmv.v.v v12, v8
@@ -326,13 +345,9 @@ define {<8 x i64>, <8 x i64>, <8 x i64>, <8 x i64>, <8 x i64>, <8 x i64>} @load_
; RV32-NEXT: lui a5, %hi(.LCPI8_3)
; RV32-NEXT: addi a5, a5, %lo(.LCPI8_3)
; RV32-NEXT: addi a1, a1, 48
-; 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 v16, a4
; RV32-NEXT: vsetvli zero, a2, e16, m4, ta, ma
-; RV32-NEXT: vle16.v v4, (a3)
+; RV32-NEXT: vle16.v v20, (a3)
; RV32-NEXT: vsetivli zero, 16, e32, m4, ta, ma
; RV32-NEXT: vle16.v v8, (a5)
; RV32-NEXT: csrr a3, vlenb
@@ -341,49 +356,52 @@ define {<8 x i64>, <8 x i64>, <8 x i64>, <8 x i64>, <8 x i64>, <8 x i64>} @load_
; RV32-NEXT: add a3, sp, a3
; RV32-NEXT: addi a3, a3, 16
; RV32-NEXT: vs2r.v v8, (a3) # Unknown-size Folded Spill
-; RV32-NEXT: vmv.s.x v22, a1
+; RV32-NEXT: vmv.s.x v26, a1
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a3, 41
+; 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: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vl8r.v v0, (a1) # Unknown-size Folded Reload
; RV32-NEXT: vsetvli zero, a2, e32, m8, ta, mu
-; RV32-NEXT: vcompress.vm v8, v24, v22
+; RV32-NEXT: vcompress.vm v8, v0, v26
+; RV32-NEXT: vmv1r.v v0, v16
+; RV32-NEXT: vmv1r.v v6, v16
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a3, 57
+; RV32-NEXT: li a3, 53
; 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: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vrgatherei16.vv v8, v24, v20, v0.t
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a3, 49
+; RV32-NEXT: li a3, 29
; 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: vl8r.v v16, (a1) # Unknown-size Folded Reload
; RV32-NEXT: csrr a1, vlenb
; RV32-NEXT: slli a3, a1, 3
; RV32-NEXT: add a1, a3, a1
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
-; RV32-NEXT: vl2r.v v16, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vl2r.v v20, (a1) # Unknown-size Folded Reload
; RV32-NEXT: vsetivli zero, 16, e32, m4, ta, mu
-; RV32-NEXT: vrgatherei16.vv v12, v0, v16
+; RV32-NEXT: vrgatherei16.vv v12, v16, v20
; 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: vl1r.v v0, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vl1r.v v7, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vmv1r.v v0, v7
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: slli a3, a1, 5
-; 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: vl8r.v v16, (a1) # Unknown-size Folded Reload
-; RV32-NEXT: vslideup.vi v12, v16, 6, v0.t
+; RV32-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vslideup.vi v12, v24, 6, v0.t
; RV32-NEXT: vsetivli zero, 10, e32, m4, tu, ma
; RV32-NEXT: vmv.v.v v12, v8
; RV32-NEXT: csrr a1, vlenb
@@ -403,14 +421,17 @@ 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: vle16.v v12, (a4)
; RV32-NEXT: vmv.s.x v14, a1
-; RV32-NEXT: vsetvli zero, a2, e32, m8, ta, mu
-; RV32-NEXT: vcompress.vm v16, v24, v14
; RV32-NEXT: csrr a1, vlenb
+; 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: vl1r.v v0, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vsetvli zero, a2, e32, m8, ta, mu
+; RV32-NEXT: vcompress.vm v16, v24, v14
+; RV32-NEXT: vmv1r.v v0, v6
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a3, 57
+; RV32-NEXT: li a3, 53
; RV32-NEXT: mul a1, a1, a3
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
@@ -421,78 +442,73 @@ define {<8 x i64>, <8 x i64>, <8 x i64>, <8 x i64>, <8 x i64>, <8 x i64>} @load_
; RV32-NEXT: addi a1, a1, 16
; RV32-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a3, 49
+; RV32-NEXT: li a3, 29
; 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: vl8r.v v16, (a1) # Unknown-size Folded Reload
; RV32-NEXT: vsetivli zero, 16, e32, m4, ta, mu
-; RV32-NEXT: vrgatherei16.vv v4, v0, v12
+; RV32-NEXT: vrgatherei16.vv v20, v16, v12
+; RV32-NEXT: vmv1r.v v0, v7
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a3, 25
+; 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: vl1r.v v0, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vslideup.vi v20, v8, 4, v0.t
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: slli a3, a1, 5
-; 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: vl8r.v v8, (a1) # Unknown-size Folded Reload
-; RV32-NEXT: vslideup.vi v4, v8, 4, v0.t
+; RV32-NEXT: vs4r.v v20, (a1) # Unknown-size Folded Spill
; RV32-NEXT: lui a1, 768
; RV32-NEXT: lui a3, %hi(.LCPI8_6)
; RV32-NEXT: addi a3, a3, %lo(.LCPI8_6)
; RV32-NEXT: li a4, 1008
; RV32-NEXT: addi a1, a1, 768
; RV32-NEXT: vsetvli zero, a2, e32, m8, ta, mu
-; RV32-NEXT: vle16.v v8, (a3)
+; RV32-NEXT: vle16.v v20, (a3)
; RV32-NEXT: vmv.s.x v1, a4
-; RV32-NEXT: vmv.s.x v12, a1
+; RV32-NEXT: vmv.s.x v2, a1
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a3, 41
+; 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: vl8r.v v16, (a1) # Unknown-size Folded Reload
-; RV32-NEXT: vcompress.vm v24, v16, v12
+; RV32-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vcompress.vm v8, v24, v2
; RV32-NEXT: vmv1r.v v0, v1
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a3, 57
+; RV32-NEXT: li a3, 53
; 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: 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: vs8r.v v24, (a1) # Unknown-size Folded Spill
+; RV32-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vrgatherei16.vv v8, v24, v20, v0.t
; RV32-NEXT: lui a1, %hi(.LCPI8_7)
; RV32-NEXT: addi a1, a1, %lo(.LCPI8_7)
; RV32-NEXT: vsetivli zero, 16, e32, m4, ta, mu
-; RV32-NEXT: vle16.v v8, (a1)
+; RV32-NEXT: vle16.v v20, (a1)
; RV32-NEXT: lui a1, 15
; RV32-NEXT: vmv.s.x v0, a1
; RV32-NEXT: addi a1, sp, 16
; RV32-NEXT: vs1r.v v0, (a1) # Unknown-size Folded Spill
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a3, 49
+; RV32-NEXT: li a3, 29
; 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 v20, v16, 6
+; RV32-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vslideup.vi v12, v24, 6
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: slli a3, a1, 5
-; 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: vl8r.v v24, (a1) # Unknown-size Folded Reload
-; RV32-NEXT: vrgatherei16.vv v20, v24, v8, v0.t
+; RV32-NEXT: vrgatherei16.vv v12, v24, v20, v0.t
; RV32-NEXT: lui a1, 3073
; RV32-NEXT: lui a3, %hi(.LCPI8_8)
; RV32-NEXT: addi a3, a3, %lo(.LCPI8_8)
@@ -500,26 +516,26 @@ define {<8 x i64>, <8 x i64>, <8 x i64>, <8 x i64>, <8 x i64>, <8 x i64>} @load_
; RV32-NEXT: addi a4, a4, %lo(.LCPI8_9)
; RV32-NEXT: addi a1, a1, -1024
; RV32-NEXT: vsetvli zero, a2, e16, m4, ta, ma
-; RV32-NEXT: vle16.v v16, (a3)
+; RV32-NEXT: vle16.v v4, (a3)
; RV32-NEXT: vsetivli zero, 16, e32, m4, ta, ma
; RV32-NEXT: vle16.v v2, (a4)
; RV32-NEXT: vmv.s.x v0, a1
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a3, 41
+; 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: vl8r.v v24, (a1) # Unknown-size Folded Reload
; RV32-NEXT: vsetvli zero, a2, e32, m8, ta, mu
-; RV32-NEXT: vcompress.vm v8, v24, v0
+; RV32-NEXT: vcompress.vm v16, v24, v0
; RV32-NEXT: vmv1r.v v0, v1
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a2, 57
+; RV32-NEXT: li a2, 53
; RV32-NEXT: mul a1, a1, a2
; 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, v16, v0.t
+; RV32-NEXT: vrgatherei16.vv v16, v24, v4, v0.t
; RV32-NEXT: addi a1, sp, 16
; RV32-NEXT: vl1r.v v0, (a1) # Unknown-size Folded Reload
; RV32-NEXT: csrr a1, vlenb
@@ -527,36 +543,36 @@ define {<8 x i64>, <8 x i64>, <8 x i64>, <8 x i64>, <8 x i64>, <8 x i64>} @load_
; RV32-NEXT: add a1, a2, a1
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
-; RV32-NEXT: vl4r.v v12, (a1) # Unknown-size Folded Reload
+; RV32-NEXT: vl4r.v v20, (a1) # Unknown-size Folded Reload
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: slli a2, a1, 5
-; RV32-NEXT: add a1, a2, a1
+; RV32-NEXT: li a2, 37
+; RV32-NEXT: mul a1, a1, a2
; 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 v12, v24, v2, v0.t
+; RV32-NEXT: vrgatherei16.vv v20, v24, v2, v0.t
; RV32-NEXT: csrr a1, vlenb
+; RV32-NEXT: li a2, 25
+; RV32-NEXT: mul a1, a1, a2
; 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, 10, e32, m4, tu, ma
-; RV32-NEXT: vmv.v.v v4, v24
+; RV32-NEXT: vl4r.v v24, (a1) # Unknown-size Folded Reload
; RV32-NEXT: csrr a1, vlenb
-; RV32-NEXT: li a2, 25
-; RV32-NEXT: mul a1, a1, a2
; RV32-NEXT: add a1, sp, a1
; RV32-NEXT: addi a1, a1, 16
-; RV32-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
-; RV32-NEXT: vmv.v.v v20, v24
+; 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, v0
; RV32-NEXT: vmv.v.v v12, v8
+; RV32-NEXT: vmv.v.v v20, v16
; RV32-NEXT: addi a1, a0, 320
; RV32-NEXT: vsetivli zero, 16, e32, m4, ta, ma
-; RV32-NEXT: vse32.v v12, (a1)
-; RV32-NEXT: addi a1, a0, 256
; RV32-NEXT: vse32.v v20, (a1)
+; RV32-NEXT: addi a1, a0, 256
+; RV32-NEXT: vse32.v v12, (a1)
; RV32-NEXT: addi a1, a0, 192
-; RV32-NEXT: vse32.v v4, (a1)
+; RV32-NEXT: vse32.v v24, (a1)
; RV32-NEXT: addi a1, a0, 128
; RV32-NEXT: csrr a2, vlenb
; RV32-NEXT: slli a3, a2, 3
@@ -581,8 +597,8 @@ define {<8 x i64>, <8 x i64>, <8 x i64>, <8 x i64>, <8 x i64>, <8 x i64>} @load_
; RV32-NEXT: vl4r.v v8, (a1) # Unknown-size Folded Reload
; RV32-NEXT: vse32.v v8, (a0)
; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a1, a0, 6
-; RV32-NEXT: add a0, a1, a0
+; RV32-NEXT: li a1, 61
+; RV32-NEXT: mul a0, a0, a1
; RV32-NEXT: add sp, sp, a0
; RV32-NEXT: .cfi_def_cfa sp, 16
; RV32-NEXT: addi sp, sp, 16
@@ -594,331 +610,237 @@ 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, 67
+; RV64-NEXT: li a3, 53
; RV64-NEXT: mul a2, a2, a3
; RV64-NEXT: sub sp, sp, a2
-; RV64-NEXT: .cfi_escape 0x0f, 0x0e, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0xc3, 0x00, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 67 * vlenb
+; RV64-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x35, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 53 * vlenb
; RV64-NEXT: addi a2, a1, 128
; RV64-NEXT: vsetivli zero, 16, e16, m2, ta, ma
; RV64-NEXT: vle64.v v8, (a1)
; RV64-NEXT: csrr a3, vlenb
-; RV64-NEXT: li a4, 59
+; RV64-NEXT: li a4, 37
; RV64-NEXT: mul a3, a3, a4
; RV64-NEXT: add a3, sp, a3
; RV64-NEXT: addi a3, a3, 16
; RV64-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
; RV64-NEXT: addi a1, a1, 256
; RV64-NEXT: li a3, 128
-; RV64-NEXT: vle64.v v24, (a1)
+; RV64-NEXT: vle64.v v8, (a1)
; RV64-NEXT: lui a1, 1
-; RV64-NEXT: vid.v v8
-; RV64-NEXT: vmv.s.x v0, a3
-; RV64-NEXT: csrr a3, vlenb
-; RV64-NEXT: li a4, 30
-; RV64-NEXT: mul a3, a3, a4
-; RV64-NEXT: add a3, sp, a3
-; RV64-NEXT: addi a3, a3, 16
-; RV64-NEXT: vs1r.v v0, (a3) # Unknown-size Folded Spill
+; RV64-NEXT: vid.v v16
+; RV64-NEXT: vmv.s.x v1, a3
; RV64-NEXT: li a3, 6
-; RV64-NEXT: vmul.vx v6, v8, a3
+; RV64-NEXT: vmul.vx v2, v16, a3
; RV64-NEXT: li a3, 56
; RV64-NEXT: vsetivli zero, 8, e64, m4, ta, ma
-; RV64-NEXT: vrgather.vi v8, v24, 4
-; RV64-NEXT: csrr a4, vlenb
-; RV64-NEXT: li a5, 22
-; RV64-NEXT: mul a4, a4, a5
-; RV64-NEXT: add a4, sp, a4
-; RV64-NEXT: addi a4, a4, 16
-; RV64-NEXT: vs8r.v v24, (a4) # Unknown-size Folded Spill
+; RV64-NEXT: vrgather.vi v16, v8, 4
; RV64-NEXT: vsetivli zero, 8, e64, m8, ta, ma
-; RV64-NEXT: vslidedown.vi v16, v24, 8
+; RV64-NEXT: vslidedown.vi v24, v8, 8
; RV64-NEXT: csrr a4, vlenb
-; RV64-NEXT: li a5, 39
+; RV64-NEXT: li a5, 29
; RV64-NEXT: mul a4, a4, a5
; RV64-NEXT: add a4, sp, a4
; RV64-NEXT: addi a4, a4, 16
-; RV64-NEXT: vs8r.v v16, (a4) # Unknown-size Folded Spill
+; RV64-NEXT: vs8r.v v24, (a4) # Unknown-size Folded Spill
+; RV64-NEXT: vmv1r.v v0, v1
+; RV64-NEXT: addi a4, sp, 16
+; RV64-NEXT: vs1r.v v1, (a4) # Unknown-size Folded Spill
; RV64-NEXT: vsetivli zero, 8, e64, m4, ta, mu
-; RV64-NEXT: vrgather.vi v8, v16, 2, v0.t
-; RV64-NEXT: vmv.v.v v20, v8
-; RV64-NEXT: vmv.s.x v8, a3
-; RV64-NEXT: csrr a3, vlenb
-; RV64-NEXT: li a4, 55
-; RV64-NEXT: mul a3, a3, a4
-; RV64-NEXT: add a3, sp, a3
-; RV64-NEXT: addi a3, a3, 16
-; RV64-NEXT: vs1r.v v8, (a3) # Unknown-size Folded Spill
+; RV64-NEXT: vrgather.vi v16, v24, 2, v0.t
+; RV64-NEXT: vmv.v.v v4, v16
+; RV64-NEXT: vmv.s.x v12, a3
; RV64-NEXT: addi a3, a1, 65
; RV64-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV64-NEXT: vle64.v v8, (a2)
+; RV64-NEXT: vle64.v v16, (a2)
; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: li a4, 47
+; RV64-NEXT: li a4, 45
; RV64-NEXT: mul a2, a2, a4
; RV64-NEXT: add a2, sp, a2
; RV64-NEXT: addi a2, a2, 16
-; RV64-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
-; RV64-NEXT: vmv.s.x v16, a3
-; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: li a3, 35
-; RV64-NEXT: mul a2, a2, a3
-; RV64-NEXT: add a2, sp, a2
-; RV64-NEXT: addi a2, a2, 16
-; RV64-NEXT: vs1r.v v16, (a2) # Unknown-size Folded Spill
+; RV64-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
+; RV64-NEXT: vmv.s.x v13, a3
; RV64-NEXT: vsetvli zero, zero, e16, m2, ta, ma
-; RV64-NEXT: vadd.vi v16, v6, -16
-; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: slli a3, a2, 5
-; RV64-NEXT: sub a2, a3, a2
-; RV64-NEXT: add a2, sp, a2
-; RV64-NEXT: addi a2, a2, 16
-; RV64-NEXT: vs2r.v v16, (a2) # Unknown-size Folded Spill
-; RV64-NEXT: vmv2r.v v18, v6
-; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: li a3, 12
-; RV64-NEXT: mul a2, a2, a3
-; RV64-NEXT: add a2, sp, a2
-; RV64-NEXT: addi a2, a2, 16
-; RV64-NEXT: vs2r.v v6, (a2) # Unknown-size Folded Spill
+; RV64-NEXT: vadd.vi v14, v2, -16
; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: li a3, 59
+; RV64-NEXT: li a3, 37
; RV64-NEXT: mul a2, a2, a3
; RV64-NEXT: add a2, sp, a2
; RV64-NEXT: addi a2, a2, 16
-; RV64-NEXT: vl8r.v v0, (a2) # Unknown-size Folded Reload
-; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: li a3, 35
-; RV64-NEXT: mul a2, a2, a3
-; RV64-NEXT: add a2, sp, a2
-; RV64-NEXT: addi a2, a2, 16
-; RV64-NEXT: vl1r.v v16, (a2) # Unknown-size Folded Reload
+; RV64-NEXT: vl8r.v v24, (a2) # Unknown-size Folded Reload
; RV64-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; RV64-NEXT: vcompress.vm v24, v0, v16
+; RV64-NEXT: vcompress.vm v16, v24, v13
+; RV64-NEXT: vmv1r.v v0, v12
; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: li a3, 55
+; RV64-NEXT: li a3, 45
; RV64-NEXT: mul a2, a2, a3
; RV64-NEXT: add a2, sp, a2
; RV64-NEXT: addi a2, a2, 16
-; RV64-NEXT: vl1r.v v0, (a2) # Unknown-size Folded Reload
-; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: slli a3, a2, 5
-; RV64-NEXT: sub a2, a3, a2
-; RV64-NEXT: add a2, sp, a2
-; RV64-NEXT: addi a2, a2, 16
-; RV64-NEXT: vl2r.v v16, (a2) # Unknown-size Folded Reload
-; RV64-NEXT: vrgatherei16.vv v24, v8, v16, v0.t
+; RV64-NEXT: vl8r.v v24, (a2) # Unknown-size Folded Reload
+; RV64-NEXT: vrgatherei16.vv v16, v24, v14, v0.t
; RV64-NEXT: vsetivli zero, 6, e64, m4, tu, ma
-; RV64-NEXT: vmv.v.v v20, v24
-; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: li a3, 18
-; RV64-NEXT: mul a2, a2, a3
-; RV64-NEXT: add a2, sp, a2
-; RV64-NEXT: addi a2, a2, 16
-; RV64-NEXT: vs4r.v v20, (a2) # Unknown-size Folded Spill
+; RV64-NEXT: vmv.v.v v4, v16
; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: li a3, 22
+; RV64-NEXT: li a3, 13
; RV64-NEXT: mul a2, a2, a3
; RV64-NEXT: add a2, sp, a2
; RV64-NEXT: addi a2, a2, 16
-; RV64-NEXT: vl8r.v v24, (a2) # Unknown-size Folded Reload
+; RV64-NEXT: vs4r.v v4, (a2) # Unknown-size Folded Spill
; RV64-NEXT: vsetivli zero, 8, e64, m4, ta, mu
-; RV64-NEXT: vrgather.vi v8, v24, 5
-; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: li a3, 30
-; RV64-NEXT: mul a2, a2, a3
-; RV64-NEXT: add a2, sp, a2
-; RV64-NEXT: addi a2, a2, 16
-; RV64-NEXT: vl1r.v v0, (a2) # Unknown-size Folded Reload
+; RV64-NEXT: vrgather.vi v16, v8, 5
+; RV64-NEXT: vmv1r.v v0, v1
; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: li a3, 39
+; RV64-NEXT: li a3, 29
; RV64-NEXT: mul a2, a2, a3
; RV64-NEXT: add a2, sp, a2
; RV64-NEXT: addi a2, a2, 16
; RV64-NEXT: vl8r.v v24, (a2) # Unknown-size Folded Reload
-; RV64-NEXT: vrgather.vi v8, v24, 3, v0.t
-; RV64-NEXT: vmv.v.v v20, v8
+; RV64-NEXT: vrgather.vi v16, v24, 3, v0.t
+; RV64-NEXT: vmv.v.v v4, v16
; RV64-NEXT: lui a2, 2
; RV64-NEXT: addi a2, a2, 130
-; RV64-NEXT: vmv.s.x v8, a2
+; RV64-NEXT: vmv.s.x v13, a2
; RV64-NEXT: vsetivli zero, 16, e16, m2, ta, ma
-; RV64-NEXT: vadd.vi v16, v18, -15
+; RV64-NEXT: vadd.vi v14, v2, -15
; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: li a3, 59
+; RV64-NEXT: li a3, 37
; RV64-NEXT: mul a2, a2, a3
; RV64-NEXT: add a2, sp, a2
; RV64-NEXT: addi a2, a2, 16
-; RV64-NEXT: vl8r.v v0, (a2) # Unknown-size Folded Reload
+; RV64-NEXT: vl8r.v v24, (a2) # Unknown-size Folded Reload
; RV64-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; RV64-NEXT: vcompress.vm v24, v0, v8
-; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: li a3, 55
-; RV64-NEXT: mul a2, a2, a3
-; RV64-NEXT: add a2, sp, a2
-; RV64-NEXT: addi a2, a2, 16
-; RV64-NEXT: vl1r.v v0, (a2) # Unknown-size Folded Reload
+; RV64-NEXT: vcompress.vm v16, v24, v13
+; RV64-NEXT: vmv1r.v v0, v12
; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: li a3, 47
+; RV64-NEXT: li a3, 45
; RV64-NEXT: mul a2, a2, a3
; RV64-NEXT: add a2, sp, a2
; RV64-NEXT: addi a2, a2, 16
-; RV64-NEXT: vl8r.v v8, (a2) # Unknown-size Folded Reload
-; RV64-NEXT: vrgatherei16.vv v24, v8, v16, v0.t
+; RV64-NEXT: vl8r.v v24, (a2) # Unknown-size Folded Reload
+; RV64-NEXT: vrgatherei16.vv v16, v24, v14, v0.t
; RV64-NEXT: vsetivli zero, 6, e64, m4, tu, ma
-; RV64-NEXT: vmv.v.v v20, v24
+; RV64-NEXT: vmv.v.v v4, v16
; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: li a3, 14
-; RV64-NEXT: mul a2, a2, a3
+; RV64-NEXT: slli a3, a2, 3
+; RV64-NEXT: add a2, a3, a2
; RV64-NEXT: add a2, sp, a2
; RV64-NEXT: addi a2, a2, 16
-; RV64-NEXT: vs4r.v v20, (a2) # Unknown-size Folded Spill
+; RV64-NEXT: vs4r.v v4, (a2) # Unknown-size Folded Spill
; RV64-NEXT: lui a2, 16
; RV64-NEXT: addi a2, a2, 7
; RV64-NEXT: vsetivli zero, 4, e32, m1, ta, ma
-; RV64-NEXT: vmv.v.i v8, 6
-; RV64-NEXT: vmv.v.x v9, a2
-; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: li a3, 22
-; RV64-NEXT: mul a2, a2, a3
-; RV64-NEXT: add a2, sp, a2
-; RV64-NEXT: addi a2, a2, 16
-; RV64-NEXT: vl8r.v v16, (a2) # Unknown-size Folded Reload
+; RV64-NEXT: vmv.v.i v16, 6
+; RV64-NEXT: vmv.v.x v17, a2
; RV64-NEXT: vsetivli zero, 8, e64, m4, ta, ma
-; RV64-NEXT: vrgatherei16.vv v12, v16, v8
+; RV64-NEXT: vrgatherei16.vv v12, v8, v16
+; RV64-NEXT: vrgatherei16.vv v20, v8, v17
; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: li a3, 55
+; 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: vs4r.v v12, (a2) # Unknown-size Folded Spill
-; RV64-NEXT: vrgatherei16.vv v12, v16, v9
-; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: slli a3, a2, 5
-; RV64-NEXT: sub a2, a3, a2
-; RV64-NEXT: add a2, sp, a2
-; RV64-NEXT: addi a2, a2, 16
-; RV64-NEXT: vs4r.v v12, (a2) # Unknown-size Folded Spill
-; RV64-NEXT: vmv4r.v v8, v16
-; RV64-NEXT: vrgather.vi v12, v16, 2
+; RV64-NEXT: vs4r.v v20, (a2) # Unknown-size Folded Spill
+; RV64-NEXT: vrgather.vi v16, v8, 2
; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: li a3, 35
+; RV64-NEXT: li a3, 21
; RV64-NEXT: mul a2, a2, a3
; RV64-NEXT: add a2, sp, a2
; RV64-NEXT: addi a2, a2, 16
-; RV64-NEXT: vs4r.v v12, (a2) # Unknown-size Folded Spill
-; RV64-NEXT: vrgather.vi v12, v16, 3
-; 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: vs4r.v v12, (a2) # Unknown-size Folded Spill
+; RV64-NEXT: vs4r.v v16, (a2) # Unknown-size Folded Spill
+; RV64-NEXT: vrgather.vi v4, v8, 3
; RV64-NEXT: lui a2, 4
; RV64-NEXT: li a3, 24
; RV64-NEXT: addi a2, a2, 260
-; RV64-NEXT: vmv.s.x v0, a3
-; RV64-NEXT: addi a3, sp, 16
-; RV64-NEXT: vs1r.v v0, (a3) # Unknown-size Folded Spill
-; RV64-NEXT: vmv.s.x v24, a2
-; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: li a3, 12
-; RV64-NEXT: mul a2, a2, a3
-; RV64-NEXT: add a2, sp, a2
-; RV64-NEXT: addi a2, a2, 16
-; RV64-NEXT: vl2r.v v2, (a2) # Unknown-size Folded Reload
+; RV64-NEXT: vmv.s.x v1, a3
+; RV64-NEXT: vmv.s.x v8, a2
; RV64-NEXT: vsetivli zero, 16, e16, m2, ta, ma
-; RV64-NEXT: vadd.vi v6, v2, -14
+; RV64-NEXT: vadd.vi v10, v2, -14
; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: li a3, 59
+; RV64-NEXT: li a3, 37
; RV64-NEXT: mul a2, a2, a3
; RV64-NEXT: add a2, sp, a2
; RV64-NEXT: addi a2, a2, 16
; RV64-NEXT: vl8r.v v16, (a2) # Unknown-size Folded Reload
; RV64-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; RV64-NEXT: vcompress.vm v8, v16, v24
+; RV64-NEXT: vcompress.vm v24, v16, v8
+; RV64-NEXT: vmv1r.v v0, v1
; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: li a3, 47
+; RV64-NEXT: li a3, 45
; RV64-NEXT: mul a2, a2, a3
; RV64-NEXT: add a2, sp, a2
; RV64-NEXT: addi a2, a2, 16
; RV64-NEXT: vl8r.v v16, (a2) # Unknown-size Folded Reload
-; RV64-NEXT: vrgatherei16.vv v8, v16, v6, v0.t
-; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: li a3, 22
-; 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: vrgatherei16.vv v24, v16, v10, v0.t
+; RV64-NEXT: addi a2, sp, 16
+; RV64-NEXT: vl1r.v v29, (a2) # Unknown-size Folded Reload
+; RV64-NEXT: vmv1r.v v0, v29
; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: li a3, 30
+; RV64-NEXT: li a3, 29
; RV64-NEXT: mul a2, a2, a3
; RV64-NEXT: add a2, sp, a2
; RV64-NEXT: addi a2, a2, 16
-; RV64-NEXT: vl1r.v v1, (a2) # Unknown-size Folded Reload
-; RV64-NEXT: vmv1r.v v0, v1
+; RV64-NEXT: vl8r.v v16, (a2) # Unknown-size Folded Reload
+; RV64-NEXT: vsetivli zero, 8, e64, m4, ta, mu
+; RV64-NEXT: vrgather.vi v12, v16, 4, v0.t
; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: li a3, 39
-; 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: vl8r.v v24, (a2) # Unknown-size Folded Reload
+; RV64-NEXT: vs4r.v v12, (a2) # Unknown-size Folded Spill
+; RV64-NEXT: lui a2, 8
+; RV64-NEXT: addi a2, a2, 520
+; RV64-NEXT: vmv.s.x v28, a2
+; RV64-NEXT: vsetivli zero, 16, e16, m2, ta, ma
+; RV64-NEXT: vadd.vi v30, v2, -13
; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: li a3, 55
+; RV64-NEXT: li a3, 37
; RV64-NEXT: mul a2, a2, a3
; RV64-NEXT: add a2, sp, a2
; RV64-NEXT: addi a2, a2, 16
-; RV64-NEXT: vl4r.v v28, (a2) # Unknown-size Folded Reload
-; RV64-NEXT: vsetivli zero, 8, e64, m4, ta, mu
-; RV64-NEXT: vrgather.vi v28, v24, 4, v0.t
+; RV64-NEXT: vl8r.v v16, (a2) # Unknown-size Folded Reload
+; RV64-NEXT: vsetvli zero, zero, e64, m8, ta, mu
+; RV64-NEXT: vcompress.vm v8, v16, v28
+; RV64-NEXT: vmv1r.v v0, v1
; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: li a3, 55
+; RV64-NEXT: li a3, 45
; RV64-NEXT: mul a2, a2, a3
; RV64-NEXT: add a2, sp, a2
; RV64-NEXT: addi a2, a2, 16
-; RV64-NEXT: vs4r.v v28, (a2) # Unknown-size Folded Spill
-; RV64-NEXT: lui a2, 8
-; RV64-NEXT: addi a2, a2, 520
-; RV64-NEXT: vmv.s.x v7, a2
-; RV64-NEXT: vsetivli zero, 16, e16, m2, ta, ma
-; RV64-NEXT: vadd.vi v4, v2, -13
+; RV64-NEXT: vl8r.v v16, (a2) # Unknown-size Folded Reload
+; RV64-NEXT: vrgatherei16.vv v8, v16, v30, v0.t
; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: li a3, 59
-; RV64-NEXT: mul a2, a2, a3
; RV64-NEXT: add a2, sp, a2
; RV64-NEXT: addi a2, a2, 16
-; RV64-NEXT: vl8r.v v24, (a2) # Unknown-size Folded Reload
-; RV64-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; RV64-NEXT: vcompress.vm v8, v24, v7
-; RV64-NEXT: addi a2, sp, 16
-; RV64-NEXT: vl1r.v v0, (a2) # Unknown-size Folded Reload
-; RV64-NEXT: vrgatherei16.vv v8, v16, v4, v0.t
; RV64-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
-; RV64-NEXT: vmv1r.v v0, v1
+; RV64-NEXT: vmv1r.v v0, v29
; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: li a3, 39
+; RV64-NEXT: li a3, 29
; RV64-NEXT: mul a2, a2, a3
; RV64-NEXT: add a2, sp, a2
; RV64-NEXT: addi a2, a2, 16
; RV64-NEXT: vl8r.v v16, (a2) # Unknown-size Folded Reload
; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: slli a3, a2, 5
-; RV64-NEXT: sub a2, a3, a2
+; 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: vl4r.v v8, (a2) # Unknown-size Folded Reload
; RV64-NEXT: vsetivli zero, 8, e64, m4, ta, mu
; RV64-NEXT: vrgather.vi v8, v16, 5, v0.t
; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: slli a3, a2, 5
-; RV64-NEXT: sub a2, a3, a2
+; 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: vs4r.v v8, (a2) # Unknown-size Folded Spill
; RV64-NEXT: lui a2, 96
; RV64-NEXT: li a3, 192
-; RV64-NEXT: vmv.s.x v1, a3
+; RV64-NEXT: vmv.s.x v30, a3
; RV64-NEXT: vsetivli zero, 4, e32, m1, ta, ma
; RV64-NEXT: vmv.v.x v8, a2
-; RV64-NEXT: vmv1r.v v0, v1
+; RV64-NEXT: vmv1r.v v0, v30
; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: li a3, 35
+; RV64-NEXT: li a3, 21
; RV64-NEXT: mul a2, a2, a3
; RV64-NEXT: add a2, sp, a2
; RV64-NEXT: addi a2, a2, 16
@@ -926,147 +848,133 @@ define {<8 x i64>, <8 x i64>, <8 x i64>, <8 x i64>, <8 x i64>, <8 x i64>} @load_
; RV64-NEXT: vsetivli zero, 8, e64, m4, ta, mu
; RV64-NEXT: vrgatherei16.vv v12, v16, v8, v0.t
; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: li a3, 35
+; RV64-NEXT: li a3, 21
; RV64-NEXT: mul a2, a2, a3
; RV64-NEXT: add a2, sp, a2
; RV64-NEXT: addi a2, a2, 16
; RV64-NEXT: vs4r.v v12, (a2) # Unknown-size Folded Spill
; RV64-NEXT: li a2, 1040
; RV64-NEXT: li a3, 28
-; RV64-NEXT: vmv.s.x v20, a2
-; RV64-NEXT: vmv.s.x v0, a3
-; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: li a3, 30
-; RV64-NEXT: mul a2, a2, a3
-; RV64-NEXT: add a2, sp, a2
-; RV64-NEXT: addi a2, a2, 16
-; RV64-NEXT: vs1r.v v0, (a2) # Unknown-size Folded Spill
+; RV64-NEXT: vmv.s.x v31, a2
+; RV64-NEXT: vmv.s.x v1, a3
; RV64-NEXT: vsetivli zero, 16, e16, m2, ta, ma
-; RV64-NEXT: vadd.vi v22, v2, -12
+; RV64-NEXT: vadd.vi v28, v2, -12
; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: li a3, 59
+; RV64-NEXT: li a3, 37
; RV64-NEXT: mul a2, a2, a3
; RV64-NEXT: add a2, sp, a2
; RV64-NEXT: addi a2, a2, 16
-; RV64-NEXT: vl8r.v v24, (a2) # Unknown-size Folded Reload
+; RV64-NEXT: vl8r.v v16, (a2) # Unknown-size Folded Reload
; RV64-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; RV64-NEXT: vcompress.vm v8, v24, v20
+; RV64-NEXT: vcompress.vm v8, v16, v31
+; RV64-NEXT: vmv1r.v v0, v1
; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: li a3, 47
+; RV64-NEXT: li a3, 45
; RV64-NEXT: mul a2, a2, a3
; RV64-NEXT: add a2, sp, a2
; RV64-NEXT: addi a2, a2, 16
-; RV64-NEXT: vl8r.v v24, (a2) # Unknown-size Folded Reload
-; RV64-NEXT: vrgatherei16.vv v8, v24, v22, v0.t
+; RV64-NEXT: vl8r.v v16, (a2) # Unknown-size Folded Reload
+; RV64-NEXT: vrgatherei16.vv v8, v16, v28, v0.t
; RV64-NEXT: lui a2, 112
; RV64-NEXT: addi a2, a2, 1
; RV64-NEXT: vsetivli zero, 4, e32, m1, ta, ma
; RV64-NEXT: vmv.v.x v12, a2
-; RV64-NEXT: vmv1r.v v0, v1
+; RV64-NEXT: vmv1r.v v0, v30
; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: slli a2, a2, 3
+; RV64-NEXT: li a3, 29
+; RV64-NEXT: mul a2, a2, a3
; RV64-NEXT: add a2, sp, a2
; RV64-NEXT: addi a2, a2, 16
-; RV64-NEXT: vl4r.v v4, (a2) # Unknown-size Folded Reload
+; RV64-NEXT: vl8r.v v16, (a2) # Unknown-size Folded Reload
; RV64-NEXT: vsetivli zero, 8, e64, m4, ta, mu
; RV64-NEXT: vrgatherei16.vv v4, v16, v12, v0.t
; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: li a3, 55
-; 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 v12, (a2) # Unknown-size Folded Reload
-; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: li a3, 22
-; RV64-NEXT: mul a2, a2, a3
-; RV64-NEXT: add a2, sp, a2
-; RV64-NEXT: addi a2, a2, 16
-; RV64-NEXT: vl8r.v v16, (a2) # Unknown-size Folded Reload
; RV64-NEXT: vsetivli zero, 5, e64, m4, tu, ma
-; RV64-NEXT: vmv.v.v v12, v16
+; RV64-NEXT: vmv.v.v v12, v24
; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: li a3, 55
-; 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: vs4r.v v12, (a2) # Unknown-size Folded Spill
; RV64-NEXT: addi a1, a1, -2016
; RV64-NEXT: vmv.s.x v12, a1
; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: li a2, 59
+; 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 v24, (a1) # Unknown-size Folded Reload
+; RV64-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
; RV64-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; RV64-NEXT: vcompress.vm v16, v24, v12
+; RV64-NEXT: vcompress.vm v24, v16, v12
; RV64-NEXT: vsetvli zero, zero, e16, m2, ta, ma
; RV64-NEXT: vadd.vi v12, v2, -11
+; RV64-NEXT: vmv1r.v v0, v1
; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: li a2, 30
+; 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: vl8r.v v16, (a1) # Unknown-size Folded Reload
+; RV64-NEXT: vsetvli zero, zero, e64, m8, ta, mu
+; RV64-NEXT: vrgatherei16.vv v24, v16, v12, v0.t
; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: li a2, 47
+; 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 v24, (a1) # Unknown-size Folded Reload
-; RV64-NEXT: vsetvli zero, zero, e64, m8, ta, mu
-; RV64-NEXT: vrgatherei16.vv v16, v24, v12, v0.t
+; RV64-NEXT: vl4r.v v12, (a1) # Unknown-size Folded Reload
; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: slli a2, a1, 5
-; RV64-NEXT: sub a1, a2, a1
; RV64-NEXT: add a1, sp, a1
; RV64-NEXT: addi a1, a1, 16
-; RV64-NEXT: vl4r.v v12, (a1) # Unknown-size Folded Reload
-; RV64-NEXT: addi a1, sp, 16
-; RV64-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; RV64-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
; RV64-NEXT: vsetivli zero, 5, e64, m4, tu, ma
-; RV64-NEXT: vmv.v.v v12, v24
+; RV64-NEXT: vmv.v.v v12, v16
; RV64-NEXT: csrr a1, vlenb
-; RV64-NEXT: li a2, 35
+; 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 v20, (a1) # Unknown-size Folded Reload
-; RV64-NEXT: vmv.v.v v20, v8
-; RV64-NEXT: vmv4r.v v8, v4
-; RV64-NEXT: vmv.v.v v8, v16
+; RV64-NEXT: vl4r.v v16, (a1) # Unknown-size Folded Reload
+; RV64-NEXT: vmv.v.v v16, v8
+; RV64-NEXT: vmv.v.v v4, v24
; RV64-NEXT: addi a1, a0, 256
; RV64-NEXT: vsetivli zero, 8, e64, m4, ta, ma
-; RV64-NEXT: vse64.v v20, (a1)
+; RV64-NEXT: vse64.v v16, (a1)
; RV64-NEXT: addi a1, a0, 320
-; RV64-NEXT: vse64.v v8, (a1)
+; RV64-NEXT: vse64.v v4, (a1)
; RV64-NEXT: addi a1, a0, 192
; RV64-NEXT: vse64.v v12, (a1)
; RV64-NEXT: addi a1, a0, 128
; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: li a3, 55
-; 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: addi a1, a0, 64
; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: li a3, 14
-; RV64-NEXT: mul a2, a2, a3
+; RV64-NEXT: slli a3, a2, 3
+; 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: li a2, 18
+; RV64-NEXT: li a2, 13
; 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, 67
+; RV64-NEXT: li a1, 53
; RV64-NEXT: mul a0, a0, a1
; RV64-NEXT: add sp, sp, a0
; RV64-NEXT: .cfi_def_cfa sp, 16
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 46c2033d28b387..de90d250bfb299 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-nearbyint-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-nearbyint-vp.ll
@@ -578,12 +578,6 @@ define <32 x double> @vp_nearbyint_v32f64(<32 x double> %va, <32 x i1> %m, i32 z
; 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: vmv1r.v v0, v6
; CHECK-NEXT: lui a2, %hi(.LCPI26_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI26_0)(a2)
@@ -595,14 +589,10 @@ define <32 x double> @vp_nearbyint_v32f64(<32 x double> %va, <32 x i1> %m, i32 z
; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; 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: addi a2, a0, -16
; CHECK-NEXT: sltu a0, a0, a2
; CHECK-NEXT: addi a0, a0, -1
; CHECK-NEXT: and a0, a0, a2
-; 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: fsflags a1
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
@@ -620,12 +610,6 @@ define <32 x double> @vp_nearbyint_v32f64(<32 x double> %va, <32 x i1> %m, i32 z
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
; CHECK-NEXT: vfsgnj.vv v16, v24, v16, v0.t
; CHECK-NEXT: fsflags a0
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add sp, sp, a0
-; CHECK-NEXT: .cfi_def_cfa sp, 16
-; CHECK-NEXT: addi sp, sp, 16
-; CHECK-NEXT: .cfi_def_cfa_offset 0
; 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
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 8bf30f8f0d072b..b6f911b8e477a2 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-reduction-fp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-reduction-fp.ll
@@ -1945,12 +1945,6 @@ declare float @llvm.vector.reduce.fminimum.v64f32(<64 x float>)
define float @vreduce_fminimum_v64f32(ptr %x) {
; CHECK-LABEL: vreduce_fminimum_v64f32:
; 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: addi a1, a0, 128
; CHECK-NEXT: li a2, 32
; CHECK-NEXT: vsetvli zero, a2, e32, m8, ta, ma
@@ -1959,29 +1953,19 @@ define float @vreduce_fminimum_v64f32(ptr %x) {
; CHECK-NEXT: vmfeq.vv v0, v16, v16
; CHECK-NEXT: vmfeq.vv v7, v24, v24
; CHECK-NEXT: vmerge.vvm v8, v16, v24, v0
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: vmv1r.v v0, v7
-; CHECK-NEXT: vmerge.vvm v8, v24, v16, v0
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vfmin.vv v8, v8, v16
+; CHECK-NEXT: vmerge.vvm v16, v24, v16, v0
+; CHECK-NEXT: vfmin.vv v8, v16, v8
; CHECK-NEXT: vmfne.vv v16, v8, v8
; CHECK-NEXT: vcpop.m a0, v16
; CHECK-NEXT: beqz a0, .LBB119_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: lui a0, 523264
; CHECK-NEXT: fmv.w.x fa0, a0
-; CHECK-NEXT: j .LBB119_3
+; CHECK-NEXT: ret
; CHECK-NEXT: .LBB119_2:
; CHECK-NEXT: vfredmin.vs v8, v8, v8
; CHECK-NEXT: vfmv.f.s fa0, v8
-; CHECK-NEXT: .LBB119_3:
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add sp, sp, a0
-; CHECK-NEXT: .cfi_def_cfa sp, 16
-; CHECK-NEXT: addi sp, sp, 16
-; CHECK-NEXT: .cfi_def_cfa_offset 0
; CHECK-NEXT: ret
%v = load <64 x float>, ptr %x
%red = call float @llvm.vector.reduce.fminimum.v64f32(<64 x float> %v)
@@ -2019,57 +2003,57 @@ define float @vreduce_fminimum_v128f32(ptr %x) {
; CHECK-NEXT: li a1, 32
; CHECK-NEXT: addi a2, a0, 128
; CHECK-NEXT: vsetvli zero, a1, e32, m8, ta, ma
-; CHECK-NEXT: vle32.v v8, (a2)
+; CHECK-NEXT: vle32.v v24, (a2)
; CHECK-NEXT: addi a1, a0, 384
; CHECK-NEXT: vle32.v v16, (a1)
; CHECK-NEXT: addi a1, a0, 256
-; CHECK-NEXT: vle32.v v24, (a0)
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: vmfeq.vv v0, v8, v8
-; CHECK-NEXT: vmfeq.vv v7, v16, v16
-; CHECK-NEXT: vmerge.vvm v24, v8, v16, v0
+; CHECK-NEXT: vle32.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 v24, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: vle32.v v24, (a1)
+; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vmfeq.vv v0, v24, v24
+; CHECK-NEXT: vmfeq.vv v7, v16, v16
+; CHECK-NEXT: vmerge.vvm v8, v24, v16, v0
+; CHECK-NEXT: addi a0, sp, 16
+; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vle32.v v8, (a1)
; CHECK-NEXT: vmv1r.v v0, v7
-; CHECK-NEXT: vmerge.vvm v16, v16, v8, v0
+; CHECK-NEXT: vmerge.vvm v16, v16, v24, v0
+; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vfmin.vv v24, v16, 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: vfmin.vv v8, v16, v8
+; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vmfeq.vv v0, v16, v16
+; CHECK-NEXT: vmfeq.vv v7, v8, 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: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vmerge.vvm v16, v16, v8, v0
; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vmfeq.vv v0, v8, v8
-; CHECK-NEXT: vmfeq.vv v7, v24, v24
-; CHECK-NEXT: vmerge.vvm v16, v8, v24, v0
+; CHECK-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: vmv1r.v v0, v7
-; CHECK-NEXT: vmerge.vvm v8, v24, v8, v0
-; CHECK-NEXT: vfmin.vv v16, v8, v16
-; CHECK-NEXT: vmfeq.vv v0, v16, 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: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vmerge.vvm v8, v8, v16, v0
+; CHECK-NEXT: addi a0, sp, 16
+; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vfmin.vv v16, v8, v16
+; CHECK-NEXT: vmfeq.vv v0, v16, v16
; CHECK-NEXT: vmfeq.vv v7, v24, v24
; CHECK-NEXT: vmerge.vvm v8, v16, v24, v0
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: vmv1r.v v0, v7
-; CHECK-NEXT: vmerge.vvm v8, v24, v16, v0
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vfmin.vv v8, v8, v16
+; CHECK-NEXT: vmerge.vvm v16, v24, v16, v0
+; CHECK-NEXT: vfmin.vv v8, v16, v8
; CHECK-NEXT: vmfne.vv v16, v8, v8
; CHECK-NEXT: vcpop.m a0, v16
; CHECK-NEXT: beqz a0, .LBB121_2
@@ -2265,12 +2249,6 @@ declare double @llvm.vector.reduce.fminimum.v32f64(<32 x double>)
define double @vreduce_fminimum_v32f64(ptr %x) {
; CHECK-LABEL: vreduce_fminimum_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: addi a1, a0, 128
; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; CHECK-NEXT: vle64.v v16, (a0)
@@ -2278,29 +2256,19 @@ define double @vreduce_fminimum_v32f64(ptr %x) {
; CHECK-NEXT: vmfeq.vv v0, v16, v16
; CHECK-NEXT: vmfeq.vv v7, v24, v24
; CHECK-NEXT: vmerge.vvm v8, v16, v24, v0
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: vmv1r.v v0, v7
-; CHECK-NEXT: vmerge.vvm v8, v24, v16, v0
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vfmin.vv v8, v8, v16
+; CHECK-NEXT: vmerge.vvm v16, v24, v16, v0
+; CHECK-NEXT: vfmin.vv v8, v16, v8
; CHECK-NEXT: vmfne.vv v16, v8, v8
; CHECK-NEXT: vcpop.m a0, v16
; CHECK-NEXT: beqz a0, .LBB131_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: lui a0, %hi(.LCPI131_0)
; CHECK-NEXT: fld fa0, %lo(.LCPI131_0)(a0)
-; CHECK-NEXT: j .LBB131_3
+; CHECK-NEXT: ret
; CHECK-NEXT: .LBB131_2:
; CHECK-NEXT: vfredmin.vs v8, v8, v8
; CHECK-NEXT: vfmv.f.s fa0, v8
-; CHECK-NEXT: .LBB131_3:
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add sp, sp, a0
-; CHECK-NEXT: .cfi_def_cfa sp, 16
-; CHECK-NEXT: addi sp, sp, 16
-; CHECK-NEXT: .cfi_def_cfa_offset 0
; CHECK-NEXT: ret
%v = load <32 x double>, ptr %x
%red = call double @llvm.vector.reduce.fminimum.v32f64(<32 x double> %v)
@@ -2336,57 +2304,57 @@ define double @vreduce_fminimum_v64f64(ptr %x) {
; 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 v8, (a1)
+; CHECK-NEXT: vle64.v v24, (a1)
; CHECK-NEXT: addi a1, a0, 384
; CHECK-NEXT: vle64.v v16, (a1)
; CHECK-NEXT: addi a1, a0, 256
-; CHECK-NEXT: vle64.v v24, (a0)
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: vmfeq.vv v0, v8, v8
-; CHECK-NEXT: vmfeq.vv v7, v16, v16
-; CHECK-NEXT: vmerge.vvm v24, v8, v16, v0
+; CHECK-NEXT: vle64.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 v24, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: vle64.v v24, (a1)
+; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vmfeq.vv v0, v24, v24
+; CHECK-NEXT: vmfeq.vv v7, v16, v16
+; CHECK-NEXT: vmerge.vvm v8, v24, v16, v0
+; CHECK-NEXT: addi a0, sp, 16
+; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vle64.v v8, (a1)
; CHECK-NEXT: vmv1r.v v0, v7
-; CHECK-NEXT: vmerge.vvm v16, v16, v8, v0
+; CHECK-NEXT: vmerge.vvm v16, v16, v24, v0
+; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vfmin.vv v24, v16, 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: vfmin.vv v8, v16, v8
+; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vmfeq.vv v0, v16, v16
+; CHECK-NEXT: vmfeq.vv v7, v8, 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: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vmerge.vvm v16, v16, v8, v0
; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vmfeq.vv v0, v8, v8
-; CHECK-NEXT: vmfeq.vv v7, v24, v24
-; CHECK-NEXT: vmerge.vvm v16, v8, v24, v0
+; CHECK-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: vmv1r.v v0, v7
-; CHECK-NEXT: vmerge.vvm v8, v24, v8, v0
-; CHECK-NEXT: vfmin.vv v16, v8, v16
-; CHECK-NEXT: vmfeq.vv v0, v16, 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: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vmerge.vvm v8, v8, v16, v0
+; CHECK-NEXT: addi a0, sp, 16
+; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vfmin.vv v16, v8, v16
+; CHECK-NEXT: vmfeq.vv v0, v16, v16
; CHECK-NEXT: vmfeq.vv v7, v24, v24
; CHECK-NEXT: vmerge.vvm v8, v16, v24, v0
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: vmv1r.v v0, v7
-; CHECK-NEXT: vmerge.vvm v8, v24, v16, v0
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vfmin.vv v8, v8, v16
+; CHECK-NEXT: vmerge.vvm v16, v24, v16, v0
+; CHECK-NEXT: vfmin.vv v8, v16, v8
; CHECK-NEXT: vmfne.vv v16, v8, v8
; CHECK-NEXT: vcpop.m a0, v16
; CHECK-NEXT: beqz a0, .LBB133_2
@@ -2659,12 +2627,6 @@ declare float @llvm.vector.reduce.fmaximum.v64f32(<64 x float>)
define float @vreduce_fmaximum_v64f32(ptr %x) {
; CHECK-LABEL: vreduce_fmaximum_v64f32:
; 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: addi a1, a0, 128
; CHECK-NEXT: li a2, 32
; CHECK-NEXT: vsetvli zero, a2, e32, m8, ta, ma
@@ -2673,29 +2635,19 @@ define float @vreduce_fmaximum_v64f32(ptr %x) {
; CHECK-NEXT: vmfeq.vv v0, v16, v16
; CHECK-NEXT: vmfeq.vv v7, v24, v24
; CHECK-NEXT: vmerge.vvm v8, v16, v24, v0
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: vmv1r.v v0, v7
-; CHECK-NEXT: vmerge.vvm v8, v24, v16, v0
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vfmax.vv v8, v8, v16
+; CHECK-NEXT: vmerge.vvm v16, v24, v16, v0
+; CHECK-NEXT: vfmax.vv v8, v16, v8
; CHECK-NEXT: vmfne.vv v16, v8, v8
; CHECK-NEXT: vcpop.m a0, v16
; CHECK-NEXT: beqz a0, .LBB147_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: lui a0, 523264
; CHECK-NEXT: fmv.w.x fa0, a0
-; CHECK-NEXT: j .LBB147_3
+; CHECK-NEXT: ret
; CHECK-NEXT: .LBB147_2:
; CHECK-NEXT: vfredmax.vs v8, v8, v8
; CHECK-NEXT: vfmv.f.s fa0, v8
-; CHECK-NEXT: .LBB147_3:
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add sp, sp, a0
-; CHECK-NEXT: .cfi_def_cfa sp, 16
-; CHECK-NEXT: addi sp, sp, 16
-; CHECK-NEXT: .cfi_def_cfa_offset 0
; CHECK-NEXT: ret
%v = load <64 x float>, ptr %x
%red = call float @llvm.vector.reduce.fmaximum.v64f32(<64 x float> %v)
@@ -2733,57 +2685,57 @@ define float @vreduce_fmaximum_v128f32(ptr %x) {
; CHECK-NEXT: li a1, 32
; CHECK-NEXT: addi a2, a0, 128
; CHECK-NEXT: vsetvli zero, a1, e32, m8, ta, ma
-; CHECK-NEXT: vle32.v v8, (a2)
+; CHECK-NEXT: vle32.v v24, (a2)
; CHECK-NEXT: addi a1, a0, 384
; CHECK-NEXT: vle32.v v16, (a1)
; CHECK-NEXT: addi a1, a0, 256
-; CHECK-NEXT: vle32.v v24, (a0)
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: vmfeq.vv v0, v8, v8
-; CHECK-NEXT: vmfeq.vv v7, v16, v16
-; CHECK-NEXT: vmerge.vvm v24, v8, v16, v0
+; CHECK-NEXT: vle32.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 v24, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: vle32.v v24, (a1)
+; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vmfeq.vv v0, v24, v24
+; CHECK-NEXT: vmfeq.vv v7, v16, v16
+; CHECK-NEXT: vmerge.vvm v8, v24, v16, v0
+; CHECK-NEXT: addi a0, sp, 16
+; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vle32.v v8, (a1)
; CHECK-NEXT: vmv1r.v v0, v7
-; CHECK-NEXT: vmerge.vvm v16, v16, v8, v0
+; CHECK-NEXT: vmerge.vvm v16, v16, v24, v0
+; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vfmax.vv v24, v16, 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: vfmax.vv v8, v16, v8
+; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vmfeq.vv v0, v16, v16
+; CHECK-NEXT: vmfeq.vv v7, v8, 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: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vmerge.vvm v16, v16, v8, v0
; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vmfeq.vv v0, v8, v8
-; CHECK-NEXT: vmfeq.vv v7, v24, v24
-; CHECK-NEXT: vmerge.vvm v16, v8, v24, v0
+; CHECK-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: vmv1r.v v0, v7
-; CHECK-NEXT: vmerge.vvm v8, v24, v8, v0
-; CHECK-NEXT: vfmax.vv v16, v8, v16
-; CHECK-NEXT: vmfeq.vv v0, v16, 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: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vmerge.vvm v8, v8, v16, v0
+; CHECK-NEXT: addi a0, sp, 16
+; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vfmax.vv v16, v8, v16
+; CHECK-NEXT: vmfeq.vv v0, v16, v16
; CHECK-NEXT: vmfeq.vv v7, v24, v24
; CHECK-NEXT: vmerge.vvm v8, v16, v24, v0
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: vmv1r.v v0, v7
-; CHECK-NEXT: vmerge.vvm v8, v24, v16, v0
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vfmax.vv v8, v8, v16
+; CHECK-NEXT: vmerge.vvm v16, v24, v16, v0
+; CHECK-NEXT: vfmax.vv v8, v16, v8
; CHECK-NEXT: vmfne.vv v16, v8, v8
; CHECK-NEXT: vcpop.m a0, v16
; CHECK-NEXT: beqz a0, .LBB149_2
@@ -2979,12 +2931,6 @@ declare double @llvm.vector.reduce.fmaximum.v32f64(<32 x double>)
define double @vreduce_fmaximum_v32f64(ptr %x) {
; CHECK-LABEL: vreduce_fmaximum_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: addi a1, a0, 128
; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; CHECK-NEXT: vle64.v v16, (a0)
@@ -2992,29 +2938,19 @@ define double @vreduce_fmaximum_v32f64(ptr %x) {
; CHECK-NEXT: vmfeq.vv v0, v16, v16
; CHECK-NEXT: vmfeq.vv v7, v24, v24
; CHECK-NEXT: vmerge.vvm v8, v16, v24, v0
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: vmv1r.v v0, v7
-; CHECK-NEXT: vmerge.vvm v8, v24, v16, v0
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vfmax.vv v8, v8, v16
+; CHECK-NEXT: vmerge.vvm v16, v24, v16, v0
+; CHECK-NEXT: vfmax.vv v8, v16, v8
; CHECK-NEXT: vmfne.vv v16, v8, v8
; CHECK-NEXT: vcpop.m a0, v16
; CHECK-NEXT: beqz a0, .LBB159_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: lui a0, %hi(.LCPI159_0)
; CHECK-NEXT: fld fa0, %lo(.LCPI159_0)(a0)
-; CHECK-NEXT: j .LBB159_3
+; CHECK-NEXT: ret
; CHECK-NEXT: .LBB159_2:
; CHECK-NEXT: vfredmax.vs v8, v8, v8
; CHECK-NEXT: vfmv.f.s fa0, v8
-; CHECK-NEXT: .LBB159_3:
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add sp, sp, a0
-; CHECK-NEXT: .cfi_def_cfa sp, 16
-; CHECK-NEXT: addi sp, sp, 16
-; CHECK-NEXT: .cfi_def_cfa_offset 0
; CHECK-NEXT: ret
%v = load <32 x double>, ptr %x
%red = call double @llvm.vector.reduce.fmaximum.v32f64(<32 x double> %v)
@@ -3050,57 +2986,57 @@ define double @vreduce_fmaximum_v64f64(ptr %x) {
; 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 v8, (a1)
+; CHECK-NEXT: vle64.v v24, (a1)
; CHECK-NEXT: addi a1, a0, 384
; CHECK-NEXT: vle64.v v16, (a1)
; CHECK-NEXT: addi a1, a0, 256
-; CHECK-NEXT: vle64.v v24, (a0)
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: vmfeq.vv v0, v8, v8
-; CHECK-NEXT: vmfeq.vv v7, v16, v16
-; CHECK-NEXT: vmerge.vvm v24, v8, v16, v0
+; CHECK-NEXT: vle64.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 v24, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: vle64.v v24, (a1)
+; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vmfeq.vv v0, v24, v24
+; CHECK-NEXT: vmfeq.vv v7, v16, v16
+; CHECK-NEXT: vmerge.vvm v8, v24, v16, v0
+; CHECK-NEXT: addi a0, sp, 16
+; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vle64.v v8, (a1)
; CHECK-NEXT: vmv1r.v v0, v7
-; CHECK-NEXT: vmerge.vvm v16, v16, v8, v0
+; CHECK-NEXT: vmerge.vvm v16, v16, v24, v0
+; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vfmax.vv v24, v16, 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: vfmax.vv v8, v16, v8
+; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vmfeq.vv v0, v16, v16
+; CHECK-NEXT: vmfeq.vv v7, v8, 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: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vmerge.vvm v16, v16, v8, v0
; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vmfeq.vv v0, v8, v8
-; CHECK-NEXT: vmfeq.vv v7, v24, v24
-; CHECK-NEXT: vmerge.vvm v16, v8, v24, v0
+; CHECK-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: vmv1r.v v0, v7
-; CHECK-NEXT: vmerge.vvm v8, v24, v8, v0
-; CHECK-NEXT: vfmax.vv v16, v8, v16
-; CHECK-NEXT: vmfeq.vv v0, v16, 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: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vmerge.vvm v8, v8, v16, v0
+; CHECK-NEXT: addi a0, sp, 16
+; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vfmax.vv v16, v8, v16
+; CHECK-NEXT: vmfeq.vv v0, v16, v16
; CHECK-NEXT: vmfeq.vv v7, v24, v24
; CHECK-NEXT: vmerge.vvm v8, v16, v24, v0
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: vmv1r.v v0, v7
-; CHECK-NEXT: vmerge.vvm v8, v24, v16, v0
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vfmax.vv v8, v8, v16
+; CHECK-NEXT: vmerge.vvm v16, v24, v16, v0
+; CHECK-NEXT: vfmax.vv v8, v16, v8
; CHECK-NEXT: vmfne.vv v16, v8, v8
; CHECK-NEXT: vcpop.m a0, v16
; CHECK-NEXT: beqz a0, .LBB161_2
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 b8617fda3aa7ec..4c4497482e75cd 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-rint-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-rint-vp.ll
@@ -526,24 +526,14 @@ define <32 x double> @vp_rint_v32f64(<32 x double> %va, <32 x i1> %m, i32 zeroex
; 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: vmv1r.v v0, v6
; CHECK-NEXT: vsetvli zero, a1, e64, m8, ta, ma
; 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: lui a1, %hi(.LCPI26_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI26_0)(a1)
; CHECK-NEXT: addi a1, a0, -16
; CHECK-NEXT: sltu a0, a0, a1
; CHECK-NEXT: addi a0, a0, -1
-; CHECK-NEXT: addi a2, sp, 16
-; CHECK-NEXT: vl8r.v v24, (a2) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
; CHECK-NEXT: vmflt.vf v6, v24, fa5, v0.t
; CHECK-NEXT: and a0, a0, a1
@@ -564,12 +554,6 @@ define <32 x double> @vp_rint_v32f64(<32 x double> %va, <32 x i1> %m, i32 zeroex
; CHECK-NEXT: vfcvt.f.x.v v24, v24, v0.t
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, mu
; CHECK-NEXT: vfsgnj.vv v16, v24, v16, v0.t
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add sp, sp, a0
-; CHECK-NEXT: .cfi_def_cfa sp, 16
-; CHECK-NEXT: addi sp, sp, 16
-; CHECK-NEXT: .cfi_def_cfa_offset 0
; CHECK-NEXT: ret
%v = call <32 x double> @llvm.vp.rint.v32f64(<32 x double> %va, <32 x i1> %m, i32 %evl)
ret <32 x double> %v
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 820a05e3d6042b..5136505cdea9f9 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-round-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-round-vp.ll
@@ -737,29 +737,18 @@ 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: vmv1r.v v6, v0
; CHECK-NEXT: li a2, 16
; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
-; CHECK-NEXT: vslidedown.vi v24, v0, 2
+; 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: vmv1r.v v0, v25
+; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; CHECK-NEXT: vfabs.v v16, v8, v0.t
+; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: lui a1, %hi(.LCPI26_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI26_0)(a1)
; CHECK-NEXT: addi a1, a0, -16
@@ -767,43 +756,28 @@ define <32 x double> @vp_round_v32f64(<32 x double> %va, <32 x i1> %m, i32 zeroe
; CHECK-NEXT: addi a0, a0, -1
; CHECK-NEXT: and a0, a0, a1
; 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, v25
+; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
-; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
+; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: fsrm a1
-; 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: addi a1, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
-; CHECK-NEXT: vmv1r.v v0, v24
-; 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: vfsgnj.vv v8, v24, v8, v0.t
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; 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, v24
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
-; 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: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 4
-; CHECK-NEXT: add sp, sp, a0
-; CHECK-NEXT: .cfi_def_cfa sp, 16
-; CHECK-NEXT: addi sp, sp, 16
-; CHECK-NEXT: .cfi_def_cfa_offset 0
+; CHECK-NEXT: vfsgnj.vv v16, v24, v16, v0.t
; CHECK-NEXT: ret
%v = call <32 x double> @llvm.vp.round.v32f64(<32 x double> %va, <32 x i1> %m, i32 %evl)
ret <32 x double> %v
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 8391c7939180a0..726541dc6d9d44 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-roundeven-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-roundeven-vp.ll
@@ -737,29 +737,18 @@ 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: vmv1r.v v6, v0
; CHECK-NEXT: li a2, 16
; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
-; CHECK-NEXT: vslidedown.vi v24, v0, 2
+; 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: vmv1r.v v0, v25
+; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; CHECK-NEXT: vfabs.v v16, v8, v0.t
+; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: lui a1, %hi(.LCPI26_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI26_0)(a1)
; CHECK-NEXT: addi a1, a0, -16
@@ -767,43 +756,28 @@ define <32 x double> @vp_roundeven_v32f64(<32 x double> %va, <32 x i1> %m, i32 z
; CHECK-NEXT: addi a0, a0, -1
; CHECK-NEXT: and a0, a0, a1
; 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, v25
+; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
-; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
+; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: fsrm a1
-; 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: addi a1, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
-; CHECK-NEXT: vmv1r.v v0, v24
-; 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: vfsgnj.vv v8, v24, v8, v0.t
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; 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, v24
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
-; 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: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 4
-; CHECK-NEXT: add sp, sp, a0
-; CHECK-NEXT: .cfi_def_cfa sp, 16
-; CHECK-NEXT: addi sp, sp, 16
-; CHECK-NEXT: .cfi_def_cfa_offset 0
+; CHECK-NEXT: vfsgnj.vv v16, v24, v16, v0.t
; CHECK-NEXT: ret
%v = call <32 x double> @llvm.vp.roundeven.v32f64(<32 x double> %va, <32 x i1> %m, i32 %evl)
ret <32 x double> %v
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 8c38d244602655..3ac21ce8793078 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-roundtozero-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-roundtozero-vp.ll
@@ -737,29 +737,18 @@ 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: vmv1r.v v6, v0
; CHECK-NEXT: li a2, 16
; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
-; CHECK-NEXT: vslidedown.vi v24, v0, 2
+; 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: vmv1r.v v0, v25
+; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: vsetvli zero, a1, e64, m8, ta, ma
-; CHECK-NEXT: vfabs.v v16, v8, v0.t
+; CHECK-NEXT: vfabs.v v24, v8, v0.t
; CHECK-NEXT: lui a1, %hi(.LCPI26_0)
; CHECK-NEXT: fld fa5, %lo(.LCPI26_0)(a1)
; CHECK-NEXT: addi a1, a0, -16
@@ -767,43 +756,28 @@ define <32 x double> @vp_roundtozero_v32f64(<32 x double> %va, <32 x i1> %m, i32
; CHECK-NEXT: addi a0, a0, -1
; CHECK-NEXT: and a0, a0, a1
; 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, v25
+; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
-; CHECK-NEXT: vfcvt.x.f.v v16, v8, v0.t
+; CHECK-NEXT: vfcvt.x.f.v v24, v8, v0.t
; CHECK-NEXT: fsrm a1
-; 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: addi a1, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
-; CHECK-NEXT: vmv1r.v v0, v24
-; 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: vfsgnj.vv v8, v24, v8, v0.t
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; 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, v24
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
-; 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: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 4
-; CHECK-NEXT: add sp, sp, a0
-; CHECK-NEXT: .cfi_def_cfa sp, 16
-; CHECK-NEXT: addi sp, sp, 16
-; CHECK-NEXT: .cfi_def_cfa_offset 0
+; CHECK-NEXT: vfsgnj.vv v16, v24, v16, v0.t
; CHECK-NEXT: ret
%v = call <32 x double> @llvm.vp.roundtozero.v32f64(<32 x double> %va, <32 x i1> %m, i32 %evl)
ret <32 x double> %v
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 03d5762b4903ef..90d4a5e511c870 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
@@ -1063,49 +1063,39 @@ define <128 x i1> @fcmp_oeq_vv_v128f16(<128 x half> %va, <128 x half> %vb, <128
; ZVFH-NEXT: addi sp, sp, -16
; ZVFH-NEXT: .cfi_def_cfa_offset 16
; ZVFH-NEXT: csrr a1, vlenb
-; ZVFH-NEXT: slli a1, a1, 4
-; ZVFH-NEXT: sub sp, sp, a1
-; ZVFH-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
-; 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: 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: addi a1, sp, 16
; ZVFH-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
; ZVFH-NEXT: addi a1, a0, 128
; ZVFH-NEXT: li a3, 64
; ZVFH-NEXT: vsetvli zero, a3, e16, m8, ta, ma
; ZVFH-NEXT: vle16.v v16, (a1)
-; ZVFH-NEXT: addi a1, sp, 16
-; ZVFH-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
-; ZVFH-NEXT: vle16.v v16, (a0)
+; ZVFH-NEXT: vle16.v v24, (a0)
; ZVFH-NEXT: mv a0, a2
; ZVFH-NEXT: vsetivli zero, 8, e8, m1, ta, ma
-; ZVFH-NEXT: vslidedown.vi v24, v0, 8
+; ZVFH-NEXT: vslidedown.vi v7, v0, 8
; ZVFH-NEXT: bltu a2, a3, .LBB43_2
; ZVFH-NEXT: # %bb.1:
; ZVFH-NEXT: li a0, 64
; ZVFH-NEXT: .LBB43_2:
; ZVFH-NEXT: vsetvli zero, a0, e16, m8, ta, ma
-; ZVFH-NEXT: vmfeq.vv v7, v8, v16, v0.t
+; ZVFH-NEXT: vmfeq.vv v6, 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, v24
-; 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 v16, (a1) # Unknown-size Folded Reload
+; ZVFH-NEXT: vmv1r.v v0, v7
; 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: vmfeq.vv v8, v16, v24, v0.t
+; ZVFH-NEXT: vmfeq.vv v8, v24, v16, v0.t
; ZVFH-NEXT: vsetivli zero, 16, e8, m1, ta, ma
-; ZVFH-NEXT: vslideup.vi v7, v8, 8
-; ZVFH-NEXT: vmv.v.v v0, v7
+; ZVFH-NEXT: vslideup.vi v6, v8, 8
+; ZVFH-NEXT: vmv.v.v v0, v6
; ZVFH-NEXT: csrr a0, vlenb
-; ZVFH-NEXT: slli a0, a0, 4
+; ZVFH-NEXT: slli a0, a0, 3
; ZVFH-NEXT: add sp, sp, a0
; ZVFH-NEXT: .cfi_def_cfa sp, 16
; ZVFH-NEXT: addi sp, sp, 16
@@ -1143,7 +1133,7 @@ define <128 x i1> @fcmp_oeq_vv_v128f16(<128 x half> %va, <128 x half> %vb, <128
; ZVFHMIN32-NEXT: addi s0, sp, 896
; ZVFHMIN32-NEXT: .cfi_def_cfa s0, 0
; ZVFHMIN32-NEXT: csrr a1, vlenb
-; ZVFHMIN32-NEXT: li a2, 30
+; ZVFHMIN32-NEXT: li a2, 28
; ZVFHMIN32-NEXT: mul a1, a1, a2
; ZVFHMIN32-NEXT: sub sp, sp, a1
; ZVFHMIN32-NEXT: andi sp, sp, -128
@@ -1315,67 +1305,66 @@ define <128 x i1> @fcmp_oeq_vv_v128f16(<128 x half> %va, <128 x half> %vb, <128
; ZVFHMIN32-NEXT: vsetivli zero, 1, e16, m1, ta, ma
; ZVFHMIN32-NEXT: vslidedown.vi v10, v8, 7
; ZVFHMIN32-NEXT: csrr a2, vlenb
-; ZVFHMIN32-NEXT: li a3, 29
+; ZVFHMIN32-NEXT: li a3, 27
; ZVFHMIN32-NEXT: mul a2, a2, a3
; ZVFHMIN32-NEXT: add a2, sp, a2
; ZVFHMIN32-NEXT: addi a2, a2, 848
; ZVFHMIN32-NEXT: vs1r.v v10, (a2) # Unknown-size Folded Spill
; ZVFHMIN32-NEXT: vslidedown.vi v10, v8, 6
; ZVFHMIN32-NEXT: csrr a2, vlenb
-; ZVFHMIN32-NEXT: li a3, 28
+; ZVFHMIN32-NEXT: li a3, 26
; ZVFHMIN32-NEXT: mul a2, a2, a3
; ZVFHMIN32-NEXT: add a2, sp, a2
; ZVFHMIN32-NEXT: addi a2, a2, 848
; ZVFHMIN32-NEXT: vs1r.v v10, (a2) # Unknown-size Folded Spill
; ZVFHMIN32-NEXT: vslidedown.vi v10, v8, 5
; ZVFHMIN32-NEXT: csrr a2, vlenb
-; ZVFHMIN32-NEXT: li a3, 27
+; ZVFHMIN32-NEXT: li a3, 25
; ZVFHMIN32-NEXT: mul a2, a2, a3
; ZVFHMIN32-NEXT: add a2, sp, a2
; ZVFHMIN32-NEXT: addi a2, a2, 848
; ZVFHMIN32-NEXT: vs1r.v v10, (a2) # Unknown-size Folded Spill
; ZVFHMIN32-NEXT: vslidedown.vi v10, v8, 4
; ZVFHMIN32-NEXT: csrr a2, vlenb
-; ZVFHMIN32-NEXT: li a3, 26
+; ZVFHMIN32-NEXT: li a3, 24
; ZVFHMIN32-NEXT: mul a2, a2, a3
; ZVFHMIN32-NEXT: add a2, sp, a2
; ZVFHMIN32-NEXT: addi a2, a2, 848
; ZVFHMIN32-NEXT: vs1r.v v10, (a2) # Unknown-size Folded Spill
; ZVFHMIN32-NEXT: vslidedown.vi v10, v8, 3
; ZVFHMIN32-NEXT: csrr a2, vlenb
-; ZVFHMIN32-NEXT: li a3, 25
+; ZVFHMIN32-NEXT: li a3, 23
; ZVFHMIN32-NEXT: mul a2, a2, a3
; ZVFHMIN32-NEXT: add a2, sp, a2
; ZVFHMIN32-NEXT: addi a2, a2, 848
; ZVFHMIN32-NEXT: vs1r.v v10, (a2) # Unknown-size Folded Spill
; ZVFHMIN32-NEXT: vslidedown.vi v10, v8, 2
; ZVFHMIN32-NEXT: csrr a2, vlenb
-; ZVFHMIN32-NEXT: li a3, 24
+; ZVFHMIN32-NEXT: li a3, 22
; ZVFHMIN32-NEXT: mul a2, a2, a3
; ZVFHMIN32-NEXT: add a2, sp, a2
; ZVFHMIN32-NEXT: addi a2, a2, 848
; ZVFHMIN32-NEXT: vs1r.v v10, (a2) # Unknown-size Folded Spill
; ZVFHMIN32-NEXT: vslidedown.vi v10, v8, 1
; ZVFHMIN32-NEXT: csrr a2, vlenb
-; ZVFHMIN32-NEXT: li a3, 23
+; ZVFHMIN32-NEXT: li a3, 21
; ZVFHMIN32-NEXT: mul a2, a2, a3
; ZVFHMIN32-NEXT: add a2, sp, a2
; ZVFHMIN32-NEXT: addi a2, a2, 848
; ZVFHMIN32-NEXT: vs1r.v v10, (a2) # Unknown-size Folded Spill
; ZVFHMIN32-NEXT: vsetivli zero, 1, e16, m2, ta, ma
-; ZVFHMIN32-NEXT: vslidedown.vi v26, v8, 15
-; ZVFHMIN32-NEXT: vslidedown.vi v20, v8, 14
-; ZVFHMIN32-NEXT: vslidedown.vi v28, v8, 13
-; ZVFHMIN32-NEXT: vslidedown.vi v10, v8, 12
+; ZVFHMIN32-NEXT: vslidedown.vi v4, v8, 15
+; ZVFHMIN32-NEXT: vslidedown.vi v10, v8, 14
; ZVFHMIN32-NEXT: csrr a2, vlenb
-; ZVFHMIN32-NEXT: slli a2, a2, 1
; ZVFHMIN32-NEXT: add a2, sp, a2
; ZVFHMIN32-NEXT: addi a2, a2, 848
; ZVFHMIN32-NEXT: vs2r.v v10, (a2) # Unknown-size Folded Spill
-; ZVFHMIN32-NEXT: vslidedown.vi v4, v8, 11
-; ZVFHMIN32-NEXT: vslidedown.vi v2, v8, 10
-; ZVFHMIN32-NEXT: vslidedown.vi v30, v8, 9
-; ZVFHMIN32-NEXT: vslidedown.vi v22, v8, 8
+; ZVFHMIN32-NEXT: vslidedown.vi v30, v8, 13
+; ZVFHMIN32-NEXT: vslidedown.vi v6, v8, 12
+; ZVFHMIN32-NEXT: vslidedown.vi v28, v8, 11
+; ZVFHMIN32-NEXT: vslidedown.vi v26, v8, 10
+; ZVFHMIN32-NEXT: vslidedown.vi v22, v8, 9
+; ZVFHMIN32-NEXT: vslidedown.vi v20, v8, 8
; ZVFHMIN32-NEXT: vmv.x.s a4, v16
; ZVFHMIN32-NEXT: fmv.h.x fa5, a0
; ZVFHMIN32-NEXT: fmv.h.x fa4, a1
@@ -1384,52 +1373,54 @@ define <128 x i1> @fcmp_oeq_vv_v128f16(<128 x half> %va, <128 x half> %vb, <128
; ZVFHMIN32-NEXT: lh a0, 560(sp)
; ZVFHMIN32-NEXT: lh a1, 304(sp)
; ZVFHMIN32-NEXT: vsetivli zero, 1, e16, m1, ta, ma
-; ZVFHMIN32-NEXT: vslidedown.vi v3, v16, 7
-; ZVFHMIN32-NEXT: vslidedown.vi v31, v16, 6
-; ZVFHMIN32-NEXT: vslidedown.vi v5, v16, 5
-; ZVFHMIN32-NEXT: vslidedown.vi v23, v16, 4
-; ZVFHMIN32-NEXT: vslidedown.vi v8, v16, 3
+; ZVFHMIN32-NEXT: vslidedown.vi v8, v16, 7
; ZVFHMIN32-NEXT: csrr a2, vlenb
-; ZVFHMIN32-NEXT: li a3, 21
-; ZVFHMIN32-NEXT: mul a2, a2, a3
+; ZVFHMIN32-NEXT: slli a3, a2, 1
+; ZVFHMIN32-NEXT: add a2, a3, a2
; ZVFHMIN32-NEXT: add a2, sp, a2
; ZVFHMIN32-NEXT: addi a2, a2, 848
; ZVFHMIN32-NEXT: vs1r.v v8, (a2) # Unknown-size Folded Spill
-; ZVFHMIN32-NEXT: vslidedown.vi v8, v16, 2
+; ZVFHMIN32-NEXT: vslidedown.vi v21, v16, 6
+; ZVFHMIN32-NEXT: vslidedown.vi v8, v16, 5
; ZVFHMIN32-NEXT: csrr a2, vlenb
-; ZVFHMIN32-NEXT: li a3, 20
+; ZVFHMIN32-NEXT: li a3, 18
; ZVFHMIN32-NEXT: mul a2, a2, a3
; ZVFHMIN32-NEXT: add a2, sp, a2
; ZVFHMIN32-NEXT: addi a2, a2, 848
; ZVFHMIN32-NEXT: vs1r.v v8, (a2) # Unknown-size Folded Spill
-; ZVFHMIN32-NEXT: vslidedown.vi v8, v16, 1
-; ZVFHMIN32-NEXT: csrr a2, vlenb
-; ZVFHMIN32-NEXT: li a3, 22
-; ZVFHMIN32-NEXT: mul a2, a2, a3
-; ZVFHMIN32-NEXT: add a2, sp, a2
-; ZVFHMIN32-NEXT: addi a2, a2, 848
+; ZVFHMIN32-NEXT: vslidedown.vi v23, v16, 4
+; ZVFHMIN32-NEXT: vslidedown.vi v31, v16, 3
+; ZVFHMIN32-NEXT: vslidedown.vi v8, v16, 2
+; ZVFHMIN32-NEXT: addi a2, sp, 848
; ZVFHMIN32-NEXT: vs1r.v v8, (a2) # Unknown-size Folded Spill
+; ZVFHMIN32-NEXT: vslidedown.vi v7, v16, 1
; ZVFHMIN32-NEXT: vsetivli zero, 1, e16, m2, ta, ma
-; ZVFHMIN32-NEXT: vslidedown.vi v18, v16, 15
-; ZVFHMIN32-NEXT: vslidedown.vi v14, v16, 14
+; ZVFHMIN32-NEXT: vslidedown.vi v14, v16, 15
+; ZVFHMIN32-NEXT: vslidedown.vi v12, v16, 14
; ZVFHMIN32-NEXT: vslidedown.vi v8, v16, 13
-; ZVFHMIN32-NEXT: vslidedown.vi v12, v16, 12
+; ZVFHMIN32-NEXT: vslidedown.vi v18, v16, 12
; ZVFHMIN32-NEXT: vslidedown.vi v10, v16, 11
-; ZVFHMIN32-NEXT: vslidedown.vi v6, v16, 10
+; ZVFHMIN32-NEXT: vslidedown.vi v2, v16, 10
; ZVFHMIN32-NEXT: csrr a2, vlenb
-; ZVFHMIN32-NEXT: li a3, 18
+; ZVFHMIN32-NEXT: li a3, 19
; ZVFHMIN32-NEXT: mul a2, a2, a3
; ZVFHMIN32-NEXT: add a2, sp, a2
; ZVFHMIN32-NEXT: addi a2, a2, 848
-; ZVFHMIN32-NEXT: vs2r.v v6, (a2) # Unknown-size Folded Spill
-; ZVFHMIN32-NEXT: vslidedown.vi v6, v16, 9
+; ZVFHMIN32-NEXT: vs2r.v v2, (a2) # Unknown-size Folded Spill
+; ZVFHMIN32-NEXT: vslidedown.vi v2, v16, 9
; ZVFHMIN32-NEXT: csrr a2, vlenb
; ZVFHMIN32-NEXT: li a3, 14
; ZVFHMIN32-NEXT: mul a2, a2, a3
; ZVFHMIN32-NEXT: add a2, sp, a2
; ZVFHMIN32-NEXT: addi a2, a2, 848
-; ZVFHMIN32-NEXT: vs2r.v v6, (a2) # Unknown-size Folded Spill
-; ZVFHMIN32-NEXT: vslidedown.vi v6, v16, 8
+; ZVFHMIN32-NEXT: vs2r.v v2, (a2) # Unknown-size Folded Spill
+; ZVFHMIN32-NEXT: vslidedown.vi v16, v16, 8
+; ZVFHMIN32-NEXT: csrr a2, vlenb
+; ZVFHMIN32-NEXT: li a3, 12
+; ZVFHMIN32-NEXT: mul a2, a2, a3
+; ZVFHMIN32-NEXT: add a2, sp, a2
+; ZVFHMIN32-NEXT: addi a2, a2, 848
+; ZVFHMIN32-NEXT: vs2r.v v16, (a2) # Unknown-size Folded Spill
; ZVFHMIN32-NEXT: fmv.h.x fa5, a0
; ZVFHMIN32-NEXT: fmv.h.x fa4, a1
; ZVFHMIN32-NEXT: feq.h a0, fa5, fa4
@@ -1437,95 +1428,86 @@ define <128 x i1> @fcmp_oeq_vv_v128f16(<128 x half> %va, <128 x half> %vb, <128
; ZVFHMIN32-NEXT: lh a0, 558(sp)
; ZVFHMIN32-NEXT: lh a1, 302(sp)
; ZVFHMIN32-NEXT: vsetivli zero, 1, e16, m1, ta, ma
-; ZVFHMIN32-NEXT: vslidedown.vi v13, v0, 7
-; ZVFHMIN32-NEXT: vslidedown.vi v29, v0, 6
-; ZVFHMIN32-NEXT: vslidedown.vi v11, v0, 5
-; ZVFHMIN32-NEXT: vslidedown.vi v7, v0, 4
-; ZVFHMIN32-NEXT: vslidedown.vi v9, v0, 3
-; ZVFHMIN32-NEXT: vslidedown.vi v21, v0, 2
-; ZVFHMIN32-NEXT: vslidedown.vi v27, v0, 1
+; ZVFHMIN32-NEXT: vslidedown.vi v29, v0, 7
+; ZVFHMIN32-NEXT: vslidedown.vi v19, v0, 6
+; ZVFHMIN32-NEXT: vslidedown.vi v27, v0, 5
+; ZVFHMIN32-NEXT: vslidedown.vi v13, v0, 4
+; ZVFHMIN32-NEXT: vslidedown.vi v15, v0, 3
+; ZVFHMIN32-NEXT: vslidedown.vi v9, v0, 2
+; ZVFHMIN32-NEXT: vslidedown.vi v11, v0, 1
; ZVFHMIN32-NEXT: vsetivli zero, 1, e16, m2, ta, ma
; ZVFHMIN32-NEXT: vslidedown.vi v16, v0, 15
-; ZVFHMIN32-NEXT: csrr a2, vlenb
-; ZVFHMIN32-NEXT: slli a2, a2, 2
-; ZVFHMIN32-NEXT: add a2, sp, a2
-; ZVFHMIN32-NEXT: addi a2, a2, 848
-; ZVFHMIN32-NEXT: vs2r.v v16, (a2) # Unknown-size Folded Spill
-; ZVFHMIN32-NEXT: vslidedown.vi v16, v0, 14
-; ZVFHMIN32-NEXT: csrr a2, vlenb
-; ZVFHMIN32-NEXT: slli a2, a2, 3
-; ZVFHMIN32-NEXT: add a2, sp, a2
-; ZVFHMIN32-NEXT: addi a2, a2, 848
-; ZVFHMIN32-NEXT: vs2r.v v16, (a2) # Unknown-size Folded Spill
-; ZVFHMIN32-NEXT: vslidedown.vi v16, v0, 13
+; ZVFHMIN32-NEXT: vslidedown.vi v2, v0, 14
; ZVFHMIN32-NEXT: csrr a2, vlenb
; ZVFHMIN32-NEXT: li a3, 6
; ZVFHMIN32-NEXT: mul a2, a2, a3
; ZVFHMIN32-NEXT: add a2, sp, a2
; ZVFHMIN32-NEXT: addi a2, a2, 848
-; ZVFHMIN32-NEXT: vs2r.v v16, (a2) # Unknown-size Folded Spill
-; ZVFHMIN32-NEXT: vslidedown.vi v16, v0, 12
+; ZVFHMIN32-NEXT: vs2r.v v2, (a2) # Unknown-size Folded Spill
+; ZVFHMIN32-NEXT: vslidedown.vi v2, v0, 13
; ZVFHMIN32-NEXT: csrr a2, vlenb
-; ZVFHMIN32-NEXT: li a3, 12
-; ZVFHMIN32-NEXT: mul a2, a2, a3
+; ZVFHMIN32-NEXT: slli a2, a2, 2
; ZVFHMIN32-NEXT: add a2, sp, a2
; ZVFHMIN32-NEXT: addi a2, a2, 848
-; ZVFHMIN32-NEXT: vs2r.v v16, (a2) # Unknown-size Folded Spill
-; ZVFHMIN32-NEXT: vslidedown.vi v16, v0, 11
+; ZVFHMIN32-NEXT: vs2r.v v2, (a2) # Unknown-size Folded Spill
+; ZVFHMIN32-NEXT: vslidedown.vi v2, v0, 12
; ZVFHMIN32-NEXT: csrr a2, vlenb
; ZVFHMIN32-NEXT: li a3, 10
; ZVFHMIN32-NEXT: mul a2, a2, a3
; ZVFHMIN32-NEXT: add a2, sp, a2
; ZVFHMIN32-NEXT: addi a2, a2, 848
-; ZVFHMIN32-NEXT: vs2r.v v16, (a2) # Unknown-size Folded Spill
-; ZVFHMIN32-NEXT: vslidedown.vi v16, v0, 10
+; ZVFHMIN32-NEXT: vs2r.v v2, (a2) # Unknown-size Folded Spill
+; ZVFHMIN32-NEXT: vslidedown.vi v2, v0, 11
+; ZVFHMIN32-NEXT: csrr a2, vlenb
+; ZVFHMIN32-NEXT: slli a2, a2, 3
+; ZVFHMIN32-NEXT: add a2, sp, a2
+; ZVFHMIN32-NEXT: addi a2, a2, 848
+; ZVFHMIN32-NEXT: vs2r.v v2, (a2) # Unknown-size Folded Spill
+; ZVFHMIN32-NEXT: vslidedown.vi v2, v0, 10
; ZVFHMIN32-NEXT: csrr a2, vlenb
; ZVFHMIN32-NEXT: slli a2, a2, 4
; ZVFHMIN32-NEXT: add a2, sp, a2
; ZVFHMIN32-NEXT: addi a2, a2, 848
-; ZVFHMIN32-NEXT: vs2r.v v16, (a2) # Unknown-size Folded Spill
-; ZVFHMIN32-NEXT: vslidedown.vi v16, v0, 9
+; ZVFHMIN32-NEXT: vs2r.v v2, (a2) # Unknown-size Folded Spill
+; ZVFHMIN32-NEXT: vslidedown.vi v2, v0, 9
; ZVFHMIN32-NEXT: vslidedown.vi v0, v0, 8
-; ZVFHMIN32-NEXT: addi a2, sp, 848
-; ZVFHMIN32-NEXT: vs2r.v v0, (a2) # Unknown-size Folded Spill
-; ZVFHMIN32-NEXT: vmv.x.s t4, v26
+; ZVFHMIN32-NEXT: vmv.x.s t3, v4
; ZVFHMIN32-NEXT: fmv.h.x fa5, a0
; ZVFHMIN32-NEXT: fmv.h.x fa4, a1
; ZVFHMIN32-NEXT: feq.h a0, fa5, fa4
; ZVFHMIN32-NEXT: sb a0, 215(sp)
; ZVFHMIN32-NEXT: lh a0, 556(sp)
; ZVFHMIN32-NEXT: lh a1, 300(sp)
-; ZVFHMIN32-NEXT: vmv.x.s t3, v20
-; ZVFHMIN32-NEXT: vmv.x.s t1, v28
+; ZVFHMIN32-NEXT: csrr a2, vlenb
+; ZVFHMIN32-NEXT: add a2, sp, a2
+; ZVFHMIN32-NEXT: addi a2, a2, 848
+; ZVFHMIN32-NEXT: vl2r.v v4, (a2) # Unknown-size Folded Reload
+; ZVFHMIN32-NEXT: vmv.x.s t4, v4
+; ZVFHMIN32-NEXT: vmv.x.s t2, v30
; ZVFHMIN32-NEXT: fmv.h.x fa5, a0
; ZVFHMIN32-NEXT: fmv.h.x fa4, a1
; ZVFHMIN32-NEXT: feq.h a0, fa5, fa4
; ZVFHMIN32-NEXT: sb a0, 214(sp)
; ZVFHMIN32-NEXT: lh a0, 554(sp)
; ZVFHMIN32-NEXT: lh a1, 298(sp)
-; ZVFHMIN32-NEXT: csrr a2, vlenb
-; ZVFHMIN32-NEXT: slli a2, a2, 1
-; ZVFHMIN32-NEXT: add a2, sp, a2
-; ZVFHMIN32-NEXT: addi a2, a2, 848
-; ZVFHMIN32-NEXT: vl2r.v v0, (a2) # Unknown-size Folded Reload
-; ZVFHMIN32-NEXT: vmv.x.s t2, v0
-; ZVFHMIN32-NEXT: vmv.x.s t0, v4
+; ZVFHMIN32-NEXT: vmv.x.s t1, v6
+; ZVFHMIN32-NEXT: vmv.x.s t0, v28
; ZVFHMIN32-NEXT: fmv.h.x fa5, a0
; ZVFHMIN32-NEXT: fmv.h.x fa4, a1
; ZVFHMIN32-NEXT: feq.h a0, fa5, fa4
; ZVFHMIN32-NEXT: sb a0, 213(sp)
; ZVFHMIN32-NEXT: lh a0, 552(sp)
; ZVFHMIN32-NEXT: lh a1, 296(sp)
-; ZVFHMIN32-NEXT: vmv.x.s a7, v2
-; ZVFHMIN32-NEXT: vmv.x.s a6, v30
+; ZVFHMIN32-NEXT: vmv.x.s a7, v26
+; ZVFHMIN32-NEXT: vmv.x.s a6, v22
; ZVFHMIN32-NEXT: fmv.h.x fa5, a0
; ZVFHMIN32-NEXT: fmv.h.x fa4, a1
; ZVFHMIN32-NEXT: feq.h a0, fa5, fa4
; ZVFHMIN32-NEXT: sb a0, 212(sp)
; ZVFHMIN32-NEXT: lh a0, 550(sp)
; ZVFHMIN32-NEXT: lh a1, 294(sp)
-; ZVFHMIN32-NEXT: vmv.x.s a5, v22
-; ZVFHMIN32-NEXT: vmv.x.s a2, v18
+; ZVFHMIN32-NEXT: vmv.x.s a5, v20
+; ZVFHMIN32-NEXT: vmv.x.s a2, v14
; ZVFHMIN32-NEXT: sw a2, 112(sp) # 4-byte Folded Spill
; ZVFHMIN32-NEXT: fmv.h.x fa5, a0
; ZVFHMIN32-NEXT: fmv.h.x fa4, a1
@@ -1533,7 +1515,7 @@ define <128 x i1> @fcmp_oeq_vv_v128f16(<128 x half> %va, <128 x half> %vb, <128
; ZVFHMIN32-NEXT: sb a0, 211(sp)
; ZVFHMIN32-NEXT: lh a1, 548(sp)
; ZVFHMIN32-NEXT: lh t5, 292(sp)
-; ZVFHMIN32-NEXT: vmv.x.s a0, v14
+; ZVFHMIN32-NEXT: vmv.x.s a0, v12
; ZVFHMIN32-NEXT: sw a0, 116(sp) # 4-byte Folded Spill
; ZVFHMIN32-NEXT: vmv.x.s a0, v8
; ZVFHMIN32-NEXT: sw a0, 124(sp) # 4-byte Folded Spill
@@ -1560,7 +1542,7 @@ define <128 x i1> @fcmp_oeq_vv_v128f16(<128 x half> %va, <128 x half> %vb, <128
; ZVFHMIN32-NEXT: sb a1, 208(sp)
; ZVFHMIN32-NEXT: lh t5, 738(sp)
; ZVFHMIN32-NEXT: lh t6, 482(sp)
-; ZVFHMIN32-NEXT: vmv.x.s a0, v12
+; ZVFHMIN32-NEXT: vmv.x.s a0, v18
; ZVFHMIN32-NEXT: sw a0, 108(sp) # 4-byte Folded Spill
; ZVFHMIN32-NEXT: vmv.x.s a0, v10
; ZVFHMIN32-NEXT: sw a0, 120(sp) # 4-byte Folded Spill
@@ -1571,12 +1553,12 @@ define <128 x i1> @fcmp_oeq_vv_v128f16(<128 x half> %va, <128 x half> %vb, <128
; ZVFHMIN32-NEXT: lh t5, 736(sp)
; ZVFHMIN32-NEXT: lh t6, 480(sp)
; ZVFHMIN32-NEXT: csrr a0, vlenb
-; ZVFHMIN32-NEXT: li a1, 29
+; ZVFHMIN32-NEXT: li a1, 27
; ZVFHMIN32-NEXT: mul a0, a0, a1
; ZVFHMIN32-NEXT: add a0, sp, a0
; ZVFHMIN32-NEXT: lh s5, 848(a0) # 8-byte Folded Reload
; ZVFHMIN32-NEXT: csrr a0, vlenb
-; ZVFHMIN32-NEXT: li a1, 28
+; ZVFHMIN32-NEXT: li a1, 26
; ZVFHMIN32-NEXT: mul a0, a0, a1
; ZVFHMIN32-NEXT: add a0, sp, a0
; ZVFHMIN32-NEXT: lh s6, 848(a0) # 8-byte Folded Reload
@@ -1587,12 +1569,12 @@ define <128 x i1> @fcmp_oeq_vv_v128f16(<128 x half> %va, <128 x half> %vb, <128
; ZVFHMIN32-NEXT: lh t5, 734(sp)
; ZVFHMIN32-NEXT: lh t6, 478(sp)
; ZVFHMIN32-NEXT: csrr a0, vlenb
-; ZVFHMIN32-NEXT: li a1, 27
+; ZVFHMIN32-NEXT: li a1, 25
; ZVFHMIN32-NEXT: mul a0, a0, a1
; ZVFHMIN32-NEXT: add a0, sp, a0
; ZVFHMIN32-NEXT: lh s7, 848(a0) # 8-byte Folded Reload
; ZVFHMIN32-NEXT: csrr a0, vlenb
-; ZVFHMIN32-NEXT: li a1, 26
+; ZVFHMIN32-NEXT: li a1, 24
; ZVFHMIN32-NEXT: mul a0, a0, a1
; ZVFHMIN32-NEXT: add a0, sp, a0
; ZVFHMIN32-NEXT: lh s8, 848(a0) # 8-byte Folded Reload
@@ -1603,12 +1585,12 @@ define <128 x i1> @fcmp_oeq_vv_v128f16(<128 x half> %va, <128 x half> %vb, <128
; ZVFHMIN32-NEXT: lh t5, 732(sp)
; ZVFHMIN32-NEXT: lh t6, 476(sp)
; ZVFHMIN32-NEXT: csrr a0, vlenb
-; ZVFHMIN32-NEXT: li a1, 25
+; ZVFHMIN32-NEXT: li a1, 23
; ZVFHMIN32-NEXT: mul a0, a0, a1
; ZVFHMIN32-NEXT: add a0, sp, a0
; ZVFHMIN32-NEXT: lh s4, 848(a0) # 8-byte Folded Reload
; ZVFHMIN32-NEXT: csrr a0, vlenb
-; ZVFHMIN32-NEXT: li a1, 24
+; ZVFHMIN32-NEXT: li a1, 22
; ZVFHMIN32-NEXT: mul a0, a0, a1
; ZVFHMIN32-NEXT: add a0, sp, a0
; ZVFHMIN32-NEXT: lh s3, 848(a0) # 8-byte Folded Reload
@@ -1619,43 +1601,47 @@ define <128 x i1> @fcmp_oeq_vv_v128f16(<128 x half> %va, <128 x half> %vb, <128
; ZVFHMIN32-NEXT: lh t6, 730(sp)
; ZVFHMIN32-NEXT: lh s9, 474(sp)
; ZVFHMIN32-NEXT: csrr a0, vlenb
-; ZVFHMIN32-NEXT: li a1, 23
+; ZVFHMIN32-NEXT: li a1, 21
; ZVFHMIN32-NEXT: mul a0, a0, a1
; ZVFHMIN32-NEXT: add a0, sp, a0
; ZVFHMIN32-NEXT: lh s2, 848(a0) # 8-byte Folded Reload
-; ZVFHMIN32-NEXT: vmv.x.s t5, v3
+; ZVFHMIN32-NEXT: csrr a0, vlenb
+; ZVFHMIN32-NEXT: slli a1, a0, 1
+; ZVFHMIN32-NEXT: add a0, a1, a0
+; ZVFHMIN32-NEXT: add a0, sp, a0
+; ZVFHMIN32-NEXT: lh t5, 848(a0) # 8-byte Folded Reload
; ZVFHMIN32-NEXT: fmv.h.x fa5, t6
; ZVFHMIN32-NEXT: fmv.h.x fa4, s9
; ZVFHMIN32-NEXT: feq.h t6, fa5, fa4
; ZVFHMIN32-NEXT: sb t6, 173(sp)
; ZVFHMIN32-NEXT: lh s9, 728(sp)
; ZVFHMIN32-NEXT: lh s10, 472(sp)
-; ZVFHMIN32-NEXT: vmv.x.s t6, v31
-; ZVFHMIN32-NEXT: vmv.x.s ra, v13
+; ZVFHMIN32-NEXT: vmv.x.s t6, v21
+; ZVFHMIN32-NEXT: vmv.x.s ra, v29
; ZVFHMIN32-NEXT: fmv.h.x fa5, s9
; ZVFHMIN32-NEXT: fmv.h.x fa4, s10
; ZVFHMIN32-NEXT: feq.h s9, fa5, fa4
; ZVFHMIN32-NEXT: sb s9, 172(sp)
; ZVFHMIN32-NEXT: lh s9, 726(sp)
; ZVFHMIN32-NEXT: lh s10, 470(sp)
-; ZVFHMIN32-NEXT: vmv.x.s a2, v29
-; ZVFHMIN32-NEXT: vmv.x.s a3, v11
+; ZVFHMIN32-NEXT: vmv.x.s a2, v19
+; ZVFHMIN32-NEXT: vmv.x.s a3, v27
; ZVFHMIN32-NEXT: fmv.h.x fa5, s9
; ZVFHMIN32-NEXT: fmv.h.x fa4, s10
; ZVFHMIN32-NEXT: feq.h s9, fa5, fa4
; ZVFHMIN32-NEXT: sb s9, 171(sp)
; ZVFHMIN32-NEXT: lh s10, 724(sp)
; ZVFHMIN32-NEXT: lh s11, 468(sp)
-; ZVFHMIN32-NEXT: vmv.x.s a4, v7
-; ZVFHMIN32-NEXT: vmv.x.s s9, v9
+; ZVFHMIN32-NEXT: vmv.x.s a4, v13
+; ZVFHMIN32-NEXT: vmv.x.s s9, v15
; ZVFHMIN32-NEXT: fmv.h.x fa5, s10
; ZVFHMIN32-NEXT: fmv.h.x fa4, s11
; ZVFHMIN32-NEXT: feq.h s10, fa5, fa4
; ZVFHMIN32-NEXT: sb s10, 170(sp)
; ZVFHMIN32-NEXT: lh a0, 722(sp)
; ZVFHMIN32-NEXT: lh a1, 466(sp)
-; ZVFHMIN32-NEXT: vmv.x.s s10, v21
-; ZVFHMIN32-NEXT: vmv.x.s s11, v27
+; ZVFHMIN32-NEXT: vmv.x.s s10, v9
+; ZVFHMIN32-NEXT: vmv.x.s s11, v11
; ZVFHMIN32-NEXT: fmv.h.x fa5, a0
; ZVFHMIN32-NEXT: fmv.h.x fa4, a1
; ZVFHMIN32-NEXT: feq.h a0, fa5, fa4
@@ -1738,7 +1724,11 @@ define <128 x i1> @fcmp_oeq_vv_v128f16(<128 x half> %va, <128 x half> %vb, <128
; ZVFHMIN32-NEXT: sb a1, 161(sp)
; ZVFHMIN32-NEXT: lh a0, 610(sp)
; ZVFHMIN32-NEXT: lh a1, 354(sp)
-; ZVFHMIN32-NEXT: vmv.x.s s6, v5
+; ZVFHMIN32-NEXT: csrr a2, vlenb
+; ZVFHMIN32-NEXT: li a3, 18
+; ZVFHMIN32-NEXT: mul a2, a2, a3
+; ZVFHMIN32-NEXT: add a2, sp, a2
+; ZVFHMIN32-NEXT: lh s6, 848(a2) # 8-byte Folded Reload
; ZVFHMIN32-NEXT: vmv.x.s s5, v23
; ZVFHMIN32-NEXT: fmv.h.x fa5, a0
; ZVFHMIN32-NEXT: fmv.h.x fa4, a1
@@ -1746,27 +1736,15 @@ define <128 x i1> @fcmp_oeq_vv_v128f16(<128 x half> %va, <128 x half> %vb, <128
; ZVFHMIN32-NEXT: sb a0, 241(sp)
; ZVFHMIN32-NEXT: lh a0, 608(sp)
; ZVFHMIN32-NEXT: lh a1, 352(sp)
-; ZVFHMIN32-NEXT: csrr a2, vlenb
-; ZVFHMIN32-NEXT: li a3, 21
-; ZVFHMIN32-NEXT: mul a2, a2, a3
-; ZVFHMIN32-NEXT: add a2, sp, a2
-; ZVFHMIN32-NEXT: lh s4, 848(a2) # 8-byte Folded Reload
-; ZVFHMIN32-NEXT: csrr a2, vlenb
-; ZVFHMIN32-NEXT: li a3, 20
-; ZVFHMIN32-NEXT: mul a2, a2, a3
-; ZVFHMIN32-NEXT: add a2, sp, a2
-; ZVFHMIN32-NEXT: lh s3, 848(a2) # 8-byte Folded Reload
+; ZVFHMIN32-NEXT: vmv.x.s s4, v31
+; ZVFHMIN32-NEXT: lh s3, 848(sp) # 8-byte Folded Reload
; ZVFHMIN32-NEXT: fmv.h.x fa5, a0
; ZVFHMIN32-NEXT: fmv.h.x fa4, a1
; ZVFHMIN32-NEXT: feq.h a0, fa5, fa4
; ZVFHMIN32-NEXT: sb a0, 240(sp)
; ZVFHMIN32-NEXT: lh a0, 606(sp)
; ZVFHMIN32-NEXT: lh a1, 350(sp)
-; ZVFHMIN32-NEXT: csrr a2, vlenb
-; ZVFHMIN32-NEXT: li a3, 22
-; ZVFHMIN32-NEXT: mul a2, a2, a3
-; ZVFHMIN32-NEXT: add a2, sp, a2
-; ZVFHMIN32-NEXT: lh s2, 848(a2) # 8-byte Folded Reload
+; ZVFHMIN32-NEXT: vmv.x.s s2, v7
; ZVFHMIN32-NEXT: fmv.h.x fa5, t5
; ZVFHMIN32-NEXT: fmv.h.x fa4, a0
; ZVFHMIN32-NEXT: fmv.h.x fa3, a1
@@ -1901,7 +1879,7 @@ define <128 x i1> @fcmp_oeq_vv_v128f16(<128 x half> %va, <128 x half> %vb, <128
; ZVFHMIN32-NEXT: lh a0, 766(sp)
; ZVFHMIN32-NEXT: lh a1, 510(sp)
; ZVFHMIN32-NEXT: csrr a2, vlenb
-; ZVFHMIN32-NEXT: li a3, 18
+; ZVFHMIN32-NEXT: li a3, 19
; ZVFHMIN32-NEXT: mul a2, a2, a3
; ZVFHMIN32-NEXT: add a2, sp, a2
; ZVFHMIN32-NEXT: addi a2, a2, 848
@@ -1913,20 +1891,21 @@ define <128 x i1> @fcmp_oeq_vv_v128f16(<128 x half> %va, <128 x half> %vb, <128
; ZVFHMIN32-NEXT: add a2, sp, a2
; ZVFHMIN32-NEXT: addi a2, a2, 848
; ZVFHMIN32-NEXT: vl2r.v v8, (a2) # Unknown-size Folded Reload
-; ZVFHMIN32-NEXT: vmv.x.s t6, v8
+; ZVFHMIN32-NEXT: vmv.x.s t5, v8
; ZVFHMIN32-NEXT: fmv.h.x fa5, a0
; ZVFHMIN32-NEXT: fmv.h.x fa4, a1
; ZVFHMIN32-NEXT: feq.h a0, fa5, fa4
; ZVFHMIN32-NEXT: sb a0, 191(sp)
; ZVFHMIN32-NEXT: lh a0, 764(sp)
; ZVFHMIN32-NEXT: lh a1, 508(sp)
-; ZVFHMIN32-NEXT: vmv.x.s t5, v6
; ZVFHMIN32-NEXT: csrr a2, vlenb
-; ZVFHMIN32-NEXT: slli a2, a2, 2
+; ZVFHMIN32-NEXT: li a3, 12
+; ZVFHMIN32-NEXT: mul a2, a2, a3
; ZVFHMIN32-NEXT: add a2, sp, a2
; ZVFHMIN32-NEXT: addi a2, a2, 848
; ZVFHMIN32-NEXT: vl2r.v v8, (a2) # Unknown-size Folded Reload
-; ZVFHMIN32-NEXT: vmv.x.s a2, v8
+; ZVFHMIN32-NEXT: vmv.x.s t6, v8
+; ZVFHMIN32-NEXT: vmv.x.s a2, v16
; ZVFHMIN32-NEXT: fmv.h.x fa5, a0
; ZVFHMIN32-NEXT: fmv.h.x fa4, a1
; ZVFHMIN32-NEXT: feq.h a0, fa5, fa4
@@ -1934,14 +1913,14 @@ define <128 x i1> @fcmp_oeq_vv_v128f16(<128 x half> %va, <128 x half> %vb, <128
; ZVFHMIN32-NEXT: lh a0, 762(sp)
; ZVFHMIN32-NEXT: lh a1, 506(sp)
; ZVFHMIN32-NEXT: csrr a3, vlenb
-; ZVFHMIN32-NEXT: slli a3, a3, 3
+; ZVFHMIN32-NEXT: li a4, 6
+; ZVFHMIN32-NEXT: mul a3, a3, a4
; ZVFHMIN32-NEXT: add a3, sp, a3
; ZVFHMIN32-NEXT: addi a3, a3, 848
; ZVFHMIN32-NEXT: vl2r.v v8, (a3) # Unknown-size Folded Reload
; ZVFHMIN32-NEXT: vmv.x.s a3, v8
; ZVFHMIN32-NEXT: csrr a4, vlenb
-; ZVFHMIN32-NEXT: li s3, 6
-; ZVFHMIN32-NEXT: mul a4, a4, s3
+; ZVFHMIN32-NEXT: slli a4, a4, 2
; ZVFHMIN32-NEXT: add a4, sp, a4
; ZVFHMIN32-NEXT: addi a4, a4, 848
; ZVFHMIN32-NEXT: vl2r.v v8, (a4) # Unknown-size Folded Reload
@@ -1953,15 +1932,14 @@ define <128 x i1> @fcmp_oeq_vv_v128f16(<128 x half> %va, <128 x half> %vb, <128
; ZVFHMIN32-NEXT: lh a0, 760(sp)
; ZVFHMIN32-NEXT: lh a1, 504(sp)
; ZVFHMIN32-NEXT: csrr s3, vlenb
-; ZVFHMIN32-NEXT: li s4, 12
+; ZVFHMIN32-NEXT: li s4, 10
; ZVFHMIN32-NEXT: mul s3, s3, s4
; ZVFHMIN32-NEXT: add s3, sp, s3
; ZVFHMIN32-NEXT: addi s3, s3, 848
; ZVFHMIN32-NEXT: vl2r.v v8, (s3) # Unknown-size Folded Reload
; ZVFHMIN32-NEXT: vmv.x.s s6, v8
; ZVFHMIN32-NEXT: csrr s3, vlenb
-; ZVFHMIN32-NEXT: li s4, 10
-; ZVFHMIN32-NEXT: mul s3, s3, s4
+; ZVFHMIN32-NEXT: slli s3, s3, 3
; ZVFHMIN32-NEXT: add s3, sp, s3
; ZVFHMIN32-NEXT: addi s3, s3, 848
; ZVFHMIN32-NEXT: vl2r.v v8, (s3) # Unknown-size Folded Reload
@@ -1978,38 +1956,38 @@ define <128 x i1> @fcmp_oeq_vv_v128f16(<128 x half> %va, <128 x half> %vb, <128
; ZVFHMIN32-NEXT: addi s3, s3, 848
; ZVFHMIN32-NEXT: vl2r.v v8, (s3) # Unknown-size Folded Reload
; ZVFHMIN32-NEXT: vmv.x.s s5, v8
-; ZVFHMIN32-NEXT: vmv.x.s s3, v16
+; ZVFHMIN32-NEXT: vmv.x.s s3, v2
; ZVFHMIN32-NEXT: fmv.h.x fa5, a0
; ZVFHMIN32-NEXT: fmv.h.x fa4, a1
; ZVFHMIN32-NEXT: feq.h a0, fa5, fa4
-; ZVFHMIN32-NEXT: fmv.h.x fa5, t4
+; ZVFHMIN32-NEXT: fmv.h.x fa5, t3
; ZVFHMIN32-NEXT: sb a0, 187(sp)
; ZVFHMIN32-NEXT: lh a0, 756(sp)
; ZVFHMIN32-NEXT: lh a1, 500(sp)
; ZVFHMIN32-NEXT: fmv.h.x fa4, a2
-; ZVFHMIN32-NEXT: feq.h t4, fa5, fa4
+; ZVFHMIN32-NEXT: feq.h t3, fa5, fa4
; ZVFHMIN32-NEXT: fmv.h.x fa5, a0
; ZVFHMIN32-NEXT: fmv.h.x fa4, a1
; ZVFHMIN32-NEXT: feq.h a0, fa5, fa4
-; ZVFHMIN32-NEXT: fmv.h.x fa5, t3
+; ZVFHMIN32-NEXT: fmv.h.x fa5, t4
; ZVFHMIN32-NEXT: sb a0, 186(sp)
; ZVFHMIN32-NEXT: lh a0, 754(sp)
; ZVFHMIN32-NEXT: lh a1, 498(sp)
; ZVFHMIN32-NEXT: fmv.h.x fa4, a3
-; ZVFHMIN32-NEXT: feq.h t3, fa5, fa4
+; ZVFHMIN32-NEXT: feq.h t4, fa5, fa4
; ZVFHMIN32-NEXT: fmv.h.x fa5, a0
; ZVFHMIN32-NEXT: fmv.h.x fa4, a1
; ZVFHMIN32-NEXT: feq.h a0, fa5, fa4
-; ZVFHMIN32-NEXT: fmv.h.x fa5, t1
+; ZVFHMIN32-NEXT: fmv.h.x fa5, t2
; ZVFHMIN32-NEXT: sb a0, 185(sp)
; ZVFHMIN32-NEXT: lh a0, 752(sp)
; ZVFHMIN32-NEXT: lh a1, 496(sp)
; ZVFHMIN32-NEXT: fmv.h.x fa4, a4
-; ZVFHMIN32-NEXT: feq.h t1, fa5, fa4
+; ZVFHMIN32-NEXT: feq.h t2, fa5, fa4
; ZVFHMIN32-NEXT: fmv.h.x fa5, a0
; ZVFHMIN32-NEXT: fmv.h.x fa4, a1
; ZVFHMIN32-NEXT: feq.h a0, fa5, fa4
-; ZVFHMIN32-NEXT: fmv.h.x fa5, t2
+; ZVFHMIN32-NEXT: fmv.h.x fa5, t1
; ZVFHMIN32-NEXT: sb a0, 184(sp)
; ZVFHMIN32-NEXT: lh a0, 750(sp)
; ZVFHMIN32-NEXT: lh a1, 494(sp)
@@ -2046,9 +2024,7 @@ define <128 x i1> @fcmp_oeq_vv_v128f16(<128 x half> %va, <128 x half> %vb, <128
; ZVFHMIN32-NEXT: fmv.h.x fa4, a1
; ZVFHMIN32-NEXT: feq.h a0, fa5, fa4
; ZVFHMIN32-NEXT: fmv.h.x fa5, a5
-; ZVFHMIN32-NEXT: addi a1, sp, 848
-; ZVFHMIN32-NEXT: vl2r.v v8, (a1) # Unknown-size Folded Reload
-; ZVFHMIN32-NEXT: vmv.x.s a1, v8
+; ZVFHMIN32-NEXT: vmv.x.s a1, v0
; ZVFHMIN32-NEXT: vsetivli zero, 1, e16, m2, ta, ma
; ZVFHMIN32-NEXT: vslidedown.vi v8, v24, 15
; ZVFHMIN32-NEXT: vmv.x.s a5, v8
@@ -2064,9 +2040,9 @@ define <128 x i1> @fcmp_oeq_vv_v128f16(<128 x half> %va, <128 x half> %vb, <128
; ZVFHMIN32-NEXT: lh a0, 740(sp)
; ZVFHMIN32-NEXT: lh a7, 484(sp)
; ZVFHMIN32-NEXT: sb a2, 140(sp)
-; ZVFHMIN32-NEXT: sb t1, 141(sp)
-; ZVFHMIN32-NEXT: sb t3, 142(sp)
-; ZVFHMIN32-NEXT: sb t4, 143(sp)
+; ZVFHMIN32-NEXT: sb t2, 141(sp)
+; ZVFHMIN32-NEXT: sb t4, 142(sp)
+; ZVFHMIN32-NEXT: sb t3, 143(sp)
; ZVFHMIN32-NEXT: sb a1, 136(sp)
; ZVFHMIN32-NEXT: sb a6, 137(sp)
; ZVFHMIN32-NEXT: sb a4, 138(sp)
@@ -2181,7 +2157,7 @@ define <128 x i1> @fcmp_oeq_vv_v128f16(<128 x half> %va, <128 x half> %vb, <128
; ZVFHMIN32-NEXT: fmv.h.x fa5, a0
; ZVFHMIN32-NEXT: fmv.h.x fa4, a1
; ZVFHMIN32-NEXT: feq.h a0, fa5, fa4
-; ZVFHMIN32-NEXT: fmv.h.x fa5, t6
+; ZVFHMIN32-NEXT: fmv.h.x fa5, t5
; ZVFHMIN32-NEXT: sb a0, 244(sp)
; ZVFHMIN32-NEXT: lh a0, 614(sp)
; ZVFHMIN32-NEXT: lh a1, 358(sp)
@@ -2190,7 +2166,7 @@ define <128 x i1> @fcmp_oeq_vv_v128f16(<128 x half> %va, <128 x half> %vb, <128
; ZVFHMIN32-NEXT: fmv.h.x fa5, a0
; ZVFHMIN32-NEXT: fmv.h.x fa4, a1
; ZVFHMIN32-NEXT: feq.h a0, fa5, fa4
-; ZVFHMIN32-NEXT: fmv.h.x fa5, t5
+; ZVFHMIN32-NEXT: fmv.h.x fa5, t6
; ZVFHMIN32-NEXT: vslidedown.vi v8, v24, 8
; ZVFHMIN32-NEXT: vmv.x.s a1, v8
; ZVFHMIN32-NEXT: fmv.h.x fa4, a1
@@ -2277,7 +2253,7 @@ define <128 x i1> @fcmp_oeq_vv_v128f16(<128 x half> %va, <128 x half> %vb, <128
; ZVFHMIN64-NEXT: addi s0, sp, 896
; ZVFHMIN64-NEXT: .cfi_def_cfa s0, 0
; ZVFHMIN64-NEXT: csrr a1, vlenb
-; ZVFHMIN64-NEXT: li a2, 30
+; ZVFHMIN64-NEXT: li a2, 28
; ZVFHMIN64-NEXT: mul a1, a1, a2
; ZVFHMIN64-NEXT: sub sp, sp, a1
; ZVFHMIN64-NEXT: andi sp, sp, -128
@@ -2449,67 +2425,66 @@ define <128 x i1> @fcmp_oeq_vv_v128f16(<128 x half> %va, <128 x half> %vb, <128
; ZVFHMIN64-NEXT: vsetivli zero, 1, e16, m1, ta, ma
; ZVFHMIN64-NEXT: vslidedown.vi v10, v8, 7
; ZVFHMIN64-NEXT: csrr a2, vlenb
-; ZVFHMIN64-NEXT: li a3, 29
+; ZVFHMIN64-NEXT: li a3, 27
; ZVFHMIN64-NEXT: mul a2, a2, a3
; ZVFHMIN64-NEXT: add a2, sp, a2
; ZVFHMIN64-NEXT: addi a2, a2, 800
; ZVFHMIN64-NEXT: vs1r.v v10, (a2) # Unknown-size Folded Spill
; ZVFHMIN64-NEXT: vslidedown.vi v10, v8, 6
; ZVFHMIN64-NEXT: csrr a2, vlenb
-; ZVFHMIN64-NEXT: li a3, 28
+; ZVFHMIN64-NEXT: li a3, 26
; ZVFHMIN64-NEXT: mul a2, a2, a3
; ZVFHMIN64-NEXT: add a2, sp, a2
; ZVFHMIN64-NEXT: addi a2, a2, 800
; ZVFHMIN64-NEXT: vs1r.v v10, (a2) # Unknown-size Folded Spill
; ZVFHMIN64-NEXT: vslidedown.vi v10, v8, 5
; ZVFHMIN64-NEXT: csrr a2, vlenb
-; ZVFHMIN64-NEXT: li a3, 27
+; ZVFHMIN64-NEXT: li a3, 25
; ZVFHMIN64-NEXT: mul a2, a2, a3
; ZVFHMIN64-NEXT: add a2, sp, a2
; ZVFHMIN64-NEXT: addi a2, a2, 800
; ZVFHMIN64-NEXT: vs1r.v v10, (a2) # Unknown-size Folded Spill
; ZVFHMIN64-NEXT: vslidedown.vi v10, v8, 4
; ZVFHMIN64-NEXT: csrr a2, vlenb
-; ZVFHMIN64-NEXT: li a3, 26
+; ZVFHMIN64-NEXT: li a3, 24
; ZVFHMIN64-NEXT: mul a2, a2, a3
; ZVFHMIN64-NEXT: add a2, sp, a2
; ZVFHMIN64-NEXT: addi a2, a2, 800
; ZVFHMIN64-NEXT: vs1r.v v10, (a2) # Unknown-size Folded Spill
; ZVFHMIN64-NEXT: vslidedown.vi v10, v8, 3
; ZVFHMIN64-NEXT: csrr a2, vlenb
-; ZVFHMIN64-NEXT: li a3, 25
+; ZVFHMIN64-NEXT: li a3, 23
; ZVFHMIN64-NEXT: mul a2, a2, a3
; ZVFHMIN64-NEXT: add a2, sp, a2
; ZVFHMIN64-NEXT: addi a2, a2, 800
; ZVFHMIN64-NEXT: vs1r.v v10, (a2) # Unknown-size Folded Spill
; ZVFHMIN64-NEXT: vslidedown.vi v10, v8, 2
; ZVFHMIN64-NEXT: csrr a2, vlenb
-; ZVFHMIN64-NEXT: li a3, 24
+; ZVFHMIN64-NEXT: li a3, 22
; ZVFHMIN64-NEXT: mul a2, a2, a3
; ZVFHMIN64-NEXT: add a2, sp, a2
; ZVFHMIN64-NEXT: addi a2, a2, 800
; ZVFHMIN64-NEXT: vs1r.v v10, (a2) # Unknown-size Folded Spill
; ZVFHMIN64-NEXT: vslidedown.vi v10, v8, 1
; ZVFHMIN64-NEXT: csrr a2, vlenb
-; ZVFHMIN64-NEXT: li a3, 23
+; ZVFHMIN64-NEXT: li a3, 21
; ZVFHMIN64-NEXT: mul a2, a2, a3
; ZVFHMIN64-NEXT: add a2, sp, a2
; ZVFHMIN64-NEXT: addi a2, a2, 800
; ZVFHMIN64-NEXT: vs1r.v v10, (a2) # Unknown-size Folded Spill
; ZVFHMIN64-NEXT: vsetivli zero, 1, e16, m2, ta, ma
-; ZVFHMIN64-NEXT: vslidedown.vi v26, v8, 15
-; ZVFHMIN64-NEXT: vslidedown.vi v20, v8, 14
-; ZVFHMIN64-NEXT: vslidedown.vi v28, v8, 13
-; ZVFHMIN64-NEXT: vslidedown.vi v10, v8, 12
+; ZVFHMIN64-NEXT: vslidedown.vi v4, v8, 15
+; ZVFHMIN64-NEXT: vslidedown.vi v10, v8, 14
; ZVFHMIN64-NEXT: csrr a2, vlenb
-; ZVFHMIN64-NEXT: slli a2, a2, 1
; ZVFHMIN64-NEXT: add a2, sp, a2
; ZVFHMIN64-NEXT: addi a2, a2, 800
; ZVFHMIN64-NEXT: vs2r.v v10, (a2) # Unknown-size Folded Spill
-; ZVFHMIN64-NEXT: vslidedown.vi v4, v8, 11
-; ZVFHMIN64-NEXT: vslidedown.vi v2, v8, 10
-; ZVFHMIN64-NEXT: vslidedown.vi v30, v8, 9
-; ZVFHMIN64-NEXT: vslidedown.vi v22, v8, 8
+; ZVFHMIN64-NEXT: vslidedown.vi v30, v8, 13
+; ZVFHMIN64-NEXT: vslidedown.vi v6, v8, 12
+; ZVFHMIN64-NEXT: vslidedown.vi v28, v8, 11
+; ZVFHMIN64-NEXT: vslidedown.vi v26, v8, 10
+; ZVFHMIN64-NEXT: vslidedown.vi v22, v8, 9
+; ZVFHMIN64-NEXT: vslidedown.vi v20, v8, 8
; ZVFHMIN64-NEXT: vmv.x.s a4, v16
; ZVFHMIN64-NEXT: fmv.h.x fa5, a0
; ZVFHMIN64-NEXT: fmv.h.x fa4, a1
@@ -2518,52 +2493,54 @@ define <128 x i1> @fcmp_oeq_vv_v128f16(<128 x half> %va, <128 x half> %vb, <128
; ZVFHMIN64-NEXT: lh a0, 560(sp)
; ZVFHMIN64-NEXT: lh a1, 304(sp)
; ZVFHMIN64-NEXT: vsetivli zero, 1, e16, m1, ta, ma
-; ZVFHMIN64-NEXT: vslidedown.vi v3, v16, 7
-; ZVFHMIN64-NEXT: vslidedown.vi v31, v16, 6
-; ZVFHMIN64-NEXT: vslidedown.vi v5, v16, 5
-; ZVFHMIN64-NEXT: vslidedown.vi v23, v16, 4
-; ZVFHMIN64-NEXT: vslidedown.vi v8, v16, 3
+; ZVFHMIN64-NEXT: vslidedown.vi v8, v16, 7
; ZVFHMIN64-NEXT: csrr a2, vlenb
-; ZVFHMIN64-NEXT: li a3, 21
-; ZVFHMIN64-NEXT: mul a2, a2, a3
+; ZVFHMIN64-NEXT: slli a3, a2, 1
+; ZVFHMIN64-NEXT: add a2, a3, a2
; ZVFHMIN64-NEXT: add a2, sp, a2
; ZVFHMIN64-NEXT: addi a2, a2, 800
; ZVFHMIN64-NEXT: vs1r.v v8, (a2) # Unknown-size Folded Spill
-; ZVFHMIN64-NEXT: vslidedown.vi v8, v16, 2
+; ZVFHMIN64-NEXT: vslidedown.vi v21, v16, 6
+; ZVFHMIN64-NEXT: vslidedown.vi v8, v16, 5
; ZVFHMIN64-NEXT: csrr a2, vlenb
-; ZVFHMIN64-NEXT: li a3, 20
+; ZVFHMIN64-NEXT: li a3, 18
; ZVFHMIN64-NEXT: mul a2, a2, a3
; ZVFHMIN64-NEXT: add a2, sp, a2
; ZVFHMIN64-NEXT: addi a2, a2, 800
; ZVFHMIN64-NEXT: vs1r.v v8, (a2) # Unknown-size Folded Spill
-; ZVFHMIN64-NEXT: vslidedown.vi v8, v16, 1
-; ZVFHMIN64-NEXT: csrr a2, vlenb
-; ZVFHMIN64-NEXT: li a3, 22
-; ZVFHMIN64-NEXT: mul a2, a2, a3
-; ZVFHMIN64-NEXT: add a2, sp, a2
-; ZVFHMIN64-NEXT: addi a2, a2, 800
+; ZVFHMIN64-NEXT: vslidedown.vi v23, v16, 4
+; ZVFHMIN64-NEXT: vslidedown.vi v31, v16, 3
+; ZVFHMIN64-NEXT: vslidedown.vi v8, v16, 2
+; ZVFHMIN64-NEXT: addi a2, sp, 800
; ZVFHMIN64-NEXT: vs1r.v v8, (a2) # Unknown-size Folded Spill
+; ZVFHMIN64-NEXT: vslidedown.vi v7, v16, 1
; ZVFHMIN64-NEXT: vsetivli zero, 1, e16, m2, ta, ma
-; ZVFHMIN64-NEXT: vslidedown.vi v18, v16, 15
-; ZVFHMIN64-NEXT: vslidedown.vi v14, v16, 14
+; ZVFHMIN64-NEXT: vslidedown.vi v14, v16, 15
+; ZVFHMIN64-NEXT: vslidedown.vi v12, v16, 14
; ZVFHMIN64-NEXT: vslidedown.vi v8, v16, 13
-; ZVFHMIN64-NEXT: vslidedown.vi v12, v16, 12
+; ZVFHMIN64-NEXT: vslidedown.vi v18, v16, 12
; ZVFHMIN64-NEXT: vslidedown.vi v10, v16, 11
-; ZVFHMIN64-NEXT: vslidedown.vi v6, v16, 10
+; ZVFHMIN64-NEXT: vslidedown.vi v2, v16, 10
; ZVFHMIN64-NEXT: csrr a2, vlenb
-; ZVFHMIN64-NEXT: li a3, 18
+; ZVFHMIN64-NEXT: li a3, 19
; ZVFHMIN64-NEXT: mul a2, a2, a3
; ZVFHMIN64-NEXT: add a2, sp, a2
; ZVFHMIN64-NEXT: addi a2, a2, 800
-; ZVFHMIN64-NEXT: vs2r.v v6, (a2) # Unknown-size Folded Spill
-; ZVFHMIN64-NEXT: vslidedown.vi v6, v16, 9
+; ZVFHMIN64-NEXT: vs2r.v v2, (a2) # Unknown-size Folded Spill
+; ZVFHMIN64-NEXT: vslidedown.vi v2, v16, 9
; ZVFHMIN64-NEXT: csrr a2, vlenb
; ZVFHMIN64-NEXT: li a3, 14
; ZVFHMIN64-NEXT: mul a2, a2, a3
; ZVFHMIN64-NEXT: add a2, sp, a2
; ZVFHMIN64-NEXT: addi a2, a2, 800
-; ZVFHMIN64-NEXT: vs2r.v v6, (a2) # Unknown-size Folded Spill
-; ZVFHMIN64-NEXT: vslidedown.vi v6, v16, 8
+; ZVFHMIN64-NEXT: vs2r.v v2, (a2) # Unknown-size Folded Spill
+; ZVFHMIN64-NEXT: vslidedown.vi v16, v16, 8
+; ZVFHMIN64-NEXT: csrr a2, vlenb
+; ZVFHMIN64-NEXT: li a3, 12
+; ZVFHMIN64-NEXT: mul a2, a2, a3
+; ZVFHMIN64-NEXT: add a2, sp, a2
+; ZVFHMIN64-NEXT: addi a2, a2, 800
+; ZVFHMIN64-NEXT: vs2r.v v16, (a2) # Unknown-size Folded Spill
; ZVFHMIN64-NEXT: fmv.h.x fa5, a0
; ZVFHMIN64-NEXT: fmv.h.x fa4, a1
; ZVFHMIN64-NEXT: feq.h a0, fa5, fa4
@@ -2571,95 +2548,86 @@ define <128 x i1> @fcmp_oeq_vv_v128f16(<128 x half> %va, <128 x half> %vb, <128
; ZVFHMIN64-NEXT: lh a0, 558(sp)
; ZVFHMIN64-NEXT: lh a1, 302(sp)
; ZVFHMIN64-NEXT: vsetivli zero, 1, e16, m1, ta, ma
-; ZVFHMIN64-NEXT: vslidedown.vi v13, v0, 7
-; ZVFHMIN64-NEXT: vslidedown.vi v29, v0, 6
-; ZVFHMIN64-NEXT: vslidedown.vi v11, v0, 5
-; ZVFHMIN64-NEXT: vslidedown.vi v7, v0, 4
-; ZVFHMIN64-NEXT: vslidedown.vi v9, v0, 3
-; ZVFHMIN64-NEXT: vslidedown.vi v21, v0, 2
-; ZVFHMIN64-NEXT: vslidedown.vi v27, v0, 1
+; ZVFHMIN64-NEXT: vslidedown.vi v29, v0, 7
+; ZVFHMIN64-NEXT: vslidedown.vi v19, v0, 6
+; ZVFHMIN64-NEXT: vslidedown.vi v27, v0, 5
+; ZVFHMIN64-NEXT: vslidedown.vi v13, v0, 4
+; ZVFHMIN64-NEXT: vslidedown.vi v15, v0, 3
+; ZVFHMIN64-NEXT: vslidedown.vi v9, v0, 2
+; ZVFHMIN64-NEXT: vslidedown.vi v11, v0, 1
; ZVFHMIN64-NEXT: vsetivli zero, 1, e16, m2, ta, ma
; ZVFHMIN64-NEXT: vslidedown.vi v16, v0, 15
-; ZVFHMIN64-NEXT: csrr a2, vlenb
-; ZVFHMIN64-NEXT: slli a2, a2, 2
-; ZVFHMIN64-NEXT: add a2, sp, a2
-; ZVFHMIN64-NEXT: addi a2, a2, 800
-; ZVFHMIN64-NEXT: vs2r.v v16, (a2) # Unknown-size Folded Spill
-; ZVFHMIN64-NEXT: vslidedown.vi v16, v0, 14
-; ZVFHMIN64-NEXT: csrr a2, vlenb
-; ZVFHMIN64-NEXT: slli a2, a2, 3
-; ZVFHMIN64-NEXT: add a2, sp, a2
-; ZVFHMIN64-NEXT: addi a2, a2, 800
-; ZVFHMIN64-NEXT: vs2r.v v16, (a2) # Unknown-size Folded Spill
-; ZVFHMIN64-NEXT: vslidedown.vi v16, v0, 13
+; ZVFHMIN64-NEXT: vslidedown.vi v2, v0, 14
; ZVFHMIN64-NEXT: csrr a2, vlenb
; ZVFHMIN64-NEXT: li a3, 6
; ZVFHMIN64-NEXT: mul a2, a2, a3
; ZVFHMIN64-NEXT: add a2, sp, a2
; ZVFHMIN64-NEXT: addi a2, a2, 800
-; ZVFHMIN64-NEXT: vs2r.v v16, (a2) # Unknown-size Folded Spill
-; ZVFHMIN64-NEXT: vslidedown.vi v16, v0, 12
+; ZVFHMIN64-NEXT: vs2r.v v2, (a2) # Unknown-size Folded Spill
+; ZVFHMIN64-NEXT: vslidedown.vi v2, v0, 13
; ZVFHMIN64-NEXT: csrr a2, vlenb
-; ZVFHMIN64-NEXT: li a3, 12
-; ZVFHMIN64-NEXT: mul a2, a2, a3
+; ZVFHMIN64-NEXT: slli a2, a2, 2
; ZVFHMIN64-NEXT: add a2, sp, a2
; ZVFHMIN64-NEXT: addi a2, a2, 800
-; ZVFHMIN64-NEXT: vs2r.v v16, (a2) # Unknown-size Folded Spill
-; ZVFHMIN64-NEXT: vslidedown.vi v16, v0, 11
+; ZVFHMIN64-NEXT: vs2r.v v2, (a2) # Unknown-size Folded Spill
+; ZVFHMIN64-NEXT: vslidedown.vi v2, v0, 12
; ZVFHMIN64-NEXT: csrr a2, vlenb
; ZVFHMIN64-NEXT: li a3, 10
; ZVFHMIN64-NEXT: mul a2, a2, a3
; ZVFHMIN64-NEXT: add a2, sp, a2
; ZVFHMIN64-NEXT: addi a2, a2, 800
-; ZVFHMIN64-NEXT: vs2r.v v16, (a2) # Unknown-size Folded Spill
-; ZVFHMIN64-NEXT: vslidedown.vi v16, v0, 10
+; ZVFHMIN64-NEXT: vs2r.v v2, (a2) # Unknown-size Folded Spill
+; ZVFHMIN64-NEXT: vslidedown.vi v2, v0, 11
+; ZVFHMIN64-NEXT: csrr a2, vlenb
+; ZVFHMIN64-NEXT: slli a2, a2, 3
+; ZVFHMIN64-NEXT: add a2, sp, a2
+; ZVFHMIN64-NEXT: addi a2, a2, 800
+; ZVFHMIN64-NEXT: vs2r.v v2, (a2) # Unknown-size Folded Spill
+; ZVFHMIN64-NEXT: vslidedown.vi v2, v0, 10
; ZVFHMIN64-NEXT: csrr a2, vlenb
; ZVFHMIN64-NEXT: slli a2, a2, 4
; ZVFHMIN64-NEXT: add a2, sp, a2
; ZVFHMIN64-NEXT: addi a2, a2, 800
-; ZVFHMIN64-NEXT: vs2r.v v16, (a2) # Unknown-size Folded Spill
-; ZVFHMIN64-NEXT: vslidedown.vi v16, v0, 9
+; ZVFHMIN64-NEXT: vs2r.v v2, (a2) # Unknown-size Folded Spill
+; ZVFHMIN64-NEXT: vslidedown.vi v2, v0, 9
; ZVFHMIN64-NEXT: vslidedown.vi v0, v0, 8
-; ZVFHMIN64-NEXT: addi a2, sp, 800
-; ZVFHMIN64-NEXT: vs2r.v v0, (a2) # Unknown-size Folded Spill
-; ZVFHMIN64-NEXT: vmv.x.s t4, v26
+; ZVFHMIN64-NEXT: vmv.x.s t3, v4
; ZVFHMIN64-NEXT: fmv.h.x fa5, a0
; ZVFHMIN64-NEXT: fmv.h.x fa4, a1
; ZVFHMIN64-NEXT: feq.h a0, fa5, fa4
; ZVFHMIN64-NEXT: sb a0, 215(sp)
; ZVFHMIN64-NEXT: lh a0, 556(sp)
; ZVFHMIN64-NEXT: lh a1, 300(sp)
-; ZVFHMIN64-NEXT: vmv.x.s t3, v20
-; ZVFHMIN64-NEXT: vmv.x.s t1, v28
+; ZVFHMIN64-NEXT: csrr a2, vlenb
+; ZVFHMIN64-NEXT: add a2, sp, a2
+; ZVFHMIN64-NEXT: addi a2, a2, 800
+; ZVFHMIN64-NEXT: vl2r.v v4, (a2) # Unknown-size Folded Reload
+; ZVFHMIN64-NEXT: vmv.x.s t4, v4
+; ZVFHMIN64-NEXT: vmv.x.s t2, v30
; ZVFHMIN64-NEXT: fmv.h.x fa5, a0
; ZVFHMIN64-NEXT: fmv.h.x fa4, a1
; ZVFHMIN64-NEXT: feq.h a0, fa5, fa4
; ZVFHMIN64-NEXT: sb a0, 214(sp)
; ZVFHMIN64-NEXT: lh a0, 554(sp)
; ZVFHMIN64-NEXT: lh a1, 298(sp)
-; ZVFHMIN64-NEXT: csrr a2, vlenb
-; ZVFHMIN64-NEXT: slli a2, a2, 1
-; ZVFHMIN64-NEXT: add a2, sp, a2
-; ZVFHMIN64-NEXT: addi a2, a2, 800
-; ZVFHMIN64-NEXT: vl2r.v v0, (a2) # Unknown-size Folded Reload
-; ZVFHMIN64-NEXT: vmv.x.s t2, v0
-; ZVFHMIN64-NEXT: vmv.x.s t0, v4
+; ZVFHMIN64-NEXT: vmv.x.s t1, v6
+; ZVFHMIN64-NEXT: vmv.x.s t0, v28
; ZVFHMIN64-NEXT: fmv.h.x fa5, a0
; ZVFHMIN64-NEXT: fmv.h.x fa4, a1
; ZVFHMIN64-NEXT: feq.h a0, fa5, fa4
; ZVFHMIN64-NEXT: sb a0, 213(sp)
; ZVFHMIN64-NEXT: lh a0, 552(sp)
; ZVFHMIN64-NEXT: lh a1, 296(sp)
-; ZVFHMIN64-NEXT: vmv.x.s a7, v2
-; ZVFHMIN64-NEXT: vmv.x.s a6, v30
+; ZVFHMIN64-NEXT: vmv.x.s a7, v26
+; ZVFHMIN64-NEXT: vmv.x.s a6, v22
; ZVFHMIN64-NEXT: fmv.h.x fa5, a0
; ZVFHMIN64-NEXT: fmv.h.x fa4, a1
; ZVFHMIN64-NEXT: feq.h a0, fa5, fa4
; ZVFHMIN64-NEXT: sb a0, 212(sp)
; ZVFHMIN64-NEXT: lh a0, 550(sp)
; ZVFHMIN64-NEXT: lh a1, 294(sp)
-; ZVFHMIN64-NEXT: vmv.x.s a5, v22
-; ZVFHMIN64-NEXT: vmv.x.s a2, v18
+; ZVFHMIN64-NEXT: vmv.x.s a5, v20
+; ZVFHMIN64-NEXT: vmv.x.s a2, v14
; ZVFHMIN64-NEXT: sd a2, 96(sp) # 8-byte Folded Spill
; ZVFHMIN64-NEXT: fmv.h.x fa5, a0
; ZVFHMIN64-NEXT: fmv.h.x fa4, a1
@@ -2667,7 +2635,7 @@ define <128 x i1> @fcmp_oeq_vv_v128f16(<128 x half> %va, <128 x half> %vb, <128
; ZVFHMIN64-NEXT: sb a0, 211(sp)
; ZVFHMIN64-NEXT: lh a1, 548(sp)
; ZVFHMIN64-NEXT: lh t5, 292(sp)
-; ZVFHMIN64-NEXT: vmv.x.s a0, v14
+; ZVFHMIN64-NEXT: vmv.x.s a0, v12
; ZVFHMIN64-NEXT: sd a0, 104(sp) # 8-byte Folded Spill
; ZVFHMIN64-NEXT: vmv.x.s a0, v8
; ZVFHMIN64-NEXT: sd a0, 120(sp) # 8-byte Folded Spill
@@ -2694,7 +2662,7 @@ define <128 x i1> @fcmp_oeq_vv_v128f16(<128 x half> %va, <128 x half> %vb, <128
; ZVFHMIN64-NEXT: sb a1, 208(sp)
; ZVFHMIN64-NEXT: lh t5, 738(sp)
; ZVFHMIN64-NEXT: lh t6, 482(sp)
-; ZVFHMIN64-NEXT: vmv.x.s a0, v12
+; ZVFHMIN64-NEXT: vmv.x.s a0, v18
; ZVFHMIN64-NEXT: sd a0, 88(sp) # 8-byte Folded Spill
; ZVFHMIN64-NEXT: vmv.x.s a0, v10
; ZVFHMIN64-NEXT: sd a0, 112(sp) # 8-byte Folded Spill
@@ -2705,12 +2673,12 @@ define <128 x i1> @fcmp_oeq_vv_v128f16(<128 x half> %va, <128 x half> %vb, <128
; ZVFHMIN64-NEXT: lh t5, 736(sp)
; ZVFHMIN64-NEXT: lh t6, 480(sp)
; ZVFHMIN64-NEXT: csrr a0, vlenb
-; ZVFHMIN64-NEXT: li a1, 29
+; ZVFHMIN64-NEXT: li a1, 27
; ZVFHMIN64-NEXT: mul a0, a0, a1
; ZVFHMIN64-NEXT: add a0, sp, a0
; ZVFHMIN64-NEXT: lh s5, 800(a0) # 8-byte Folded Reload
; ZVFHMIN64-NEXT: csrr a0, vlenb
-; ZVFHMIN64-NEXT: li a1, 28
+; ZVFHMIN64-NEXT: li a1, 26
; ZVFHMIN64-NEXT: mul a0, a0, a1
; ZVFHMIN64-NEXT: add a0, sp, a0
; ZVFHMIN64-NEXT: lh s6, 800(a0) # 8-byte Folded Reload
@@ -2721,12 +2689,12 @@ define <128 x i1> @fcmp_oeq_vv_v128f16(<128 x half> %va, <128 x half> %vb, <128
; ZVFHMIN64-NEXT: lh t5, 734(sp)
; ZVFHMIN64-NEXT: lh t6, 478(sp)
; ZVFHMIN64-NEXT: csrr a0, vlenb
-; ZVFHMIN64-NEXT: li a1, 27
+; ZVFHMIN64-NEXT: li a1, 25
; ZVFHMIN64-NEXT: mul a0, a0, a1
; ZVFHMIN64-NEXT: add a0, sp, a0
; ZVFHMIN64-NEXT: lh s7, 800(a0) # 8-byte Folded Reload
; ZVFHMIN64-NEXT: csrr a0, vlenb
-; ZVFHMIN64-NEXT: li a1, 26
+; ZVFHMIN64-NEXT: li a1, 24
; ZVFHMIN64-NEXT: mul a0, a0, a1
; ZVFHMIN64-NEXT: add a0, sp, a0
; ZVFHMIN64-NEXT: lh s8, 800(a0) # 8-byte Folded Reload
@@ -2737,12 +2705,12 @@ define <128 x i1> @fcmp_oeq_vv_v128f16(<128 x half> %va, <128 x half> %vb, <128
; ZVFHMIN64-NEXT: lh t5, 732(sp)
; ZVFHMIN64-NEXT: lh t6, 476(sp)
; ZVFHMIN64-NEXT: csrr a0, vlenb
-; ZVFHMIN64-NEXT: li a1, 25
+; ZVFHMIN64-NEXT: li a1, 23
; ZVFHMIN64-NEXT: mul a0, a0, a1
; ZVFHMIN64-NEXT: add a0, sp, a0
; ZVFHMIN64-NEXT: lh s4, 800(a0) # 8-byte Folded Reload
; ZVFHMIN64-NEXT: csrr a0, vlenb
-; ZVFHMIN64-NEXT: li a1, 24
+; ZVFHMIN64-NEXT: li a1, 22
; ZVFHMIN64-NEXT: mul a0, a0, a1
; ZVFHMIN64-NEXT: add a0, sp, a0
; ZVFHMIN64-NEXT: lh s3, 800(a0) # 8-byte Folded Reload
@@ -2753,43 +2721,47 @@ define <128 x i1> @fcmp_oeq_vv_v128f16(<128 x half> %va, <128 x half> %vb, <128
; ZVFHMIN64-NEXT: lh t6, 730(sp)
; ZVFHMIN64-NEXT: lh s9, 474(sp)
; ZVFHMIN64-NEXT: csrr a0, vlenb
-; ZVFHMIN64-NEXT: li a1, 23
+; ZVFHMIN64-NEXT: li a1, 21
; ZVFHMIN64-NEXT: mul a0, a0, a1
; ZVFHMIN64-NEXT: add a0, sp, a0
; ZVFHMIN64-NEXT: lh s2, 800(a0) # 8-byte Folded Reload
-; ZVFHMIN64-NEXT: vmv.x.s t5, v3
+; ZVFHMIN64-NEXT: csrr a0, vlenb
+; ZVFHMIN64-NEXT: slli a1, a0, 1
+; ZVFHMIN64-NEXT: add a0, a1, a0
+; ZVFHMIN64-NEXT: add a0, sp, a0
+; ZVFHMIN64-NEXT: lh t5, 800(a0) # 8-byte Folded Reload
; ZVFHMIN64-NEXT: fmv.h.x fa5, t6
; ZVFHMIN64-NEXT: fmv.h.x fa4, s9
; ZVFHMIN64-NEXT: feq.h t6, fa5, fa4
; ZVFHMIN64-NEXT: sb t6, 173(sp)
; ZVFHMIN64-NEXT: lh s9, 728(sp)
; ZVFHMIN64-NEXT: lh s10, 472(sp)
-; ZVFHMIN64-NEXT: vmv.x.s t6, v31
-; ZVFHMIN64-NEXT: vmv.x.s ra, v13
+; ZVFHMIN64-NEXT: vmv.x.s t6, v21
+; ZVFHMIN64-NEXT: vmv.x.s ra, v29
; ZVFHMIN64-NEXT: fmv.h.x fa5, s9
; ZVFHMIN64-NEXT: fmv.h.x fa4, s10
; ZVFHMIN64-NEXT: feq.h s9, fa5, fa4
; ZVFHMIN64-NEXT: sb s9, 172(sp)
; ZVFHMIN64-NEXT: lh s9, 726(sp)
; ZVFHMIN64-NEXT: lh s10, 470(sp)
-; ZVFHMIN64-NEXT: vmv.x.s a2, v29
-; ZVFHMIN64-NEXT: vmv.x.s a3, v11
+; ZVFHMIN64-NEXT: vmv.x.s a2, v19
+; ZVFHMIN64-NEXT: vmv.x.s a3, v27
; ZVFHMIN64-NEXT: fmv.h.x fa5, s9
; ZVFHMIN64-NEXT: fmv.h.x fa4, s10
; ZVFHMIN64-NEXT: feq.h s9, fa5, fa4
; ZVFHMIN64-NEXT: sb s9, 171(sp)
; ZVFHMIN64-NEXT: lh s10, 724(sp)
; ZVFHMIN64-NEXT: lh s11, 468(sp)
-; ZVFHMIN64-NEXT: vmv.x.s a4, v7
-; ZVFHMIN64-NEXT: vmv.x.s s9, v9
+; ZVFHMIN64-NEXT: vmv.x.s a4, v13
+; ZVFHMIN64-NEXT: vmv.x.s s9, v15
; ZVFHMIN64-NEXT: fmv.h.x fa5, s10
; ZVFHMIN64-NEXT: fmv.h.x fa4, s11
; ZVFHMIN64-NEXT: feq.h s10, fa5, fa4
; ZVFHMIN64-NEXT: sb s10, 170(sp)
; ZVFHMIN64-NEXT: lh a0, 722(sp)
; ZVFHMIN64-NEXT: lh a1, 466(sp)
-; ZVFHMIN64-NEXT: vmv.x.s s10, v21
-; ZVFHMIN64-NEXT: vmv.x.s s11, v27
+; ZVFHMIN64-NEXT: vmv.x.s s10, v9
+; ZVFHMIN64-NEXT: vmv.x.s s11, v11
; ZVFHMIN64-NEXT: fmv.h.x fa5, a0
; ZVFHMIN64-NEXT: fmv.h.x fa4, a1
; ZVFHMIN64-NEXT: feq.h a0, fa5, fa4
@@ -2872,7 +2844,11 @@ define <128 x i1> @fcmp_oeq_vv_v128f16(<128 x half> %va, <128 x half> %vb, <128
; ZVFHMIN64-NEXT: sb a1, 161(sp)
; ZVFHMIN64-NEXT: lh a0, 610(sp)
; ZVFHMIN64-NEXT: lh a1, 354(sp)
-; ZVFHMIN64-NEXT: vmv.x.s s6, v5
+; ZVFHMIN64-NEXT: csrr a2, vlenb
+; ZVFHMIN64-NEXT: li a3, 18
+; ZVFHMIN64-NEXT: mul a2, a2, a3
+; ZVFHMIN64-NEXT: add a2, sp, a2
+; ZVFHMIN64-NEXT: lh s6, 800(a2) # 8-byte Folded Reload
; ZVFHMIN64-NEXT: vmv.x.s s5, v23
; ZVFHMIN64-NEXT: fmv.h.x fa5, a0
; ZVFHMIN64-NEXT: fmv.h.x fa4, a1
@@ -2880,27 +2856,15 @@ define <128 x i1> @fcmp_oeq_vv_v128f16(<128 x half> %va, <128 x half> %vb, <128
; ZVFHMIN64-NEXT: sb a0, 241(sp)
; ZVFHMIN64-NEXT: lh a0, 608(sp)
; ZVFHMIN64-NEXT: lh a1, 352(sp)
-; ZVFHMIN64-NEXT: csrr a2, vlenb
-; ZVFHMIN64-NEXT: li a3, 21
-; ZVFHMIN64-NEXT: mul a2, a2, a3
-; ZVFHMIN64-NEXT: add a2, sp, a2
-; ZVFHMIN64-NEXT: lh s4, 800(a2) # 8-byte Folded Reload
-; ZVFHMIN64-NEXT: csrr a2, vlenb
-; ZVFHMIN64-NEXT: li a3, 20
-; ZVFHMIN64-NEXT: mul a2, a2, a3
-; ZVFHMIN64-NEXT: add a2, sp, a2
-; ZVFHMIN64-NEXT: lh s3, 800(a2) # 8-byte Folded Reload
+; ZVFHMIN64-NEXT: vmv.x.s s4, v31
+; ZVFHMIN64-NEXT: lh s3, 800(sp) # 8-byte Folded Reload
; ZVFHMIN64-NEXT: fmv.h.x fa5, a0
; ZVFHMIN64-NEXT: fmv.h.x fa4, a1
; ZVFHMIN64-NEXT: feq.h a0, fa5, fa4
; ZVFHMIN64-NEXT: sb a0, 240(sp)
; ZVFHMIN64-NEXT: lh a0, 606(sp)
; ZVFHMIN64-NEXT: lh a1, 350(sp)
-; ZVFHMIN64-NEXT: csrr a2, vlenb
-; ZVFHMIN64-NEXT: li a3, 22
-; ZVFHMIN64-NEXT: mul a2, a2, a3
-; ZVFHMIN64-NEXT: add a2, sp, a2
-; ZVFHMIN64-NEXT: lh s2, 800(a2) # 8-byte Folded Reload
+; ZVFHMIN64-NEXT: vmv.x.s s2, v7
; ZVFHMIN64-NEXT: fmv.h.x fa5, t5
; ZVFHMIN64-NEXT: fmv.h.x fa4, a0
; ZVFHMIN64-NEXT: fmv.h.x fa3, a1
@@ -3035,7 +2999,7 @@ define <128 x i1> @fcmp_oeq_vv_v128f16(<128 x half> %va, <128 x half> %vb, <128
; ZVFHMIN64-NEXT: lh a0, 766(sp)
; ZVFHMIN64-NEXT: lh a1, 510(sp)
; ZVFHMIN64-NEXT: csrr a2, vlenb
-; ZVFHMIN64-NEXT: li a3, 18
+; ZVFHMIN64-NEXT: li a3, 19
; ZVFHMIN64-NEXT: mul a2, a2, a3
; ZVFHMIN64-NEXT: add a2, sp, a2
; ZVFHMIN64-NEXT: addi a2, a2, 800
@@ -3047,20 +3011,21 @@ define <128 x i1> @fcmp_oeq_vv_v128f16(<128 x half> %va, <128 x half> %vb, <128
; ZVFHMIN64-NEXT: add a2, sp, a2
; ZVFHMIN64-NEXT: addi a2, a2, 800
; ZVFHMIN64-NEXT: vl2r.v v8, (a2) # Unknown-size Folded Reload
-; ZVFHMIN64-NEXT: vmv.x.s t6, v8
+; ZVFHMIN64-NEXT: vmv.x.s t5, v8
; ZVFHMIN64-NEXT: fmv.h.x fa5, a0
; ZVFHMIN64-NEXT: fmv.h.x fa4, a1
; ZVFHMIN64-NEXT: feq.h a0, fa5, fa4
; ZVFHMIN64-NEXT: sb a0, 191(sp)
; ZVFHMIN64-NEXT: lh a0, 764(sp)
; ZVFHMIN64-NEXT: lh a1, 508(sp)
-; ZVFHMIN64-NEXT: vmv.x.s t5, v6
; ZVFHMIN64-NEXT: csrr a2, vlenb
-; ZVFHMIN64-NEXT: slli a2, a2, 2
+; ZVFHMIN64-NEXT: li a3, 12
+; ZVFHMIN64-NEXT: mul a2, a2, a3
; ZVFHMIN64-NEXT: add a2, sp, a2
; ZVFHMIN64-NEXT: addi a2, a2, 800
; ZVFHMIN64-NEXT: vl2r.v v8, (a2) # Unknown-size Folded Reload
-; ZVFHMIN64-NEXT: vmv.x.s a2, v8
+; ZVFHMIN64-NEXT: vmv.x.s t6, v8
+; ZVFHMIN64-NEXT: vmv.x.s a2, v16
; ZVFHMIN64-NEXT: fmv.h.x fa5, a0
; ZVFHMIN64-NEXT: fmv.h.x fa4, a1
; ZVFHMIN64-NEXT: feq.h a0, fa5, fa4
@@ -3068,14 +3033,14 @@ define <128 x i1> @fcmp_oeq_vv_v128f16(<128 x half> %va, <128 x half> %vb, <128
; ZVFHMIN64-NEXT: lh a0, 762(sp)
; ZVFHMIN64-NEXT: lh a1, 506(sp)
; ZVFHMIN64-NEXT: csrr a3, vlenb
-; ZVFHMIN64-NEXT: slli a3, a3, 3
+; ZVFHMIN64-NEXT: li a4, 6
+; ZVFHMIN64-NEXT: mul a3, a3, a4
; ZVFHMIN64-NEXT: add a3, sp, a3
; ZVFHMIN64-NEXT: addi a3, a3, 800
; ZVFHMIN64-NEXT: vl2r.v v8, (a3) # Unknown-size Folded Reload
; ZVFHMIN64-NEXT: vmv.x.s a3, v8
; ZVFHMIN64-NEXT: csrr a4, vlenb
-; ZVFHMIN64-NEXT: li s3, 6
-; ZVFHMIN64-NEXT: mul a4, a4, s3
+; ZVFHMIN64-NEXT: slli a4, a4, 2
; ZVFHMIN64-NEXT: add a4, sp, a4
; ZVFHMIN64-NEXT: addi a4, a4, 800
; ZVFHMIN64-NEXT: vl2r.v v8, (a4) # Unknown-size Folded Reload
@@ -3087,15 +3052,14 @@ define <128 x i1> @fcmp_oeq_vv_v128f16(<128 x half> %va, <128 x half> %vb, <128
; ZVFHMIN64-NEXT: lh a0, 760(sp)
; ZVFHMIN64-NEXT: lh a1, 504(sp)
; ZVFHMIN64-NEXT: csrr s3, vlenb
-; ZVFHMIN64-NEXT: li s4, 12
+; ZVFHMIN64-NEXT: li s4, 10
; ZVFHMIN64-NEXT: mul s3, s3, s4
; ZVFHMIN64-NEXT: add s3, sp, s3
; ZVFHMIN64-NEXT: addi s3, s3, 800
; ZVFHMIN64-NEXT: vl2r.v v8, (s3) # Unknown-size Folded Reload
; ZVFHMIN64-NEXT: vmv.x.s s6, v8
; ZVFHMIN64-NEXT: csrr s3, vlenb
-; ZVFHMIN64-NEXT: li s4, 10
-; ZVFHMIN64-NEXT: mul s3, s3, s4
+; ZVFHMIN64-NEXT: slli s3, s3, 3
; ZVFHMIN64-NEXT: add s3, sp, s3
; ZVFHMIN64-NEXT: addi s3, s3, 800
; ZVFHMIN64-NEXT: vl2r.v v8, (s3) # Unknown-size Folded Reload
@@ -3112,38 +3076,38 @@ define <128 x i1> @fcmp_oeq_vv_v128f16(<128 x half> %va, <128 x half> %vb, <128
; ZVFHMIN64-NEXT: addi s3, s3, 800
; ZVFHMIN64-NEXT: vl2r.v v8, (s3) # Unknown-size Folded Reload
; ZVFHMIN64-NEXT: vmv.x.s s5, v8
-; ZVFHMIN64-NEXT: vmv.x.s s3, v16
+; ZVFHMIN64-NEXT: vmv.x.s s3, v2
; ZVFHMIN64-NEXT: fmv.h.x fa5, a0
; ZVFHMIN64-NEXT: fmv.h.x fa4, a1
; ZVFHMIN64-NEXT: feq.h a0, fa5, fa4
-; ZVFHMIN64-NEXT: fmv.h.x fa5, t4
+; ZVFHMIN64-NEXT: fmv.h.x fa5, t3
; ZVFHMIN64-NEXT: sb a0, 187(sp)
; ZVFHMIN64-NEXT: lh a0, 756(sp)
; ZVFHMIN64-NEXT: lh a1, 500(sp)
; ZVFHMIN64-NEXT: fmv.h.x fa4, a2
-; ZVFHMIN64-NEXT: feq.h t4, fa5, fa4
+; ZVFHMIN64-NEXT: feq.h t3, fa5, fa4
; ZVFHMIN64-NEXT: fmv.h.x fa5, a0
; ZVFHMIN64-NEXT: fmv.h.x fa4, a1
; ZVFHMIN64-NEXT: feq.h a0, fa5, fa4
-; ZVFHMIN64-NEXT: fmv.h.x fa5, t3
+; ZVFHMIN64-NEXT: fmv.h.x fa5, t4
; ZVFHMIN64-NEXT: sb a0, 186(sp)
; ZVFHMIN64-NEXT: lh a0, 754(sp)
; ZVFHMIN64-NEXT: lh a1, 498(sp)
; ZVFHMIN64-NEXT: fmv.h.x fa4, a3
-; ZVFHMIN64-NEXT: feq.h t3, fa5, fa4
+; ZVFHMIN64-NEXT: feq.h t4, fa5, fa4
; ZVFHMIN64-NEXT: fmv.h.x fa5, a0
; ZVFHMIN64-NEXT: fmv.h.x fa4, a1
; ZVFHMIN64-NEXT: feq.h a0, fa5, fa4
-; ZVFHMIN64-NEXT: fmv.h.x fa5, t1
+; ZVFHMIN64-NEXT: fmv.h.x fa5, t2
; ZVFHMIN64-NEXT: sb a0, 185(sp)
; ZVFHMIN64-NEXT: lh a0, 752(sp)
; ZVFHMIN64-NEXT: lh a1, 496(sp)
; ZVFHMIN64-NEXT: fmv.h.x fa4, a4
-; ZVFHMIN64-NEXT: feq.h t1, fa5, fa4
+; ZVFHMIN64-NEXT: feq.h t2, fa5, fa4
; ZVFHMIN64-NEXT: fmv.h.x fa5, a0
; ZVFHMIN64-NEXT: fmv.h.x fa4, a1
; ZVFHMIN64-NEXT: feq.h a0, fa5, fa4
-; ZVFHMIN64-NEXT: fmv.h.x fa5, t2
+; ZVFHMIN64-NEXT: fmv.h.x fa5, t1
; ZVFHMIN64-NEXT: sb a0, 184(sp)
; ZVFHMIN64-NEXT: lh a0, 750(sp)
; ZVFHMIN64-NEXT: lh a1, 494(sp)
@@ -3180,9 +3144,7 @@ define <128 x i1> @fcmp_oeq_vv_v128f16(<128 x half> %va, <128 x half> %vb, <128
; ZVFHMIN64-NEXT: fmv.h.x fa4, a1
; ZVFHMIN64-NEXT: feq.h a0, fa5, fa4
; ZVFHMIN64-NEXT: fmv.h.x fa5, a5
-; ZVFHMIN64-NEXT: addi a1, sp, 800
-; ZVFHMIN64-NEXT: vl2r.v v8, (a1) # Unknown-size Folded Reload
-; ZVFHMIN64-NEXT: vmv.x.s a1, v8
+; ZVFHMIN64-NEXT: vmv.x.s a1, v0
; ZVFHMIN64-NEXT: vsetivli zero, 1, e16, m2, ta, ma
; ZVFHMIN64-NEXT: vslidedown.vi v8, v24, 15
; ZVFHMIN64-NEXT: vmv.x.s a5, v8
@@ -3198,9 +3160,9 @@ define <128 x i1> @fcmp_oeq_vv_v128f16(<128 x half> %va, <128 x half> %vb, <128
; ZVFHMIN64-NEXT: lh a0, 740(sp)
; ZVFHMIN64-NEXT: lh a7, 484(sp)
; ZVFHMIN64-NEXT: sb a2, 140(sp)
-; ZVFHMIN64-NEXT: sb t1, 141(sp)
-; ZVFHMIN64-NEXT: sb t3, 142(sp)
-; ZVFHMIN64-NEXT: sb t4, 143(sp)
+; ZVFHMIN64-NEXT: sb t2, 141(sp)
+; ZVFHMIN64-NEXT: sb t4, 142(sp)
+; ZVFHMIN64-NEXT: sb t3, 143(sp)
; ZVFHMIN64-NEXT: sb a1, 136(sp)
; ZVFHMIN64-NEXT: sb a6, 137(sp)
; ZVFHMIN64-NEXT: sb a4, 138(sp)
@@ -3315,7 +3277,7 @@ define <128 x i1> @fcmp_oeq_vv_v128f16(<128 x half> %va, <128 x half> %vb, <128
; ZVFHMIN64-NEXT: fmv.h.x fa5, a0
; ZVFHMIN64-NEXT: fmv.h.x fa4, a1
; ZVFHMIN64-NEXT: feq.h a0, fa5, fa4
-; ZVFHMIN64-NEXT: fmv.h.x fa5, t6
+; ZVFHMIN64-NEXT: fmv.h.x fa5, t5
; ZVFHMIN64-NEXT: sb a0, 244(sp)
; ZVFHMIN64-NEXT: lh a0, 614(sp)
; ZVFHMIN64-NEXT: lh a1, 358(sp)
@@ -3324,7 +3286,7 @@ define <128 x i1> @fcmp_oeq_vv_v128f16(<128 x half> %va, <128 x half> %vb, <128
; ZVFHMIN64-NEXT: fmv.h.x fa5, a0
; ZVFHMIN64-NEXT: fmv.h.x fa4, a1
; ZVFHMIN64-NEXT: feq.h a0, fa5, fa4
-; ZVFHMIN64-NEXT: fmv.h.x fa5, t5
+; ZVFHMIN64-NEXT: fmv.h.x fa5, t6
; ZVFHMIN64-NEXT: vslidedown.vi v8, v24, 8
; ZVFHMIN64-NEXT: vmv.x.s a1, v8
; ZVFHMIN64-NEXT: fmv.h.x fa4, a1
@@ -3944,49 +3906,39 @@ define <32 x i1> @fcmp_oeq_vv_v32f64(<32 x double> %va, <32 x double> %vb, <32 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: 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: csrr a1, vlenb
; CHECK-NEXT: slli a1, a1, 3
-; CHECK-NEXT: add a1, sp, a1
-; CHECK-NEXT: addi a1, a1, 16
+; 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: addi a1, sp, 16
; CHECK-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
; 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, sp, 16
-; CHECK-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
-; CHECK-NEXT: vle64.v v16, (a0)
+; CHECK-NEXT: vle64.v v24, (a0)
; CHECK-NEXT: li a1, 16
; CHECK-NEXT: mv a0, a2
; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
-; CHECK-NEXT: vslidedown.vi v24, v0, 2
+; CHECK-NEXT: vslidedown.vi v7, v0, 2
; CHECK-NEXT: bltu a2, a1, .LBB87_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: li a0, 16
; CHECK-NEXT: .LBB87_2:
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: vmfeq.vv v7, v8, v16, v0.t
+; CHECK-NEXT: vmfeq.vv v6, 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, v24
-; 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: vmv1r.v v0, v7
; 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: vmfeq.vv v8, v16, v24, v0.t
+; CHECK-NEXT: vmfeq.vv v8, v24, v16, v0.t
; CHECK-NEXT: vsetivli zero, 4, e8, mf4, ta, ma
-; CHECK-NEXT: vslideup.vi v7, v8, 2
-; CHECK-NEXT: vmv1r.v v0, v7
+; CHECK-NEXT: vslideup.vi v6, v8, 2
+; CHECK-NEXT: vmv1r.v v0, v6
; 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: .cfi_def_cfa sp, 16
; CHECK-NEXT: addi sp, sp, 16
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 d52c42891fcc3b..cf7f284d7cd085 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
@@ -595,47 +595,37 @@ define <256 x i1> @icmp_eq_vv_v256i8(<256 x i8> %va, <256 x i8> %vb, <256 x i1>
; 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: .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: slli a1, a1, 3
-; CHECK-NEXT: add a1, sp, a1
-; CHECK-NEXT: addi a1, a1, 16
+; CHECK-NEXT: addi a1, sp, 16
; CHECK-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
; CHECK-NEXT: li a1, 128
; CHECK-NEXT: addi a4, a0, 128
; CHECK-NEXT: vsetvli zero, a1, e8, m8, ta, ma
; CHECK-NEXT: vlm.v v0, (a2)
; CHECK-NEXT: addi a2, a3, -128
-; CHECK-NEXT: vle8.v v8, (a4)
+; CHECK-NEXT: vle8.v v24, (a4)
; CHECK-NEXT: sltu a4, a3, a2
-; 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 v8, (a0)
; 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: vmseq.vv v6, v16, v24, v0.t
; CHECK-NEXT: bltu a3, a1, .LBB51_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: li a3, 128
; CHECK-NEXT: .LBB51_2:
; 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: 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: vmseq.vv v16, v24, v8, v0.t
; CHECK-NEXT: vmv1r.v v0, v16
; CHECK-NEXT: vmv1r.v v8, v6
; 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: .cfi_def_cfa sp, 16
; CHECK-NEXT: addi sp, sp, 16
@@ -1250,49 +1240,39 @@ define <64 x i1> @icmp_eq_vv_v64i32(<64 x i32> %va, <64 x i32> %vb, <64 x i1> %m
; 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: csrr a1, vlenb
; CHECK-NEXT: slli a1, a1, 3
-; CHECK-NEXT: add a1, sp, a1
-; CHECK-NEXT: addi a1, a1, 16
+; 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: addi a1, sp, 16
; CHECK-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
; CHECK-NEXT: addi a1, a0, 128
; CHECK-NEXT: li a3, 32
; CHECK-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; CHECK-NEXT: vle32.v v16, (a1)
-; CHECK-NEXT: addi a1, sp, 16
-; CHECK-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
-; CHECK-NEXT: vle32.v v16, (a0)
+; CHECK-NEXT: vle32.v v24, (a0)
; CHECK-NEXT: mv a0, a2
; CHECK-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
-; CHECK-NEXT: vslidedown.vi v24, v0, 4
+; CHECK-NEXT: vslidedown.vi v7, v0, 4
; CHECK-NEXT: bltu a2, a3, .LBB99_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: li a0, 32
; CHECK-NEXT: .LBB99_2:
; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
-; CHECK-NEXT: vmseq.vv v7, v8, v16, v0.t
+; CHECK-NEXT: vmseq.vv v6, 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, v24
-; 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: vmv1r.v v0, v7
; 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: vmseq.vv v8, v16, v24, v0.t
+; CHECK-NEXT: vmseq.vv v8, v24, v16, v0.t
; CHECK-NEXT: vsetivli zero, 8, e8, mf2, ta, ma
-; CHECK-NEXT: vslideup.vi v7, v8, 4
-; CHECK-NEXT: vmv1r.v v0, v7
+; CHECK-NEXT: vslideup.vi v6, v8, 4
+; CHECK-NEXT: vmv1r.v v0, v6
; 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: .cfi_def_cfa sp, 16
; CHECK-NEXT: addi sp, sp, 16
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 12893ec55cda76..f2fde30a74ab46 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-trunc-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-trunc-vp.ll
@@ -227,19 +227,18 @@ define <128 x i32> @vtrunc_v128i32_v128i64(<128 x i64> %a, <128 x i1> %m, i32 ze
; CHECK-NEXT: addi sp, sp, -16
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: csrr a2, vlenb
-; CHECK-NEXT: li a3, 72
-; CHECK-NEXT: mul a2, a2, a3
+; CHECK-NEXT: slli a2, a2, 6
; CHECK-NEXT: sub sp, sp, a2
-; CHECK-NEXT: .cfi_escape 0x0f, 0x0e, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0xc8, 0x00, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 72 * vlenb
+; CHECK-NEXT: .cfi_escape 0x0f, 0x0e, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0xc0, 0x00, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 64 * vlenb
; CHECK-NEXT: vmv1r.v v7, v0
; CHECK-NEXT: csrr a2, vlenb
-; CHECK-NEXT: li a3, 24
-; CHECK-NEXT: mul a2, a2, a3
+; 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: csrr a2, vlenb
-; CHECK-NEXT: slli a2, a2, 5
+; 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: vs8r.v v8, (a2) # Unknown-size Folded Spill
@@ -249,12 +248,12 @@ define <128 x i32> @vtrunc_v128i32_v128i64(<128 x i64> %a, <128 x i1> %m, i32 ze
; CHECK-NEXT: addi a3, a1, 640
; CHECK-NEXT: addi a4, a7, -64
; CHECK-NEXT: vsetivli zero, 4, e8, mf2, ta, ma
-; CHECK-NEXT: vslidedown.vi v27, v6, 4
+; CHECK-NEXT: vslidedown.vi v30, v6, 4
; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; CHECK-NEXT: vle64.v v8, (a3)
; CHECK-NEXT: sltu a3, a7, a4
; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
-; CHECK-NEXT: vslidedown.vi v0, v27, 2
+; CHECK-NEXT: vslidedown.vi v0, v30, 2
; CHECK-NEXT: addi a3, a3, -1
; CHECK-NEXT: and a4, a3, a4
; CHECK-NEXT: addi a3, a4, -32
@@ -268,55 +267,44 @@ define <128 x i32> @vtrunc_v128i32_v128i64(<128 x i64> %a, <128 x i1> %m, i32 ze
; 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: slli a5, a5, 4
+; CHECK-NEXT: slli a5, a5, 3
; 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: vle64.v v16, (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: vslidedown.vi v28, v7, 4
; CHECK-NEXT: bltu a3, a2, .LBB16_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: li a3, 16
; CHECK-NEXT: .LBB16_2:
-; CHECK-NEXT: vmv1r.v v0, v27
+; CHECK-NEXT: vmv1r.v v0, v30
; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; CHECK-NEXT: vle64.v v16, (a5)
-; CHECK-NEXT: csrr a5, 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: vle64.v v8, (a5)
; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
-; CHECK-NEXT: vslidedown.vi v27, v26, 2
+; CHECK-NEXT: vslidedown.vi v29, v28, 2
; CHECK-NEXT: li a5, 64
; CHECK-NEXT: vsetvli zero, a3, e32, m4, ta, ma
-; CHECK-NEXT: vnsrl.wi v16, v8, 0, v0.t
+; CHECK-NEXT: vnsrl.wi v24, v16, 0, v0.t
; CHECK-NEXT: csrr a3, vlenb
-; CHECK-NEXT: slli a3, a3, 6
+; CHECK-NEXT: li a6, 56
+; 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: vs8r.v v24, (a3) # Unknown-size Folded Spill
; CHECK-NEXT: mv a6, a7
; CHECK-NEXT: bltu a7, a5, .LBB16_4
; CHECK-NEXT: # %bb.3:
; CHECK-NEXT: li a6, 64
; CHECK-NEXT: .LBB16_4:
-; CHECK-NEXT: vmv1r.v v0, v27
+; CHECK-NEXT: vmv1r.v v0, v29
; CHECK-NEXT: addi a5, a1, 384
; CHECK-NEXT: li a3, 32
; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; CHECK-NEXT: vle64.v v8, (a1)
-; CHECK-NEXT: csrr t0, vlenb
-; CHECK-NEXT: li t1, 48
-; CHECK-NEXT: mul t0, t0, t1
-; CHECK-NEXT: add t0, sp, t0
-; CHECK-NEXT: addi t0, t0, 16
-; CHECK-NEXT: vs8r.v v8, (t0) # Unknown-size Folded Spill
+; CHECK-NEXT: vle64.v v16, (a1)
; CHECK-NEXT: addi t0, a6, -32
; CHECK-NEXT: sltu a6, a6, t0
; CHECK-NEXT: addi a6, a6, -1
@@ -325,41 +313,24 @@ 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: csrr t1, vlenb
-; CHECK-NEXT: li t2, 56
-; 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, 3
-; CHECK-NEXT: add t0, sp, t0
-; CHECK-NEXT: addi t0, t0, 16
-; CHECK-NEXT: vs8r.v v8, (t0) # Unknown-size Folded Spill
+; CHECK-NEXT: vnsrl.wi v24, v8, 0, v0.t
+; CHECK-NEXT: addi t0, sp, 16
+; CHECK-NEXT: vs8r.v v24, (t0) # Unknown-size Folded Spill
; CHECK-NEXT: bltu a6, a2, .LBB16_6
; CHECK-NEXT: # %bb.5:
; CHECK-NEXT: li a6, 16
; CHECK-NEXT: .LBB16_6:
-; CHECK-NEXT: vmv1r.v v0, v26
+; CHECK-NEXT: vmv1r.v v0, v28
; 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: vle64.v v24, (a5)
; CHECK-NEXT: addi a1, a1, 256
; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
-; CHECK-NEXT: vslidedown.vi v26, v6, 2
-; CHECK-NEXT: csrr a5, vlenb
-; CHECK-NEXT: li t0, 48
-; CHECK-NEXT: mul a5, a5, t0
-; CHECK-NEXT: add a5, sp, a5
-; CHECK-NEXT: addi a5, a5, 16
-; CHECK-NEXT: vl8r.v v16, (a5) # Unknown-size Folded Reload
+; CHECK-NEXT: vslidedown.vi v12, v6, 2
; CHECK-NEXT: vsetvli zero, a6, e32, m4, ta, ma
; CHECK-NEXT: vnsrl.wi v8, v16, 0, v0.t
; CHECK-NEXT: csrr a5, vlenb
-; CHECK-NEXT: li a6, 56
+; CHECK-NEXT: li a6, 48
; CHECK-NEXT: mul a5, a5, a6
; CHECK-NEXT: add a5, sp, a5
; CHECK-NEXT: addi a5, a5, 16
@@ -375,14 +346,11 @@ define <128 x i32> @vtrunc_v128i32_v128i64(<128 x i64> %a, <128 x i1> %m, i32 ze
; CHECK-NEXT: sltu a5, a5, a1
; CHECK-NEXT: addi a5, a5, -1
; CHECK-NEXT: and a1, a5, a1
-; CHECK-NEXT: vmv1r.v v0, v26
-; CHECK-NEXT: addi a5, sp, 16
-; CHECK-NEXT: vl8r.v v24, (a5) # Unknown-size Folded Reload
+; CHECK-NEXT: vmv1r.v v0, v12
; CHECK-NEXT: vsetvli zero, a1, e32, m4, ta, ma
; CHECK-NEXT: vnsrl.wi v8, v24, 0, v0.t
; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: li a5, 40
-; CHECK-NEXT: mul a1, a1, a5
+; CHECK-NEXT: slli a1, a1, 5
; CHECK-NEXT: add a1, sp, a1
; CHECK-NEXT: addi a1, a1, 16
; CHECK-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
@@ -392,11 +360,11 @@ define <128 x i32> @vtrunc_v128i32_v128i64(<128 x i64> %a, <128 x i1> %m, i32 ze
; CHECK-NEXT: .LBB16_10:
; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
-; CHECK-NEXT: vslidedown.vi v25, v7, 2
+; CHECK-NEXT: vslidedown.vi v12, v7, 2
; CHECK-NEXT: vsetvli zero, a4, e32, m4, ta, ma
; CHECK-NEXT: vnsrl.wi v8, v16, 0, v0.t
; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: li a4, 48
+; CHECK-NEXT: li a4, 40
; CHECK-NEXT: mul a1, a1, a4
; CHECK-NEXT: add a1, sp, a1
; CHECK-NEXT: addi a1, a1, 16
@@ -406,71 +374,67 @@ define <128 x i32> @vtrunc_v128i32_v128i64(<128 x i64> %a, <128 x i1> %m, i32 ze
; CHECK-NEXT: # %bb.11:
; CHECK-NEXT: li a1, 32
; CHECK-NEXT: .LBB16_12:
-; CHECK-NEXT: vmv1r.v v0, v25
+; CHECK-NEXT: vmv1r.v v0, v12
; CHECK-NEXT: csrr a4, vlenb
-; CHECK-NEXT: slli a4, a4, 4
+; CHECK-NEXT: slli a4, a4, 3
; CHECK-NEXT: add a4, sp, a4
; CHECK-NEXT: addi a4, a4, 16
; CHECK-NEXT: vl8r.v v16, (a4) # Unknown-size Folded Reload
; CHECK-NEXT: vmv4r.v v24, v16
-; 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: addi a4, sp, 16
; CHECK-NEXT: vl8r.v v8, (a4) # Unknown-size Folded Reload
; CHECK-NEXT: csrr a4, vlenb
-; CHECK-NEXT: li a5, 40
-; CHECK-NEXT: mul a4, a4, a5
+; 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: csrr a4, vlenb
-; CHECK-NEXT: li a5, 40
-; CHECK-NEXT: mul a4, a4, a5
+; CHECK-NEXT: slli a4, a4, 5
; CHECK-NEXT: add a4, sp, a4
; CHECK-NEXT: addi a4, a4, 16
; CHECK-NEXT: vs8r.v v16, (a4) # Unknown-size Folded Spill
; CHECK-NEXT: csrr a4, vlenb
-; CHECK-NEXT: slli a4, a4, 6
+; 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 v16, (a4) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; CHECK-NEXT: vslideup.vi v16, v24, 16
; CHECK-NEXT: csrr a4, vlenb
-; CHECK-NEXT: slli a4, a4, 6
+; 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: vs8r.v v16, (a4) # Unknown-size Folded Spill
; CHECK-NEXT: addi a4, a1, -16
; CHECK-NEXT: csrr a5, vlenb
-; CHECK-NEXT: li a6, 56
+; 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 v16, (a5) # Unknown-size Folded Reload
; CHECK-NEXT: vslideup.vi v16, v8, 16
; CHECK-NEXT: csrr a5, vlenb
-; CHECK-NEXT: li a6, 56
+; 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: vs8r.v v16, (a5) # Unknown-size Folded Spill
; CHECK-NEXT: csrr a5, vlenb
-; CHECK-NEXT: li a6, 48
+; 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 v8, (a5) # Unknown-size Folded Reload
; CHECK-NEXT: csrr a5, vlenb
-; CHECK-NEXT: li a6, 40
-; CHECK-NEXT: mul a5, a5, a6
+; CHECK-NEXT: slli a5, a5, 5
; CHECK-NEXT: add a5, sp, a5
; CHECK-NEXT: addi a5, a5, 16
; CHECK-NEXT: vl8r.v v16, (a5) # Unknown-size Folded Reload
; CHECK-NEXT: vslideup.vi v8, v16, 16
; CHECK-NEXT: csrr a5, vlenb
-; CHECK-NEXT: li a6, 48
+; CHECK-NEXT: li a6, 40
; CHECK-NEXT: mul a5, a5, a6
; CHECK-NEXT: add a5, sp, a5
; CHECK-NEXT: addi a5, a5, 16
@@ -479,8 +443,7 @@ define <128 x i32> @vtrunc_v128i32_v128i64(<128 x i64> %a, <128 x i1> %m, i32 ze
; CHECK-NEXT: addi a1, a1, -1
; CHECK-NEXT: and a1, a1, a4
; CHECK-NEXT: csrr a4, vlenb
-; CHECK-NEXT: li a5, 24
-; CHECK-NEXT: mul a4, a4, a5
+; CHECK-NEXT: slli a4, a4, 4
; CHECK-NEXT: add a4, sp, a4
; CHECK-NEXT: addi a4, a4, 16
; CHECK-NEXT: vl8r.v v16, (a4) # Unknown-size Folded Reload
@@ -492,18 +455,19 @@ define <128 x i32> @vtrunc_v128i32_v128i64(<128 x i64> %a, <128 x i1> %m, i32 ze
; CHECK-NEXT: .LBB16_14:
; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a1, a1, 5
+; 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 v16, (a1) # Unknown-size Folded Reload
+; CHECK-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, a7, e32, m4, ta, ma
-; CHECK-NEXT: vnsrl.wi v24, v16, 0, v0.t
+; CHECK-NEXT: vnsrl.wi v16, v24, 0, v0.t
; CHECK-NEXT: vsetvli zero, a3, e32, m8, ta, ma
-; CHECK-NEXT: vslideup.vi v24, v8, 16
-; CHECK-NEXT: vse32.v v24, (a0)
+; CHECK-NEXT: vslideup.vi v16, v8, 16
+; CHECK-NEXT: vse32.v v16, (a0)
; CHECK-NEXT: addi a1, a0, 256
; CHECK-NEXT: csrr a2, vlenb
-; CHECK-NEXT: li a3, 48
+; CHECK-NEXT: li a3, 40
; CHECK-NEXT: mul a2, a2, a3
; CHECK-NEXT: add a2, sp, a2
; CHECK-NEXT: addi a2, a2, 16
@@ -511,7 +475,7 @@ define <128 x i32> @vtrunc_v128i32_v128i64(<128 x i64> %a, <128 x i1> %m, i32 ze
; CHECK-NEXT: vse32.v v8, (a1)
; CHECK-NEXT: addi a1, a0, 128
; CHECK-NEXT: csrr a2, vlenb
-; CHECK-NEXT: li a3, 56
+; CHECK-NEXT: li a3, 48
; CHECK-NEXT: mul a2, a2, a3
; CHECK-NEXT: add a2, sp, a2
; CHECK-NEXT: addi a2, a2, 16
@@ -519,14 +483,14 @@ define <128 x i32> @vtrunc_v128i32_v128i64(<128 x i64> %a, <128 x i1> %m, i32 ze
; CHECK-NEXT: vse32.v v8, (a1)
; CHECK-NEXT: addi a0, a0, 384
; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a1, a1, 6
+; CHECK-NEXT: li a2, 56
+; 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: vse32.v v8, (a0)
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a1, 72
-; CHECK-NEXT: mul a0, a0, a1
+; CHECK-NEXT: slli a0, a0, 6
; CHECK-NEXT: add sp, sp, a0
; CHECK-NEXT: .cfi_def_cfa sp, 16
; CHECK-NEXT: addi sp, sp, 16
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 fa82065f3b4131..68f9f23e6d3095 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vcopysign-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vcopysign-vp.ll
@@ -298,46 +298,36 @@ 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, 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: csrr a1, vlenb
; CHECK-NEXT: slli a1, a1, 3
-; CHECK-NEXT: add a1, sp, a1
-; CHECK-NEXT: addi a1, a1, 16
+; 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: addi a1, sp, 16
; CHECK-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
; 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, sp, 16
-; CHECK-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
-; CHECK-NEXT: vle64.v v16, (a0)
+; CHECK-NEXT: vle64.v v24, (a0)
; CHECK-NEXT: li a1, 16
; CHECK-NEXT: mv a0, a2
; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
-; CHECK-NEXT: vslidedown.vi v24, v0, 2
+; CHECK-NEXT: vslidedown.vi v7, v0, 2
; CHECK-NEXT: bltu a2, a1, .LBB26_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: li a0, 16
; CHECK-NEXT: .LBB26_2:
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: vfsgnj.vv v8, v8, v16, v0.t
+; 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: vmv1r.v v0, v24
-; 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: vmv1r.v v0, v7
; 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 v16, v16, v24, v0.t
+; 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: .cfi_def_cfa sp, 16
; CHECK-NEXT: addi sp, sp, 16
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 cad7adbc19f3c8..696ff70d271bb6 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfmax-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfmax-vp.ll
@@ -390,46 +390,36 @@ 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, 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: csrr a1, vlenb
; CHECK-NEXT: slli a1, a1, 3
-; CHECK-NEXT: add a1, sp, a1
-; CHECK-NEXT: addi a1, a1, 16
+; 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: addi a1, sp, 16
; CHECK-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
; 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, sp, 16
-; CHECK-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
-; CHECK-NEXT: vle64.v v16, (a0)
+; CHECK-NEXT: vle64.v v24, (a0)
; CHECK-NEXT: li a1, 16
; CHECK-NEXT: mv a0, a2
; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
-; CHECK-NEXT: vslidedown.vi v24, v0, 2
+; CHECK-NEXT: vslidedown.vi v7, v0, 2
; CHECK-NEXT: bltu a2, a1, .LBB26_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: li a0, 16
; CHECK-NEXT: .LBB26_2:
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: vfmax.vv v8, v8, v16, v0.t
+; 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: vmv1r.v v0, v24
-; 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: vmv1r.v v0, v7
; 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 v16, v16, v24, v0.t
+; CHECK-NEXT: vfmax.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: .cfi_def_cfa sp, 16
; CHECK-NEXT: addi sp, sp, 16
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 d8ee7a7044b49c..fdbf00ed75c5a6 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfmin-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfmin-vp.ll
@@ -390,46 +390,36 @@ 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, 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: csrr a1, vlenb
; CHECK-NEXT: slli a1, a1, 3
-; CHECK-NEXT: add a1, sp, a1
-; CHECK-NEXT: addi a1, a1, 16
+; 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: addi a1, sp, 16
; CHECK-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
; 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, sp, 16
-; CHECK-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
-; CHECK-NEXT: vle64.v v16, (a0)
+; CHECK-NEXT: vle64.v v24, (a0)
; CHECK-NEXT: li a1, 16
; CHECK-NEXT: mv a0, a2
; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
-; CHECK-NEXT: vslidedown.vi v24, v0, 2
+; CHECK-NEXT: vslidedown.vi v7, v0, 2
; CHECK-NEXT: bltu a2, a1, .LBB26_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: li a0, 16
; CHECK-NEXT: .LBB26_2:
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: vfmin.vv v8, v8, v16, v0.t
+; 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: vmv1r.v v0, v24
-; 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: vmv1r.v v0, v7
; 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 v16, v16, v24, v0.t
+; CHECK-NEXT: vfmin.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: .cfi_def_cfa sp, 16
; CHECK-NEXT: addi sp, sp, 16
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 1d8af4c46cc078..828460539b0e51 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vselect-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vselect-vp.ll
@@ -205,51 +205,30 @@ define <256 x i8> @select_evl_v256i8(<256 x i1> %a, <256 x i8> %b, <256 x i8> %c
; CHECK-NEXT: addi sp, sp, -16
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: csrr a2, vlenb
-; CHECK-NEXT: li a3, 24
-; CHECK-NEXT: mul a2, a2, a3
+; CHECK-NEXT: slli a2, a2, 3
; CHECK-NEXT: sub sp, sp, a2
-; 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: csrr a2, vlenb
-; CHECK-NEXT: slli a2, a2, 4
-; CHECK-NEXT: add a2, sp, a2
-; CHECK-NEXT: addi a2, a2, 16
+; 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: addi a2, sp, 16
; CHECK-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
+; CHECK-NEXT: vmv1r.v v7, v8
; CHECK-NEXT: li a2, 128
; CHECK-NEXT: vsetvli zero, a2, e8, m8, ta, ma
-; CHECK-NEXT: vle8.v v16, (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 v16, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: addi a0, a1, 128
; CHECK-NEXT: vle8.v v24, (a0)
+; CHECK-NEXT: addi a0, a1, 128
+; CHECK-NEXT: vle8.v v8, (a0)
; CHECK-NEXT: vle8.v v16, (a1)
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: vmv1r.v v9, v0
-; CHECK-NEXT: vmv1r.v 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 v16, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vmv1r.v v6, v0
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetivli zero, 1, e8, m8, ta, ma
-; CHECK-NEXT: vmerge.vvm v24, v24, v16, v0
-; CHECK-NEXT: vmv1r.v v0, v9
-; 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: vmerge.vvm v24, v8, v24, v0
+; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, a2, e8, m8, ta, ma
; CHECK-NEXT: vmerge.vvm v8, v16, v8, v0
; CHECK-NEXT: vmv8r.v v16, v24
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a1, 24
-; CHECK-NEXT: mul a0, a0, a1
+; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add sp, sp, a0
; CHECK-NEXT: .cfi_def_cfa sp, 16
; CHECK-NEXT: addi sp, sp, 16
@@ -455,48 +434,27 @@ define <32 x i64> @select_evl_v32i64(<32 x i1> %a, <32 x i64> %b, <32 x i64> %c)
; CHECK-NEXT: addi sp, sp, -16
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: li a2, 24
-; CHECK-NEXT: mul a1, a1, a2
+; CHECK-NEXT: slli a1, a1, 3
; 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: csrr a1, vlenb
-; CHECK-NEXT: slli a1, a1, 4
-; CHECK-NEXT: add a1, sp, a1
-; CHECK-NEXT: addi a1, a1, 16
+; 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: addi a1, sp, 16
; CHECK-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
-; 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: vmv8r.v v16, v8
; CHECK-NEXT: addi a1, a0, 128
; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; CHECK-NEXT: vle64.v v8, (a0)
-; CHECK-NEXT: vle64.v v16, (a1)
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vle64.v v24, (a1)
; CHECK-NEXT: vsetivli zero, 2, e8, mf4, ta, ma
-; CHECK-NEXT: vslidedown.vi v24, v0, 2
-; 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: vslidedown.vi v7, v0, 2
; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; CHECK-NEXT: vmerge.vvm v8, v8, v16, v0
-; CHECK-NEXT: vmv1r.v v0, v24
-; 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 v24, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vsetivli zero, 1, e64, m8, ta, ma
-; CHECK-NEXT: vmerge.vvm v16, v16, v24, v0
+; CHECK-NEXT: vmerge.vvm v16, v24, v16, v0
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a1, 24
-; CHECK-NEXT: mul a0, a0, a1
+; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add sp, sp, a0
; CHECK-NEXT: .cfi_def_cfa sp, 16
; CHECK-NEXT: addi sp, sp, 16
diff --git a/llvm/test/CodeGen/RISCV/rvv/floor-vp.ll b/llvm/test/CodeGen/RISCV/rvv/floor-vp.ll
index e6dfe5e78cdb4b..b43a4772df0b12 100644
--- a/llvm/test/CodeGen/RISCV/rvv/floor-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/floor-vp.ll
@@ -273,12 +273,6 @@ declare <vscale x 32 x bfloat> @llvm.vp.floor.nxv32bf16(<vscale x 32 x bfloat>,
define <vscale x 32 x bfloat> @vp_floor_nxv32bf16(<vscale x 32 x bfloat> %va, <vscale x 32 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_floor_nxv32bf16:
; 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 v7, v0
; CHECK-NEXT: csrr a2, vlenb
; CHECK-NEXT: vsetvli a1, zero, e16, m4, ta, ma
@@ -302,11 +296,7 @@ define <vscale x 32 x bfloat> @vp_floor_nxv32bf16(<vscale x 32 x bfloat> %va, <v
; CHECK-NEXT: vmv1r.v v0, v12
; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v16, v24, v0.t
-; CHECK-NEXT: addi a3, sp, 16
-; CHECK-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
; CHECK-NEXT: fsrm a2
-; CHECK-NEXT: addi a2, sp, 16
-; CHECK-NEXT: vl8r.v v16, (a2) # Unknown-size Folded Reload
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, mu
; CHECK-NEXT: vfsgnj.vv v24, v16, v24, v0.t
@@ -332,12 +322,6 @@ define <vscale x 32 x bfloat> @vp_floor_nxv32bf16(<vscale x 32 x bfloat> %va, <v
; CHECK-NEXT: vfsgnj.vv v24, v16, v24, v0.t
; CHECK-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; CHECK-NEXT: vfncvtbf16.f.f.w v8, v24
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add sp, sp, a0
-; CHECK-NEXT: .cfi_def_cfa sp, 16
-; CHECK-NEXT: addi sp, sp, 16
-; CHECK-NEXT: .cfi_def_cfa_offset 0
; CHECK-NEXT: ret
%v = call <vscale x 32 x bfloat> @llvm.vp.floor.nxv32bf16(<vscale x 32 x bfloat> %va, <vscale x 32 x i1> %m, i32 %evl)
ret <vscale x 32 x bfloat> %v
@@ -840,12 +824,6 @@ define <vscale x 32 x half> @vp_floor_nxv32f16(<vscale x 32 x half> %va, <vscale
;
; ZVFHMIN-LABEL: vp_floor_nxv32f16:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: addi sp, sp, -16
-; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
-; ZVFHMIN-NEXT: csrr a1, vlenb
-; 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 v7, v0
; ZVFHMIN-NEXT: csrr a2, vlenb
; ZVFHMIN-NEXT: vsetvli a1, zero, e16, m4, ta, ma
@@ -869,11 +847,7 @@ define <vscale x 32 x half> @vp_floor_nxv32f16(<vscale x 32 x half> %va, <vscale
; ZVFHMIN-NEXT: vmv1r.v v0, v12
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfcvt.x.f.v v16, v24, v0.t
-; ZVFHMIN-NEXT: addi a3, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: fsrm a2
-; ZVFHMIN-NEXT: addi a2, sp, 16
-; ZVFHMIN-NEXT: vl8r.v v16, (a2) # 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, v16, v24, v0.t
@@ -899,12 +873,6 @@ define <vscale x 32 x half> @vp_floor_nxv32f16(<vscale x 32 x half> %va, <vscale
; ZVFHMIN-NEXT: vfsgnj.vv v24, v16, v24, v0.t
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; 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
-; ZVFHMIN-NEXT: .cfi_def_cfa sp, 16
-; ZVFHMIN-NEXT: addi sp, sp, 16
-; ZVFHMIN-NEXT: .cfi_def_cfa_offset 0
; ZVFHMIN-NEXT: ret
%v = call <vscale x 32 x half> @llvm.vp.floor.nxv32f16(<vscale x 32 x half> %va, <vscale x 32 x i1> %m, i32 %evl)
ret <vscale x 32 x half> %v
@@ -1419,12 +1387,6 @@ declare <vscale x 16 x double> @llvm.vp.floor.nxv16f64(<vscale x 16 x double>, <
define <vscale x 16 x double> @vp_floor_nxv16f64(<vscale x 16 x double> %va, <vscale x 16 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_floor_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, 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: csrr a1, vlenb
; CHECK-NEXT: lui a2, %hi(.LCPI44_0)
@@ -1445,40 +1407,27 @@ define <vscale x 16 x double> @vp_floor_nxv16f64(<vscale x 16 x double> %va, <vs
; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; 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: 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, v24, v16, v0.t
-; CHECK-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
; CHECK-NEXT: bltu a0, a1, .LBB44_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB44_2:
; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; 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 v7, 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: 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: addi a0, sp, 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 sp, sp, a0
-; CHECK-NEXT: .cfi_def_cfa sp, 16
-; CHECK-NEXT: addi sp, sp, 16
-; CHECK-NEXT: .cfi_def_cfa_offset 0
+; CHECK-NEXT: vfsgnj.vv v8, v24, v8, v0.t
; CHECK-NEXT: ret
%v = call <vscale x 16 x double> @llvm.vp.floor.nxv16f64(<vscale x 16 x double> %va, <vscale x 16 x i1> %m, i32 %evl)
ret <vscale x 16 x double> %v
diff --git a/llvm/test/CodeGen/RISCV/rvv/fmaximum-sdnode.ll b/llvm/test/CodeGen/RISCV/rvv/fmaximum-sdnode.ll
index a1cdbd4be25794..4dc1b840c2728b 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fmaximum-sdnode.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fmaximum-sdnode.ll
@@ -113,12 +113,6 @@ declare <vscale x 16 x bfloat> @llvm.maximum.nxv16bf16(<vscale x 16 x bfloat>, <
define <vscale x 16 x bfloat> @vfmax_nxv16bf16_vv(<vscale x 16 x bfloat> %a, <vscale x 16 x bfloat> %b) {
; CHECK-LABEL: vfmax_nxv16bf16_vv:
; CHECK: # %bb.0:
-; CHECK-NEXT: addi sp, sp, -16
-; CHECK-NEXT: .cfi_def_cfa_offset 16
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; 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: vsetvli a0, zero, e16, m4, ta, ma
; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v12
; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v8
@@ -126,20 +120,11 @@ define <vscale x 16 x bfloat> @vfmax_nxv16bf16_vv(<vscale x 16 x bfloat> %a, <vs
; CHECK-NEXT: vmfeq.vv v0, v24, v24
; CHECK-NEXT: vmfeq.vv v7, v16, v16
; CHECK-NEXT: vmerge.vvm v8, v24, v16, v0
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: vmv1r.v v0, v7
-; CHECK-NEXT: vmerge.vvm v8, v16, v24, v0
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vfmax.vv v16, v8, v16
+; CHECK-NEXT: vmerge.vvm v16, v16, v24, v0
+; CHECK-NEXT: vfmax.vv v16, v16, v8
; CHECK-NEXT: vsetvli zero, zero, e16, m4, ta, ma
; CHECK-NEXT: vfncvtbf16.f.f.w v8, v16
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add sp, sp, a0
-; CHECK-NEXT: .cfi_def_cfa sp, 16
-; CHECK-NEXT: addi sp, sp, 16
-; CHECK-NEXT: .cfi_def_cfa_offset 0
; CHECK-NEXT: ret
%v = call <vscale x 16 x bfloat> @llvm.maximum.nxv16bf16(<vscale x 16 x bfloat> %a, <vscale x 16 x bfloat> %b)
ret <vscale x 16 x bfloat> %v
@@ -448,12 +433,6 @@ define <vscale x 16 x half> @vfmax_nxv16f16_vv(<vscale x 16 x half> %a, <vscale
;
; ZVFHMIN-LABEL: vfmax_nxv16f16_vv:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: addi sp, sp, -16
-; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
-; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 3
-; ZVFHMIN-NEXT: sub sp, sp, a0
-; 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: vsetvli a0, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v8
@@ -461,20 +440,11 @@ define <vscale x 16 x half> @vfmax_nxv16f16_vv(<vscale x 16 x half> %a, <vscale
; ZVFHMIN-NEXT: vmfeq.vv v0, v24, v24
; ZVFHMIN-NEXT: vmfeq.vv v7, v16, v16
; ZVFHMIN-NEXT: vmerge.vvm v8, v24, v16, v0
-; ZVFHMIN-NEXT: addi a0, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vmv1r.v v0, v7
-; ZVFHMIN-NEXT: vmerge.vvm v8, v16, v24, v0
-; ZVFHMIN-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfmax.vv v16, v8, v16
+; ZVFHMIN-NEXT: vmerge.vvm v16, v16, v24, v0
+; ZVFHMIN-NEXT: vfmax.vv v16, v16, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v16
-; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 3
-; ZVFHMIN-NEXT: add sp, sp, a0
-; ZVFHMIN-NEXT: .cfi_def_cfa sp, 16
-; ZVFHMIN-NEXT: addi sp, sp, 16
-; ZVFHMIN-NEXT: .cfi_def_cfa_offset 0
; ZVFHMIN-NEXT: ret
%v = call <vscale x 16 x half> @llvm.maximum.nxv16f16(<vscale x 16 x half> %a, <vscale x 16 x half> %b)
ret <vscale x 16 x half> %v
diff --git a/llvm/test/CodeGen/RISCV/rvv/fmaximum-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fmaximum-vp.ll
index 33e793a691b81d..4990f31963fd21 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fmaximum-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fmaximum-vp.ll
@@ -252,12 +252,6 @@ define <vscale x 16 x bfloat> @vfmax_vv_nxv16bf16(<vscale x 16 x bfloat> %va, <v
define <vscale x 16 x bfloat> @vfmax_vv_nxv16bf16_unmasked(<vscale x 16 x bfloat> %va, <vscale x 16 x bfloat> %vb, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_nxv16bf16_unmasked:
; 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: vsetvli a1, zero, e16, m4, ta, ma
; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v8
; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v12
@@ -265,20 +259,11 @@ define <vscale x 16 x bfloat> @vfmax_vv_nxv16bf16_unmasked(<vscale x 16 x bfloat
; CHECK-NEXT: vmfeq.vv v0, v16, v16
; CHECK-NEXT: vmfeq.vv v7, v24, v24
; CHECK-NEXT: vmerge.vvm v8, v16, v24, v0
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: vmv1r.v v0, v7
-; CHECK-NEXT: vmerge.vvm v8, v24, v16, v0
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vfmax.vv v16, v8, v16
+; CHECK-NEXT: vmerge.vvm v16, v24, v16, v0
+; CHECK-NEXT: vfmax.vv v16, v16, v8
; CHECK-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; CHECK-NEXT: vfncvtbf16.f.f.w v8, v16
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add sp, sp, a0
-; CHECK-NEXT: .cfi_def_cfa sp, 16
-; CHECK-NEXT: addi sp, sp, 16
-; CHECK-NEXT: .cfi_def_cfa_offset 0
; CHECK-NEXT: ret
%v = call <vscale x 16 x bfloat> @llvm.vp.maximum.nxv16bf16(<vscale x 16 x bfloat> %va, <vscale x 16 x bfloat> %vb, <vscale x 16 x i1> splat (i1 true), i32 %evl)
ret <vscale x 16 x bfloat> %v
@@ -292,16 +277,16 @@ define <vscale x 32 x bfloat> @vfmax_vv_nxv32bf16(<vscale x 32 x bfloat> %va, <v
; CHECK-NEXT: addi sp, sp, -16
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a2, a1, 5
-; CHECK-NEXT: add a1, a2, a1
+; CHECK-NEXT: li a2, 25
+; CHECK-NEXT: mul a1, a1, a2
; CHECK-NEXT: sub sp, sp, a1
-; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x21, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 33 * vlenb
+; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x19, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 25 * vlenb
; CHECK-NEXT: csrr a2, vlenb
; CHECK-NEXT: vsetvli a1, zero, e16, m4, ta, ma
; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v12
; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: li a3, 24
-; CHECK-NEXT: mul a1, a1, a3
+; CHECK-NEXT: slli a3, a1, 4
+; CHECK-NEXT: add a1, a3, a1
; CHECK-NEXT: add a1, sp, a1
; CHECK-NEXT: addi a1, a1, 16
; CHECK-NEXT: vs8r.v v24, (a1) # Unknown-size Folded Spill
@@ -309,7 +294,7 @@ define <vscale x 32 x bfloat> @vfmax_vv_nxv32bf16(<vscale x 32 x bfloat> %va, <v
; CHECK-NEXT: srli a2, a2, 2
; CHECK-NEXT: sub a3, a0, a1
; CHECK-NEXT: csrr a4, vlenb
-; CHECK-NEXT: slli a4, a4, 5
+; CHECK-NEXT: slli a4, a4, 3
; CHECK-NEXT: add a4, sp, a4
; CHECK-NEXT: addi a4, a4, 16
; CHECK-NEXT: vs1r.v v0, (a4) # Unknown-size Folded Spill
@@ -323,7 +308,8 @@ define <vscale x 32 x bfloat> @vfmax_vv_nxv32bf16(<vscale x 32 x bfloat> %va, <v
; CHECK-NEXT: vmfeq.vv v13, v24, v24, v0.t
; CHECK-NEXT: vmv8r.v v0, v16
; CHECK-NEXT: csrr a3, vlenb
-; CHECK-NEXT: slli a3, a3, 4
+; CHECK-NEXT: slli a4, a3, 3
+; CHECK-NEXT: add a3, a4, a3
; CHECK-NEXT: add a3, sp, a3
; CHECK-NEXT: addi a3, a3, 16
; CHECK-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
@@ -331,102 +317,96 @@ define <vscale x 32 x bfloat> @vfmax_vv_nxv32bf16(<vscale x 32 x bfloat> %va, <v
; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v4
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: csrr a3, vlenb
-; CHECK-NEXT: li a4, 24
-; CHECK-NEXT: mul a3, a3, a4
+; CHECK-NEXT: slli a4, a3, 4
+; CHECK-NEXT: add a3, a4, a3
; CHECK-NEXT: add a3, sp, a3
; CHECK-NEXT: addi a3, a3, 16
; CHECK-NEXT: vl8r.v v24, (a3) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, a2, e32, m8, ta, ma
; CHECK-NEXT: vmerge.vvm v24, v24, v16, v0
-; 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: addi a2, sp, 16
; CHECK-NEXT: vs8r.v v24, (a2) # Unknown-size Folded Spill
; CHECK-NEXT: vmv1r.v v0, v12
; CHECK-NEXT: vmfeq.vv v13, v16, v16, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: csrr a2, vlenb
-; CHECK-NEXT: li a3, 24
-; CHECK-NEXT: mul a2, a2, a3
+; CHECK-NEXT: slli a3, a2, 4
+; CHECK-NEXT: add a2, a3, a2
; CHECK-NEXT: add a2, sp, a2
; CHECK-NEXT: addi a2, a2, 16
; CHECK-NEXT: vl8r.v v24, (a2) # Unknown-size Folded Reload
; CHECK-NEXT: vmerge.vvm v16, v16, v24, v0
; CHECK-NEXT: vmv1r.v v0, v12
-; 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: addi a2, sp, 16
; CHECK-NEXT: vl8r.v v24, (a2) # Unknown-size Folded Reload
; CHECK-NEXT: vfmax.vv v16, v16, v24, v0.t
; CHECK-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; CHECK-NEXT: vfncvtbf16.f.f.w v12, 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 v8, (a2) # Unknown-size Folded Spill
; CHECK-NEXT: bltu a0, a1, .LBB10_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB10_2:
; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v8
; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a1, a1, 5
+; CHECK-NEXT: slli a2, a1, 4
+; CHECK-NEXT: add a1, a2, a1
; CHECK-NEXT: add a1, sp, a1
; CHECK-NEXT: addi a1, a1, 16
-; CHECK-NEXT: vl1r.v v0, (a1) # Unknown-size Folded Reload
-; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
-; CHECK-NEXT: vmfeq.vv v24, v16, v16, v0.t
-; CHECK-NEXT: vmv8r.v v8, v16
-; CHECK-NEXT: addi a1, sp, 16
; CHECK-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a1, a1, 4
+; CHECK-NEXT: slli a1, a1, 3
+; CHECK-NEXT: add a1, sp, a1
+; CHECK-NEXT: addi a1, a1, 16
+; CHECK-NEXT: vl1r.v v9, (a1) # Unknown-size Folded Reload
+; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
+; CHECK-NEXT: vmfeq.vv v8, v16, v16, v0.t
+; CHECK-NEXT: csrr a1, vlenb
+; CHECK-NEXT: slli a2, a1, 3
+; CHECK-NEXT: add a1, a2, a1
; CHECK-NEXT: add a1, sp, a1
; CHECK-NEXT: addi a1, a1, 16
; CHECK-NEXT: vl8r.v v0, (a1) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli a1, zero, e16, m4, ta, ma
; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v0
-; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: csrr a1, vlenb
+; CHECK-NEXT: slli a2, a1, 4
+; CHECK-NEXT: add a1, a2, a1
+; 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: vmerge.vvm v24, v8, v16, v0
+; CHECK-NEXT: vmerge.vvm v24, v24, v16, v0
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a1, 24
-; CHECK-NEXT: mul a0, a0, a1
+; CHECK-NEXT: slli a1, a0, 3
+; CHECK-NEXT: add a0, a1, a0
; 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, 5
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl1r.v v9, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vmv1r.v v0, v9
; CHECK-NEXT: vmfeq.vv v8, v16, v16, v0.t
; CHECK-NEXT: vmv1r.v v0, v8
-; CHECK-NEXT: addi a0, sp, 16
+; CHECK-NEXT: csrr a0, vlenb
+; CHECK-NEXT: slli a1, a0, 4
+; CHECK-NEXT: add a0, a1, a0
+; CHECK-NEXT: add a0, sp, a0
+; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vmerge.vvm v16, v16, v24, v0
; CHECK-NEXT: vmv1r.v v0, v9
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a1, 24
-; CHECK-NEXT: mul a0, a0, a1
+; CHECK-NEXT: slli a1, a0, 3
+; CHECK-NEXT: add a0, a1, a0
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vfmax.vv v16, v16, v24, 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 v8, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; CHECK-NEXT: vfncvtbf16.f.f.w v8, v16
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a1, a0, 5
-; CHECK-NEXT: add a0, a1, a0
+; CHECK-NEXT: li a1, 25
+; CHECK-NEXT: mul a0, a0, a1
; CHECK-NEXT: add sp, sp, a0
; CHECK-NEXT: .cfi_def_cfa sp, 16
; CHECK-NEXT: addi sp, sp, 16
@@ -921,12 +901,6 @@ define <vscale x 16 x half> @vfmax_vv_nxv16f16_unmasked(<vscale x 16 x half> %va
;
; ZVFHMIN-LABEL: vfmax_vv_nxv16f16_unmasked:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: addi sp, sp, -16
-; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
-; ZVFHMIN-NEXT: csrr a1, vlenb
-; 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: vsetvli a1, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
@@ -934,20 +908,11 @@ define <vscale x 16 x half> @vfmax_vv_nxv16f16_unmasked(<vscale x 16 x half> %va
; ZVFHMIN-NEXT: vmfeq.vv v0, v16, v16
; ZVFHMIN-NEXT: vmfeq.vv v7, v24, v24
; ZVFHMIN-NEXT: vmerge.vvm v8, v16, v24, v0
-; ZVFHMIN-NEXT: addi a0, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vmv1r.v v0, v7
-; ZVFHMIN-NEXT: vmerge.vvm v8, v24, v16, v0
-; ZVFHMIN-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfmax.vv v16, v8, v16
+; ZVFHMIN-NEXT: vmerge.vvm v16, v24, v16, v0
+; ZVFHMIN-NEXT: vfmax.vv v16, v16, v8
; 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, 3
-; ZVFHMIN-NEXT: add sp, sp, a0
-; ZVFHMIN-NEXT: .cfi_def_cfa sp, 16
-; ZVFHMIN-NEXT: addi sp, sp, 16
-; ZVFHMIN-NEXT: .cfi_def_cfa_offset 0
; ZVFHMIN-NEXT: ret
%v = call <vscale x 16 x half> @llvm.vp.maximum.nxv16f16(<vscale x 16 x half> %va, <vscale x 16 x half> %vb, <vscale x 16 x i1> splat (i1 true), i32 %evl)
ret <vscale x 16 x half> %v
@@ -991,16 +956,16 @@ define <vscale x 32 x half> @vfmax_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
; ZVFHMIN-NEXT: addi sp, sp, -16
; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a2, a1, 5
-; ZVFHMIN-NEXT: add a1, a2, a1
+; ZVFHMIN-NEXT: li a2, 25
+; ZVFHMIN-NEXT: mul a1, a1, a2
; ZVFHMIN-NEXT: sub sp, sp, a1
-; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x21, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 33 * vlenb
+; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x19, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 25 * vlenb
; ZVFHMIN-NEXT: csrr a2, vlenb
; ZVFHMIN-NEXT: vsetvli a1, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: li a3, 24
-; ZVFHMIN-NEXT: mul a1, a1, a3
+; ZVFHMIN-NEXT: slli a3, a1, 4
+; ZVFHMIN-NEXT: add a1, a3, a1
; ZVFHMIN-NEXT: add a1, sp, a1
; ZVFHMIN-NEXT: addi a1, a1, 16
; ZVFHMIN-NEXT: vs8r.v v24, (a1) # Unknown-size Folded Spill
@@ -1008,7 +973,7 @@ define <vscale x 32 x half> @vfmax_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
; ZVFHMIN-NEXT: srli a2, a2, 2
; ZVFHMIN-NEXT: sub a3, a0, a1
; ZVFHMIN-NEXT: csrr a4, vlenb
-; ZVFHMIN-NEXT: slli a4, a4, 5
+; ZVFHMIN-NEXT: slli a4, a4, 3
; ZVFHMIN-NEXT: add a4, sp, a4
; ZVFHMIN-NEXT: addi a4, a4, 16
; ZVFHMIN-NEXT: vs1r.v v0, (a4) # Unknown-size Folded Spill
@@ -1022,7 +987,8 @@ define <vscale x 32 x half> @vfmax_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
; ZVFHMIN-NEXT: vmfeq.vv v13, v24, v24, v0.t
; ZVFHMIN-NEXT: vmv8r.v v0, v16
; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: slli a3, a3, 4
+; ZVFHMIN-NEXT: slli a4, a3, 3
+; ZVFHMIN-NEXT: add a3, a4, a3
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
; ZVFHMIN-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
@@ -1030,102 +996,96 @@ define <vscale x 32 x half> @vfmax_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v4
; ZVFHMIN-NEXT: vmv1r.v v0, v13
; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: li a4, 24
-; ZVFHMIN-NEXT: mul a3, a3, a4
+; ZVFHMIN-NEXT: slli a4, a3, 4
+; ZVFHMIN-NEXT: add a3, a4, a3
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
; ZVFHMIN-NEXT: vl8r.v v24, (a3) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, a2, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmerge.vvm v24, v24, v16, v0
-; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a2, a2, 3
-; ZVFHMIN-NEXT: add a2, sp, a2
-; ZVFHMIN-NEXT: addi a2, a2, 16
+; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vs8r.v v24, (a2) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vmv1r.v v0, v12
; ZVFHMIN-NEXT: vmfeq.vv v13, v16, v16, v0.t
; ZVFHMIN-NEXT: vmv1r.v v0, v13
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: li a3, 24
-; ZVFHMIN-NEXT: mul a2, a2, a3
+; ZVFHMIN-NEXT: slli a3, a2, 4
+; ZVFHMIN-NEXT: add a2, a3, a2
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
; ZVFHMIN-NEXT: vl8r.v v24, (a2) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vmerge.vvm v16, v16, v24, v0
; ZVFHMIN-NEXT: vmv1r.v v0, v12
-; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a2, a2, 3
-; ZVFHMIN-NEXT: add a2, sp, a2
-; ZVFHMIN-NEXT: addi a2, a2, 16
+; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vl8r.v v24, (a2) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfmax.vv v16, v16, v24, v0.t
; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v12, v16
-; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a2, a2, 3
-; ZVFHMIN-NEXT: add a2, sp, a2
-; ZVFHMIN-NEXT: addi a2, a2, 16
-; ZVFHMIN-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: bltu a0, a1, .LBB22_2
; ZVFHMIN-NEXT: # %bb.1:
; ZVFHMIN-NEXT: mv a0, a1
; ZVFHMIN-NEXT: .LBB22_2:
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a1, a1, 5
+; ZVFHMIN-NEXT: slli a2, a1, 4
+; ZVFHMIN-NEXT: add a1, a2, a1
; 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 v24, v16, v16, v0.t
-; ZVFHMIN-NEXT: vmv8r.v v8, v16
-; ZVFHMIN-NEXT: addi a1, sp, 16
; ZVFHMIN-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a1, a1, 4
+; ZVFHMIN-NEXT: slli a1, a1, 3
+; ZVFHMIN-NEXT: add a1, sp, a1
+; ZVFHMIN-NEXT: addi a1, a1, 16
+; ZVFHMIN-NEXT: vl1r.v v9, (a1) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vmv1r.v v0, v9
+; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
+; ZVFHMIN-NEXT: vmfeq.vv v8, v16, v16, v0.t
+; ZVFHMIN-NEXT: csrr a1, vlenb
+; ZVFHMIN-NEXT: slli a2, a1, 3
+; ZVFHMIN-NEXT: add a1, a2, a1
; 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, v24
+; ZVFHMIN-NEXT: vmv1r.v v0, v8
+; ZVFHMIN-NEXT: csrr a1, vlenb
+; ZVFHMIN-NEXT: slli a2, a1, 4
+; ZVFHMIN-NEXT: add a1, a2, a1
+; 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 zero, a0, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vmerge.vvm v24, v8, v16, v0
+; ZVFHMIN-NEXT: vmerge.vvm v24, v24, v16, v0
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: li a1, 24
-; ZVFHMIN-NEXT: mul a0, a0, a1
+; ZVFHMIN-NEXT: slli a1, a0, 3
+; ZVFHMIN-NEXT: add a0, a1, a0
; 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, 5
-; ZVFHMIN-NEXT: add a0, sp, a0
-; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vl1r.v v9, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vmv1r.v v0, v9
; ZVFHMIN-NEXT: vmfeq.vv v8, v16, v16, v0.t
; ZVFHMIN-NEXT: vmv1r.v v0, v8
-; ZVFHMIN-NEXT: addi a0, sp, 16
+; ZVFHMIN-NEXT: csrr a0, vlenb
+; ZVFHMIN-NEXT: slli a1, a0, 4
+; ZVFHMIN-NEXT: add a0, a1, a0
+; ZVFHMIN-NEXT: add a0, sp, a0
+; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vmerge.vvm v16, v16, v24, v0
; ZVFHMIN-NEXT: vmv1r.v v0, v9
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: li a1, 24
-; ZVFHMIN-NEXT: mul a0, a0, a1
+; ZVFHMIN-NEXT: slli a1, a0, 3
+; ZVFHMIN-NEXT: add a0, a1, a0
; 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: 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: slli a1, a0, 5
-; ZVFHMIN-NEXT: add a0, a1, a0
+; ZVFHMIN-NEXT: li a1, 25
+; ZVFHMIN-NEXT: mul a0, a0, a1
; ZVFHMIN-NEXT: add sp, sp, a0
; ZVFHMIN-NEXT: .cfi_def_cfa sp, 16
; ZVFHMIN-NEXT: addi sp, sp, 16
@@ -1578,8 +1538,6 @@ define <vscale x 16 x double> @vfmax_vv_nxv16f64(<vscale x 16 x double> %va, <vs
; 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: 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: slli a1, a1, 4
; CHECK-NEXT: add a1, sp, a1
@@ -1606,20 +1564,19 @@ define <vscale x 16 x double> @vfmax_vv_nxv16f64(<vscale x 16 x double> %va, <vs
; CHECK-NEXT: add a3, sp, a3
; CHECK-NEXT: addi a3, a3, 16
; CHECK-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
+; CHECK-NEXT: vmv8r.v v24, v16
; CHECK-NEXT: vmv1r.v v0, v6
-; CHECK-NEXT: vmfeq.vv v26, v8, v8, v0.t
+; CHECK-NEXT: vmfeq.vv v5, v8, v8, v0.t
; CHECK-NEXT: vl8re64.v v16, (a0)
-; CHECK-NEXT: vmv1r.v v0, v26
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vmerge.vvm v24, v8, v24, v0
+; CHECK-NEXT: vmv1r.v v0, v5
+; CHECK-NEXT: vmerge.vvm v8, v8, v24, v0
; CHECK-NEXT: vmv1r.v v0, v6
; 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: vfmax.vv v8, v24, v8, 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, 3
; CHECK-NEXT: add a0, sp, a0
diff --git a/llvm/test/CodeGen/RISCV/rvv/fminimum-sdnode.ll b/llvm/test/CodeGen/RISCV/rvv/fminimum-sdnode.ll
index d41da7b6a2af96..aa5ebd008e1d99 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fminimum-sdnode.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fminimum-sdnode.ll
@@ -113,12 +113,6 @@ declare <vscale x 16 x bfloat> @llvm.minimum.nxv16bf16(<vscale x 16 x bfloat>, <
define <vscale x 16 x bfloat> @vfmin_nxv16bf16_vv(<vscale x 16 x bfloat> %a, <vscale x 16 x bfloat> %b) {
; CHECK-LABEL: vfmin_nxv16bf16_vv:
; CHECK: # %bb.0:
-; CHECK-NEXT: addi sp, sp, -16
-; CHECK-NEXT: .cfi_def_cfa_offset 16
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; 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: vsetvli a0, zero, e16, m4, ta, ma
; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v12
; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v8
@@ -126,20 +120,11 @@ define <vscale x 16 x bfloat> @vfmin_nxv16bf16_vv(<vscale x 16 x bfloat> %a, <vs
; CHECK-NEXT: vmfeq.vv v0, v24, v24
; CHECK-NEXT: vmfeq.vv v7, v16, v16
; CHECK-NEXT: vmerge.vvm v8, v24, v16, v0
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: vmv1r.v v0, v7
-; CHECK-NEXT: vmerge.vvm v8, v16, v24, v0
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vfmin.vv v16, v8, v16
+; CHECK-NEXT: vmerge.vvm v16, v16, v24, v0
+; CHECK-NEXT: vfmin.vv v16, v16, v8
; CHECK-NEXT: vsetvli zero, zero, e16, m4, ta, ma
; CHECK-NEXT: vfncvtbf16.f.f.w v8, v16
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add sp, sp, a0
-; CHECK-NEXT: .cfi_def_cfa sp, 16
-; CHECK-NEXT: addi sp, sp, 16
-; CHECK-NEXT: .cfi_def_cfa_offset 0
; CHECK-NEXT: ret
%v = call <vscale x 16 x bfloat> @llvm.minimum.nxv16bf16(<vscale x 16 x bfloat> %a, <vscale x 16 x bfloat> %b)
ret <vscale x 16 x bfloat> %v
@@ -448,12 +433,6 @@ define <vscale x 16 x half> @vfmin_nxv16f16_vv(<vscale x 16 x half> %a, <vscale
;
; ZVFHMIN-LABEL: vfmin_nxv16f16_vv:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: addi sp, sp, -16
-; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
-; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 3
-; ZVFHMIN-NEXT: sub sp, sp, a0
-; 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: vsetvli a0, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v8
@@ -461,20 +440,11 @@ define <vscale x 16 x half> @vfmin_nxv16f16_vv(<vscale x 16 x half> %a, <vscale
; ZVFHMIN-NEXT: vmfeq.vv v0, v24, v24
; ZVFHMIN-NEXT: vmfeq.vv v7, v16, v16
; ZVFHMIN-NEXT: vmerge.vvm v8, v24, v16, v0
-; ZVFHMIN-NEXT: addi a0, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vmv1r.v v0, v7
-; ZVFHMIN-NEXT: vmerge.vvm v8, v16, v24, v0
-; ZVFHMIN-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfmin.vv v16, v8, v16
+; ZVFHMIN-NEXT: vmerge.vvm v16, v16, v24, v0
+; ZVFHMIN-NEXT: vfmin.vv v16, v16, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v16
-; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 3
-; ZVFHMIN-NEXT: add sp, sp, a0
-; ZVFHMIN-NEXT: .cfi_def_cfa sp, 16
-; ZVFHMIN-NEXT: addi sp, sp, 16
-; ZVFHMIN-NEXT: .cfi_def_cfa_offset 0
; ZVFHMIN-NEXT: ret
%v = call <vscale x 16 x half> @llvm.minimum.nxv16f16(<vscale x 16 x half> %a, <vscale x 16 x half> %b)
ret <vscale x 16 x half> %v
diff --git a/llvm/test/CodeGen/RISCV/rvv/fminimum-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fminimum-vp.ll
index ef6f33de4ce636..898d248181e200 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fminimum-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fminimum-vp.ll
@@ -252,12 +252,6 @@ define <vscale x 16 x bfloat> @vfmin_vv_nxv16bf16(<vscale x 16 x bfloat> %va, <v
define <vscale x 16 x bfloat> @vfmin_vv_nxv16bf16_unmasked(<vscale x 16 x bfloat> %va, <vscale x 16 x bfloat> %vb, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_nxv16bf16_unmasked:
; 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: vsetvli a1, zero, e16, m4, ta, ma
; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v8
; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v12
@@ -265,20 +259,11 @@ define <vscale x 16 x bfloat> @vfmin_vv_nxv16bf16_unmasked(<vscale x 16 x bfloat
; CHECK-NEXT: vmfeq.vv v0, v16, v16
; CHECK-NEXT: vmfeq.vv v7, v24, v24
; CHECK-NEXT: vmerge.vvm v8, v16, v24, v0
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: vmv1r.v v0, v7
-; CHECK-NEXT: vmerge.vvm v8, v24, v16, v0
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vfmin.vv v16, v8, v16
+; CHECK-NEXT: vmerge.vvm v16, v24, v16, v0
+; CHECK-NEXT: vfmin.vv v16, v16, v8
; CHECK-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; CHECK-NEXT: vfncvtbf16.f.f.w v8, v16
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add sp, sp, a0
-; CHECK-NEXT: .cfi_def_cfa sp, 16
-; CHECK-NEXT: addi sp, sp, 16
-; CHECK-NEXT: .cfi_def_cfa_offset 0
; CHECK-NEXT: ret
%v = call <vscale x 16 x bfloat> @llvm.vp.minimum.nxv16bf16(<vscale x 16 x bfloat> %va, <vscale x 16 x bfloat> %vb, <vscale x 16 x i1> splat (i1 true), i32 %evl)
ret <vscale x 16 x bfloat> %v
@@ -292,16 +277,16 @@ define <vscale x 32 x bfloat> @vfmin_vv_nxv32bf16(<vscale x 32 x bfloat> %va, <v
; CHECK-NEXT: addi sp, sp, -16
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a2, a1, 5
-; CHECK-NEXT: add a1, a2, a1
+; CHECK-NEXT: li a2, 25
+; CHECK-NEXT: mul a1, a1, a2
; CHECK-NEXT: sub sp, sp, a1
-; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x21, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 33 * vlenb
+; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x19, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 25 * vlenb
; CHECK-NEXT: csrr a2, vlenb
; CHECK-NEXT: vsetvli a1, zero, e16, m4, ta, ma
; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v12
; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: li a3, 24
-; CHECK-NEXT: mul a1, a1, a3
+; CHECK-NEXT: slli a3, a1, 4
+; CHECK-NEXT: add a1, a3, a1
; CHECK-NEXT: add a1, sp, a1
; CHECK-NEXT: addi a1, a1, 16
; CHECK-NEXT: vs8r.v v24, (a1) # Unknown-size Folded Spill
@@ -309,7 +294,7 @@ define <vscale x 32 x bfloat> @vfmin_vv_nxv32bf16(<vscale x 32 x bfloat> %va, <v
; CHECK-NEXT: srli a2, a2, 2
; CHECK-NEXT: sub a3, a0, a1
; CHECK-NEXT: csrr a4, vlenb
-; CHECK-NEXT: slli a4, a4, 5
+; CHECK-NEXT: slli a4, a4, 3
; CHECK-NEXT: add a4, sp, a4
; CHECK-NEXT: addi a4, a4, 16
; CHECK-NEXT: vs1r.v v0, (a4) # Unknown-size Folded Spill
@@ -323,7 +308,8 @@ define <vscale x 32 x bfloat> @vfmin_vv_nxv32bf16(<vscale x 32 x bfloat> %va, <v
; CHECK-NEXT: vmfeq.vv v13, v24, v24, v0.t
; CHECK-NEXT: vmv8r.v v0, v16
; CHECK-NEXT: csrr a3, vlenb
-; CHECK-NEXT: slli a3, a3, 4
+; CHECK-NEXT: slli a4, a3, 3
+; CHECK-NEXT: add a3, a4, a3
; CHECK-NEXT: add a3, sp, a3
; CHECK-NEXT: addi a3, a3, 16
; CHECK-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
@@ -331,102 +317,96 @@ define <vscale x 32 x bfloat> @vfmin_vv_nxv32bf16(<vscale x 32 x bfloat> %va, <v
; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v4
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: csrr a3, vlenb
-; CHECK-NEXT: li a4, 24
-; CHECK-NEXT: mul a3, a3, a4
+; CHECK-NEXT: slli a4, a3, 4
+; CHECK-NEXT: add a3, a4, a3
; CHECK-NEXT: add a3, sp, a3
; CHECK-NEXT: addi a3, a3, 16
; CHECK-NEXT: vl8r.v v24, (a3) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, a2, e32, m8, ta, ma
; CHECK-NEXT: vmerge.vvm v24, v24, v16, v0
-; 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: addi a2, sp, 16
; CHECK-NEXT: vs8r.v v24, (a2) # Unknown-size Folded Spill
; CHECK-NEXT: vmv1r.v v0, v12
; CHECK-NEXT: vmfeq.vv v13, v16, v16, v0.t
; CHECK-NEXT: vmv1r.v v0, v13
; CHECK-NEXT: csrr a2, vlenb
-; CHECK-NEXT: li a3, 24
-; CHECK-NEXT: mul a2, a2, a3
+; CHECK-NEXT: slli a3, a2, 4
+; CHECK-NEXT: add a2, a3, a2
; CHECK-NEXT: add a2, sp, a2
; CHECK-NEXT: addi a2, a2, 16
; CHECK-NEXT: vl8r.v v24, (a2) # Unknown-size Folded Reload
; CHECK-NEXT: vmerge.vvm v16, v16, v24, v0
; CHECK-NEXT: vmv1r.v v0, v12
-; 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: addi a2, sp, 16
; CHECK-NEXT: vl8r.v v24, (a2) # Unknown-size Folded Reload
; CHECK-NEXT: vfmin.vv v16, v16, v24, v0.t
; CHECK-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; CHECK-NEXT: vfncvtbf16.f.f.w v12, 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 v8, (a2) # Unknown-size Folded Spill
; CHECK-NEXT: bltu a0, a1, .LBB10_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB10_2:
; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v8
; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a1, a1, 5
+; CHECK-NEXT: slli a2, a1, 4
+; CHECK-NEXT: add a1, a2, a1
; CHECK-NEXT: add a1, sp, a1
; CHECK-NEXT: addi a1, a1, 16
-; CHECK-NEXT: vl1r.v v0, (a1) # Unknown-size Folded Reload
-; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
-; CHECK-NEXT: vmfeq.vv v24, v16, v16, v0.t
-; CHECK-NEXT: vmv8r.v v8, v16
-; CHECK-NEXT: addi a1, sp, 16
; CHECK-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a1, a1, 4
+; CHECK-NEXT: slli a1, a1, 3
+; CHECK-NEXT: add a1, sp, a1
+; CHECK-NEXT: addi a1, a1, 16
+; CHECK-NEXT: vl1r.v v9, (a1) # Unknown-size Folded Reload
+; CHECK-NEXT: vmv1r.v v0, v9
+; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
+; CHECK-NEXT: vmfeq.vv v8, v16, v16, v0.t
+; CHECK-NEXT: csrr a1, vlenb
+; CHECK-NEXT: slli a2, a1, 3
+; CHECK-NEXT: add a1, a2, a1
; CHECK-NEXT: add a1, sp, a1
; CHECK-NEXT: addi a1, a1, 16
; CHECK-NEXT: vl8r.v v0, (a1) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli a1, zero, e16, m4, ta, ma
; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v0
-; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vmv1r.v v0, v8
+; CHECK-NEXT: csrr a1, vlenb
+; CHECK-NEXT: slli a2, a1, 4
+; CHECK-NEXT: add a1, a2, a1
+; 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: vmerge.vvm v24, v8, v16, v0
+; CHECK-NEXT: vmerge.vvm v24, v24, v16, v0
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a1, 24
-; CHECK-NEXT: mul a0, a0, a1
+; CHECK-NEXT: slli a1, a0, 3
+; CHECK-NEXT: add a0, a1, a0
; 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, 5
-; CHECK-NEXT: add a0, sp, a0
-; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl1r.v v9, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vmv1r.v v0, v9
; CHECK-NEXT: vmfeq.vv v8, v16, v16, v0.t
; CHECK-NEXT: vmv1r.v v0, v8
-; CHECK-NEXT: addi a0, sp, 16
+; CHECK-NEXT: csrr a0, vlenb
+; CHECK-NEXT: slli a1, a0, 4
+; CHECK-NEXT: add a0, a1, a0
+; CHECK-NEXT: add a0, sp, a0
+; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vmerge.vvm v16, v16, v24, v0
; CHECK-NEXT: vmv1r.v v0, v9
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a1, 24
-; CHECK-NEXT: mul a0, a0, a1
+; CHECK-NEXT: slli a1, a0, 3
+; CHECK-NEXT: add a0, a1, a0
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vfmin.vv v16, v16, v24, 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 v8, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; CHECK-NEXT: vfncvtbf16.f.f.w v8, v16
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a1, a0, 5
-; CHECK-NEXT: add a0, a1, a0
+; CHECK-NEXT: li a1, 25
+; CHECK-NEXT: mul a0, a0, a1
; CHECK-NEXT: add sp, sp, a0
; CHECK-NEXT: .cfi_def_cfa sp, 16
; CHECK-NEXT: addi sp, sp, 16
@@ -921,12 +901,6 @@ define <vscale x 16 x half> @vfmin_vv_nxv16f16_unmasked(<vscale x 16 x half> %va
;
; ZVFHMIN-LABEL: vfmin_vv_nxv16f16_unmasked:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: addi sp, sp, -16
-; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
-; ZVFHMIN-NEXT: csrr a1, vlenb
-; 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: vsetvli a1, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
@@ -934,20 +908,11 @@ define <vscale x 16 x half> @vfmin_vv_nxv16f16_unmasked(<vscale x 16 x half> %va
; ZVFHMIN-NEXT: vmfeq.vv v0, v16, v16
; ZVFHMIN-NEXT: vmfeq.vv v7, v24, v24
; ZVFHMIN-NEXT: vmerge.vvm v8, v16, v24, v0
-; ZVFHMIN-NEXT: addi a0, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vmv1r.v v0, v7
-; ZVFHMIN-NEXT: vmerge.vvm v8, v24, v16, v0
-; ZVFHMIN-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfmin.vv v16, v8, v16
+; ZVFHMIN-NEXT: vmerge.vvm v16, v24, v16, v0
+; ZVFHMIN-NEXT: vfmin.vv v16, v16, v8
; 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, 3
-; ZVFHMIN-NEXT: add sp, sp, a0
-; ZVFHMIN-NEXT: .cfi_def_cfa sp, 16
-; ZVFHMIN-NEXT: addi sp, sp, 16
-; ZVFHMIN-NEXT: .cfi_def_cfa_offset 0
; ZVFHMIN-NEXT: ret
%v = call <vscale x 16 x half> @llvm.vp.minimum.nxv16f16(<vscale x 16 x half> %va, <vscale x 16 x half> %vb, <vscale x 16 x i1> splat (i1 true), i32 %evl)
ret <vscale x 16 x half> %v
@@ -991,16 +956,16 @@ define <vscale x 32 x half> @vfmin_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
; ZVFHMIN-NEXT: addi sp, sp, -16
; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a2, a1, 5
-; ZVFHMIN-NEXT: add a1, a2, a1
+; ZVFHMIN-NEXT: li a2, 25
+; ZVFHMIN-NEXT: mul a1, a1, a2
; ZVFHMIN-NEXT: sub sp, sp, a1
-; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x21, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 33 * vlenb
+; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x19, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 25 * vlenb
; ZVFHMIN-NEXT: csrr a2, vlenb
; ZVFHMIN-NEXT: vsetvli a1, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: li a3, 24
-; ZVFHMIN-NEXT: mul a1, a1, a3
+; ZVFHMIN-NEXT: slli a3, a1, 4
+; ZVFHMIN-NEXT: add a1, a3, a1
; ZVFHMIN-NEXT: add a1, sp, a1
; ZVFHMIN-NEXT: addi a1, a1, 16
; ZVFHMIN-NEXT: vs8r.v v24, (a1) # Unknown-size Folded Spill
@@ -1008,7 +973,7 @@ define <vscale x 32 x half> @vfmin_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
; ZVFHMIN-NEXT: srli a2, a2, 2
; ZVFHMIN-NEXT: sub a3, a0, a1
; ZVFHMIN-NEXT: csrr a4, vlenb
-; ZVFHMIN-NEXT: slli a4, a4, 5
+; ZVFHMIN-NEXT: slli a4, a4, 3
; ZVFHMIN-NEXT: add a4, sp, a4
; ZVFHMIN-NEXT: addi a4, a4, 16
; ZVFHMIN-NEXT: vs1r.v v0, (a4) # Unknown-size Folded Spill
@@ -1022,7 +987,8 @@ define <vscale x 32 x half> @vfmin_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
; ZVFHMIN-NEXT: vmfeq.vv v13, v24, v24, v0.t
; ZVFHMIN-NEXT: vmv8r.v v0, v16
; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: slli a3, a3, 4
+; ZVFHMIN-NEXT: slli a4, a3, 3
+; ZVFHMIN-NEXT: add a3, a4, a3
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
; ZVFHMIN-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
@@ -1030,102 +996,96 @@ define <vscale x 32 x half> @vfmin_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v4
; ZVFHMIN-NEXT: vmv1r.v v0, v13
; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: li a4, 24
-; ZVFHMIN-NEXT: mul a3, a3, a4
+; ZVFHMIN-NEXT: slli a4, a3, 4
+; ZVFHMIN-NEXT: add a3, a4, a3
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
; ZVFHMIN-NEXT: vl8r.v v24, (a3) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, a2, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmerge.vvm v24, v24, v16, v0
-; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a2, a2, 3
-; ZVFHMIN-NEXT: add a2, sp, a2
-; ZVFHMIN-NEXT: addi a2, a2, 16
+; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vs8r.v v24, (a2) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vmv1r.v v0, v12
; ZVFHMIN-NEXT: vmfeq.vv v13, v16, v16, v0.t
; ZVFHMIN-NEXT: vmv1r.v v0, v13
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: li a3, 24
-; ZVFHMIN-NEXT: mul a2, a2, a3
+; ZVFHMIN-NEXT: slli a3, a2, 4
+; ZVFHMIN-NEXT: add a2, a3, a2
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
; ZVFHMIN-NEXT: vl8r.v v24, (a2) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vmerge.vvm v16, v16, v24, v0
; ZVFHMIN-NEXT: vmv1r.v v0, v12
-; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a2, a2, 3
-; ZVFHMIN-NEXT: add a2, sp, a2
-; ZVFHMIN-NEXT: addi a2, a2, 16
+; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vl8r.v v24, (a2) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfmin.vv v16, v16, v24, v0.t
; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v12, v16
-; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a2, a2, 3
-; ZVFHMIN-NEXT: add a2, sp, a2
-; ZVFHMIN-NEXT: addi a2, a2, 16
-; ZVFHMIN-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: bltu a0, a1, .LBB22_2
; ZVFHMIN-NEXT: # %bb.1:
; ZVFHMIN-NEXT: mv a0, a1
; ZVFHMIN-NEXT: .LBB22_2:
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a1, a1, 5
+; ZVFHMIN-NEXT: slli a2, a1, 4
+; ZVFHMIN-NEXT: add a1, a2, a1
; 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 v24, v16, v16, v0.t
-; ZVFHMIN-NEXT: vmv8r.v v8, v16
-; ZVFHMIN-NEXT: addi a1, sp, 16
; ZVFHMIN-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a1, a1, 4
+; ZVFHMIN-NEXT: slli a1, a1, 3
+; ZVFHMIN-NEXT: add a1, sp, a1
+; ZVFHMIN-NEXT: addi a1, a1, 16
+; ZVFHMIN-NEXT: vl1r.v v9, (a1) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vmv1r.v v0, v9
+; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
+; ZVFHMIN-NEXT: vmfeq.vv v8, v16, v16, v0.t
+; ZVFHMIN-NEXT: csrr a1, vlenb
+; ZVFHMIN-NEXT: slli a2, a1, 3
+; ZVFHMIN-NEXT: add a1, a2, a1
; 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, v24
+; ZVFHMIN-NEXT: vmv1r.v v0, v8
+; ZVFHMIN-NEXT: csrr a1, vlenb
+; ZVFHMIN-NEXT: slli a2, a1, 4
+; ZVFHMIN-NEXT: add a1, a2, a1
+; 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 zero, a0, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vmerge.vvm v24, v8, v16, v0
+; ZVFHMIN-NEXT: vmerge.vvm v24, v24, v16, v0
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: li a1, 24
-; ZVFHMIN-NEXT: mul a0, a0, a1
+; ZVFHMIN-NEXT: slli a1, a0, 3
+; ZVFHMIN-NEXT: add a0, a1, a0
; 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, 5
-; ZVFHMIN-NEXT: add a0, sp, a0
-; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vl1r.v v9, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vmv1r.v v0, v9
; ZVFHMIN-NEXT: vmfeq.vv v8, v16, v16, v0.t
; ZVFHMIN-NEXT: vmv1r.v v0, v8
-; ZVFHMIN-NEXT: addi a0, sp, 16
+; ZVFHMIN-NEXT: csrr a0, vlenb
+; ZVFHMIN-NEXT: slli a1, a0, 4
+; ZVFHMIN-NEXT: add a0, a1, a0
+; ZVFHMIN-NEXT: add a0, sp, a0
+; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vmerge.vvm v16, v16, v24, v0
; ZVFHMIN-NEXT: vmv1r.v v0, v9
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: li a1, 24
-; ZVFHMIN-NEXT: mul a0, a0, a1
+; ZVFHMIN-NEXT: slli a1, a0, 3
+; ZVFHMIN-NEXT: add a0, a1, a0
; 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: 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: slli a1, a0, 5
-; ZVFHMIN-NEXT: add a0, a1, a0
+; ZVFHMIN-NEXT: li a1, 25
+; ZVFHMIN-NEXT: mul a0, a0, a1
; ZVFHMIN-NEXT: add sp, sp, a0
; ZVFHMIN-NEXT: .cfi_def_cfa sp, 16
; ZVFHMIN-NEXT: addi sp, sp, 16
@@ -1578,8 +1538,6 @@ define <vscale x 16 x double> @vfmin_vv_nxv16f64(<vscale x 16 x double> %va, <vs
; 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: 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: slli a1, a1, 4
; CHECK-NEXT: add a1, sp, a1
@@ -1606,20 +1564,19 @@ define <vscale x 16 x double> @vfmin_vv_nxv16f64(<vscale x 16 x double> %va, <vs
; CHECK-NEXT: add a3, sp, a3
; CHECK-NEXT: addi a3, a3, 16
; CHECK-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
+; CHECK-NEXT: vmv8r.v v24, v16
; CHECK-NEXT: vmv1r.v v0, v6
-; CHECK-NEXT: vmfeq.vv v26, v8, v8, v0.t
+; CHECK-NEXT: vmfeq.vv v5, v8, v8, v0.t
; CHECK-NEXT: vl8re64.v v16, (a0)
-; CHECK-NEXT: vmv1r.v v0, v26
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vmerge.vvm v24, v8, v24, v0
+; CHECK-NEXT: vmv1r.v v0, v5
+; CHECK-NEXT: vmerge.vvm v8, v8, v24, v0
; CHECK-NEXT: vmv1r.v v0, v6
; 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: vfmin.vv v8, v24, v8, 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, 3
; CHECK-NEXT: add a0, sp, a0
diff --git a/llvm/test/CodeGen/RISCV/rvv/fshr-fshl-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fshr-fshl-vp.ll
index c7e3c8cb519829..976c66d80d472f 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fshr-fshl-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fshr-fshl-vp.ll
@@ -984,19 +984,18 @@ 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, 48
-; CHECK-NEXT: mul a1, a1, a3
+; CHECK-NEXT: slli a1, a1, 5
; CHECK-NEXT: sub sp, sp, a1
-; 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: .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 v7, v0
; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: li a3, 24
-; CHECK-NEXT: mul a1, a1, a3
+; 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: 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
@@ -1005,97 +1004,41 @@ define <vscale x 16 x i64> @fshr_v16i64(<vscale x 16 x i64> %a, <vscale x 16 x i
; CHECK-NEXT: sub a5, a4, a1
; CHECK-NEXT: add a6, a2, a3
; CHECK-NEXT: vl8re64.v v8, (a6)
-; CHECK-NEXT: csrr a6, vlenb
-; CHECK-NEXT: li a7, 40
-; CHECK-NEXT: mul a6, a6, a7
-; CHECK-NEXT: add a6, sp, a6
-; CHECK-NEXT: addi a6, a6, 16
-; CHECK-NEXT: vs8r.v v8, (a6) # Unknown-size Folded Spill
; 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 v16, (a3)
-; CHECK-NEXT: csrr a3, vlenb
-; 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: vsetvli a3, zero, e8, mf4, ta, ma
; CHECK-NEXT: vslidedown.vx v0, v0, a6
; CHECK-NEXT: li a3, 63
-; CHECK-NEXT: csrr a6, vlenb
-; CHECK-NEXT: li a7, 40
-; CHECK-NEXT: mul a6, a6, a7
-; CHECK-NEXT: add a6, sp, a6
-; CHECK-NEXT: addi a6, a6, 16
-; CHECK-NEXT: vl8r.v v8, (a6) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, a5, e64, m8, ta, ma
-; CHECK-NEXT: vand.vx v8, v8, a3, 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: csrr a5, vlenb
-; CHECK-NEXT: slli a5, a5, 4
-; CHECK-NEXT: add a5, sp, a5
-; CHECK-NEXT: addi a5, a5, 16
-; CHECK-NEXT: vl8r.v v16, (a5) # 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: vsrl.vv v16, 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: vand.vx v24, v8, a3, v0.t
+; CHECK-NEXT: vsrl.vv v16, v16, v24, v0.t
+; CHECK-NEXT: addi a5, sp, 16
; CHECK-NEXT: vs8r.v v16, (a5) # Unknown-size Folded Spill
-; 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 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: vand.vx v16, v8, a3, v0.t
; 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 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: 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: vl8re64.v v24, (a2)
; 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: vl8r.v v8, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vsll.vi v16, v8, 1, v0.t
+; CHECK-NEXT: vsll.vi v8, v8, 1, v0.t
+; CHECK-NEXT: vsll.vv v8, v8, v16, 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: 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: vor.vv v8, v16, v8, v0.t
+; 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, 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
@@ -1103,56 +1046,33 @@ define <vscale x 16 x i64> @fshr_v16i64(<vscale x 16 x i64> %a, <vscale x 16 x i
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a4, a1
; CHECK-NEXT: .LBB46_2:
-; CHECK-NEXT: vmv1r.v v0, v24
-; CHECK-NEXT: csrr a0, vlenb
-; 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 v16, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, a4, e64, m8, ta, ma
-; CHECK-NEXT: vand.vx v8, v16, a3, v0.t
+; CHECK-NEXT: vand.vx v8, v24, a3, v0.t
; 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 v16, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vsrl.vv v8, v16, v8, 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: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vsrl.vv v16, v16, v8, v0.t
+; CHECK-NEXT: vnot.v v24, v24, v0.t
+; CHECK-NEXT: vand.vx v24, v24, a3, 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 v8, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vnot.v v16, v8, 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
-; 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, 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: vsll.vv v8, v8, v24, v0.t
; CHECK-NEXT: vor.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: csrr a0, vlenb
-; CHECK-NEXT: li a1, 48
-; CHECK-NEXT: mul a0, a0, a1
+; CHECK-NEXT: slli a0, a0, 5
; CHECK-NEXT: add sp, sp, a0
; CHECK-NEXT: .cfi_def_cfa sp, 16
; CHECK-NEXT: addi sp, sp, 16
@@ -1169,151 +1089,77 @@ 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, 40
+; CHECK-NEXT: li a3, 24
; CHECK-NEXT: mul a1, a1, a3
; CHECK-NEXT: sub sp, sp, a1
-; 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: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x18, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 24 * vlenb
+; CHECK-NEXT: vmv1r.v v7, v0
; 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: 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: 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: csrr a3, vlenb
-; CHECK-NEXT: slli a5, a3, 3
-; CHECK-NEXT: srli a1, a3, 3
-; CHECK-NEXT: sub a6, a4, a3
+; CHECK-NEXT: csrr a1, vlenb
+; CHECK-NEXT: slli a5, a1, 3
+; CHECK-NEXT: srli a3, a1, 3
+; CHECK-NEXT: sub a6, a4, a1
; CHECK-NEXT: vsetvli a7, zero, e8, mf4, ta, ma
-; CHECK-NEXT: vslidedown.vx v0, v0, a1
-; CHECK-NEXT: add a1, a2, a5
-; CHECK-NEXT: vl8re64.v v8, (a1)
-; CHECK-NEXT: addi a1, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
-; CHECK-NEXT: sltu a1, a4, a6
-; CHECK-NEXT: addi a1, a1, -1
-; CHECK-NEXT: and a6, a1, a6
-; CHECK-NEXT: li a1, 63
+; CHECK-NEXT: vslidedown.vx v0, v0, a3
+; CHECK-NEXT: add a3, a2, a5
+; CHECK-NEXT: vl8re64.v v8, (a3)
+; CHECK-NEXT: sltu a3, a4, a6
+; CHECK-NEXT: addi a3, a3, -1
+; CHECK-NEXT: and a6, a3, a6
+; CHECK-NEXT: li a3, 63
; 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, 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: 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: 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: 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: vand.vx v24, v8, a3, v0.t
+; CHECK-NEXT: vsll.vv v16, v16, v24, v0.t
+; CHECK-NEXT: addi a6, sp, 16
; CHECK-NEXT: vs8r.v v16, (a6) # Unknown-size Folded Spill
; CHECK-NEXT: add a5, a0, a5
-; CHECK-NEXT: addi a6, sp, 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: 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: vl8re64.v v8, (a2)
+; CHECK-NEXT: vand.vx v8, v8, a3, v0.t
+; CHECK-NEXT: vl8re64.v v24, (a0)
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 5
+; 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: vl8re64.v v24, (a2)
; CHECK-NEXT: vsrl.vi v16, v16, 1, v0.t
+; CHECK-NEXT: vsrl.vv v8, v16, v8, 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: 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: 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: vs8r.v v8, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: bltu a4, a3, .LBB47_2
+; CHECK-NEXT: bltu a4, a1, .LBB47_2
; CHECK-NEXT: # %bb.1:
-; CHECK-NEXT: mv a4, a3
+; CHECK-NEXT: mv a4, a1
; CHECK-NEXT: .LBB47_2:
-; CHECK-NEXT: vmv1r.v v0, v24
-; 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 v8, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, a4, e64, m8, ta, ma
-; 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: vand.vx v8, v24, a3, 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: 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: li a2, 24
-; 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: 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 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: vsll.vv v16, v16, v8, v0.t
+; CHECK-NEXT: vnot.v v24, v24, v0.t
+; CHECK-NEXT: vand.vx v24, v24, a3, v0.t
; 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: vsrl.vi v8, v8, 1, v0.t
-; CHECK-NEXT: vsrl.vv v8, v8, v16, v0.t
-; 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 v16, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vsrl.vv v8, v8, v24, v0.t
; CHECK-NEXT: vor.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: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a1, 40
+; CHECK-NEXT: li a1, 24
; CHECK-NEXT: mul a0, a0, a1
; CHECK-NEXT: add sp, sp, a0
; CHECK-NEXT: .cfi_def_cfa sp, 16
diff --git a/llvm/test/CodeGen/RISCV/rvv/mscatter-sdnode.ll b/llvm/test/CodeGen/RISCV/rvv/mscatter-sdnode.ll
index 72c251ce985cbf..3540bb5516b537 100644
--- a/llvm/test/CodeGen/RISCV/rvv/mscatter-sdnode.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/mscatter-sdnode.ll
@@ -1894,57 +1894,26 @@ 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, 5
+; RV64-NEXT: slli a2, a2, 3
; RV64-NEXT: sub sp, sp, a2
-; RV64-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x20, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 32 * vlenb
-; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: li a3, 24
-; RV64-NEXT: mul a2, a2, a3
-; RV64-NEXT: add a2, sp, a2
-; RV64-NEXT: addi a2, a2, 16
+; 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: addi a2, sp, 16
; RV64-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
-; RV64-NEXT: csrr a2, vlenb
-; RV64-NEXT: slli a2, a2, 4
-; RV64-NEXT: add a2, sp, a2
-; RV64-NEXT: addi a2, a2, 16
-; RV64-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
-; RV64-NEXT: vl8re64.v v8, (a0)
-; 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 v8, (a0) # Unknown-size Folded Spill
+; RV64-NEXT: vmv8r.v v16, v8
+; RV64-NEXT: vl8re64.v v24, (a0)
; RV64-NEXT: csrr a0, vlenb
; RV64-NEXT: vl8re64.v v8, (a1)
-; RV64-NEXT: addi a1, sp, 16
-; RV64-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
; RV64-NEXT: srli a0, a0, 3
; RV64-NEXT: vsetvli a1, zero, e8, mf4, ta, ma
-; RV64-NEXT: vslidedown.vx v24, v0, a0
-; RV64-NEXT: csrr a0, vlenb
-; RV64-NEXT: slli a0, a0, 4
-; RV64-NEXT: add a0, sp, a0
-; RV64-NEXT: addi a0, a0, 16
-; RV64-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; 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: vslidedown.vx v7, v0, a0
; RV64-NEXT: vsetvli a0, zero, e64, m8, ta, ma
-; RV64-NEXT: vsoxei64.v v16, (zero), v8, v0.t
-; RV64-NEXT: vmv1r.v v0, v24
-; RV64-NEXT: csrr a0, vlenb
-; RV64-NEXT: li a1, 24
-; RV64-NEXT: mul a0, a0, a1
-; RV64-NEXT: add a0, sp, a0
-; RV64-NEXT: addi a0, a0, 16
-; RV64-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; RV64-NEXT: vsoxei64.v v16, (zero), v24, v0.t
+; RV64-NEXT: vmv1r.v v0, v7
; RV64-NEXT: addi a0, sp, 16
-; RV64-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; RV64-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; RV64-NEXT: vsoxei64.v v16, (zero), v8, v0.t
; RV64-NEXT: csrr a0, vlenb
-; RV64-NEXT: slli a0, a0, 5
+; RV64-NEXT: slli a0, a0, 3
; RV64-NEXT: add sp, sp, a0
; RV64-NEXT: .cfi_def_cfa sp, 16
; RV64-NEXT: addi sp, sp, 16
@@ -2001,35 +1970,20 @@ define void @mscatter_baseidx_nxv16i8_nxv16f64(<vscale x 8 x double> %val0, <vsc
define void @mscatter_baseidx_nxv16i16_nxv16f64(<vscale x 8 x double> %val0, <vscale x 8 x double> %val1, ptr %base, <vscale x 16 x i16> %idxs, <vscale x 16 x i1> %m) {
; RV32-LABEL: mscatter_baseidx_nxv16i16_nxv16f64:
; RV32: # %bb.0:
-; RV32-NEXT: addi sp, sp, -16
-; RV32-NEXT: .cfi_def_cfa_offset 16
-; RV32-NEXT: csrr a2, vlenb
-; RV32-NEXT: slli a2, a2, 3
-; RV32-NEXT: sub sp, sp, a2
-; RV32-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
-; RV32-NEXT: addi a2, sp, 16
-; RV32-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
-; RV32-NEXT: vmv8r.v v16, v8
-; RV32-NEXT: vl4re16.v v8, (a1)
+; RV32-NEXT: vmv1r.v v6, v0
+; RV32-NEXT: vl4re16.v v0, (a1)
; RV32-NEXT: csrr a1, vlenb
; RV32-NEXT: srli a1, a1, 3
; RV32-NEXT: vsetvli a2, zero, e8, mf4, ta, ma
-; RV32-NEXT: vslidedown.vx v7, v0, a1
+; RV32-NEXT: vslidedown.vx v7, v6, a1
; RV32-NEXT: vsetvli a1, zero, e32, m8, ta, ma
-; RV32-NEXT: vsext.vf2 v24, v8
-; RV32-NEXT: vsll.vi v8, v24, 3
+; RV32-NEXT: vsext.vf2 v24, v0
+; RV32-NEXT: vsll.vi v24, v24, 3
+; RV32-NEXT: vmv1r.v v0, v6
; RV32-NEXT: vsetvli a1, zero, e64, m8, ta, ma
-; RV32-NEXT: vsoxei32.v v16, (a0), v8, v0.t
+; RV32-NEXT: vsoxei32.v v8, (a0), v24, v0.t
; RV32-NEXT: vmv1r.v v0, v7
-; RV32-NEXT: addi a1, sp, 16
-; RV32-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
-; RV32-NEXT: vsoxei32.v v16, (a0), v12, v0.t
-; RV32-NEXT: csrr a0, vlenb
-; RV32-NEXT: slli a0, a0, 3
-; RV32-NEXT: add sp, sp, a0
-; RV32-NEXT: .cfi_def_cfa sp, 16
-; RV32-NEXT: addi sp, sp, 16
-; RV32-NEXT: .cfi_def_cfa_offset 0
+; RV32-NEXT: vsoxei32.v v16, (a0), v28, v0.t
; RV32-NEXT: ret
;
; RV64-LABEL: mscatter_baseidx_nxv16i16_nxv16f64:
diff --git a/llvm/test/CodeGen/RISCV/rvv/rint-vp.ll b/llvm/test/CodeGen/RISCV/rvv/rint-vp.ll
index 2a69dd31118bd8..ad6b28e04025c5 100644
--- a/llvm/test/CodeGen/RISCV/rvv/rint-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/rint-vp.ll
@@ -253,12 +253,6 @@ declare <vscale x 32 x bfloat> @llvm.vp.rint.nxv32bf16(<vscale x 32 x bfloat>, <
define <vscale x 32 x bfloat> @vp_rint_nxv32bf16(<vscale x 32 x bfloat> %va, <vscale x 32 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_rint_nxv32bf16:
; 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 v7, v0
; CHECK-NEXT: csrr a2, vlenb
; CHECK-NEXT: vsetvli a1, zero, e16, m4, ta, ma
@@ -276,9 +270,6 @@ define <vscale x 32 x bfloat> @vp_rint_nxv32bf16(<vscale x 32 x bfloat> %va, <vs
; CHECK-NEXT: vmv1r.v v0, v12
; CHECK-NEXT: vsetvli zero, a2, e32, m8, ta, ma
; CHECK-NEXT: vfabs.v v16, v24, v0.t
-; CHECK-NEXT: addi a2, sp, 16
-; CHECK-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
-; CHECK-NEXT: vl8r.v v16, (a2) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, mu
; CHECK-NEXT: vmflt.vf v12, v16, fa5, v0.t
; CHECK-NEXT: vmv1r.v v0, v12
@@ -307,12 +298,6 @@ define <vscale x 32 x bfloat> @vp_rint_nxv32bf16(<vscale x 32 x bfloat> %va, <vs
; CHECK-NEXT: vfsgnj.vv v24, v16, v24, v0.t
; CHECK-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; CHECK-NEXT: vfncvtbf16.f.f.w v8, v24
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add sp, sp, a0
-; CHECK-NEXT: .cfi_def_cfa sp, 16
-; CHECK-NEXT: addi sp, sp, 16
-; CHECK-NEXT: .cfi_def_cfa_offset 0
; CHECK-NEXT: ret
%v = call <vscale x 32 x bfloat> @llvm.vp.rint.nxv32bf16(<vscale x 32 x bfloat> %va, <vscale x 32 x i1> %m, i32 %evl)
ret <vscale x 32 x bfloat> %v
@@ -769,12 +754,6 @@ define <vscale x 32 x half> @vp_rint_nxv32f16(<vscale x 32 x half> %va, <vscale
;
; ZVFHMIN-LABEL: vp_rint_nxv32f16:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: addi sp, sp, -16
-; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
-; ZVFHMIN-NEXT: csrr a1, vlenb
-; 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 v7, v0
; ZVFHMIN-NEXT: csrr a2, vlenb
; ZVFHMIN-NEXT: vsetvli a1, zero, e16, m4, ta, ma
@@ -792,9 +771,6 @@ define <vscale x 32 x half> @vp_rint_nxv32f16(<vscale x 32 x half> %va, <vscale
; ZVFHMIN-NEXT: vmv1r.v v0, v12
; ZVFHMIN-NEXT: vsetvli zero, a2, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfabs.v v16, v24, v0.t
-; ZVFHMIN-NEXT: addi a2, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vl8r.v v16, (a2) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, mu
; ZVFHMIN-NEXT: vmflt.vf v12, v16, fa5, v0.t
; ZVFHMIN-NEXT: vmv1r.v v0, v12
@@ -823,12 +799,6 @@ define <vscale x 32 x half> @vp_rint_nxv32f16(<vscale x 32 x half> %va, <vscale
; ZVFHMIN-NEXT: vfsgnj.vv v24, v16, v24, v0.t
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; 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
-; ZVFHMIN-NEXT: .cfi_def_cfa sp, 16
-; ZVFHMIN-NEXT: addi sp, sp, 16
-; ZVFHMIN-NEXT: .cfi_def_cfa_offset 0
; ZVFHMIN-NEXT: ret
%v = call <vscale x 32 x half> @llvm.vp.rint.nxv32f16(<vscale x 32 x half> %va, <vscale x 32 x i1> %m, i32 %evl)
ret <vscale x 32 x half> %v
@@ -1297,12 +1267,6 @@ declare <vscale x 16 x double> @llvm.vp.rint.nxv16f64(<vscale x 16 x double>, <v
define <vscale x 16 x double> @vp_rint_nxv16f64(<vscale x 16 x double> %va, <vscale x 16 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_rint_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, 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: csrr a1, vlenb
; CHECK-NEXT: lui a2, %hi(.LCPI44_0)
@@ -1317,9 +1281,6 @@ define <vscale x 16 x double> @vp_rint_nxv16f64(<vscale x 16 x double> %va, <vsc
; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: vsetvli zero, a2, e64, m8, ta, ma
; 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 v6, v24, fa5, v0.t
; CHECK-NEXT: vmv1r.v v0, v6
@@ -1343,12 +1304,6 @@ define <vscale x 16 x double> @vp_rint_nxv16f64(<vscale x 16 x double> %va, <vsc
; 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: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add sp, sp, a0
-; CHECK-NEXT: .cfi_def_cfa sp, 16
-; CHECK-NEXT: addi sp, sp, 16
-; CHECK-NEXT: .cfi_def_cfa_offset 0
; CHECK-NEXT: ret
%v = call <vscale x 16 x double> @llvm.vp.rint.nxv16f64(<vscale x 16 x double> %va, <vscale x 16 x i1> %m, i32 %evl)
ret <vscale x 16 x double> %v
diff --git a/llvm/test/CodeGen/RISCV/rvv/round-vp.ll b/llvm/test/CodeGen/RISCV/rvv/round-vp.ll
index 8a10e75333ad0a..d5879ea519201b 100644
--- a/llvm/test/CodeGen/RISCV/rvv/round-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/round-vp.ll
@@ -273,12 +273,6 @@ declare <vscale x 32 x bfloat> @llvm.vp.round.nxv32bf16(<vscale x 32 x bfloat>,
define <vscale x 32 x bfloat> @vp_round_nxv32bf16(<vscale x 32 x bfloat> %va, <vscale x 32 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_round_nxv32bf16:
; 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 v7, v0
; CHECK-NEXT: csrr a2, vlenb
; CHECK-NEXT: vsetvli a1, zero, e16, m4, ta, ma
@@ -302,11 +296,7 @@ define <vscale x 32 x bfloat> @vp_round_nxv32bf16(<vscale x 32 x bfloat> %va, <v
; CHECK-NEXT: vmv1r.v v0, v12
; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v16, v24, v0.t
-; CHECK-NEXT: addi a3, sp, 16
-; CHECK-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
; CHECK-NEXT: fsrm a2
-; CHECK-NEXT: addi a2, sp, 16
-; CHECK-NEXT: vl8r.v v16, (a2) # Unknown-size Folded Reload
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, mu
; CHECK-NEXT: vfsgnj.vv v24, v16, v24, v0.t
@@ -332,12 +322,6 @@ define <vscale x 32 x bfloat> @vp_round_nxv32bf16(<vscale x 32 x bfloat> %va, <v
; CHECK-NEXT: vfsgnj.vv v24, v16, v24, v0.t
; CHECK-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; CHECK-NEXT: vfncvtbf16.f.f.w v8, v24
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add sp, sp, a0
-; CHECK-NEXT: .cfi_def_cfa sp, 16
-; CHECK-NEXT: addi sp, sp, 16
-; CHECK-NEXT: .cfi_def_cfa_offset 0
; CHECK-NEXT: ret
%v = call <vscale x 32 x bfloat> @llvm.vp.round.nxv32bf16(<vscale x 32 x bfloat> %va, <vscale x 32 x i1> %m, i32 %evl)
ret <vscale x 32 x bfloat> %v
@@ -840,12 +824,6 @@ define <vscale x 32 x half> @vp_round_nxv32f16(<vscale x 32 x half> %va, <vscale
;
; ZVFHMIN-LABEL: vp_round_nxv32f16:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: addi sp, sp, -16
-; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
-; ZVFHMIN-NEXT: csrr a1, vlenb
-; 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 v7, v0
; ZVFHMIN-NEXT: csrr a2, vlenb
; ZVFHMIN-NEXT: vsetvli a1, zero, e16, m4, ta, ma
@@ -869,11 +847,7 @@ define <vscale x 32 x half> @vp_round_nxv32f16(<vscale x 32 x half> %va, <vscale
; ZVFHMIN-NEXT: vmv1r.v v0, v12
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfcvt.x.f.v v16, v24, v0.t
-; ZVFHMIN-NEXT: addi a3, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: fsrm a2
-; ZVFHMIN-NEXT: addi a2, sp, 16
-; ZVFHMIN-NEXT: vl8r.v v16, (a2) # 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, v16, v24, v0.t
@@ -899,12 +873,6 @@ define <vscale x 32 x half> @vp_round_nxv32f16(<vscale x 32 x half> %va, <vscale
; ZVFHMIN-NEXT: vfsgnj.vv v24, v16, v24, v0.t
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; 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
-; ZVFHMIN-NEXT: .cfi_def_cfa sp, 16
-; ZVFHMIN-NEXT: addi sp, sp, 16
-; ZVFHMIN-NEXT: .cfi_def_cfa_offset 0
; ZVFHMIN-NEXT: ret
%v = call <vscale x 32 x half> @llvm.vp.round.nxv32f16(<vscale x 32 x half> %va, <vscale x 32 x i1> %m, i32 %evl)
ret <vscale x 32 x half> %v
@@ -1419,12 +1387,6 @@ declare <vscale x 16 x double> @llvm.vp.round.nxv16f64(<vscale x 16 x double>, <
define <vscale x 16 x double> @vp_round_nxv16f64(<vscale x 16 x double> %va, <vscale x 16 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_round_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, 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: csrr a1, vlenb
; CHECK-NEXT: lui a2, %hi(.LCPI44_0)
@@ -1445,40 +1407,27 @@ define <vscale x 16 x double> @vp_round_nxv16f64(<vscale x 16 x double> %va, <vs
; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; 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: 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, v24, v16, v0.t
-; CHECK-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
; CHECK-NEXT: bltu a0, a1, .LBB44_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB44_2:
; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; 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 v7, 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: 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: addi a0, sp, 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 sp, sp, a0
-; CHECK-NEXT: .cfi_def_cfa sp, 16
-; CHECK-NEXT: addi sp, sp, 16
-; CHECK-NEXT: .cfi_def_cfa_offset 0
+; CHECK-NEXT: vfsgnj.vv v8, v24, v8, v0.t
; CHECK-NEXT: ret
%v = call <vscale x 16 x double> @llvm.vp.round.nxv16f64(<vscale x 16 x double> %va, <vscale x 16 x i1> %m, i32 %evl)
ret <vscale x 16 x double> %v
diff --git a/llvm/test/CodeGen/RISCV/rvv/roundeven-vp.ll b/llvm/test/CodeGen/RISCV/rvv/roundeven-vp.ll
index 4cd909e4b0a637..d746a96596f8d2 100644
--- a/llvm/test/CodeGen/RISCV/rvv/roundeven-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/roundeven-vp.ll
@@ -273,12 +273,6 @@ declare <vscale x 32 x bfloat> @llvm.vp.roundeven.nxv32bf16(<vscale x 32 x bfloa
define <vscale x 32 x bfloat> @vp_roundeven_nxv32bf16(<vscale x 32 x bfloat> %va, <vscale x 32 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_roundeven_nxv32bf16:
; 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 v7, v0
; CHECK-NEXT: csrr a2, vlenb
; CHECK-NEXT: vsetvli a1, zero, e16, m4, ta, ma
@@ -302,11 +296,7 @@ define <vscale x 32 x bfloat> @vp_roundeven_nxv32bf16(<vscale x 32 x bfloat> %va
; CHECK-NEXT: vmv1r.v v0, v12
; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v16, v24, v0.t
-; CHECK-NEXT: addi a3, sp, 16
-; CHECK-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
; CHECK-NEXT: fsrm a2
-; CHECK-NEXT: addi a2, sp, 16
-; CHECK-NEXT: vl8r.v v16, (a2) # Unknown-size Folded Reload
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, mu
; CHECK-NEXT: vfsgnj.vv v24, v16, v24, v0.t
@@ -332,12 +322,6 @@ define <vscale x 32 x bfloat> @vp_roundeven_nxv32bf16(<vscale x 32 x bfloat> %va
; CHECK-NEXT: vfsgnj.vv v24, v16, v24, v0.t
; CHECK-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; CHECK-NEXT: vfncvtbf16.f.f.w v8, v24
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add sp, sp, a0
-; CHECK-NEXT: .cfi_def_cfa sp, 16
-; CHECK-NEXT: addi sp, sp, 16
-; CHECK-NEXT: .cfi_def_cfa_offset 0
; CHECK-NEXT: ret
%v = call <vscale x 32 x bfloat> @llvm.vp.roundeven.nxv32bf16(<vscale x 32 x bfloat> %va, <vscale x 32 x i1> %m, i32 %evl)
ret <vscale x 32 x bfloat> %v
@@ -840,12 +824,6 @@ define <vscale x 32 x half> @vp_roundeven_nxv32f16(<vscale x 32 x half> %va, <vs
;
; ZVFHMIN-LABEL: vp_roundeven_nxv32f16:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: addi sp, sp, -16
-; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
-; ZVFHMIN-NEXT: csrr a1, vlenb
-; 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 v7, v0
; ZVFHMIN-NEXT: csrr a2, vlenb
; ZVFHMIN-NEXT: vsetvli a1, zero, e16, m4, ta, ma
@@ -869,11 +847,7 @@ define <vscale x 32 x half> @vp_roundeven_nxv32f16(<vscale x 32 x half> %va, <vs
; ZVFHMIN-NEXT: vmv1r.v v0, v12
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfcvt.x.f.v v16, v24, v0.t
-; ZVFHMIN-NEXT: addi a3, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: fsrm a2
-; ZVFHMIN-NEXT: addi a2, sp, 16
-; ZVFHMIN-NEXT: vl8r.v v16, (a2) # 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, v16, v24, v0.t
@@ -899,12 +873,6 @@ define <vscale x 32 x half> @vp_roundeven_nxv32f16(<vscale x 32 x half> %va, <vs
; ZVFHMIN-NEXT: vfsgnj.vv v24, v16, v24, v0.t
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; 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
-; ZVFHMIN-NEXT: .cfi_def_cfa sp, 16
-; ZVFHMIN-NEXT: addi sp, sp, 16
-; ZVFHMIN-NEXT: .cfi_def_cfa_offset 0
; ZVFHMIN-NEXT: ret
%v = call <vscale x 32 x half> @llvm.vp.roundeven.nxv32f16(<vscale x 32 x half> %va, <vscale x 32 x i1> %m, i32 %evl)
ret <vscale x 32 x half> %v
@@ -1419,12 +1387,6 @@ declare <vscale x 16 x double> @llvm.vp.roundeven.nxv16f64(<vscale x 16 x double
define <vscale x 16 x double> @vp_roundeven_nxv16f64(<vscale x 16 x double> %va, <vscale x 16 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_roundeven_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, 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: csrr a1, vlenb
; CHECK-NEXT: lui a2, %hi(.LCPI44_0)
@@ -1445,40 +1407,27 @@ define <vscale x 16 x double> @vp_roundeven_nxv16f64(<vscale x 16 x double> %va,
; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; 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: 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, v24, v16, v0.t
-; CHECK-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
; CHECK-NEXT: bltu a0, a1, .LBB44_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB44_2:
; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; 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 v7, 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: 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: addi a0, sp, 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 sp, sp, a0
-; CHECK-NEXT: .cfi_def_cfa sp, 16
-; CHECK-NEXT: addi sp, sp, 16
-; CHECK-NEXT: .cfi_def_cfa_offset 0
+; CHECK-NEXT: vfsgnj.vv v8, v24, v8, v0.t
; CHECK-NEXT: ret
%v = call <vscale x 16 x double> @llvm.vp.roundeven.nxv16f64(<vscale x 16 x double> %va, <vscale x 16 x i1> %m, i32 %evl)
ret <vscale x 16 x double> %v
diff --git a/llvm/test/CodeGen/RISCV/rvv/roundtozero-vp.ll b/llvm/test/CodeGen/RISCV/rvv/roundtozero-vp.ll
index 96c821a76ae84e..e11df1338d8713 100644
--- a/llvm/test/CodeGen/RISCV/rvv/roundtozero-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/roundtozero-vp.ll
@@ -273,12 +273,6 @@ declare <vscale x 32 x bfloat> @llvm.vp.roundtozero.nxv32bf16(<vscale x 32 x bfl
define <vscale x 32 x bfloat> @vp_roundtozero_nxv32bf16(<vscale x 32 x bfloat> %va, <vscale x 32 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_roundtozero_nxv32bf16:
; 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 v7, v0
; CHECK-NEXT: csrr a2, vlenb
; CHECK-NEXT: vsetvli a1, zero, e16, m4, ta, ma
@@ -302,11 +296,7 @@ define <vscale x 32 x bfloat> @vp_roundtozero_nxv32bf16(<vscale x 32 x bfloat> %
; CHECK-NEXT: vmv1r.v v0, v12
; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; CHECK-NEXT: vfcvt.x.f.v v16, v24, v0.t
-; CHECK-NEXT: addi a3, sp, 16
-; CHECK-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
; CHECK-NEXT: fsrm a2
-; CHECK-NEXT: addi a2, sp, 16
-; CHECK-NEXT: vl8r.v v16, (a2) # Unknown-size Folded Reload
; CHECK-NEXT: vfcvt.f.x.v v16, v16, v0.t
; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, mu
; CHECK-NEXT: vfsgnj.vv v24, v16, v24, v0.t
@@ -332,12 +322,6 @@ define <vscale x 32 x bfloat> @vp_roundtozero_nxv32bf16(<vscale x 32 x bfloat> %
; CHECK-NEXT: vfsgnj.vv v24, v16, v24, v0.t
; CHECK-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; CHECK-NEXT: vfncvtbf16.f.f.w v8, v24
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add sp, sp, a0
-; CHECK-NEXT: .cfi_def_cfa sp, 16
-; CHECK-NEXT: addi sp, sp, 16
-; CHECK-NEXT: .cfi_def_cfa_offset 0
; CHECK-NEXT: ret
%v = call <vscale x 32 x bfloat> @llvm.vp.roundtozero.nxv32bf16(<vscale x 32 x bfloat> %va, <vscale x 32 x i1> %m, i32 %evl)
ret <vscale x 32 x bfloat> %v
@@ -840,12 +824,6 @@ define <vscale x 32 x half> @vp_roundtozero_nxv32f16(<vscale x 32 x half> %va, <
;
; ZVFHMIN-LABEL: vp_roundtozero_nxv32f16:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: addi sp, sp, -16
-; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
-; ZVFHMIN-NEXT: csrr a1, vlenb
-; 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 v7, v0
; ZVFHMIN-NEXT: csrr a2, vlenb
; ZVFHMIN-NEXT: vsetvli a1, zero, e16, m4, ta, ma
@@ -869,11 +847,7 @@ define <vscale x 32 x half> @vp_roundtozero_nxv32f16(<vscale x 32 x half> %va, <
; ZVFHMIN-NEXT: vmv1r.v v0, v12
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfcvt.x.f.v v16, v24, v0.t
-; ZVFHMIN-NEXT: addi a3, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: fsrm a2
-; ZVFHMIN-NEXT: addi a2, sp, 16
-; ZVFHMIN-NEXT: vl8r.v v16, (a2) # 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, v16, v24, v0.t
@@ -899,12 +873,6 @@ define <vscale x 32 x half> @vp_roundtozero_nxv32f16(<vscale x 32 x half> %va, <
; ZVFHMIN-NEXT: vfsgnj.vv v24, v16, v24, v0.t
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; 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
-; ZVFHMIN-NEXT: .cfi_def_cfa sp, 16
-; ZVFHMIN-NEXT: addi sp, sp, 16
-; ZVFHMIN-NEXT: .cfi_def_cfa_offset 0
; ZVFHMIN-NEXT: ret
%v = call <vscale x 32 x half> @llvm.vp.roundtozero.nxv32f16(<vscale x 32 x half> %va, <vscale x 32 x i1> %m, i32 %evl)
ret <vscale x 32 x half> %v
@@ -1419,12 +1387,6 @@ declare <vscale x 16 x double> @llvm.vp.roundtozero.nxv16f64(<vscale x 16 x doub
define <vscale x 16 x double> @vp_roundtozero_nxv16f64(<vscale x 16 x double> %va, <vscale x 16 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vp_roundtozero_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, 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: csrr a1, vlenb
; CHECK-NEXT: lui a2, %hi(.LCPI44_0)
@@ -1445,40 +1407,27 @@ define <vscale x 16 x double> @vp_roundtozero_nxv16f64(<vscale x 16 x double> %v
; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: vsetvli zero, zero, e64, m8, ta, ma
; 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: 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, v24, v16, v0.t
-; CHECK-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
; CHECK-NEXT: bltu a0, a1, .LBB44_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB44_2:
; CHECK-NEXT: vmv1r.v v0, v7
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; 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 v7, 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: 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: addi a0, sp, 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 sp, sp, a0
-; CHECK-NEXT: .cfi_def_cfa sp, 16
-; CHECK-NEXT: addi sp, sp, 16
-; CHECK-NEXT: .cfi_def_cfa_offset 0
+; CHECK-NEXT: vfsgnj.vv v8, v24, v8, v0.t
; CHECK-NEXT: ret
%v = call <vscale x 16 x double> @llvm.vp.roundtozero.nxv16f64(<vscale x 16 x double> %va, <vscale x 16 x i1> %m, i32 %evl)
ret <vscale x 16 x double> %v
diff --git a/llvm/test/CodeGen/RISCV/rvv/setcc-fp-vp.ll b/llvm/test/CodeGen/RISCV/rvv/setcc-fp-vp.ll
index 6c11e9413525e0..e9fd42566f43a6 100644
--- a/llvm/test/CodeGen/RISCV/rvv/setcc-fp-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/setcc-fp-vp.ll
@@ -1467,18 +1467,21 @@ define <vscale x 64 x i1> @fcmp_oeq_vv_nxv64bf16(<vscale x 64 x bfloat> %va, <vs
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: csrr a1, vlenb
; CHECK-NEXT: mv a3, a1
-; CHECK-NEXT: slli a1, a1, 3
+; CHECK-NEXT: slli a1, a1, 1
; CHECK-NEXT: add a3, a3, a1
; CHECK-NEXT: slli a1, a1, 2
+; CHECK-NEXT: add a3, a3, a1
+; CHECK-NEXT: slli a1, a1, 1
; CHECK-NEXT: add a1, a1, a3
; CHECK-NEXT: sub sp, sp, a1
-; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x29, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 41 * vlenb
+; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x1b, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 27 * vlenb
; CHECK-NEXT: vmv1r.v v24, v0
+; CHECK-NEXT: vmv8r.v v0, v16
; CHECK-NEXT: csrr a1, vlenb
; CHECK-NEXT: mv a3, a1
-; CHECK-NEXT: slli a1, a1, 3
-; CHECK-NEXT: add a3, a3, a1
; CHECK-NEXT: slli a1, a1, 1
+; CHECK-NEXT: add a3, a3, a1
+; CHECK-NEXT: slli a1, a1, 3
; CHECK-NEXT: add a1, a1, a3
; CHECK-NEXT: add a1, sp, a1
; CHECK-NEXT: addi a1, a1, 16
@@ -1489,7 +1492,7 @@ define <vscale x 64 x i1> @fcmp_oeq_vv_nxv64bf16(<vscale x 64 x bfloat> %va, <vs
; CHECK-NEXT: slli a4, a3, 1
; CHECK-NEXT: add a1, a0, a1
; CHECK-NEXT: sub a6, a2, a5
-; CHECK-NEXT: vl8re16.v v0, (a1)
+; CHECK-NEXT: vl8re16.v v16, (a1)
; CHECK-NEXT: sltu a1, a2, a6
; CHECK-NEXT: addi a1, a1, -1
; CHECK-NEXT: and a6, a1, a6
@@ -1500,66 +1503,62 @@ define <vscale x 64 x i1> @fcmp_oeq_vv_nxv64bf16(<vscale x 64 x bfloat> %va, <vs
; CHECK-NEXT: srli a1, a3, 1
; CHECK-NEXT: srli a3, a3, 2
; CHECK-NEXT: csrr t0, vlenb
-; CHECK-NEXT: slli t0, t0, 3
+; CHECK-NEXT: slli t0, t0, 1
; CHECK-NEXT: add t0, sp, t0
; CHECK-NEXT: addi t0, t0, 16
; CHECK-NEXT: vs1r.v v24, (t0) # Unknown-size Folded Spill
; CHECK-NEXT: vsetvli t0, zero, e8, m1, ta, ma
-; CHECK-NEXT: vslidedown.vx v25, v24, a1
+; CHECK-NEXT: vslidedown.vx v8, v24, a1
+; CHECK-NEXT: csrr t0, vlenb
+; CHECK-NEXT: add t0, sp, t0
+; CHECK-NEXT: addi t0, t0, 16
+; CHECK-NEXT: vs1r.v v8, (t0) # Unknown-size Folded Spill
; CHECK-NEXT: vsetvli t0, zero, e8, mf2, ta, ma
-; CHECK-NEXT: vslidedown.vx v24, v25, a3
+; CHECK-NEXT: vslidedown.vx v8, v8, a3
+; CHECK-NEXT: addi t0, sp, 16
+; CHECK-NEXT: vs1r.v v8, (t0) # Unknown-size Folded Spill
; CHECK-NEXT: vl8re16.v v8, (a0)
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli t0, a0, 5
-; CHECK-NEXT: add a0, t0, a0
+; CHECK-NEXT: mv t0, a0
+; CHECK-NEXT: slli a0, a0, 1
+; CHECK-NEXT: add t0, t0, a0
+; CHECK-NEXT: slli a0, a0, 2
+; CHECK-NEXT: add a0, a0, t0
; 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: slli t0, a0, 3
-; CHECK-NEXT: add a0, t0, 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: vsetvli a0, zero, e16, m4, ta, ma
-; CHECK-NEXT: vfwcvtbf16.f.f.v v8, v4
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: vmv8r.v v8, v16
+; CHECK-NEXT: vfwcvtbf16.f.f.v v8, v20
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli t0, a0, 4
+; CHECK-NEXT: slli t0, a0, 1
; CHECK-NEXT: add a0, t0, a0
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v12
-; CHECK-NEXT: vmv1r.v v0, v24
+; CHECK-NEXT: vs8r.v v0, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v4
; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vl1r.v v0, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, a7, e32, m8, ta, ma
-; CHECK-NEXT: vmfeq.vv v6, v16, v8, v0.t
+; CHECK-NEXT: vmfeq.vv v4, v24, v8, v0.t
; CHECK-NEXT: bltu a6, a4, .LBB85_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a6, a4
; CHECK-NEXT: .LBB85_2:
-; CHECK-NEXT: vmv1r.v v0, v25
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a7, a0, 3
-; CHECK-NEXT: add a0, a7, a0
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
-; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vl1r.v v0, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; CHECK-NEXT: vfwcvtbf16.f.f.v v8, v16
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a7, a0, 4
+; CHECK-NEXT: slli a7, a0, 1
; CHECK-NEXT: add a0, a7, a0
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v16
; CHECK-NEXT: vsetvli zero, a6, e32, m8, ta, ma
-; CHECK-NEXT: vmfeq.vv v5, v24, v8, v0.t
+; CHECK-NEXT: vmfeq.vv v6, v24, v8, v0.t
; CHECK-NEXT: add a0, a3, a3
; CHECK-NEXT: bltu a2, a5, .LBB85_4
; CHECK-NEXT: # %bb.3:
@@ -1567,7 +1566,7 @@ define <vscale x 64 x i1> @fcmp_oeq_vv_nxv64bf16(<vscale x 64 x bfloat> %va, <vs
; CHECK-NEXT: .LBB85_4:
; CHECK-NEXT: sub a5, a2, a4
; CHECK-NEXT: csrr a6, vlenb
-; CHECK-NEXT: slli a6, a6, 3
+; CHECK-NEXT: slli a6, a6, 1
; CHECK-NEXT: add a6, sp, a6
; CHECK-NEXT: addi a6, a6, 16
; CHECK-NEXT: vl1r.v v7, (a6) # Unknown-size Folded Reload
@@ -1575,51 +1574,54 @@ define <vscale x 64 x i1> @fcmp_oeq_vv_nxv64bf16(<vscale x 64 x bfloat> %va, <vs
; CHECK-NEXT: vslidedown.vx v0, v7, a3
; CHECK-NEXT: csrr a6, vlenb
; CHECK-NEXT: mv a7, a6
-; CHECK-NEXT: slli a6, a6, 3
-; CHECK-NEXT: add a7, a7, a6
; CHECK-NEXT: slli a6, a6, 1
+; CHECK-NEXT: add a7, a7, a6
+; CHECK-NEXT: slli a6, a6, 3
; CHECK-NEXT: add a6, a6, a7
; CHECK-NEXT: add a6, sp, a6
; CHECK-NEXT: addi a6, a6, 16
-; CHECK-NEXT: vl8r.v v16, (a6) # Unknown-size Folded Reload
+; CHECK-NEXT: vl8r.v v8, (a6) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli a6, zero, e16, m4, ta, ma
-; CHECK-NEXT: vfwcvtbf16.f.f.v v8, v20
+; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v12
; CHECK-NEXT: csrr a6, vlenb
-; CHECK-NEXT: slli a7, a6, 4
-; CHECK-NEXT: add a6, a7, a6
-; CHECK-NEXT: add a6, sp, a6
-; CHECK-NEXT: addi a6, a6, 16
-; CHECK-NEXT: vs8r.v v8, (a6) # Unknown-size Folded Spill
-; CHECK-NEXT: csrr a6, vlenb
-; CHECK-NEXT: slli a7, a6, 5
-; CHECK-NEXT: add a6, a7, a6
+; CHECK-NEXT: mv a7, a6
+; CHECK-NEXT: slli a6, a6, 1
+; CHECK-NEXT: add a7, a7, a6
+; CHECK-NEXT: slli a6, a6, 2
+; CHECK-NEXT: add a6, a6, a7
; CHECK-NEXT: add a6, sp, a6
; CHECK-NEXT: addi a6, a6, 16
-; CHECK-NEXT: vl8r.v v24, (a6) # Unknown-size Folded Reload
-; CHECK-NEXT: vfwcvtbf16.f.f.v v8, v28
+; CHECK-NEXT: vl8r.v v16, (a6) # Unknown-size Folded Reload
+; CHECK-NEXT: vfwcvtbf16.f.f.v v8, v20
; CHECK-NEXT: sltu a6, a2, a5
; CHECK-NEXT: addi a6, a6, -1
; CHECK-NEXT: and a5, a6, a5
-; CHECK-NEXT: csrr a6, vlenb
-; CHECK-NEXT: slli a7, a6, 4
-; CHECK-NEXT: add a6, a7, a6
-; CHECK-NEXT: add a6, sp, a6
-; CHECK-NEXT: addi a6, a6, 16
-; CHECK-NEXT: vl8r.v v24, (a6) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, a5, e32, m8, ta, ma
-; CHECK-NEXT: vmfeq.vv v4, v24, v8, v0.t
+; CHECK-NEXT: vmfeq.vv v5, v24, v8, v0.t
; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, ma
-; CHECK-NEXT: vslideup.vx v5, v6, a3
+; CHECK-NEXT: vslideup.vx v6, v4, a3
; CHECK-NEXT: bltu a2, a4, .LBB85_6
; CHECK-NEXT: # %bb.5:
; CHECK-NEXT: mv a2, a4
; CHECK-NEXT: .LBB85_6:
; CHECK-NEXT: vmv1r.v v0, v7
+; CHECK-NEXT: csrr a4, vlenb
+; CHECK-NEXT: mv a5, a4
+; CHECK-NEXT: slli a4, a4, 1
+; CHECK-NEXT: add a5, a5, a4
+; CHECK-NEXT: slli a4, a4, 3
+; CHECK-NEXT: add 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 a4, zero, e16, m4, ta, ma
-; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v16
+; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v8
; CHECK-NEXT: csrr a4, vlenb
-; CHECK-NEXT: slli a5, a4, 5
-; CHECK-NEXT: add a4, a5, a4
+; CHECK-NEXT: mv a5, a4
+; CHECK-NEXT: slli a4, a4, 1
+; CHECK-NEXT: add a5, a5, a4
+; CHECK-NEXT: slli a4, a4, 2
+; CHECK-NEXT: add 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
@@ -1627,16 +1629,18 @@ define <vscale x 64 x i1> @fcmp_oeq_vv_nxv64bf16(<vscale x 64 x bfloat> %va, <vs
; CHECK-NEXT: vsetvli zero, a2, e32, m8, ta, ma
; CHECK-NEXT: vmfeq.vv v8, v24, v16, v0.t
; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, ma
-; CHECK-NEXT: vslideup.vx v8, v4, a3
+; CHECK-NEXT: vslideup.vx v8, v5, a3
; CHECK-NEXT: add a0, a1, a1
; CHECK-NEXT: vsetvli zero, a0, e8, m1, ta, ma
-; CHECK-NEXT: vslideup.vx v8, v5, a1
+; CHECK-NEXT: vslideup.vx v8, v6, a1
; CHECK-NEXT: vmv.v.v v0, v8
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: mv a1, a0
-; CHECK-NEXT: slli a0, a0, 3
+; CHECK-NEXT: slli a0, a0, 1
; CHECK-NEXT: add a1, a1, a0
; CHECK-NEXT: slli a0, a0, 2
+; CHECK-NEXT: add a1, a1, a0
+; CHECK-NEXT: slli a0, a0, 1
; CHECK-NEXT: add a0, a0, a1
; CHECK-NEXT: add sp, sp, a0
; CHECK-NEXT: .cfi_def_cfa sp, 16
@@ -3717,14 +3721,11 @@ define <vscale x 64 x i1> @fcmp_oeq_vv_nxv64f16(<vscale x 64 x half> %va, <vscal
; ZVFH-NEXT: addi sp, sp, -16
; ZVFH-NEXT: .cfi_def_cfa_offset 16
; ZVFH-NEXT: csrr a1, vlenb
-; ZVFH-NEXT: slli a1, a1, 4
-; ZVFH-NEXT: sub sp, sp, a1
-; ZVFH-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
-; ZVFH-NEXT: vmv1r.v v24, v0
-; 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: 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 v8, (a1) # Unknown-size Folded Spill
; ZVFH-NEXT: csrr a3, vlenb
; ZVFH-NEXT: srli a1, a3, 1
@@ -3732,37 +3733,30 @@ define <vscale x 64 x i1> @fcmp_oeq_vv_nxv64f16(<vscale x 64 x half> %va, <vscal
; ZVFH-NEXT: slli a3, a3, 2
; ZVFH-NEXT: add a4, a0, a4
; ZVFH-NEXT: sub a5, a2, a3
-; ZVFH-NEXT: vl8re16.v v8, (a4)
+; ZVFH-NEXT: vl8re16.v v24, (a4)
; ZVFH-NEXT: sltu a4, a2, a5
; ZVFH-NEXT: addi a4, a4, -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: vl8re16.v v8, (a0)
; ZVFH-NEXT: vsetvli a0, zero, e8, m1, ta, ma
-; ZVFH-NEXT: vslidedown.vx v0, v24, a1
+; ZVFH-NEXT: vslidedown.vx v0, v0, a1
; ZVFH-NEXT: and a4, a4, a5
; ZVFH-NEXT: vsetvli zero, a4, e16, m8, ta, ma
-; ZVFH-NEXT: vmfeq.vv v7, v16, v8, v0.t
+; ZVFH-NEXT: vmfeq.vv v6, v16, v24, v0.t
; ZVFH-NEXT: bltu a2, a3, .LBB171_2
; ZVFH-NEXT: # %bb.1:
; ZVFH-NEXT: mv a2, a3
; ZVFH-NEXT: .LBB171_2:
-; ZVFH-NEXT: vmv1r.v v0, v24
-; 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: vmv1r.v v0, v7
; 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: vmfeq.vv v16, v24, v8, v0.t
; ZVFH-NEXT: add a0, a1, a1
; ZVFH-NEXT: vsetvli zero, a0, e8, m1, ta, ma
-; ZVFH-NEXT: vslideup.vx v16, v7, a1
+; ZVFH-NEXT: vslideup.vx v16, v6, a1
; ZVFH-NEXT: vmv.v.v v0, v16
; ZVFH-NEXT: csrr a0, vlenb
-; ZVFH-NEXT: slli a0, a0, 4
+; ZVFH-NEXT: slli a0, a0, 3
; ZVFH-NEXT: add sp, sp, a0
; ZVFH-NEXT: .cfi_def_cfa sp, 16
; ZVFH-NEXT: addi sp, sp, 16
@@ -3775,18 +3769,21 @@ define <vscale x 64 x i1> @fcmp_oeq_vv_nxv64f16(<vscale x 64 x half> %va, <vscal
; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
; ZVFHMIN-NEXT: csrr a1, vlenb
; ZVFHMIN-NEXT: mv a3, a1
-; ZVFHMIN-NEXT: slli a1, a1, 3
+; ZVFHMIN-NEXT: slli a1, a1, 1
; ZVFHMIN-NEXT: add a3, a3, a1
; ZVFHMIN-NEXT: slli a1, a1, 2
+; ZVFHMIN-NEXT: add a3, a3, a1
+; ZVFHMIN-NEXT: slli a1, a1, 1
; ZVFHMIN-NEXT: add a1, a1, a3
; ZVFHMIN-NEXT: sub sp, sp, a1
-; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x29, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 41 * vlenb
+; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x1b, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 27 * vlenb
; ZVFHMIN-NEXT: vmv1r.v v24, v0
+; ZVFHMIN-NEXT: vmv8r.v v0, v16
; ZVFHMIN-NEXT: csrr a1, vlenb
; ZVFHMIN-NEXT: mv a3, a1
-; ZVFHMIN-NEXT: slli a1, a1, 3
-; ZVFHMIN-NEXT: add a3, a3, a1
; ZVFHMIN-NEXT: slli a1, a1, 1
+; ZVFHMIN-NEXT: add a3, a3, a1
+; ZVFHMIN-NEXT: slli a1, a1, 3
; ZVFHMIN-NEXT: add a1, a1, a3
; ZVFHMIN-NEXT: add a1, sp, a1
; ZVFHMIN-NEXT: addi a1, a1, 16
@@ -3797,7 +3794,7 @@ define <vscale x 64 x i1> @fcmp_oeq_vv_nxv64f16(<vscale x 64 x half> %va, <vscal
; ZVFHMIN-NEXT: slli a4, a3, 1
; ZVFHMIN-NEXT: add a1, a0, a1
; ZVFHMIN-NEXT: sub a6, a2, a5
-; ZVFHMIN-NEXT: vl8re16.v v0, (a1)
+; ZVFHMIN-NEXT: vl8re16.v v16, (a1)
; ZVFHMIN-NEXT: sltu a1, a2, a6
; ZVFHMIN-NEXT: addi a1, a1, -1
; ZVFHMIN-NEXT: and a6, a1, a6
@@ -3808,66 +3805,62 @@ define <vscale x 64 x i1> @fcmp_oeq_vv_nxv64f16(<vscale x 64 x half> %va, <vscal
; ZVFHMIN-NEXT: srli a1, a3, 1
; ZVFHMIN-NEXT: srli a3, a3, 2
; ZVFHMIN-NEXT: csrr t0, vlenb
-; ZVFHMIN-NEXT: slli t0, t0, 3
+; ZVFHMIN-NEXT: slli t0, t0, 1
; ZVFHMIN-NEXT: add t0, sp, t0
; ZVFHMIN-NEXT: addi t0, t0, 16
; ZVFHMIN-NEXT: vs1r.v v24, (t0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vsetvli t0, zero, e8, m1, ta, ma
-; ZVFHMIN-NEXT: vslidedown.vx v25, v24, a1
+; ZVFHMIN-NEXT: vslidedown.vx v8, v24, a1
+; ZVFHMIN-NEXT: csrr t0, vlenb
+; ZVFHMIN-NEXT: add t0, sp, t0
+; ZVFHMIN-NEXT: addi t0, t0, 16
+; ZVFHMIN-NEXT: vs1r.v v8, (t0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vsetvli t0, zero, e8, mf2, ta, ma
-; ZVFHMIN-NEXT: vslidedown.vx v24, v25, a3
+; ZVFHMIN-NEXT: vslidedown.vx v8, v8, a3
+; ZVFHMIN-NEXT: addi t0, sp, 16
+; ZVFHMIN-NEXT: vs1r.v v8, (t0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vl8re16.v v8, (a0)
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli t0, a0, 5
-; ZVFHMIN-NEXT: add a0, t0, a0
+; ZVFHMIN-NEXT: mv t0, a0
+; ZVFHMIN-NEXT: slli a0, a0, 1
+; ZVFHMIN-NEXT: add t0, t0, a0
+; ZVFHMIN-NEXT: slli a0, a0, 2
+; ZVFHMIN-NEXT: add a0, a0, t0
; 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 t0, a0, 3
-; ZVFHMIN-NEXT: add a0, t0, a0
-; ZVFHMIN-NEXT: add a0, sp, a0
-; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vs8r.v v0, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
-; 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: vmv8r.v v8, v16
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v20
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli t0, a0, 4
+; ZVFHMIN-NEXT: slli t0, a0, 1
; ZVFHMIN-NEXT: add a0, t0, a0
; 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 v16, v12
-; ZVFHMIN-NEXT: vmv1r.v v0, v24
+; ZVFHMIN-NEXT: vs8r.v v0, (a0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v4
; ZVFHMIN-NEXT: addi a0, sp, 16
-; ZVFHMIN-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vl1r.v v0, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, a7, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vmfeq.vv v6, v16, v8, v0.t
+; ZVFHMIN-NEXT: vmfeq.vv v4, v24, v8, v0.t
; ZVFHMIN-NEXT: bltu a6, a4, .LBB171_2
; ZVFHMIN-NEXT: # %bb.1:
; ZVFHMIN-NEXT: mv a6, a4
; ZVFHMIN-NEXT: .LBB171_2:
-; ZVFHMIN-NEXT: vmv1r.v v0, v25
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a7, a0, 3
-; ZVFHMIN-NEXT: add a0, a7, a0
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vl1r.v v0, (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: slli a7, a0, 4
+; ZVFHMIN-NEXT: slli a7, a0, 1
; ZVFHMIN-NEXT: add a0, a7, a0
; 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 v24, v16
; ZVFHMIN-NEXT: vsetvli zero, a6, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vmfeq.vv v5, v24, v8, v0.t
+; ZVFHMIN-NEXT: vmfeq.vv v6, v24, v8, v0.t
; ZVFHMIN-NEXT: add a0, a3, a3
; ZVFHMIN-NEXT: bltu a2, a5, .LBB171_4
; ZVFHMIN-NEXT: # %bb.3:
@@ -3875,7 +3868,7 @@ define <vscale x 64 x i1> @fcmp_oeq_vv_nxv64f16(<vscale x 64 x half> %va, <vscal
; ZVFHMIN-NEXT: .LBB171_4:
; ZVFHMIN-NEXT: sub a5, a2, a4
; ZVFHMIN-NEXT: csrr a6, vlenb
-; ZVFHMIN-NEXT: slli a6, a6, 3
+; ZVFHMIN-NEXT: slli a6, a6, 1
; ZVFHMIN-NEXT: add a6, sp, a6
; ZVFHMIN-NEXT: addi a6, a6, 16
; ZVFHMIN-NEXT: vl1r.v v7, (a6) # Unknown-size Folded Reload
@@ -3883,51 +3876,54 @@ define <vscale x 64 x i1> @fcmp_oeq_vv_nxv64f16(<vscale x 64 x half> %va, <vscal
; ZVFHMIN-NEXT: vslidedown.vx v0, v7, a3
; ZVFHMIN-NEXT: csrr a6, vlenb
; ZVFHMIN-NEXT: mv a7, a6
-; ZVFHMIN-NEXT: slli a6, a6, 3
-; ZVFHMIN-NEXT: add a7, a7, a6
; ZVFHMIN-NEXT: slli a6, a6, 1
+; ZVFHMIN-NEXT: add a7, a7, a6
+; ZVFHMIN-NEXT: slli a6, a6, 3
; ZVFHMIN-NEXT: add 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: vl8r.v v8, (a6) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli a6, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v20
-; ZVFHMIN-NEXT: csrr a6, vlenb
-; ZVFHMIN-NEXT: slli a7, a6, 4
-; ZVFHMIN-NEXT: add a6, a7, a6
-; ZVFHMIN-NEXT: add a6, sp, a6
-; ZVFHMIN-NEXT: addi a6, a6, 16
-; ZVFHMIN-NEXT: vs8r.v v8, (a6) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
; ZVFHMIN-NEXT: csrr a6, vlenb
-; ZVFHMIN-NEXT: slli a7, a6, 5
-; ZVFHMIN-NEXT: add a6, a7, a6
+; ZVFHMIN-NEXT: mv a7, a6
+; ZVFHMIN-NEXT: slli a6, a6, 1
+; ZVFHMIN-NEXT: add a7, a7, a6
+; ZVFHMIN-NEXT: slli a6, a6, 2
+; ZVFHMIN-NEXT: add 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: vfwcvt.f.f.v v8, v28
+; ZVFHMIN-NEXT: vl8r.v v16, (a6) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v20
; ZVFHMIN-NEXT: sltu a6, a2, a5
; ZVFHMIN-NEXT: addi a6, a6, -1
; ZVFHMIN-NEXT: and a5, a6, a5
-; ZVFHMIN-NEXT: csrr a6, vlenb
-; ZVFHMIN-NEXT: slli a7, a6, 4
-; ZVFHMIN-NEXT: add a6, a7, a6
-; 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 zero, a5, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vmfeq.vv v4, v24, v8, v0.t
+; ZVFHMIN-NEXT: vmfeq.vv v5, v24, v8, v0.t
; ZVFHMIN-NEXT: vsetvli zero, a0, e8, mf2, ta, ma
-; ZVFHMIN-NEXT: vslideup.vx v5, v6, a3
+; ZVFHMIN-NEXT: vslideup.vx v6, v4, a3
; ZVFHMIN-NEXT: bltu a2, a4, .LBB171_6
; ZVFHMIN-NEXT: # %bb.5:
; ZVFHMIN-NEXT: mv a2, a4
; ZVFHMIN-NEXT: .LBB171_6:
; ZVFHMIN-NEXT: vmv1r.v v0, v7
+; ZVFHMIN-NEXT: csrr a4, vlenb
+; ZVFHMIN-NEXT: mv a5, a4
+; ZVFHMIN-NEXT: slli a4, a4, 1
+; ZVFHMIN-NEXT: add a5, a5, a4
+; ZVFHMIN-NEXT: slli a4, a4, 3
+; ZVFHMIN-NEXT: add a4, a4, a5
+; ZVFHMIN-NEXT: add a4, sp, a4
+; ZVFHMIN-NEXT: addi a4, a4, 16
+; ZVFHMIN-NEXT: vl8r.v v8, (a4) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli a4, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v16
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v8
; ZVFHMIN-NEXT: csrr a4, vlenb
-; ZVFHMIN-NEXT: slli a5, a4, 5
-; ZVFHMIN-NEXT: add a4, a5, a4
+; ZVFHMIN-NEXT: mv a5, a4
+; ZVFHMIN-NEXT: slli a4, a4, 1
+; ZVFHMIN-NEXT: add a5, a5, a4
+; ZVFHMIN-NEXT: slli a4, a4, 2
+; ZVFHMIN-NEXT: add a4, a4, a5
; ZVFHMIN-NEXT: add a4, sp, a4
; ZVFHMIN-NEXT: addi a4, a4, 16
; ZVFHMIN-NEXT: vl8r.v v8, (a4) # Unknown-size Folded Reload
@@ -3935,16 +3931,18 @@ define <vscale x 64 x i1> @fcmp_oeq_vv_nxv64f16(<vscale x 64 x half> %va, <vscal
; ZVFHMIN-NEXT: vsetvli zero, a2, e32, m8, ta, ma
; ZVFHMIN-NEXT: vmfeq.vv v8, v24, v16, v0.t
; ZVFHMIN-NEXT: vsetvli zero, a0, e8, mf2, ta, ma
-; ZVFHMIN-NEXT: vslideup.vx v8, v4, a3
+; ZVFHMIN-NEXT: vslideup.vx v8, v5, a3
; ZVFHMIN-NEXT: add a0, a1, a1
; ZVFHMIN-NEXT: vsetvli zero, a0, e8, m1, ta, ma
-; ZVFHMIN-NEXT: vslideup.vx v8, v5, a1
+; ZVFHMIN-NEXT: vslideup.vx v8, v6, a1
; ZVFHMIN-NEXT: vmv.v.v v0, v8
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: mv a1, a0
-; ZVFHMIN-NEXT: slli a0, a0, 3
+; ZVFHMIN-NEXT: slli a0, a0, 1
; ZVFHMIN-NEXT: add a1, a1, a0
; ZVFHMIN-NEXT: slli a0, a0, 2
+; ZVFHMIN-NEXT: add a1, a1, a0
+; ZVFHMIN-NEXT: slli a0, a0, 1
; ZVFHMIN-NEXT: add a0, a0, a1
; ZVFHMIN-NEXT: add sp, sp, a0
; ZVFHMIN-NEXT: .cfi_def_cfa sp, 16
diff --git a/llvm/test/CodeGen/RISCV/rvv/setcc-int-vp.ll b/llvm/test/CodeGen/RISCV/rvv/setcc-int-vp.ll
index e8099c2d08a5f8..d06281eb62a78d 100644
--- a/llvm/test/CodeGen/RISCV/rvv/setcc-int-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/setcc-int-vp.ll
@@ -1089,14 +1089,11 @@ define <vscale x 128 x i1> @icmp_eq_vv_nxv128i8(<vscale x 128 x i8> %va, <vscale
; 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: .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: slli a1, a1, 3
-; CHECK-NEXT: add a1, sp, a1
-; CHECK-NEXT: addi a1, a1, 16
+; CHECK-NEXT: addi a1, sp, 16
; CHECK-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
; CHECK-NEXT: csrr a1, vlenb
; CHECK-NEXT: vsetvli a4, zero, e8, m8, ta, ma
@@ -1104,33 +1101,26 @@ 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 a2, a0, a1
; CHECK-NEXT: sub a4, a3, a1
-; CHECK-NEXT: vl8r.v v8, (a2)
+; CHECK-NEXT: vl8r.v v24, (a2)
; CHECK-NEXT: sltu a2, a3, a4
-; CHECK-NEXT: vl8r.v v24, (a0)
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vl8r.v v8, (a0)
; CHECK-NEXT: addi a2, a2, -1
; CHECK-NEXT: and a2, a2, a4
; CHECK-NEXT: vsetvli zero, a2, e8, m8, ta, ma
-; CHECK-NEXT: vmseq.vv v6, v16, v8, v0.t
+; CHECK-NEXT: vmseq.vv v6, v16, v24, v0.t
; CHECK-NEXT: bltu a3, a1, .LBB96_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a3, a1
; CHECK-NEXT: .LBB96_2:
; 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: 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: vmseq.vv v16, v24, v8, v0.t
; CHECK-NEXT: vmv1r.v v0, v16
; CHECK-NEXT: vmv1r.v v8, v6
; 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: .cfi_def_cfa sp, 16
; CHECK-NEXT: addi sp, sp, 16
@@ -2241,14 +2231,11 @@ define <vscale x 32 x i1> @icmp_eq_vv_nxv32i32(<vscale x 32 x i32> %va, <vscale
; 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: csrr a1, vlenb
; CHECK-NEXT: slli a1, a1, 3
-; CHECK-NEXT: add a1, sp, a1
-; CHECK-NEXT: addi a1, a1, 16
+; 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 v8, (a1) # Unknown-size Folded Spill
; CHECK-NEXT: csrr a3, vlenb
; CHECK-NEXT: srli a1, a3, 2
@@ -2256,37 +2243,30 @@ define <vscale x 32 x i1> @icmp_eq_vv_nxv32i32(<vscale x 32 x i32> %va, <vscale
; CHECK-NEXT: slli a3, a3, 1
; CHECK-NEXT: add a4, a0, a4
; CHECK-NEXT: sub a5, a2, a3
-; CHECK-NEXT: vl8re32.v v8, (a4)
+; CHECK-NEXT: vl8re32.v v24, (a4)
; CHECK-NEXT: sltu a4, a2, a5
; CHECK-NEXT: addi a4, a4, -1
-; CHECK-NEXT: vl8re32.v v0, (a0)
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vs8r.v v0, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vl8re32.v v8, (a0)
; CHECK-NEXT: vsetvli a0, zero, e8, mf2, ta, ma
-; CHECK-NEXT: vslidedown.vx v0, v24, a1
+; CHECK-NEXT: vslidedown.vx v0, v0, a1
; CHECK-NEXT: and a4, a4, a5
; CHECK-NEXT: vsetvli zero, a4, e32, m8, ta, ma
-; CHECK-NEXT: vmseq.vv v7, v16, v8, v0.t
+; CHECK-NEXT: vmseq.vv v6, v16, v24, v0.t
; CHECK-NEXT: bltu a2, a3, .LBB189_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a2, a3
; CHECK-NEXT: .LBB189_2:
-; 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: vmv1r.v v0, v7
; 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: vmseq.vv v16, v24, v8, v0.t
; CHECK-NEXT: add a0, a1, a1
; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, ma
-; CHECK-NEXT: vslideup.vx v16, v7, a1
+; CHECK-NEXT: vslideup.vx v16, v6, a1
; CHECK-NEXT: vmv1r.v v0, v16
; 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: .cfi_def_cfa sp, 16
; CHECK-NEXT: addi sp, sp, 16
diff --git a/llvm/test/CodeGen/RISCV/rvv/strided-vpstore.ll b/llvm/test/CodeGen/RISCV/rvv/strided-vpstore.ll
index 98ec99bcfea33e..a9f10f4ff424fe 100644
--- a/llvm/test/CodeGen/RISCV/rvv/strided-vpstore.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/strided-vpstore.ll
@@ -609,15 +609,7 @@ declare void @llvm.experimental.vp.strided.store.nxv16f64.p0.i32(<vscale x 16 x
define void @strided_store_nxv17f64(<vscale x 17 x double> %v, ptr %ptr, i32 signext %stride, <vscale x 17 x i1> %mask, i32 zeroext %evl) {
; CHECK-LABEL: strided_store_nxv17f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: addi sp, sp, -16
-; CHECK-NEXT: .cfi_def_cfa_offset 16
-; CHECK-NEXT: csrr a4, vlenb
-; CHECK-NEXT: slli a4, a4, 3
-; CHECK-NEXT: sub sp, sp, a4
-; 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 v24, v0
-; CHECK-NEXT: addi a4, sp, 16
-; CHECK-NEXT: vs8r.v v16, (a4) # Unknown-size Folded Spill
+; CHECK-NEXT: vmv1r.v v7, v0
; CHECK-NEXT: csrr a4, vlenb
; CHECK-NEXT: slli a6, a4, 1
; CHECK-NEXT: mv a5, a3
@@ -630,8 +622,8 @@ define void @strided_store_nxv17f64(<vscale x 17 x double> %v, ptr %ptr, i32 sig
; CHECK-NEXT: # %bb.3:
; CHECK-NEXT: mv a7, a4
; CHECK-NEXT: .LBB48_4:
-; CHECK-NEXT: vmv1r.v v0, v24
-; CHECK-NEXT: vl8re64.v v16, (a0)
+; CHECK-NEXT: vmv1r.v v0, v7
+; CHECK-NEXT: vl8re64.v v24, (a0)
; CHECK-NEXT: vsetvli zero, a7, e64, m8, ta, ma
; CHECK-NEXT: vsse64.v v8, (a1), a2, v0.t
; CHECK-NEXT: sub a0, a5, a4
@@ -639,7 +631,7 @@ define void @strided_store_nxv17f64(<vscale x 17 x double> %v, ptr %ptr, i32 sig
; CHECK-NEXT: srli t0, a4, 3
; CHECK-NEXT: sub a6, a3, a6
; CHECK-NEXT: vsetvli t1, zero, e8, mf4, ta, ma
-; CHECK-NEXT: vslidedown.vx v0, v24, t0
+; CHECK-NEXT: vslidedown.vx v0, v7, t0
; CHECK-NEXT: sltu t0, a5, a0
; CHECK-NEXT: add a7, a1, a7
; CHECK-NEXT: sltu a3, a3, a6
@@ -647,10 +639,8 @@ define void @strided_store_nxv17f64(<vscale x 17 x double> %v, ptr %ptr, i32 sig
; CHECK-NEXT: addi a3, a3, -1
; CHECK-NEXT: and t0, t0, a0
; CHECK-NEXT: and a0, a3, a6
-; CHECK-NEXT: addi a3, sp, 16
-; CHECK-NEXT: vl8r.v v8, (a3) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, t0, e64, m8, ta, ma
-; CHECK-NEXT: vsse64.v v8, (a7), a2, v0.t
+; CHECK-NEXT: vsse64.v v16, (a7), a2, v0.t
; CHECK-NEXT: bltu a0, a4, .LBB48_6
; CHECK-NEXT: # %bb.5:
; CHECK-NEXT: mv a0, a4
@@ -658,16 +648,10 @@ define void @strided_store_nxv17f64(<vscale x 17 x double> %v, ptr %ptr, i32 sig
; CHECK-NEXT: mul a3, a5, a2
; CHECK-NEXT: srli a4, a4, 2
; CHECK-NEXT: vsetvli a5, zero, e8, mf2, ta, ma
-; CHECK-NEXT: vslidedown.vx v0, v24, a4
+; CHECK-NEXT: vslidedown.vx v0, v7, a4
; CHECK-NEXT: add a1, a1, a3
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: vsse64.v v16, (a1), a2, v0.t
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add sp, sp, a0
-; CHECK-NEXT: .cfi_def_cfa sp, 16
-; CHECK-NEXT: addi sp, sp, 16
-; CHECK-NEXT: .cfi_def_cfa_offset 0
+; CHECK-NEXT: vsse64.v v24, (a1), a2, v0.t
; CHECK-NEXT: ret
call void @llvm.experimental.vp.strided.store.nxv17f64.p0.i32(<vscale x 17 x double> %v, ptr %ptr, i32 %stride, <vscale x 17 x i1> %mask, i32 %evl)
ret void
diff --git a/llvm/test/CodeGen/RISCV/rvv/vector-deinterleave-load.ll b/llvm/test/CodeGen/RISCV/rvv/vector-deinterleave-load.ll
index 4338d1f61af728..5da63c070921bc 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vector-deinterleave-load.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vector-deinterleave-load.ll
@@ -106,56 +106,43 @@ define {<vscale x 8 x i64>, <vscale x 8 x i64>} @vector_deinterleave_load_nxv8i6
; CHECK-NEXT: addi sp, sp, -16
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: li a2, 24
-; CHECK-NEXT: mul a1, a1, a2
+; CHECK-NEXT: slli a1, a1, 4
; 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: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: vl8re64.v v16, (a0)
+; CHECK-NEXT: vl8re64.v v24, (a0)
; CHECK-NEXT: vsetvli a2, zero, e64, m8, ta, ma
; CHECK-NEXT: vid.v v8
; CHECK-NEXT: slli a1, a1, 3
; CHECK-NEXT: vand.vi v8, v8, 1
; CHECK-NEXT: add a0, a0, a1
-; CHECK-NEXT: vmseq.vi v24, v8, 0
-; 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 v8, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: vmnot.m v6, v24
-; CHECK-NEXT: vcompress.vm v8, v16, v24
-; CHECK-NEXT: vmv1r.v v13, v24
-; CHECK-NEXT: vcompress.vm v24, v16, v6
-; CHECK-NEXT: vmv1r.v v12, v6
-; 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: vcompress.vm v0, v16, v13
+; CHECK-NEXT: vmseq.vi v16, v8, 0
+; CHECK-NEXT: vl8re64.v v0, (a0)
+; CHECK-NEXT: vmnot.m v17, v16
+; CHECK-NEXT: vcompress.vm v8, v24, v16
+; CHECK-NEXT: vmv1r.v v12, v16
+; CHECK-NEXT: vmv1r.v v13, v17
+; CHECK-NEXT: vcompress.vm v16, v24, v13
+; CHECK-NEXT: vcompress.vm v24, v0, v12
; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vs8r.v v0, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: vcompress.vm v0, v16, v12
+; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vcompress.vm v24, v0, v13
; 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 v0, (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: vmv4r.v v12, v16
+; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vmv4r.v v12, 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: vmv4r.v v28, v16
-; CHECK-NEXT: vmv8r.v v16, v24
+; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vmv4r.v v20, v24
; 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 sp, sp, a0
; CHECK-NEXT: .cfi_def_cfa sp, 16
; CHECK-NEXT: addi sp, sp, 16
diff --git a/llvm/test/CodeGen/RISCV/rvv/vector-deinterleave.ll b/llvm/test/CodeGen/RISCV/rvv/vector-deinterleave.ll
index 99743066c79a82..4933979b7ec4c7 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vector-deinterleave.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vector-deinterleave.ll
@@ -182,50 +182,39 @@ define {<vscale x 8 x i64>, <vscale x 8 x i64>} @vector_deinterleave_nxv8i64_nxv
; CHECK-NEXT: addi sp, sp, -16
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a1, 24
-; CHECK-NEXT: mul a0, a0, a1
-; CHECK-NEXT: sub sp, sp, a0
-; 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: 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: 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: vmv8r.v v24, v8
; CHECK-NEXT: vsetvli a0, zero, e64, m8, ta, ma
-; CHECK-NEXT: vid.v v16
-; CHECK-NEXT: vand.vi v24, v16, 1
-; CHECK-NEXT: vmseq.vi v16, v24, 0
-; CHECK-NEXT: vcompress.vm v24, v8, v16
+; CHECK-NEXT: vid.v v8
+; CHECK-NEXT: vand.vi v8, v8, 1
+; CHECK-NEXT: vmseq.vi v7, v8, 0
+; CHECK-NEXT: vcompress.vm v8, v24, v7
+; CHECK-NEXT: vmnot.m v12, v7
+; CHECK-NEXT: vmv1r.v v13, v7
+; CHECK-NEXT: vcompress.vm v0, v24, v12
+; CHECK-NEXT: vcompress.vm v24, v16, v13
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: vmnot.m v17, v16
-; CHECK-NEXT: vcompress.vm v0, v8, v17
-; 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: vcompress.vm v24, v8, v16
+; CHECK-NEXT: vcompress.vm v24, v16, v12
; 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: vcompress.vm v24, v8, v17
+; CHECK-NEXT: addi a0, sp, 16
+; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vmv4r.v v12, 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: 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: vmv4r.v v20, v8
-; CHECK-NEXT: vmv4r.v v4, v24
-; CHECK-NEXT: vmv8r.v v8, v16
+; CHECK-NEXT: vmv4r.v v4, v16
; CHECK-NEXT: vmv8r.v v16, v0
; 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 sp, sp, a0
; CHECK-NEXT: .cfi_def_cfa sp, 16
; CHECK-NEXT: addi sp, sp, 16
@@ -423,50 +412,39 @@ define {<vscale x 8 x double>, <vscale x 8 x double>} @vector_deinterleave_nxv8f
; CHECK-NEXT: addi sp, sp, -16
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: li a1, 24
-; CHECK-NEXT: mul a0, a0, a1
-; CHECK-NEXT: sub sp, sp, a0
-; 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: 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: 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: vmv8r.v v24, v8
; CHECK-NEXT: vsetvli a0, zero, e64, m8, ta, ma
-; CHECK-NEXT: vid.v v16
-; CHECK-NEXT: vand.vi v24, v16, 1
-; CHECK-NEXT: vmseq.vi v16, v24, 0
-; CHECK-NEXT: vcompress.vm v24, v8, v16
+; CHECK-NEXT: vid.v v8
+; CHECK-NEXT: vand.vi v8, v8, 1
+; CHECK-NEXT: vmseq.vi v7, v8, 0
+; CHECK-NEXT: vcompress.vm v8, v24, v7
+; CHECK-NEXT: vmnot.m v12, v7
+; CHECK-NEXT: vmv1r.v v13, v7
+; CHECK-NEXT: vcompress.vm v0, v24, v12
+; CHECK-NEXT: vcompress.vm v24, v16, v13
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: vmnot.m v17, v16
-; CHECK-NEXT: vcompress.vm v0, v8, v17
-; 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: vcompress.vm v24, v8, v16
+; CHECK-NEXT: vcompress.vm v24, v16, v12
; 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: vcompress.vm v24, v8, v17
+; CHECK-NEXT: addi a0, sp, 16
+; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vmv4r.v v12, 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: 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: vmv4r.v v20, v8
-; CHECK-NEXT: vmv4r.v v4, v24
-; CHECK-NEXT: vmv8r.v v8, v16
+; CHECK-NEXT: vmv4r.v v4, v16
; CHECK-NEXT: vmv8r.v v16, v0
; 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 sp, sp, a0
; CHECK-NEXT: .cfi_def_cfa sp, 16
; CHECK-NEXT: addi sp, sp, 16
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfadd-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vfadd-vp.ll
index 1953cfd2a0169f..676f405ec82710 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfadd-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfadd-vp.ll
@@ -514,75 +514,57 @@ define <vscale x 32 x bfloat> @vfadd_vf_nxv32bf16(<vscale x 32 x bfloat> %va, bf
; CHECK-NEXT: addi sp, sp, -16
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a2, a1, 4
+; CHECK-NEXT: slli a2, a1, 3
; CHECK-NEXT: add a1, a2, a1
; CHECK-NEXT: sub sp, sp, a1
-; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x11, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 17 * vlenb
-; CHECK-NEXT: vmv8r.v v16, v8
+; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x09, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 9 * vlenb
+; CHECK-NEXT: vmv1r.v v24, v0
; CHECK-NEXT: fmv.x.h a1, fa0
; CHECK-NEXT: csrr a2, vlenb
-; CHECK-NEXT: addi a3, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
; CHECK-NEXT: vsetvli a3, zero, e16, m4, ta, ma
-; CHECK-NEXT: vfwcvtbf16.f.f.v v8, v20
+; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v12
; CHECK-NEXT: vsetvli a3, zero, e16, m8, ta, ma
-; CHECK-NEXT: vmv.v.x v16, a1
-; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a3, a1, 3
-; CHECK-NEXT: add a1, a3, a1
-; CHECK-NEXT: add a1, sp, a1
-; CHECK-NEXT: addi a1, a1, 16
-; CHECK-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
+; CHECK-NEXT: vmv.v.x v0, a1
; CHECK-NEXT: slli a1, a2, 1
; CHECK-NEXT: srli a2, a2, 2
; CHECK-NEXT: sub a3, a0, a1
-; 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: vs1r.v v0, (a4) # Unknown-size Folded Spill
+; CHECK-NEXT: addi a4, sp, 16
+; CHECK-NEXT: vs1r.v v24, (a4) # Unknown-size Folded Spill
; CHECK-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
-; CHECK-NEXT: vslidedown.vx v0, v0, a2
+; CHECK-NEXT: vslidedown.vx v12, v24, a2
; CHECK-NEXT: sltu a2, a0, a3
; CHECK-NEXT: addi a2, a2, -1
; CHECK-NEXT: and a2, a2, a3
+; CHECK-NEXT: vmv4r.v v24, v0
; CHECK-NEXT: csrr a3, vlenb
-; CHECK-NEXT: slli a4, a3, 3
-; CHECK-NEXT: add a3, a4, a3
; CHECK-NEXT: add a3, sp, a3
; CHECK-NEXT: addi a3, a3, 16
-; CHECK-NEXT: vl8r.v v24, (a3) # Unknown-size Folded Reload
+; CHECK-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
; CHECK-NEXT: vsetvli a3, zero, e16, m4, ta, ma
-; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v28
+; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v4
+; CHECK-NEXT: vmv1r.v v0, v12
; CHECK-NEXT: vsetvli zero, a2, e32, m8, ta, ma
-; CHECK-NEXT: vfadd.vv v16, v8, v16, v0.t
+; CHECK-NEXT: vfadd.vv v16, v16, v24, v0.t
; CHECK-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; CHECK-NEXT: vfncvtbf16.f.f.w v12, v16
; CHECK-NEXT: bltu a0, a1, .LBB24_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB24_2:
-; CHECK-NEXT: addi a1, sp, 16
-; CHECK-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
-; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v24
+; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v8
; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a2, a1, 3
-; CHECK-NEXT: add a1, a2, a1
; CHECK-NEXT: add a1, sp, a1
; CHECK-NEXT: addi a1, a1, 16
; CHECK-NEXT: vl8r.v v0, (a1) # Unknown-size Folded Reload
; CHECK-NEXT: vfwcvtbf16.f.f.v v24, 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: addi a1, sp, 16
; CHECK-NEXT: vl1r.v v0, (a1) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; CHECK-NEXT: vfadd.vv v16, v16, v24, v0.t
; CHECK-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; CHECK-NEXT: vfncvtbf16.f.f.w v8, v16
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a1, a0, 4
+; CHECK-NEXT: slli a1, a0, 3
; CHECK-NEXT: add a0, a1, a0
; CHECK-NEXT: add sp, sp, a0
; CHECK-NEXT: .cfi_def_cfa sp, 16
@@ -601,55 +583,41 @@ define <vscale x 32 x bfloat> @vfadd_vf_nxv32bf16_unmasked(<vscale x 32 x bfloat
; 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: vmv8r.v v16, v8
+; 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: fmv.x.h a1, fa0
; CHECK-NEXT: csrr a2, vlenb
; CHECK-NEXT: vsetvli a3, zero, e8, m4, ta, ma
-; CHECK-NEXT: vmset.m v7
-; CHECK-NEXT: addi a3, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
+; CHECK-NEXT: vmset.m v16
; CHECK-NEXT: vsetvli a3, zero, e16, m4, ta, ma
-; CHECK-NEXT: vfwcvtbf16.f.f.v v8, v20
+; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v12
; CHECK-NEXT: vsetvli a3, zero, e16, m8, ta, ma
-; CHECK-NEXT: vmv.v.x v16, 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 v16, (a1) # Unknown-size Folded Spill
+; CHECK-NEXT: vmv.v.x v0, a1
; CHECK-NEXT: slli a1, a2, 1
; CHECK-NEXT: srli a2, a2, 2
; CHECK-NEXT: sub a3, a0, a1
; CHECK-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
-; CHECK-NEXT: vslidedown.vx v0, v7, a2
+; CHECK-NEXT: vslidedown.vx v12, v16, a2
; CHECK-NEXT: sltu a2, a0, a3
; CHECK-NEXT: addi a2, a2, -1
; CHECK-NEXT: and a2, a2, a3
-; CHECK-NEXT: csrr a3, vlenb
-; CHECK-NEXT: slli a3, a3, 3
-; CHECK-NEXT: add a3, sp, a3
-; CHECK-NEXT: addi a3, a3, 16
-; CHECK-NEXT: vl8r.v v24, (a3) # Unknown-size Folded Reload
+; CHECK-NEXT: vmv4r.v v16, v0
+; CHECK-NEXT: addi a3, sp, 16
+; CHECK-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
; CHECK-NEXT: vsetvli a3, zero, e16, m4, ta, ma
-; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v28
+; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v4
+; CHECK-NEXT: vmv1r.v v0, v12
; CHECK-NEXT: vsetvli zero, a2, e32, m8, ta, ma
-; CHECK-NEXT: vfadd.vv v16, v8, v16, v0.t
+; CHECK-NEXT: vfadd.vv v16, v24, v16, v0.t
; CHECK-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; CHECK-NEXT: vfncvtbf16.f.f.w v12, v16
; CHECK-NEXT: bltu a0, a1, .LBB25_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB25_2:
+; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v8
; CHECK-NEXT: addi a1, sp, 16
-; CHECK-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
-; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v24
-; 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 v0, (a1) # Unknown-size Folded Reload
; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
@@ -657,7 +625,7 @@ define <vscale x 32 x bfloat> @vfadd_vf_nxv32bf16_unmasked(<vscale x 32 x bfloat
; CHECK-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; CHECK-NEXT: vfncvtbf16.f.f.w v8, v16
; 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: .cfi_def_cfa sp, 16
; CHECK-NEXT: addi sp, sp, 16
@@ -1320,75 +1288,57 @@ define <vscale x 32 x half> @vfadd_vf_nxv32f16(<vscale x 32 x half> %va, half %b
; ZVFHMIN-NEXT: addi sp, sp, -16
; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a2, a1, 4
+; ZVFHMIN-NEXT: slli a2, a1, 3
; ZVFHMIN-NEXT: add a1, a2, a1
; ZVFHMIN-NEXT: sub sp, sp, a1
-; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x11, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 17 * vlenb
-; ZVFHMIN-NEXT: vmv8r.v v16, v8
+; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x09, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 9 * vlenb
+; ZVFHMIN-NEXT: vmv1r.v v24, v0
; ZVFHMIN-NEXT: fmv.x.h a1, fa0
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: addi a3, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v20
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m8, ta, ma
-; ZVFHMIN-NEXT: vmv.v.x v16, a1
-; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a3, a1, 3
-; ZVFHMIN-NEXT: add a1, a3, a1
-; ZVFHMIN-NEXT: add a1, sp, a1
-; ZVFHMIN-NEXT: addi a1, a1, 16
-; ZVFHMIN-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vmv.v.x v0, a1
; ZVFHMIN-NEXT: slli a1, a2, 1
; ZVFHMIN-NEXT: srli a2, a2, 2
; ZVFHMIN-NEXT: sub a3, a0, a1
-; ZVFHMIN-NEXT: csrr a4, vlenb
-; ZVFHMIN-NEXT: slli a4, a4, 3
-; ZVFHMIN-NEXT: add a4, sp, a4
-; ZVFHMIN-NEXT: addi a4, a4, 16
-; ZVFHMIN-NEXT: vs1r.v v0, (a4) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: addi a4, sp, 16
+; ZVFHMIN-NEXT: vs1r.v v24, (a4) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
-; ZVFHMIN-NEXT: vslidedown.vx v0, v0, a2
+; ZVFHMIN-NEXT: vslidedown.vx v12, v24, a2
; ZVFHMIN-NEXT: sltu a2, a0, a3
; ZVFHMIN-NEXT: addi a2, a2, -1
; ZVFHMIN-NEXT: and a2, a2, a3
+; ZVFHMIN-NEXT: vmv4r.v v24, v0
; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: slli a4, a3, 3
-; ZVFHMIN-NEXT: add a3, a4, a3
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
-; ZVFHMIN-NEXT: vl8r.v v24, (a3) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v28
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v4
+; ZVFHMIN-NEXT: vmv1r.v v0, v12
; ZVFHMIN-NEXT: vsetvli zero, a2, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfadd.vv v16, v8, v16, v0.t
+; ZVFHMIN-NEXT: vfadd.vv v16, v16, v24, v0.t
; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v12, v16
; ZVFHMIN-NEXT: bltu a0, a1, .LBB50_2
; ZVFHMIN-NEXT: # %bb.1:
; ZVFHMIN-NEXT: mv a0, a1
; ZVFHMIN-NEXT: .LBB50_2:
-; ZVFHMIN-NEXT: addi a1, sp, 16
-; ZVFHMIN-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v24
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a2, a1, 3
-; ZVFHMIN-NEXT: add a1, a2, a1
; ZVFHMIN-NEXT: add a1, sp, a1
; ZVFHMIN-NEXT: addi a1, a1, 16
; ZVFHMIN-NEXT: vl8r.v v0, (a1) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v0
-; 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: addi a1, sp, 16
; ZVFHMIN-NEXT: vl1r.v v0, (a1) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; 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 a1, a0, 4
+; ZVFHMIN-NEXT: slli a1, a0, 3
; ZVFHMIN-NEXT: add a0, a1, a0
; ZVFHMIN-NEXT: add sp, sp, a0
; ZVFHMIN-NEXT: .cfi_def_cfa sp, 16
@@ -1413,55 +1363,41 @@ define <vscale x 32 x half> @vfadd_vf_nxv32f16_unmasked(<vscale x 32 x half> %va
; ZVFHMIN-NEXT: addi sp, sp, -16
; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a1, a1, 4
+; ZVFHMIN-NEXT: slli a1, a1, 3
; ZVFHMIN-NEXT: sub sp, sp, a1
-; 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 v16, v8
+; 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: fmv.x.h a1, fa0
; ZVFHMIN-NEXT: csrr a2, vlenb
; ZVFHMIN-NEXT: vsetvli a3, zero, e8, m4, ta, ma
-; ZVFHMIN-NEXT: vmset.m v7
-; ZVFHMIN-NEXT: addi a3, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vmset.m v16
; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v20
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m8, ta, ma
-; ZVFHMIN-NEXT: vmv.v.x v16, a1
-; 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: vmv.v.x v0, a1
; ZVFHMIN-NEXT: slli a1, a2, 1
; ZVFHMIN-NEXT: srli a2, a2, 2
; ZVFHMIN-NEXT: sub a3, a0, a1
; ZVFHMIN-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
-; ZVFHMIN-NEXT: vslidedown.vx v0, v7, a2
+; ZVFHMIN-NEXT: vslidedown.vx v12, v16, a2
; ZVFHMIN-NEXT: sltu a2, a0, a3
; ZVFHMIN-NEXT: addi a2, a2, -1
; ZVFHMIN-NEXT: and a2, a2, a3
-; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: slli a3, a3, 3
-; ZVFHMIN-NEXT: add a3, sp, a3
-; ZVFHMIN-NEXT: addi a3, a3, 16
-; ZVFHMIN-NEXT: vl8r.v v24, (a3) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vmv4r.v v16, v0
+; ZVFHMIN-NEXT: addi a3, sp, 16
+; ZVFHMIN-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v28
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v4
+; ZVFHMIN-NEXT: vmv1r.v v0, v12
; ZVFHMIN-NEXT: vsetvli zero, a2, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfadd.vv v16, v8, v16, v0.t
+; ZVFHMIN-NEXT: vfadd.vv v16, v24, v16, v0.t
; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v12, v16
; ZVFHMIN-NEXT: bltu a0, a1, .LBB51_2
; ZVFHMIN-NEXT: # %bb.1:
; ZVFHMIN-NEXT: mv a0, a1
; ZVFHMIN-NEXT: .LBB51_2:
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8
; ZVFHMIN-NEXT: addi a1, sp, 16
-; ZVFHMIN-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v24
-; 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: vl8r.v v0, (a1) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v0
; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
@@ -1469,7 +1405,7 @@ define <vscale x 32 x half> @vfadd_vf_nxv32f16_unmasked(<vscale x 32 x half> %va
; 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, 4
+; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: add sp, sp, a0
; ZVFHMIN-NEXT: .cfi_def_cfa sp, 16
; ZVFHMIN-NEXT: addi sp, sp, 16
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfdiv-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vfdiv-vp.ll
index ccd286b7ee5fd3..740fc453a4f96c 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfdiv-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfdiv-vp.ll
@@ -476,75 +476,57 @@ define <vscale x 32 x bfloat> @vfdiv_vf_nxv32bf16(<vscale x 32 x bfloat> %va, bf
; CHECK-NEXT: addi sp, sp, -16
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a2, a1, 4
+; CHECK-NEXT: slli a2, a1, 3
; CHECK-NEXT: add a1, a2, a1
; CHECK-NEXT: sub sp, sp, a1
-; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x11, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 17 * vlenb
-; CHECK-NEXT: vmv8r.v v16, v8
+; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x09, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 9 * vlenb
+; CHECK-NEXT: vmv1r.v v24, v0
; CHECK-NEXT: fmv.x.h a1, fa0
; CHECK-NEXT: csrr a2, vlenb
-; CHECK-NEXT: addi a3, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
; CHECK-NEXT: vsetvli a3, zero, e16, m4, ta, ma
-; CHECK-NEXT: vfwcvtbf16.f.f.v v8, v20
+; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v12
; CHECK-NEXT: vsetvli a3, zero, e16, m8, ta, ma
-; CHECK-NEXT: vmv.v.x v16, a1
-; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a3, a1, 3
-; CHECK-NEXT: add a1, a3, a1
-; CHECK-NEXT: add a1, sp, a1
-; CHECK-NEXT: addi a1, a1, 16
-; CHECK-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
+; CHECK-NEXT: vmv.v.x v0, a1
; CHECK-NEXT: slli a1, a2, 1
; CHECK-NEXT: srli a2, a2, 2
; CHECK-NEXT: sub a3, a0, a1
-; 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: vs1r.v v0, (a4) # Unknown-size Folded Spill
+; CHECK-NEXT: addi a4, sp, 16
+; CHECK-NEXT: vs1r.v v24, (a4) # Unknown-size Folded Spill
; CHECK-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
-; CHECK-NEXT: vslidedown.vx v0, v0, a2
+; CHECK-NEXT: vslidedown.vx v12, v24, a2
; CHECK-NEXT: sltu a2, a0, a3
; CHECK-NEXT: addi a2, a2, -1
; CHECK-NEXT: and a2, a2, a3
+; CHECK-NEXT: vmv4r.v v24, v0
; CHECK-NEXT: csrr a3, vlenb
-; CHECK-NEXT: slli a4, a3, 3
-; CHECK-NEXT: add a3, a4, a3
; CHECK-NEXT: add a3, sp, a3
; CHECK-NEXT: addi a3, a3, 16
-; CHECK-NEXT: vl8r.v v24, (a3) # Unknown-size Folded Reload
+; CHECK-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
; CHECK-NEXT: vsetvli a3, zero, e16, m4, ta, ma
-; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v28
+; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v4
+; CHECK-NEXT: vmv1r.v v0, v12
; CHECK-NEXT: vsetvli zero, a2, e32, m8, ta, ma
-; CHECK-NEXT: vfdiv.vv v16, v8, v16, v0.t
+; CHECK-NEXT: vfdiv.vv v16, v16, v24, v0.t
; CHECK-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; CHECK-NEXT: vfncvtbf16.f.f.w v12, v16
; CHECK-NEXT: bltu a0, a1, .LBB22_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB22_2:
-; CHECK-NEXT: addi a1, sp, 16
-; CHECK-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
-; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v24
+; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v8
; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a2, a1, 3
-; CHECK-NEXT: add a1, a2, a1
; CHECK-NEXT: add a1, sp, a1
; CHECK-NEXT: addi a1, a1, 16
; CHECK-NEXT: vl8r.v v0, (a1) # Unknown-size Folded Reload
; CHECK-NEXT: vfwcvtbf16.f.f.v v24, 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: addi a1, sp, 16
; CHECK-NEXT: vl1r.v v0, (a1) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; CHECK-NEXT: vfdiv.vv v16, v16, v24, v0.t
; CHECK-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; CHECK-NEXT: vfncvtbf16.f.f.w v8, v16
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a1, a0, 4
+; CHECK-NEXT: slli a1, a0, 3
; CHECK-NEXT: add a0, a1, a0
; CHECK-NEXT: add sp, sp, a0
; CHECK-NEXT: .cfi_def_cfa sp, 16
@@ -563,55 +545,41 @@ define <vscale x 32 x bfloat> @vfdiv_vf_nxv32bf16_unmasked(<vscale x 32 x bfloat
; 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: vmv8r.v v16, v8
+; 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: fmv.x.h a1, fa0
; CHECK-NEXT: csrr a2, vlenb
; CHECK-NEXT: vsetvli a3, zero, e8, m4, ta, ma
-; CHECK-NEXT: vmset.m v7
-; CHECK-NEXT: addi a3, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
+; CHECK-NEXT: vmset.m v16
; CHECK-NEXT: vsetvli a3, zero, e16, m4, ta, ma
-; CHECK-NEXT: vfwcvtbf16.f.f.v v8, v20
+; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v12
; CHECK-NEXT: vsetvli a3, zero, e16, m8, ta, ma
-; CHECK-NEXT: vmv.v.x v16, 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 v16, (a1) # Unknown-size Folded Spill
+; CHECK-NEXT: vmv.v.x v0, a1
; CHECK-NEXT: slli a1, a2, 1
; CHECK-NEXT: srli a2, a2, 2
; CHECK-NEXT: sub a3, a0, a1
; CHECK-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
-; CHECK-NEXT: vslidedown.vx v0, v7, a2
+; CHECK-NEXT: vslidedown.vx v12, v16, a2
; CHECK-NEXT: sltu a2, a0, a3
; CHECK-NEXT: addi a2, a2, -1
; CHECK-NEXT: and a2, a2, a3
-; CHECK-NEXT: csrr a3, vlenb
-; CHECK-NEXT: slli a3, a3, 3
-; CHECK-NEXT: add a3, sp, a3
-; CHECK-NEXT: addi a3, a3, 16
-; CHECK-NEXT: vl8r.v v24, (a3) # Unknown-size Folded Reload
+; CHECK-NEXT: vmv4r.v v16, v0
+; CHECK-NEXT: addi a3, sp, 16
+; CHECK-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
; CHECK-NEXT: vsetvli a3, zero, e16, m4, ta, ma
-; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v28
+; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v4
+; CHECK-NEXT: vmv1r.v v0, v12
; CHECK-NEXT: vsetvli zero, a2, e32, m8, ta, ma
-; CHECK-NEXT: vfdiv.vv v16, v8, v16, v0.t
+; CHECK-NEXT: vfdiv.vv v16, v24, v16, v0.t
; CHECK-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; CHECK-NEXT: vfncvtbf16.f.f.w v12, v16
; CHECK-NEXT: bltu a0, a1, .LBB23_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB23_2:
+; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v8
; CHECK-NEXT: addi a1, sp, 16
-; CHECK-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
-; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v24
-; 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 v0, (a1) # Unknown-size Folded Reload
; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
@@ -619,7 +587,7 @@ define <vscale x 32 x bfloat> @vfdiv_vf_nxv32bf16_unmasked(<vscale x 32 x bfloat
; CHECK-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; CHECK-NEXT: vfncvtbf16.f.f.w v8, v16
; 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: .cfi_def_cfa sp, 16
; CHECK-NEXT: addi sp, sp, 16
@@ -1232,75 +1200,57 @@ define <vscale x 32 x half> @vfdiv_vf_nxv32f16(<vscale x 32 x half> %va, half %b
; ZVFHMIN-NEXT: addi sp, sp, -16
; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a2, a1, 4
+; ZVFHMIN-NEXT: slli a2, a1, 3
; ZVFHMIN-NEXT: add a1, a2, a1
; ZVFHMIN-NEXT: sub sp, sp, a1
-; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x11, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 17 * vlenb
-; ZVFHMIN-NEXT: vmv8r.v v16, v8
+; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x09, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 9 * vlenb
+; ZVFHMIN-NEXT: vmv1r.v v24, v0
; ZVFHMIN-NEXT: fmv.x.h a1, fa0
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: addi a3, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v20
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m8, ta, ma
-; ZVFHMIN-NEXT: vmv.v.x v16, a1
-; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a3, a1, 3
-; ZVFHMIN-NEXT: add a1, a3, a1
-; ZVFHMIN-NEXT: add a1, sp, a1
-; ZVFHMIN-NEXT: addi a1, a1, 16
-; ZVFHMIN-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vmv.v.x v0, a1
; ZVFHMIN-NEXT: slli a1, a2, 1
; ZVFHMIN-NEXT: srli a2, a2, 2
; ZVFHMIN-NEXT: sub a3, a0, a1
-; ZVFHMIN-NEXT: csrr a4, vlenb
-; ZVFHMIN-NEXT: slli a4, a4, 3
-; ZVFHMIN-NEXT: add a4, sp, a4
-; ZVFHMIN-NEXT: addi a4, a4, 16
-; ZVFHMIN-NEXT: vs1r.v v0, (a4) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: addi a4, sp, 16
+; ZVFHMIN-NEXT: vs1r.v v24, (a4) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
-; ZVFHMIN-NEXT: vslidedown.vx v0, v0, a2
+; ZVFHMIN-NEXT: vslidedown.vx v12, v24, a2
; ZVFHMIN-NEXT: sltu a2, a0, a3
; ZVFHMIN-NEXT: addi a2, a2, -1
; ZVFHMIN-NEXT: and a2, a2, a3
+; ZVFHMIN-NEXT: vmv4r.v v24, v0
; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: slli a4, a3, 3
-; ZVFHMIN-NEXT: add a3, a4, a3
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
-; ZVFHMIN-NEXT: vl8r.v v24, (a3) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v28
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v4
+; ZVFHMIN-NEXT: vmv1r.v v0, v12
; ZVFHMIN-NEXT: vsetvli zero, a2, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfdiv.vv v16, v8, v16, v0.t
+; ZVFHMIN-NEXT: vfdiv.vv v16, v16, v24, v0.t
; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v12, v16
; ZVFHMIN-NEXT: bltu a0, a1, .LBB46_2
; ZVFHMIN-NEXT: # %bb.1:
; ZVFHMIN-NEXT: mv a0, a1
; ZVFHMIN-NEXT: .LBB46_2:
-; ZVFHMIN-NEXT: addi a1, sp, 16
-; ZVFHMIN-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v24
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a2, a1, 3
-; ZVFHMIN-NEXT: add a1, a2, a1
; ZVFHMIN-NEXT: add a1, sp, a1
; ZVFHMIN-NEXT: addi a1, a1, 16
; ZVFHMIN-NEXT: vl8r.v v0, (a1) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v0
-; 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: addi a1, sp, 16
; ZVFHMIN-NEXT: vl1r.v v0, (a1) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; 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 a1, a0, 4
+; ZVFHMIN-NEXT: slli a1, a0, 3
; ZVFHMIN-NEXT: add a0, a1, a0
; ZVFHMIN-NEXT: add sp, sp, a0
; ZVFHMIN-NEXT: .cfi_def_cfa sp, 16
@@ -1325,55 +1275,41 @@ define <vscale x 32 x half> @vfdiv_vf_nxv32f16_unmasked(<vscale x 32 x half> %va
; ZVFHMIN-NEXT: addi sp, sp, -16
; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a1, a1, 4
+; ZVFHMIN-NEXT: slli a1, a1, 3
; ZVFHMIN-NEXT: sub sp, sp, a1
-; 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 v16, v8
+; 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: fmv.x.h a1, fa0
; ZVFHMIN-NEXT: csrr a2, vlenb
; ZVFHMIN-NEXT: vsetvli a3, zero, e8, m4, ta, ma
-; ZVFHMIN-NEXT: vmset.m v7
-; ZVFHMIN-NEXT: addi a3, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vmset.m v16
; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v20
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m8, ta, ma
-; ZVFHMIN-NEXT: vmv.v.x v16, a1
-; 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: vmv.v.x v0, a1
; ZVFHMIN-NEXT: slli a1, a2, 1
; ZVFHMIN-NEXT: srli a2, a2, 2
; ZVFHMIN-NEXT: sub a3, a0, a1
; ZVFHMIN-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
-; ZVFHMIN-NEXT: vslidedown.vx v0, v7, a2
+; ZVFHMIN-NEXT: vslidedown.vx v12, v16, a2
; ZVFHMIN-NEXT: sltu a2, a0, a3
; ZVFHMIN-NEXT: addi a2, a2, -1
; ZVFHMIN-NEXT: and a2, a2, a3
-; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: slli a3, a3, 3
-; ZVFHMIN-NEXT: add a3, sp, a3
-; ZVFHMIN-NEXT: addi a3, a3, 16
-; ZVFHMIN-NEXT: vl8r.v v24, (a3) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vmv4r.v v16, v0
+; ZVFHMIN-NEXT: addi a3, sp, 16
+; ZVFHMIN-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v28
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v4
+; ZVFHMIN-NEXT: vmv1r.v v0, v12
; ZVFHMIN-NEXT: vsetvli zero, a2, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfdiv.vv v16, v8, v16, v0.t
+; ZVFHMIN-NEXT: vfdiv.vv v16, v24, v16, v0.t
; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v12, v16
; ZVFHMIN-NEXT: bltu a0, a1, .LBB47_2
; ZVFHMIN-NEXT: # %bb.1:
; ZVFHMIN-NEXT: mv a0, a1
; ZVFHMIN-NEXT: .LBB47_2:
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8
; ZVFHMIN-NEXT: addi a1, sp, 16
-; ZVFHMIN-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v24
-; 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: vl8r.v v0, (a1) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v0
; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
@@ -1381,7 +1317,7 @@ define <vscale x 32 x half> @vfdiv_vf_nxv32f16_unmasked(<vscale x 32 x half> %va
; 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, 4
+; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: add sp, sp, a0
; ZVFHMIN-NEXT: .cfi_def_cfa sp, 16
; ZVFHMIN-NEXT: addi sp, sp, 16
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfma-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vfma-vp.ll
index fd518d9be786de..7d3eefd3a29d52 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfma-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfma-vp.ll
@@ -473,29 +473,16 @@ declare <vscale x 16 x bfloat> @llvm.vp.fma.nxv16bf16(<vscale x 16 x bfloat>, <v
define <vscale x 16 x bfloat> @vfma_vv_nxv16bf16(<vscale x 16 x bfloat> %va, <vscale x 16 x bfloat> %b, <vscale x 16 x bfloat> %c, <vscale x 16 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfma_vv_nxv16bf16:
; 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: vmv4r.v v4, v12
+; CHECK-NEXT: vmv4r.v v20, v8
; CHECK-NEXT: vsetvli a1, zero, e16, m4, ta, ma
; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v16
-; CHECK-NEXT: addi a1, sp, 16
-; CHECK-NEXT: vs8r.v v24, (a1) # Unknown-size Folded Spill
-; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v8
-; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v12
-; CHECK-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
+; CHECK-NEXT: vfwcvtbf16.f.f.v v8, v20
+; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v4
; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
-; CHECK-NEXT: vfmadd.vv v16, v24, v8, v0.t
+; CHECK-NEXT: vfmadd.vv v16, v8, v24, v0.t
; CHECK-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; CHECK-NEXT: vfncvtbf16.f.f.w v8, v16
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add sp, sp, a0
-; CHECK-NEXT: .cfi_def_cfa sp, 16
-; CHECK-NEXT: addi sp, sp, 16
-; CHECK-NEXT: .cfi_def_cfa_offset 0
; CHECK-NEXT: ret
%v = call <vscale x 16 x bfloat> @llvm.vp.fma.nxv16bf16(<vscale x 16 x bfloat> %va, <vscale x 16 x bfloat> %b, <vscale x 16 x bfloat> %c, <vscale x 16 x i1> %m, i32 %evl)
ret <vscale x 16 x bfloat> %v
@@ -520,32 +507,18 @@ define <vscale x 16 x bfloat> @vfma_vv_nxv16bf16_unmasked(<vscale x 16 x bfloat>
define <vscale x 16 x bfloat> @vfma_vf_nxv16bf16(<vscale x 16 x bfloat> %va, bfloat %b, <vscale x 16 x bfloat> %vc, <vscale x 16 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfma_vf_nxv16bf16:
; 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: vmv4r.v v16, v12
+; CHECK-NEXT: vmv4r.v v20, v8
; CHECK-NEXT: fmv.x.h a1, fa0
; CHECK-NEXT: vsetvli a2, zero, e16, m4, ta, ma
-; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v12
-; CHECK-NEXT: addi a2, sp, 16
-; CHECK-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
-; CHECK-NEXT: vmv.v.x v12, a1
-; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v8
-; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v12
-; CHECK-NEXT: addi a1, sp, 16
-; CHECK-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
+; CHECK-NEXT: vfwcvtbf16.f.f.v v8, v16
+; CHECK-NEXT: vmv.v.x v4, a1
+; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v20
+; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v4
; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; CHECK-NEXT: vfmadd.vv v16, v24, v8, v0.t
; CHECK-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; CHECK-NEXT: vfncvtbf16.f.f.w v8, v16
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add sp, sp, a0
-; CHECK-NEXT: .cfi_def_cfa sp, 16
-; CHECK-NEXT: addi sp, sp, 16
-; CHECK-NEXT: .cfi_def_cfa_offset 0
; CHECK-NEXT: ret
%elt.head = insertelement <vscale x 16 x bfloat> poison, bfloat %b, i32 0
%vb = shufflevector <vscale x 16 x bfloat> %elt.head, <vscale x 16 x bfloat> poison, <vscale x 16 x i32> zeroinitializer
@@ -628,19 +601,27 @@ define <vscale x 32 x bfloat> @vfma_vv_nxv32bf16(<vscale x 32 x bfloat> %va, <vs
; CHECK-NEXT: add a2, a2, a3
; CHECK-NEXT: sub sp, sp, a2
; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x29, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 41 * vlenb
-; CHECK-NEXT: vmv1r.v v24, v0
-; CHECK-NEXT: vl8re16.v v0, (a0)
+; CHECK-NEXT: vl8re16.v v24, (a0)
+; CHECK-NEXT: csrr a0, vlenb
+; CHECK-NEXT: slli a2, a0, 5
+; CHECK-NEXT: add a0, a2, a0
+; 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 a2, vlenb
; CHECK-NEXT: slli a0, a2, 1
; CHECK-NEXT: srli a2, a2, 2
; CHECK-NEXT: sub a3, a1, a0
; CHECK-NEXT: csrr a4, vlenb
-; CHECK-NEXT: slli a4, a4, 4
+; CHECK-NEXT: slli a4, a4, 3
+; CHECK-NEXT: mv a5, a4
+; CHECK-NEXT: slli a4, a4, 1
+; CHECK-NEXT: add a4, a4, a5
; CHECK-NEXT: add a4, sp, a4
; CHECK-NEXT: addi a4, a4, 16
-; CHECK-NEXT: vs1r.v v24, (a4) # Unknown-size Folded Spill
+; CHECK-NEXT: vs1r.v v0, (a4) # Unknown-size Folded Spill
; CHECK-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
-; CHECK-NEXT: vslidedown.vx v24, v24, a2
+; CHECK-NEXT: vslidedown.vx v24, v0, a2
; CHECK-NEXT: csrr a2, vlenb
; CHECK-NEXT: slli a2, a2, 3
; CHECK-NEXT: add a2, sp, a2
@@ -650,41 +631,35 @@ define <vscale x 32 x bfloat> @vfma_vv_nxv32bf16(<vscale x 32 x bfloat> %va, <vs
; CHECK-NEXT: addi a2, a2, -1
; CHECK-NEXT: and a2, a2, a3
; CHECK-NEXT: csrr a3, vlenb
-; CHECK-NEXT: mv a4, a3
-; CHECK-NEXT: slli a3, a3, 3
-; CHECK-NEXT: add a4, a4, a3
-; CHECK-NEXT: slli a3, a3, 1
-; CHECK-NEXT: add a3, a3, a4
+; CHECK-NEXT: slli a3, a3, 4
; CHECK-NEXT: add a3, sp, a3
; CHECK-NEXT: addi a3, a3, 16
; CHECK-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
; CHECK-NEXT: vsetvli a3, zero, e16, m4, ta, ma
; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v12
-; CHECK-NEXT: addi a3, sp, 16
-; CHECK-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
-; CHECK-NEXT: vmv4r.v v8, v16
-; CHECK-NEXT: vmv8r.v v24, v16
+; CHECK-NEXT: vmv8r.v v0, v16
; CHECK-NEXT: csrr a3, vlenb
-; CHECK-NEXT: slli a4, a3, 5
-; CHECK-NEXT: add a3, a4, a3
+; CHECK-NEXT: mv a4, a3
+; CHECK-NEXT: slli a3, a3, 3
+; CHECK-NEXT: add a4, a4, a3
+; CHECK-NEXT: slli a3, a3, 1
+; CHECK-NEXT: add a3, a3, a4
; CHECK-NEXT: add a3, sp, a3
; CHECK-NEXT: addi a3, a3, 16
-; CHECK-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
-; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v28
+; CHECK-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
+; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v4
; CHECK-NEXT: csrr a3, vlenb
-; CHECK-NEXT: slli a4, a3, 4
+; CHECK-NEXT: slli a4, a3, 5
; CHECK-NEXT: add a3, a4, a3
; CHECK-NEXT: add a3, sp, a3
; CHECK-NEXT: addi a3, a3, 16
-; CHECK-NEXT: vs8r.v v0, (a3) # Unknown-size Folded Spill
+; CHECK-NEXT: vl8r.v v0, (a3) # Unknown-size Folded Reload
; CHECK-NEXT: vfwcvtbf16.f.f.v v8, v4
; CHECK-NEXT: csrr a3, vlenb
; CHECK-NEXT: slli a3, a3, 3
; CHECK-NEXT: add a3, sp, a3
; CHECK-NEXT: addi a3, a3, 16
; CHECK-NEXT: vl1r.v v0, (a3) # Unknown-size Folded Reload
-; CHECK-NEXT: addi a3, sp, 16
-; CHECK-NEXT: vl8r.v v24, (a3) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, a2, e32, m8, ta, ma
; CHECK-NEXT: vfmadd.vv v16, v24, v8, v0.t
; CHECK-NEXT: vsetvli a2, zero, e16, m4, ta, ma
@@ -696,11 +671,7 @@ define <vscale x 32 x bfloat> @vfma_vv_nxv32bf16(<vscale x 32 x bfloat> %va, <vs
; CHECK-NEXT: mv a1, a0
; CHECK-NEXT: .LBB30_2:
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: mv a2, a0
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add a2, a2, a0
-; CHECK-NEXT: slli a0, a0, 1
-; CHECK-NEXT: add 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 v16, (a0) # Unknown-size Folded Reload
@@ -711,27 +682,36 @@ define <vscale x 32 x bfloat> @vfma_vv_nxv32bf16(<vscale x 32 x bfloat> %va, <vs
; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a2, a0, 5
-; CHECK-NEXT: add a0, a2, a0
+; CHECK-NEXT: mv a2, a0
+; CHECK-NEXT: slli a0, a0, 3
+; CHECK-NEXT: add a2, a2, a0
+; CHECK-NEXT: slli a0, a0, 1
+; CHECK-NEXT: add 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: vfwcvtbf16.f.f.v v8, v16
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a2, a0, 4
+; CHECK-NEXT: slli a2, a0, 5
; CHECK-NEXT: add a0, a2, a0
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vl8r.v v0, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v0
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a2, a0, 5
-; CHECK-NEXT: add a0, a2, a0
+; CHECK-NEXT: mv a2, a0
+; CHECK-NEXT: slli a0, a0, 3
+; CHECK-NEXT: add a2, a2, a0
+; CHECK-NEXT: slli a0, a0, 1
+; CHECK-NEXT: add 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: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 4
+; CHECK-NEXT: slli a0, a0, 3
+; CHECK-NEXT: mv a2, a0
+; CHECK-NEXT: slli a0, a0, 1
+; CHECK-NEXT: add a0, a0, a2
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vl1r.v v0, (a0) # Unknown-size Folded Reload
@@ -741,8 +721,11 @@ define <vscale x 32 x bfloat> @vfma_vv_nxv32bf16(<vscale x 32 x bfloat> %va, <vs
; CHECK-NEXT: addi a0, a0, 16
; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a2, a0, 5
-; CHECK-NEXT: add a0, a2, a0
+; CHECK-NEXT: mv a2, a0
+; CHECK-NEXT: slli a0, a0, 3
+; CHECK-NEXT: add a2, a2, a0
+; CHECK-NEXT: slli a0, a0, 1
+; CHECK-NEXT: add 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
@@ -777,7 +760,8 @@ define <vscale x 32 x bfloat> @vfma_vv_nxv32bf16_unmasked(<vscale x 32 x bfloat>
; CHECK-NEXT: slli a2, a2, 5
; CHECK-NEXT: sub sp, sp, a2
; 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: vl8re16.v v24, (a0)
+; CHECK-NEXT: vmv8r.v v0, v16
+; CHECK-NEXT: vl8re16.v v16, (a0)
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: mv a2, a0
@@ -785,15 +769,17 @@ define <vscale x 32 x bfloat> @vfma_vv_nxv32bf16_unmasked(<vscale x 32 x bfloat>
; CHECK-NEXT: add a0, a0, a2
; 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 v16, (a0) # Unknown-size Folded Spill
; CHECK-NEXT: csrr a2, vlenb
; CHECK-NEXT: vsetvli a0, zero, e8, m4, ta, ma
-; CHECK-NEXT: vmset.m v7
+; CHECK-NEXT: vmset.m v16
; CHECK-NEXT: slli a0, a2, 1
; CHECK-NEXT: srli a2, a2, 2
; CHECK-NEXT: sub a3, a1, a0
; CHECK-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
-; CHECK-NEXT: vslidedown.vx v0, v7, a2
+; CHECK-NEXT: vslidedown.vx v16, v16, a2
+; CHECK-NEXT: addi a2, sp, 16
+; CHECK-NEXT: vs1r.v v16, (a2) # Unknown-size Folded Spill
; CHECK-NEXT: sltu a2, a1, a3
; CHECK-NEXT: addi a2, a2, -1
; CHECK-NEXT: and a2, a2, a3
@@ -804,16 +790,12 @@ define <vscale x 32 x bfloat> @vfma_vv_nxv32bf16_unmasked(<vscale x 32 x bfloat>
; CHECK-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
; CHECK-NEXT: vsetvli a3, zero, e16, m4, ta, ma
; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v12
-; CHECK-NEXT: addi a3, sp, 16
-; CHECK-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
-; CHECK-NEXT: vmv4r.v v8, v16
-; CHECK-NEXT: vmv8r.v v24, v16
; CHECK-NEXT: csrr a3, vlenb
; CHECK-NEXT: slli a3, a3, 4
; CHECK-NEXT: add a3, sp, a3
; CHECK-NEXT: addi a3, a3, 16
-; CHECK-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
-; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v28
+; CHECK-NEXT: vs8r.v v0, (a3) # Unknown-size Folded Spill
+; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v4
; CHECK-NEXT: csrr a3, vlenb
; CHECK-NEXT: slli a3, a3, 3
; CHECK-NEXT: mv a4, a3
@@ -821,10 +803,10 @@ define <vscale x 32 x bfloat> @vfma_vv_nxv32bf16_unmasked(<vscale x 32 x bfloat>
; CHECK-NEXT: add 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: vfwcvtbf16.f.f.v v8, v28
+; CHECK-NEXT: vl8r.v v0, (a3) # Unknown-size Folded Reload
+; CHECK-NEXT: vfwcvtbf16.f.f.v v8, v4
; CHECK-NEXT: addi a3, sp, 16
-; CHECK-NEXT: vl8r.v v24, (a3) # Unknown-size Folded Reload
+; CHECK-NEXT: vl1r.v v0, (a3) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, a2, e32, m8, ta, ma
; CHECK-NEXT: vfmadd.vv v16, v24, v8, v0.t
; CHECK-NEXT: vsetvli a2, zero, e16, m4, ta, ma
@@ -2000,29 +1982,16 @@ define <vscale x 16 x half> @vfma_vv_nxv16f16(<vscale x 16 x half> %va, <vscale
;
; ZVFHMIN-LABEL: vfma_vv_nxv16f16:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: addi sp, sp, -16
-; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
-; ZVFHMIN-NEXT: csrr a1, vlenb
-; 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: vmv4r.v v4, v12
+; ZVFHMIN-NEXT: vmv4r.v v20, v8
; ZVFHMIN-NEXT: vsetvli a1, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v16
-; ZVFHMIN-NEXT: addi a1, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v24, (a1) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v8
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
-; ZVFHMIN-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v20
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v4
; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfmadd.vv v16, v24, v8, v0.t
+; ZVFHMIN-NEXT: vfmadd.vv v16, v8, 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, 3
-; ZVFHMIN-NEXT: add sp, sp, a0
-; ZVFHMIN-NEXT: .cfi_def_cfa sp, 16
-; ZVFHMIN-NEXT: addi sp, sp, 16
-; ZVFHMIN-NEXT: .cfi_def_cfa_offset 0
; ZVFHMIN-NEXT: ret
%v = call <vscale x 16 x half> @llvm.vp.fma.nxv16f16(<vscale x 16 x half> %va, <vscale x 16 x half> %b, <vscale x 16 x half> %c, <vscale x 16 x i1> %m, i32 %evl)
ret <vscale x 16 x half> %v
@@ -2059,32 +2028,18 @@ define <vscale x 16 x half> @vfma_vf_nxv16f16(<vscale x 16 x half> %va, half %b,
;
; ZVFHMIN-LABEL: vfma_vf_nxv16f16:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: addi sp, sp, -16
-; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
-; ZVFHMIN-NEXT: csrr a1, vlenb
-; 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: vmv4r.v v16, v12
+; ZVFHMIN-NEXT: vmv4r.v v20, v8
; ZVFHMIN-NEXT: fmv.x.h a1, fa0
; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
-; ZVFHMIN-NEXT: addi a2, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vmv.v.x v12, a1
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v8
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
-; ZVFHMIN-NEXT: addi a1, sp, 16
-; ZVFHMIN-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v16
+; ZVFHMIN-NEXT: vmv.v.x v4, a1
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v4
; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfmadd.vv v16, v24, v8, 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, 3
-; ZVFHMIN-NEXT: add sp, sp, a0
-; ZVFHMIN-NEXT: .cfi_def_cfa sp, 16
-; ZVFHMIN-NEXT: addi sp, sp, 16
-; ZVFHMIN-NEXT: .cfi_def_cfa_offset 0
; 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
@@ -2193,19 +2148,27 @@ define <vscale x 32 x half> @vfma_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
; ZVFHMIN-NEXT: add a2, a2, a3
; ZVFHMIN-NEXT: sub sp, sp, a2
; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x29, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 41 * vlenb
-; ZVFHMIN-NEXT: vmv1r.v v24, v0
-; ZVFHMIN-NEXT: vl8re16.v v0, (a0)
+; ZVFHMIN-NEXT: vl8re16.v v24, (a0)
+; ZVFHMIN-NEXT: csrr a0, vlenb
+; ZVFHMIN-NEXT: slli a2, a0, 5
+; ZVFHMIN-NEXT: add a0, a2, a0
+; 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 a2, vlenb
; ZVFHMIN-NEXT: slli a0, a2, 1
; ZVFHMIN-NEXT: srli a2, a2, 2
; ZVFHMIN-NEXT: sub a3, a1, a0
; ZVFHMIN-NEXT: csrr a4, vlenb
-; ZVFHMIN-NEXT: slli a4, a4, 4
+; ZVFHMIN-NEXT: slli a4, a4, 3
+; ZVFHMIN-NEXT: mv a5, a4
+; ZVFHMIN-NEXT: slli a4, a4, 1
+; ZVFHMIN-NEXT: add a4, a4, a5
; ZVFHMIN-NEXT: add a4, sp, a4
; ZVFHMIN-NEXT: addi a4, a4, 16
-; ZVFHMIN-NEXT: vs1r.v v24, (a4) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vs1r.v v0, (a4) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
-; ZVFHMIN-NEXT: vslidedown.vx v24, v24, a2
+; ZVFHMIN-NEXT: vslidedown.vx v24, v0, a2
; ZVFHMIN-NEXT: csrr a2, vlenb
; ZVFHMIN-NEXT: slli a2, a2, 3
; ZVFHMIN-NEXT: add a2, sp, a2
@@ -2215,41 +2178,35 @@ define <vscale x 32 x half> @vfma_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
; ZVFHMIN-NEXT: addi a2, a2, -1
; ZVFHMIN-NEXT: and a2, a2, a3
; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: mv a4, a3
-; ZVFHMIN-NEXT: slli a3, a3, 3
-; ZVFHMIN-NEXT: add a4, a4, a3
-; ZVFHMIN-NEXT: slli a3, a3, 1
-; ZVFHMIN-NEXT: add a3, a3, a4
+; ZVFHMIN-NEXT: slli a3, a3, 4
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
-; ZVFHMIN-NEXT: addi a3, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vmv4r.v v8, v16
-; ZVFHMIN-NEXT: vmv8r.v v24, v16
+; ZVFHMIN-NEXT: vmv8r.v v0, v16
; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: slli a4, a3, 5
-; ZVFHMIN-NEXT: add a3, a4, a3
+; ZVFHMIN-NEXT: mv a4, a3
+; ZVFHMIN-NEXT: slli a3, a3, 3
+; ZVFHMIN-NEXT: add a4, a4, a3
+; ZVFHMIN-NEXT: slli a3, a3, 1
+; ZVFHMIN-NEXT: add a3, a3, a4
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
-; ZVFHMIN-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v28
+; ZVFHMIN-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v4
; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: slli a4, a3, 4
+; ZVFHMIN-NEXT: slli a4, a3, 5
; ZVFHMIN-NEXT: add a3, a4, a3
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
-; ZVFHMIN-NEXT: vs8r.v v0, (a3) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vl8r.v v0, (a3) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v4
; ZVFHMIN-NEXT: csrr a3, vlenb
; ZVFHMIN-NEXT: slli a3, a3, 3
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
; ZVFHMIN-NEXT: vl1r.v v0, (a3) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: addi a3, sp, 16
-; ZVFHMIN-NEXT: vl8r.v v24, (a3) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, a2, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfmadd.vv v16, v24, v8, v0.t
; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
@@ -2261,11 +2218,7 @@ define <vscale x 32 x half> @vfma_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
; ZVFHMIN-NEXT: mv a1, a0
; ZVFHMIN-NEXT: .LBB66_2:
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: mv a2, a0
-; ZVFHMIN-NEXT: slli a0, a0, 3
-; ZVFHMIN-NEXT: add a2, a2, a0
-; ZVFHMIN-NEXT: slli a0, a0, 1
-; ZVFHMIN-NEXT: add a0, a0, a2
+; ZVFHMIN-NEXT: slli a0, a0, 4
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
@@ -2276,27 +2229,36 @@ define <vscale x 32 x half> @vfma_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a2, a0, 5
-; ZVFHMIN-NEXT: add a0, a2, a0
+; ZVFHMIN-NEXT: mv a2, a0
+; ZVFHMIN-NEXT: slli a0, a0, 3
+; ZVFHMIN-NEXT: add a2, a2, a0
+; ZVFHMIN-NEXT: slli a0, a0, 1
+; ZVFHMIN-NEXT: add a0, a0, a2
; 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 v8, v16
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a2, a0, 4
+; ZVFHMIN-NEXT: slli a2, a0, 5
; ZVFHMIN-NEXT: add a0, a2, a0
; 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 v16, v0
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a2, a0, 5
-; ZVFHMIN-NEXT: add a0, a2, a0
+; ZVFHMIN-NEXT: mv a2, a0
+; ZVFHMIN-NEXT: slli a0, a0, 3
+; ZVFHMIN-NEXT: add a2, a2, a0
+; ZVFHMIN-NEXT: slli a0, a0, 1
+; ZVFHMIN-NEXT: add a0, a0, a2
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 4
+; ZVFHMIN-NEXT: slli a0, a0, 3
+; ZVFHMIN-NEXT: mv a2, a0
+; ZVFHMIN-NEXT: slli a0, a0, 1
+; ZVFHMIN-NEXT: add a0, a0, a2
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vl1r.v v0, (a0) # Unknown-size Folded Reload
@@ -2306,8 +2268,11 @@ define <vscale x 32 x half> @vfma_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a2, a0, 5
-; ZVFHMIN-NEXT: add a0, a2, a0
+; ZVFHMIN-NEXT: mv a2, a0
+; ZVFHMIN-NEXT: slli a0, a0, 3
+; ZVFHMIN-NEXT: add a2, a2, a0
+; ZVFHMIN-NEXT: slli a0, a0, 1
+; ZVFHMIN-NEXT: add a0, a0, a2
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
@@ -2349,7 +2314,8 @@ define <vscale x 32 x half> @vfma_vv_nxv32f16_unmasked(<vscale x 32 x half> %va,
; ZVFHMIN-NEXT: slli a2, a2, 5
; ZVFHMIN-NEXT: sub sp, sp, a2
; 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 v24, (a0)
+; ZVFHMIN-NEXT: vmv8r.v v0, v16
+; ZVFHMIN-NEXT: vl8re16.v v16, (a0)
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: mv a2, a0
@@ -2357,15 +2323,17 @@ define <vscale x 32 x half> @vfma_vv_nxv32f16_unmasked(<vscale x 32 x half> %va,
; ZVFHMIN-NEXT: add a0, a0, a2
; 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: csrr a2, vlenb
; ZVFHMIN-NEXT: vsetvli a0, zero, e8, m4, ta, ma
-; ZVFHMIN-NEXT: vmset.m v7
+; ZVFHMIN-NEXT: vmset.m v16
; ZVFHMIN-NEXT: slli a0, a2, 1
; ZVFHMIN-NEXT: srli a2, a2, 2
; ZVFHMIN-NEXT: sub a3, a1, a0
; ZVFHMIN-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
-; ZVFHMIN-NEXT: vslidedown.vx v0, v7, a2
+; ZVFHMIN-NEXT: vslidedown.vx v16, v16, a2
+; ZVFHMIN-NEXT: addi a2, sp, 16
+; ZVFHMIN-NEXT: vs1r.v v16, (a2) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: sltu a2, a1, a3
; ZVFHMIN-NEXT: addi a2, a2, -1
; ZVFHMIN-NEXT: and a2, a2, a3
@@ -2376,16 +2344,12 @@ define <vscale x 32 x half> @vfma_vv_nxv32f16_unmasked(<vscale x 32 x half> %va,
; ZVFHMIN-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
-; ZVFHMIN-NEXT: addi a3, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vmv4r.v v8, v16
-; ZVFHMIN-NEXT: vmv8r.v v24, v16
; ZVFHMIN-NEXT: csrr a3, vlenb
; ZVFHMIN-NEXT: slli a3, a3, 4
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
-; ZVFHMIN-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v28
+; ZVFHMIN-NEXT: vs8r.v v0, (a3) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v4
; ZVFHMIN-NEXT: csrr a3, vlenb
; ZVFHMIN-NEXT: slli a3, a3, 3
; ZVFHMIN-NEXT: mv a4, a3
@@ -2393,10 +2357,10 @@ define <vscale x 32 x half> @vfma_vv_nxv32f16_unmasked(<vscale x 32 x half> %va,
; ZVFHMIN-NEXT: add a3, a3, a4
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
-; ZVFHMIN-NEXT: vl8r.v v24, (a3) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v28
+; ZVFHMIN-NEXT: vl8r.v v0, (a3) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v4
; ZVFHMIN-NEXT: addi a3, sp, 16
-; ZVFHMIN-NEXT: vl8r.v v24, (a3) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vl1r.v v0, (a3) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, a2, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfmadd.vv v16, v24, v8, v0.t
; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
@@ -7825,33 +7789,19 @@ define <vscale x 16 x half> @vfnmadd_vv_nxv16f16_commuted(<vscale x 16 x half> %
;
; ZVFHMIN-LABEL: vfnmadd_vv_nxv16f16_commuted:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: addi sp, sp, -16
-; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
-; ZVFHMIN-NEXT: csrr a1, vlenb
-; 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: vmv4r.v v4, v8
; ZVFHMIN-NEXT: lui a1, 8
; ZVFHMIN-NEXT: vsetvli zero, a0, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vxor.vx v12, v12, a1, v0.t
+; ZVFHMIN-NEXT: vxor.vx v20, v12, a1, v0.t
; ZVFHMIN-NEXT: vxor.vx v16, v16, a1, v0.t
; ZVFHMIN-NEXT: vsetvli a1, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v16
-; ZVFHMIN-NEXT: addi a1, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v24, (a1) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8
-; ZVFHMIN-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v20
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v4
; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfmadd.vv v16, v24, v8, v0.t
+; ZVFHMIN-NEXT: vfmadd.vv v16, v8, 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, 3
-; ZVFHMIN-NEXT: add sp, sp, a0
-; ZVFHMIN-NEXT: .cfi_def_cfa sp, 16
-; ZVFHMIN-NEXT: addi sp, sp, 16
-; ZVFHMIN-NEXT: .cfi_def_cfa_offset 0
; ZVFHMIN-NEXT: ret
%negb = call <vscale x 16 x half> @llvm.vp.fneg.nxv16f16(<vscale x 16 x half> %b, <vscale x 16 x i1> %m, i32 %evl)
%negc = call <vscale x 16 x half> @llvm.vp.fneg.nxv16f16(<vscale x 16 x half> %c, <vscale x 16 x i1> %m, i32 %evl)
@@ -7924,36 +7874,21 @@ define <vscale x 16 x half> @vfnmadd_vf_nxv16f16(<vscale x 16 x half> %va, half
;
; ZVFHMIN-LABEL: vfnmadd_vf_nxv16f16:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: addi sp, sp, -16
-; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
-; ZVFHMIN-NEXT: csrr a1, vlenb
-; 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: fmv.x.h a1, fa0
; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vmv.v.x v4, a1
; ZVFHMIN-NEXT: lui a1, 8
; ZVFHMIN-NEXT: vsetvli zero, a0, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vxor.vx v8, v8, a1, v0.t
+; ZVFHMIN-NEXT: vxor.vx v16, v8, a1, v0.t
; ZVFHMIN-NEXT: vxor.vx v12, v12, a1, v0.t
; ZVFHMIN-NEXT: vsetvli a1, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
-; ZVFHMIN-NEXT: addi a1, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v8
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v16
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v4
-; ZVFHMIN-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfmadd.vv v16, v24, v8, v0.t
+; ZVFHMIN-NEXT: vfmadd.vv v16, v8, 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, 3
-; ZVFHMIN-NEXT: add sp, sp, a0
-; ZVFHMIN-NEXT: .cfi_def_cfa sp, 16
-; ZVFHMIN-NEXT: addi sp, sp, 16
-; ZVFHMIN-NEXT: .cfi_def_cfa_offset 0
; 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
@@ -8152,22 +8087,37 @@ define <vscale x 16 x half> @vfnmadd_vf_nxv16f16_neg_splat_commute(<vscale x 16
;
; ZVFHMIN-LABEL: vfnmadd_vf_nxv16f16_neg_splat_commute:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: vmv4r.v v4, v8
+; 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: addi a1, sp, 16
+; ZVFHMIN-NEXT: vs4r.v v8, (a1) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: fmv.x.h a1, fa0
; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vmv.v.x v16, a1
; ZVFHMIN-NEXT: lui a1, 8
; ZVFHMIN-NEXT: vsetvli zero, a0, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vxor.vx v16, v16, a1, v0.t
+; ZVFHMIN-NEXT: vxor.vx v4, v16, a1, v0.t
; ZVFHMIN-NEXT: vxor.vx v12, v12, a1, v0.t
; ZVFHMIN-NEXT: vsetvli a1, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v16
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v4
+; ZVFHMIN-NEXT: addi a1, sp, 16
+; ZVFHMIN-NEXT: vl4r.v v4, (a1) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v4
; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfmadd.vv v16, v8, 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: .cfi_def_cfa sp, 16
+; ZVFHMIN-NEXT: addi sp, sp, 16
+; ZVFHMIN-NEXT: .cfi_def_cfa_offset 0
; 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
@@ -8282,33 +8232,19 @@ define <vscale x 16 x half> @vfnmsub_vv_nxv16f16_commuted(<vscale x 16 x half> %
;
; ZVFHMIN-LABEL: vfnmsub_vv_nxv16f16_commuted:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: addi sp, sp, -16
-; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
-; ZVFHMIN-NEXT: csrr a1, vlenb
-; 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: vmv4r.v v4, v8
; ZVFHMIN-NEXT: lui a1, 8
; ZVFHMIN-NEXT: vsetvli zero, a0, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vxor.vx v12, v12, a1, v0.t
+; ZVFHMIN-NEXT: vxor.vx v20, v12, a1, v0.t
; ZVFHMIN-NEXT: vxor.vx v16, v16, a1, v0.t
; ZVFHMIN-NEXT: vsetvli a1, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v16
-; ZVFHMIN-NEXT: addi a1, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v24, (a1) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8
-; ZVFHMIN-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v20
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v4
; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfmadd.vv v16, v24, v8, v0.t
+; ZVFHMIN-NEXT: vfmadd.vv v16, v8, 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, 3
-; ZVFHMIN-NEXT: add sp, sp, a0
-; ZVFHMIN-NEXT: .cfi_def_cfa sp, 16
-; ZVFHMIN-NEXT: addi sp, sp, 16
-; ZVFHMIN-NEXT: .cfi_def_cfa_offset 0
; ZVFHMIN-NEXT: ret
%negb = call <vscale x 16 x half> @llvm.vp.fneg.nxv16f16(<vscale x 16 x half> %b, <vscale x 16 x i1> %m, i32 %evl)
%negc = call <vscale x 16 x half> @llvm.vp.fneg.nxv16f16(<vscale x 16 x half> %c, <vscale x 16 x i1> %m, i32 %evl)
@@ -8851,17 +8787,10 @@ define <vscale x 32 x half> @vfmsub_vv_nxv32f16_unmasked(<vscale x 32 x half> %v
; ZVFHMIN-NEXT: addi sp, sp, -16
; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a2, a2, 3
-; ZVFHMIN-NEXT: mv a3, a2
-; ZVFHMIN-NEXT: slli a2, a2, 2
-; ZVFHMIN-NEXT: add a2, a2, a3
+; ZVFHMIN-NEXT: slli a3, a2, 5
+; ZVFHMIN-NEXT: add a2, a3, a2
; ZVFHMIN-NEXT: sub sp, sp, a2
-; 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 a2, vlenb
-; ZVFHMIN-NEXT: slli a2, a2, 5
-; ZVFHMIN-NEXT: add a2, sp, a2
-; ZVFHMIN-NEXT: addi a2, a2, 16
-; ZVFHMIN-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x21, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 33 * vlenb
; ZVFHMIN-NEXT: vl8re16.v v24, (a0)
; ZVFHMIN-NEXT: lui a2, 8
; ZVFHMIN-NEXT: vsetvli a0, zero, e8, m4, ta, ma
@@ -8871,7 +8800,12 @@ define <vscale x 32 x half> @vfmsub_vv_nxv32f16_unmasked(<vscale x 32 x half> %v
; ZVFHMIN-NEXT: srli a3, a3, 2
; ZVFHMIN-NEXT: sub a4, a1, a0
; ZVFHMIN-NEXT: vsetvli a5, zero, e8, mf2, ta, ma
-; ZVFHMIN-NEXT: vslidedown.vx v20, v7, a3
+; ZVFHMIN-NEXT: vslidedown.vx v7, v7, a3
+; ZVFHMIN-NEXT: csrr a3, vlenb
+; ZVFHMIN-NEXT: slli a3, a3, 3
+; ZVFHMIN-NEXT: add a3, sp, a3
+; ZVFHMIN-NEXT: addi a3, a3, 16
+; ZVFHMIN-NEXT: vs1r.v v7, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: sltu a3, a1, a4
; ZVFHMIN-NEXT: vsetvli zero, a1, e16, m8, ta, ma
; ZVFHMIN-NEXT: vxor.vx v0, v24, a2
@@ -8880,43 +8814,41 @@ define <vscale x 32 x half> @vfmsub_vv_nxv32f16_unmasked(<vscale x 32 x half> %v
; ZVFHMIN-NEXT: addi a3, a3, -1
; ZVFHMIN-NEXT: and a3, a3, a4
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a2, a2, 3
-; ZVFHMIN-NEXT: mv a4, a2
-; ZVFHMIN-NEXT: slli a2, a2, 1
-; ZVFHMIN-NEXT: add a2, a2, a4
+; ZVFHMIN-NEXT: slli a4, a2, 4
+; ZVFHMIN-NEXT: add a2, a4, a2
; 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: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a2, a2, 4
+; ZVFHMIN-NEXT: slli a4, a2, 3
+; ZVFHMIN-NEXT: add a2, a4, a2
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
; ZVFHMIN-NEXT: vs8r.v v24, (a2) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a2, a2, 5
+; ZVFHMIN-NEXT: mv a4, a2
+; ZVFHMIN-NEXT: slli a2, a2, 3
+; ZVFHMIN-NEXT: add a4, a4, a2
+; ZVFHMIN-NEXT: slli a2, a2, 1
+; ZVFHMIN-NEXT: add a2, a2, a4
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
-; ZVFHMIN-NEXT: vl8r.v v24, (a2) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v28
+; ZVFHMIN-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v20
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v4
; ZVFHMIN-NEXT: csrr a2, vlenb
; ZVFHMIN-NEXT: slli a2, a2, 3
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
-; ZVFHMIN-NEXT: vs8r.v v24, (a2) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vmv1r.v v0, v20
+; ZVFHMIN-NEXT: vl1r.v v0, (a2) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a2, a2, 4
+; ZVFHMIN-NEXT: slli a4, a2, 3
+; ZVFHMIN-NEXT: add a2, a4, a2
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
; ZVFHMIN-NEXT: vl8r.v v16, (a2) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a2, a2, 3
-; ZVFHMIN-NEXT: add a2, sp, a2
-; ZVFHMIN-NEXT: addi a2, a2, 16
-; ZVFHMIN-NEXT: vl8r.v v24, (a2) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfmadd.vv v8, v16, v24, v0.t
; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
@@ -8925,7 +8857,8 @@ define <vscale x 32 x half> @vfmsub_vv_nxv32f16_unmasked(<vscale x 32 x half> %v
; ZVFHMIN-NEXT: vl8r.v v24, (a2) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v24
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a2, a2, 4
+; ZVFHMIN-NEXT: slli a3, a2, 3
+; ZVFHMIN-NEXT: add a2, a3, a2
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
@@ -8934,22 +8867,25 @@ define <vscale x 32 x half> @vfmsub_vv_nxv32f16_unmasked(<vscale x 32 x half> %v
; ZVFHMIN-NEXT: mv a1, a0
; ZVFHMIN-NEXT: .LBB281_2:
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 3
-; ZVFHMIN-NEXT: mv a2, a0
-; ZVFHMIN-NEXT: slli a0, a0, 1
-; ZVFHMIN-NEXT: add a0, a0, a2
+; ZVFHMIN-NEXT: slli a2, a0, 4
+; ZVFHMIN-NEXT: add a0, a2, a0
; 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 v8, v24
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 5
+; ZVFHMIN-NEXT: mv a2, a0
+; ZVFHMIN-NEXT: slli a0, a0, 3
+; ZVFHMIN-NEXT: add a2, a2, a0
+; ZVFHMIN-NEXT: slli a0, a0, 1
+; ZVFHMIN-NEXT: add a0, a0, a2
; 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: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 4
+; ZVFHMIN-NEXT: slli a2, a0, 3
+; ZVFHMIN-NEXT: add a0, a2, a0
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vl8r.v v0, (a0) # Unknown-size Folded Reload
@@ -8959,10 +8895,8 @@ define <vscale x 32 x half> @vfmsub_vv_nxv32f16_unmasked(<vscale x 32 x half> %v
; ZVFHMIN-NEXT: vfncvt.f.f.w v16, v24
; ZVFHMIN-NEXT: vmv8r.v v8, v16
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 3
-; ZVFHMIN-NEXT: mv a1, a0
-; ZVFHMIN-NEXT: slli a0, a0, 2
-; ZVFHMIN-NEXT: add a0, a0, a1
+; ZVFHMIN-NEXT: slli a1, a0, 5
+; ZVFHMIN-NEXT: add a0, a1, a0
; ZVFHMIN-NEXT: add sp, sp, a0
; ZVFHMIN-NEXT: .cfi_def_cfa sp, 16
; ZVFHMIN-NEXT: addi sp, sp, 16
@@ -9020,21 +8954,23 @@ define <vscale x 32 x half> @vfmsub_vf_nxv32f16(<vscale x 32 x half> %va, half %
; ZVFHMIN-NEXT: # %bb.1:
; ZVFHMIN-NEXT: mv a3, a2
; ZVFHMIN-NEXT: .LBB282_2:
-; ZVFHMIN-NEXT: vmv8r.v v16, v8
-; ZVFHMIN-NEXT: addi a4, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v8, (a4) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v16
; ZVFHMIN-NEXT: csrr a4, vlenb
; ZVFHMIN-NEXT: slli a4, a4, 4
; ZVFHMIN-NEXT: add a4, sp, a4
; ZVFHMIN-NEXT: addi a4, a4, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a4) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8
+; ZVFHMIN-NEXT: csrr a4, vlenb
+; ZVFHMIN-NEXT: slli a4, a4, 3
+; ZVFHMIN-NEXT: add a4, sp, a4
+; ZVFHMIN-NEXT: addi a4, a4, 16
+; ZVFHMIN-NEXT: vs8r.v v16, (a4) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a4, vlenb
; ZVFHMIN-NEXT: slli a4, a4, 5
; ZVFHMIN-NEXT: add a4, sp, a4
; ZVFHMIN-NEXT: addi a4, a4, 16
-; ZVFHMIN-NEXT: vl8r.v v24, (a4) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v24
+; ZVFHMIN-NEXT: vl8r.v v8, (a4) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v8
; ZVFHMIN-NEXT: csrr a4, vlenb
; ZVFHMIN-NEXT: slli a4, a4, 3
; ZVFHMIN-NEXT: mv a5, a4
@@ -9042,19 +8978,16 @@ define <vscale x 32 x half> @vfmsub_vf_nxv32f16(<vscale x 32 x half> %va, half %
; ZVFHMIN-NEXT: add a4, a4, a5
; ZVFHMIN-NEXT: add a4, sp, a4
; ZVFHMIN-NEXT: addi a4, a4, 16
-; ZVFHMIN-NEXT: vl8r.v v24, (a4) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vl8r.v v8, (a4) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: csrr a4, vlenb
-; ZVFHMIN-NEXT: slli a4, a4, 4
+; ZVFHMIN-NEXT: slli a4, a4, 3
; ZVFHMIN-NEXT: add a4, sp, a4
; ZVFHMIN-NEXT: addi a4, a4, 16
; ZVFHMIN-NEXT: vl8r.v v16, (a4) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfmadd.vv v8, v16, v24, v0.t
-; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: slli a3, a3, 3
-; ZVFHMIN-NEXT: add a3, sp, a3
-; ZVFHMIN-NEXT: addi a3, a3, 16
-; ZVFHMIN-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vfmadd.vv v24, v16, v8, v0.t
+; ZVFHMIN-NEXT: addi a3, sp, 16
+; ZVFHMIN-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v4
; ZVFHMIN-NEXT: csrr a3, vlenb
@@ -9067,11 +9000,14 @@ define <vscale x 32 x half> @vfmsub_vf_nxv32f16(<vscale x 32 x half> %va, half %
; ZVFHMIN-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: sub a2, a0, a2
; ZVFHMIN-NEXT: srli a1, a1, 2
-; ZVFHMIN-NEXT: addi a3, sp, 16
+; ZVFHMIN-NEXT: csrr a3, vlenb
+; ZVFHMIN-NEXT: slli a3, a3, 4
+; ZVFHMIN-NEXT: add a3, sp, a3
+; ZVFHMIN-NEXT: addi a3, a3, 16
; ZVFHMIN-NEXT: vl8r.v v8, (a3) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: slli a3, a3, 4
+; ZVFHMIN-NEXT: slli a3, a3, 3
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
; ZVFHMIN-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
@@ -9095,16 +9031,13 @@ define <vscale x 32 x half> @vfmsub_vf_nxv32f16(<vscale x 32 x half> %va, half %
; ZVFHMIN-NEXT: addi a1, a1, 16
; ZVFHMIN-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a1, a1, 4
+; ZVFHMIN-NEXT: slli a1, a1, 3
; ZVFHMIN-NEXT: add a1, sp, a1
; ZVFHMIN-NEXT: addi a1, a1, 16
; ZVFHMIN-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfmadd.vv v24, v16, v8, v0.t
-; 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: vl8r.v v16, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v16
@@ -9158,11 +9091,7 @@ define <vscale x 32 x half> @vfmsub_vf_nxv32f16_commute(<vscale x 32 x half> %va
; ZVFHMIN-NEXT: vxor.vx v16, v16, a3, v0.t
; ZVFHMIN-NEXT: slli a2, a1, 1
; ZVFHMIN-NEXT: mv a3, a0
-; ZVFHMIN-NEXT: csrr a4, vlenb
-; ZVFHMIN-NEXT: slli a4, a4, 3
-; ZVFHMIN-NEXT: add a4, sp, a4
-; ZVFHMIN-NEXT: addi a4, a4, 16
-; ZVFHMIN-NEXT: vs8r.v v16, (a4) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vmv4r.v v4, v20
; ZVFHMIN-NEXT: vsetvli a4, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v16
; ZVFHMIN-NEXT: csrr a4, vlenb
@@ -9175,7 +9104,11 @@ define <vscale x 32 x half> @vfmsub_vf_nxv32f16_commute(<vscale x 32 x half> %va
; ZVFHMIN-NEXT: mv a3, a2
; ZVFHMIN-NEXT: .LBB283_2:
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8
-; ZVFHMIN-NEXT: vmv4r.v v4, v12
+; ZVFHMIN-NEXT: csrr a4, vlenb
+; ZVFHMIN-NEXT: slli a4, a4, 3
+; ZVFHMIN-NEXT: add a4, sp, a4
+; ZVFHMIN-NEXT: addi a4, a4, 16
+; ZVFHMIN-NEXT: vs8r.v v8, (a4) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a4, vlenb
; ZVFHMIN-NEXT: slli a4, a4, 3
; ZVFHMIN-NEXT: mv a5, a4
@@ -9194,13 +9127,8 @@ define <vscale x 32 x half> @vfmsub_vf_nxv32f16_commute(<vscale x 32 x half> %va
; ZVFHMIN-NEXT: vfmadd.vv v16, v8, v24, v0.t
; ZVFHMIN-NEXT: addi a3, sp, 16
; ZVFHMIN-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: slli a3, a3, 3
-; ZVFHMIN-NEXT: add a3, sp, a3
-; ZVFHMIN-NEXT: addi a3, a3, 16
-; ZVFHMIN-NEXT: vl8r.v v24, (a3) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v28
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v4
; ZVFHMIN-NEXT: csrr a3, vlenb
; ZVFHMIN-NEXT: slli a3, a3, 4
; ZVFHMIN-NEXT: add a3, sp, a3
@@ -9208,7 +9136,12 @@ define <vscale x 32 x half> @vfmsub_vf_nxv32f16_commute(<vscale x 32 x half> %va
; ZVFHMIN-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: sub a2, a0, a2
; ZVFHMIN-NEXT: srli a1, a1, 2
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v4
+; ZVFHMIN-NEXT: csrr a3, vlenb
+; ZVFHMIN-NEXT: slli a3, a3, 3
+; ZVFHMIN-NEXT: add a3, sp, a3
+; ZVFHMIN-NEXT: addi a3, a3, 16
+; ZVFHMIN-NEXT: vl8r.v v8, (a3) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
; ZVFHMIN-NEXT: csrr a3, vlenb
; ZVFHMIN-NEXT: slli a3, a3, 3
; ZVFHMIN-NEXT: mv a4, a3
@@ -9303,12 +9236,12 @@ define <vscale x 32 x half> @vfmsub_vf_nxv32f16_unmasked(<vscale x 32 x half> %v
; ZVFHMIN-NEXT: srli a3, a3, 2
; ZVFHMIN-NEXT: sub a4, a0, a1
; ZVFHMIN-NEXT: vsetvli a5, zero, e8, mf2, ta, ma
-; ZVFHMIN-NEXT: vslidedown.vx v0, v7, a3
+; ZVFHMIN-NEXT: vslidedown.vx v16, v7, a3
; ZVFHMIN-NEXT: sltu a3, a0, a4
; ZVFHMIN-NEXT: addi a3, a3, -1
; ZVFHMIN-NEXT: and a3, a3, a4
; ZVFHMIN-NEXT: vsetvli a4, zero, e16, m8, ta, ma
-; ZVFHMIN-NEXT: vmv.v.x v24, a2
+; ZVFHMIN-NEXT: vmv.v.x v0, a2
; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v20
; ZVFHMIN-NEXT: csrr a2, vlenb
@@ -9316,7 +9249,7 @@ define <vscale x 32 x half> @vfmsub_vf_nxv32f16_unmasked(<vscale x 32 x half> %v
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vmv4r.v v8, v24
+; ZVFHMIN-NEXT: vmv4r.v v8, v0
; ZVFHMIN-NEXT: csrr a2, vlenb
; ZVFHMIN-NEXT: slli a2, a2, 3
; ZVFHMIN-NEXT: mv a4, a2
@@ -9325,7 +9258,8 @@ define <vscale x 32 x half> @vfmsub_vf_nxv32f16_unmasked(<vscale x 32 x half> %v
; 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 v8, v28
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v4
+; ZVFHMIN-NEXT: vmv1r.v v0, v16
; ZVFHMIN-NEXT: csrr a2, vlenb
; ZVFHMIN-NEXT: slli a2, a2, 4
; ZVFHMIN-NEXT: add a2, sp, a2
@@ -9406,12 +9340,12 @@ define <vscale x 32 x half> @vfmsub_vf_nxv32f16_unmasked_commute(<vscale x 32 x
; ZVFHMIN-NEXT: addi sp, sp, -16
; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a1, a1, 3
+; ZVFHMIN-NEXT: slli a1, a1, 4
; ZVFHMIN-NEXT: mv a2, a1
-; ZVFHMIN-NEXT: slli a1, a1, 2
+; ZVFHMIN-NEXT: slli a1, a1, 1
; ZVFHMIN-NEXT: add a1, a1, a2
; 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: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x30, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 48 * vlenb
; ZVFHMIN-NEXT: fmv.x.h a2, fa0
; ZVFHMIN-NEXT: lui a1, 8
; ZVFHMIN-NEXT: vsetvli a3, zero, e8, m4, ta, ma
@@ -9420,7 +9354,7 @@ define <vscale x 32 x half> @vfmsub_vf_nxv32f16_unmasked_commute(<vscale x 32 x
; ZVFHMIN-NEXT: csrr a4, vlenb
; ZVFHMIN-NEXT: slli a4, a4, 3
; ZVFHMIN-NEXT: mv a5, a4
-; ZVFHMIN-NEXT: slli a4, a4, 1
+; ZVFHMIN-NEXT: slli a4, a4, 2
; ZVFHMIN-NEXT: add a4, a4, a5
; ZVFHMIN-NEXT: add a4, sp, a4
; ZVFHMIN-NEXT: addi a4, a4, 16
@@ -9435,17 +9369,20 @@ define <vscale x 32 x half> @vfmsub_vf_nxv32f16_unmasked_commute(<vscale x 32 x
; ZVFHMIN-NEXT: srli a3, a3, 2
; ZVFHMIN-NEXT: sub a4, a0, a1
; ZVFHMIN-NEXT: vsetvli a5, zero, e8, mf2, ta, ma
-; ZVFHMIN-NEXT: vslidedown.vx v0, v7, a3
+; ZVFHMIN-NEXT: vslidedown.vx v8, v7, a3
+; ZVFHMIN-NEXT: csrr a3, vlenb
+; ZVFHMIN-NEXT: slli a3, a3, 3
+; ZVFHMIN-NEXT: mv a5, a3
+; ZVFHMIN-NEXT: slli a3, a3, 1
+; ZVFHMIN-NEXT: add a3, a3, a5
+; ZVFHMIN-NEXT: add a3, sp, a3
+; ZVFHMIN-NEXT: addi a3, a3, 16
+; ZVFHMIN-NEXT: vs1r.v v8, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: sltu a3, a0, a4
; ZVFHMIN-NEXT: addi a3, a3, -1
; ZVFHMIN-NEXT: and a3, a3, a4
; ZVFHMIN-NEXT: vsetvli a4, zero, e16, m8, ta, ma
-; ZVFHMIN-NEXT: vmv.v.x v8, a2
-; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a2, a2, 5
-; ZVFHMIN-NEXT: add a2, sp, a2
-; ZVFHMIN-NEXT: addi a2, a2, 16
-; ZVFHMIN-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vmv.v.x v0, a2
; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v20
; ZVFHMIN-NEXT: csrr a2, vlenb
@@ -9453,18 +9390,27 @@ define <vscale x 32 x half> @vfmsub_vf_nxv32f16_unmasked_commute(<vscale x 32 x
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vmv4r.v v8, v0
; ZVFHMIN-NEXT: csrr a2, vlenb
; ZVFHMIN-NEXT: slli a2, a2, 5
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
-; ZVFHMIN-NEXT: vl8r.v v16, (a2) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v20
+; ZVFHMIN-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v4
; ZVFHMIN-NEXT: csrr a2, vlenb
; ZVFHMIN-NEXT: slli a2, a2, 3
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a2, vlenb
+; ZVFHMIN-NEXT: slli a2, a2, 3
+; ZVFHMIN-NEXT: mv a4, a2
+; ZVFHMIN-NEXT: slli a2, a2, 1
+; ZVFHMIN-NEXT: add a2, a2, a4
+; ZVFHMIN-NEXT: add a2, sp, a2
+; ZVFHMIN-NEXT: addi a2, a2, 16
+; ZVFHMIN-NEXT: vl1r.v v0, (a2) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: csrr a2, vlenb
; ZVFHMIN-NEXT: slli a2, a2, 4
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
@@ -9482,7 +9428,10 @@ define <vscale x 32 x half> @vfmsub_vf_nxv32f16_unmasked_commute(<vscale x 32 x
; ZVFHMIN-NEXT: vl8r.v v16, (a2) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v16
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a2, a2, 4
+; ZVFHMIN-NEXT: slli a2, a2, 3
+; ZVFHMIN-NEXT: mv a3, a2
+; ZVFHMIN-NEXT: slli a2, a2, 1
+; ZVFHMIN-NEXT: add a2, a2, a3
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
; ZVFHMIN-NEXT: vs8r.v v24, (a2) # Unknown-size Folded Spill
@@ -9493,7 +9442,7 @@ define <vscale x 32 x half> @vfmsub_vf_nxv32f16_unmasked_commute(<vscale x 32 x
; ZVFHMIN-NEXT: csrr a1, vlenb
; ZVFHMIN-NEXT: slli a1, a1, 3
; ZVFHMIN-NEXT: mv a2, a1
-; ZVFHMIN-NEXT: slli a1, a1, 1
+; ZVFHMIN-NEXT: slli a1, a1, 2
; ZVFHMIN-NEXT: add a1, a1, a2
; ZVFHMIN-NEXT: add a1, sp, a1
; ZVFHMIN-NEXT: addi a1, a1, 16
@@ -9506,7 +9455,10 @@ define <vscale x 32 x half> @vfmsub_vf_nxv32f16_unmasked_commute(<vscale x 32 x
; ZVFHMIN-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v24
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a1, a1, 4
+; ZVFHMIN-NEXT: slli a1, a1, 3
+; ZVFHMIN-NEXT: mv a2, a1
+; ZVFHMIN-NEXT: slli a1, a1, 1
+; ZVFHMIN-NEXT: add 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
@@ -9515,9 +9467,9 @@ define <vscale x 32 x half> @vfmsub_vf_nxv32f16_unmasked_commute(<vscale x 32 x
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v0
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 3
+; ZVFHMIN-NEXT: slli a0, a0, 4
; ZVFHMIN-NEXT: mv a1, a0
-; ZVFHMIN-NEXT: slli a0, a0, 2
+; ZVFHMIN-NEXT: slli a0, a0, 1
; ZVFHMIN-NEXT: add a0, a0, a1
; ZVFHMIN-NEXT: add sp, sp, a0
; ZVFHMIN-NEXT: .cfi_def_cfa sp, 16
@@ -9819,124 +9771,115 @@ define <vscale x 32 x half> @vfnmadd_vv_nxv32f16_unmasked(<vscale x 32 x half> %
; ZVFHMIN-NEXT: addi sp, sp, -16
; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
; ZVFHMIN-NEXT: csrr a2, vlenb
+; ZVFHMIN-NEXT: slli a2, a2, 4
; ZVFHMIN-NEXT: mv a3, a2
-; ZVFHMIN-NEXT: slli a2, a2, 3
-; ZVFHMIN-NEXT: add a3, a3, a2
-; ZVFHMIN-NEXT: slli a2, a2, 2
+; ZVFHMIN-NEXT: slli a2, a2, 1
; ZVFHMIN-NEXT: add a2, a2, a3
; ZVFHMIN-NEXT: sub sp, sp, a2
-; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x29, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 41 * vlenb
+; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x30, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 48 * vlenb
; ZVFHMIN-NEXT: vl8re16.v v24, (a0)
+; ZVFHMIN-NEXT: lui a2, 8
+; ZVFHMIN-NEXT: vsetvli a0, zero, e8, m4, ta, ma
+; ZVFHMIN-NEXT: vmset.m v7
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: mv a2, a0
; ZVFHMIN-NEXT: slli a0, a0, 3
-; ZVFHMIN-NEXT: add a2, a2, a0
-; ZVFHMIN-NEXT: slli a0, a0, 1
-; ZVFHMIN-NEXT: add a0, a0, a2
+; ZVFHMIN-NEXT: mv a3, a0
+; ZVFHMIN-NEXT: slli a0, a0, 2
+; ZVFHMIN-NEXT: add a0, a0, a3
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: lui a2, 8
-; ZVFHMIN-NEXT: vsetvli a0, zero, e8, m4, ta, ma
-; ZVFHMIN-NEXT: vmset.m v7
+; ZVFHMIN-NEXT: vs1r.v v7, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a3, vlenb
; ZVFHMIN-NEXT: vsetvli zero, a1, e16, m8, ta, ma
-; ZVFHMIN-NEXT: vxor.vx v16, v16, a2
+; ZVFHMIN-NEXT: vxor.vx v0, v16, a2
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a4, a0, 5
-; ZVFHMIN-NEXT: add a0, a4, a0
+; ZVFHMIN-NEXT: slli a0, a0, 3
+; ZVFHMIN-NEXT: mv a4, a0
+; ZVFHMIN-NEXT: slli a0, a0, 1
+; ZVFHMIN-NEXT: add a0, a0, a4
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vs8r.v v0, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: slli a0, a3, 1
; ZVFHMIN-NEXT: srli a3, a3, 2
; ZVFHMIN-NEXT: sub a4, a1, a0
+; ZVFHMIN-NEXT: csrr a5, vlenb
+; ZVFHMIN-NEXT: slli a5, a5, 3
+; ZVFHMIN-NEXT: mv a6, a5
+; ZVFHMIN-NEXT: slli a5, a5, 2
+; ZVFHMIN-NEXT: add a5, a5, a6
+; ZVFHMIN-NEXT: add a5, sp, a5
+; ZVFHMIN-NEXT: addi a5, a5, 16
+; ZVFHMIN-NEXT: vl1r.v v16, (a5) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli a5, zero, e8, mf2, ta, ma
-; ZVFHMIN-NEXT: vslidedown.vx v16, v7, a3
+; ZVFHMIN-NEXT: vslidedown.vx v16, v16, a3
; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: slli a3, a3, 4
+; ZVFHMIN-NEXT: slli a3, a3, 5
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
; ZVFHMIN-NEXT: vs1r.v v16, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: sltu a3, a1, a4
-; ZVFHMIN-NEXT: csrr a5, vlenb
-; ZVFHMIN-NEXT: mv a6, a5
-; ZVFHMIN-NEXT: slli a5, a5, 3
-; ZVFHMIN-NEXT: add a6, a6, a5
-; ZVFHMIN-NEXT: slli a5, a5, 1
-; ZVFHMIN-NEXT: add 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: vsetvli zero, a1, e16, m8, ta, ma
-; ZVFHMIN-NEXT: vxor.vx v0, v16, a2
-; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a2, a2, 3
-; ZVFHMIN-NEXT: add a2, sp, a2
-; ZVFHMIN-NEXT: addi a2, a2, 16
-; ZVFHMIN-NEXT: vs8r.v v0, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vxor.vx v16, v24, a2
+; ZVFHMIN-NEXT: addi a2, sp, 16
+; ZVFHMIN-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: addi a3, a3, -1
; ZVFHMIN-NEXT: and a3, a3, a4
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: mv a4, a2
; ZVFHMIN-NEXT: slli a2, a2, 3
-; ZVFHMIN-NEXT: add a4, a4, a2
-; ZVFHMIN-NEXT: slli a2, a2, 1
+; ZVFHMIN-NEXT: mv a4, a2
+; ZVFHMIN-NEXT: slli a2, a2, 2
; ZVFHMIN-NEXT: add a2, a2, a4
; 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 v16, v12
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a4, a2, 4
-; ZVFHMIN-NEXT: add a2, a4, a2
+; ZVFHMIN-NEXT: slli a2, a2, 4
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
-; ZVFHMIN-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vs8r.v v24, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v4
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v20
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a4, a2, 5
-; ZVFHMIN-NEXT: add a2, a4, a2
+; ZVFHMIN-NEXT: slli a2, a2, 3
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
-; ZVFHMIN-NEXT: vl8r.v v8, (a2) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v4
-; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a2, a2, 4
+; ZVFHMIN-NEXT: slli a2, a2, 5
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
; ZVFHMIN-NEXT: vl1r.v v0, (a2) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a4, a2, 4
-; ZVFHMIN-NEXT: add a2, a4, a2
+; ZVFHMIN-NEXT: slli a2, a2, 4
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
-; ZVFHMIN-NEXT: vl8r.v v16, (a2) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vl8r.v v8, (a2) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: csrr a2, vlenb
+; ZVFHMIN-NEXT: slli a2, a2, 3
+; ZVFHMIN-NEXT: add a2, sp, a2
+; ZVFHMIN-NEXT: addi a2, a2, 16
+; ZVFHMIN-NEXT: vl8r.v v16, (a2) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfmadd.vv v24, v16, v8, v0.t
+; ZVFHMIN-NEXT: vfmadd.vv v24, v8, v16, v0.t
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a3, a2, 5
-; ZVFHMIN-NEXT: add a2, a3, a2
+; ZVFHMIN-NEXT: slli a2, a2, 3
+; ZVFHMIN-NEXT: mv a3, a2
+; ZVFHMIN-NEXT: slli a2, a2, 1
+; ZVFHMIN-NEXT: add a2, a2, a3
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
; ZVFHMIN-NEXT: vl8r.v v8, (a2) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v8
; ZVFHMIN-NEXT: vfncvt.f.f.w v12, v24
-; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a2, a2, 3
-; ZVFHMIN-NEXT: add a2, sp, a2
-; ZVFHMIN-NEXT: addi a2, a2, 16
+; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vl8r.v v16, (a2) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v16
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a3, a2, 5
-; ZVFHMIN-NEXT: add a2, a3, a2
+; ZVFHMIN-NEXT: slli a2, a2, 5
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
; ZVFHMIN-NEXT: vs8r.v v24, (a2) # Unknown-size Folded Spill
@@ -9945,18 +9888,16 @@ define <vscale x 32 x half> @vfnmadd_vv_nxv32f16_unmasked(<vscale x 32 x half> %
; ZVFHMIN-NEXT: mv a1, a0
; ZVFHMIN-NEXT: .LBB288_2:
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: mv a2, a0
; ZVFHMIN-NEXT: slli a0, a0, 3
-; ZVFHMIN-NEXT: add a2, a2, a0
-; ZVFHMIN-NEXT: slli a0, a0, 1
+; ZVFHMIN-NEXT: mv a2, a0
+; ZVFHMIN-NEXT: slli a0, a0, 2
; ZVFHMIN-NEXT: add a0, a0, a2
; 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: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a2, a0, 5
-; ZVFHMIN-NEXT: add a0, a2, a0
+; ZVFHMIN-NEXT: slli a0, a0, 5
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
@@ -9965,10 +9906,9 @@ define <vscale x 32 x half> @vfnmadd_vv_nxv32f16_unmasked(<vscale x 32 x half> %
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v0
; ZVFHMIN-NEXT: csrr a0, vlenb
+; ZVFHMIN-NEXT: slli a0, a0, 4
; ZVFHMIN-NEXT: mv a1, a0
-; ZVFHMIN-NEXT: slli a0, a0, 3
-; ZVFHMIN-NEXT: add a1, a1, a0
-; ZVFHMIN-NEXT: slli a0, a0, 2
+; ZVFHMIN-NEXT: slli a0, a0, 1
; ZVFHMIN-NEXT: add a0, a0, a1
; ZVFHMIN-NEXT: add sp, sp, a0
; ZVFHMIN-NEXT: .cfi_def_cfa sp, 16
@@ -9994,109 +9934,101 @@ define <vscale x 32 x half> @vfnmadd_vv_nxv32f16_unmasked_commuted(<vscale x 32
; ZVFHMIN-NEXT: addi sp, sp, -16
; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
; ZVFHMIN-NEXT: csrr a2, vlenb
+; ZVFHMIN-NEXT: slli a2, a2, 4
; ZVFHMIN-NEXT: mv a3, a2
-; ZVFHMIN-NEXT: slli a2, a2, 3
-; ZVFHMIN-NEXT: add a3, a3, a2
-; ZVFHMIN-NEXT: slli a2, a2, 2
+; ZVFHMIN-NEXT: slli a2, a2, 1
; ZVFHMIN-NEXT: add a2, a2, a3
; ZVFHMIN-NEXT: sub sp, sp, a2
-; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x29, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 41 * vlenb
-; ZVFHMIN-NEXT: vmv8r.v v0, v16
-; ZVFHMIN-NEXT: vmv8r.v v16, v8
+; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x30, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 48 * vlenb
+; ZVFHMIN-NEXT: vmv8r.v v0, v8
; ZVFHMIN-NEXT: vl8re16.v v24, (a0)
; ZVFHMIN-NEXT: lui a2, 8
; ZVFHMIN-NEXT: vsetvli a0, zero, e8, m4, ta, ma
; ZVFHMIN-NEXT: vmset.m v8
; ZVFHMIN-NEXT: csrr a3, vlenb
; ZVFHMIN-NEXT: vsetvli zero, a1, e16, m8, ta, ma
-; ZVFHMIN-NEXT: vxor.vx v0, v0, a2
+; ZVFHMIN-NEXT: vxor.vx v16, v16, a2
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a4, a0, 5
-; ZVFHMIN-NEXT: add a0, a4, a0
+; ZVFHMIN-NEXT: slli a0, a0, 3
+; ZVFHMIN-NEXT: mv a4, a0
+; ZVFHMIN-NEXT: slli a0, a0, 1
+; ZVFHMIN-NEXT: add a0, a0, a4
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vs8r.v v0, (a0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: slli a0, a3, 1
; ZVFHMIN-NEXT: srli a3, a3, 2
; ZVFHMIN-NEXT: sub a4, a1, a0
; ZVFHMIN-NEXT: vsetvli a5, zero, e8, mf2, ta, ma
; ZVFHMIN-NEXT: vslidedown.vx v8, v8, a3
; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: slli a3, a3, 4
+; ZVFHMIN-NEXT: slli a3, a3, 5
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
; ZVFHMIN-NEXT: vs1r.v v8, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: sltu a3, a1, a4
; ZVFHMIN-NEXT: vsetvli zero, a1, e16, m8, ta, ma
-; ZVFHMIN-NEXT: vxor.vx v0, v24, a2
-; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a2, a2, 3
-; ZVFHMIN-NEXT: add a2, sp, a2
-; ZVFHMIN-NEXT: addi a2, a2, 16
-; ZVFHMIN-NEXT: vs8r.v v0, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vxor.vx v24, v24, a2
+; ZVFHMIN-NEXT: addi a2, sp, 16
+; ZVFHMIN-NEXT: vs8r.v v24, (a2) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: addi a3, a3, -1
; ZVFHMIN-NEXT: and a3, a3, a4
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: mv a4, a2
; ZVFHMIN-NEXT: slli a2, a2, 3
-; ZVFHMIN-NEXT: add a4, a4, a2
-; ZVFHMIN-NEXT: slli a2, a2, 1
+; ZVFHMIN-NEXT: mv a4, a2
+; ZVFHMIN-NEXT: slli a2, a2, 2
; ZVFHMIN-NEXT: add a2, a2, a4
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
-; ZVFHMIN-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
+; 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 v8, v20
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v4
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v20
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a4, a2, 5
-; ZVFHMIN-NEXT: add a2, a4, a2
+; ZVFHMIN-NEXT: slli a2, a2, 4
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
-; ZVFHMIN-NEXT: vl8r.v v16, (a2) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20
+; ZVFHMIN-NEXT: vs8r.v v0, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v28
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a4, a2, 4
-; ZVFHMIN-NEXT: add a2, a4, a2
+; ZVFHMIN-NEXT: slli a2, a2, 3
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
-; ZVFHMIN-NEXT: vs8r.v v24, (a2) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v4
-; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a2, a2, 4
+; ZVFHMIN-NEXT: slli a2, a2, 5
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
; ZVFHMIN-NEXT: vl1r.v v0, (a2) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a4, a2, 4
-; ZVFHMIN-NEXT: add a2, a4, a2
+; ZVFHMIN-NEXT: slli a2, a2, 4
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
-; ZVFHMIN-NEXT: vl8r.v v24, (a2) # Unknown-size Folded Reload
-; 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: vfmadd.vv v8, v24, v16, v0.t
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a3, a2, 5
-; ZVFHMIN-NEXT: add a2, a3, a2
+; ZVFHMIN-NEXT: slli a2, a2, 3
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
-; ZVFHMIN-NEXT: vl8r.v v16, (a2) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v16
+; ZVFHMIN-NEXT: vl8r.v v24, (a2) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
+; ZVFHMIN-NEXT: vfmadd.vv v8, v16, v24, v0.t
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a3, a2, 4
-; ZVFHMIN-NEXT: add a2, a3, a2
+; ZVFHMIN-NEXT: slli a2, a2, 3
+; ZVFHMIN-NEXT: mv a3, a2
+; ZVFHMIN-NEXT: slli a2, a2, 1
+; ZVFHMIN-NEXT: add a2, a2, a3
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
-; ZVFHMIN-NEXT: vs8r.v v0, (a2) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vfncvt.f.f.w v28, v8
+; ZVFHMIN-NEXT: vl8r.v v0, (a2) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v0
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a2, a2, 3
+; ZVFHMIN-NEXT: slli a2, a2, 5
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
+; ZVFHMIN-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vfncvt.f.f.w v28, v8
+; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vl8r.v v8, (a2) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8
; ZVFHMIN-NEXT: bltu a1, a0, .LBB289_2
@@ -10104,18 +10036,16 @@ define <vscale x 32 x half> @vfnmadd_vv_nxv32f16_unmasked_commuted(<vscale x 32
; ZVFHMIN-NEXT: mv a1, a0
; ZVFHMIN-NEXT: .LBB289_2:
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: mv a2, a0
; ZVFHMIN-NEXT: slli a0, a0, 3
-; ZVFHMIN-NEXT: add a2, a2, a0
-; ZVFHMIN-NEXT: slli a0, a0, 1
+; ZVFHMIN-NEXT: mv a2, a0
+; ZVFHMIN-NEXT: slli a0, a0, 2
; ZVFHMIN-NEXT: add a0, a0, a2
; 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 v8, v0
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a2, a0, 4
-; ZVFHMIN-NEXT: add a0, a2, a0
+; ZVFHMIN-NEXT: slli a0, a0, 5
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vl8r.v v0, (a0) # Unknown-size Folded Reload
@@ -10125,10 +10055,9 @@ define <vscale x 32 x half> @vfnmadd_vv_nxv32f16_unmasked_commuted(<vscale x 32
; ZVFHMIN-NEXT: vfncvt.f.f.w v24, v8
; ZVFHMIN-NEXT: vmv8r.v v8, v24
; ZVFHMIN-NEXT: csrr a0, vlenb
+; ZVFHMIN-NEXT: slli a0, a0, 4
; ZVFHMIN-NEXT: mv a1, a0
-; ZVFHMIN-NEXT: slli a0, a0, 3
-; ZVFHMIN-NEXT: add a1, a1, a0
-; ZVFHMIN-NEXT: slli a0, a0, 2
+; ZVFHMIN-NEXT: slli a0, a0, 1
; ZVFHMIN-NEXT: add a0, a0, a1
; ZVFHMIN-NEXT: add sp, sp, a0
; ZVFHMIN-NEXT: .cfi_def_cfa sp, 16
@@ -10174,8 +10103,7 @@ define <vscale x 32 x half> @vfnmadd_vf_nxv32f16(<vscale x 32 x half> %va, half
; ZVFHMIN-NEXT: vxor.vx v8, v8, a4, v0.t
; ZVFHMIN-NEXT: vxor.vx v16, v16, a4, v0.t
; ZVFHMIN-NEXT: slli a2, a1, 1
-; ZVFHMIN-NEXT: addi a4, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v16, (a4) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vmv4r.v v4, v20
; ZVFHMIN-NEXT: vsetvli a4, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v16
; ZVFHMIN-NEXT: csrr a4, vlenb
@@ -10186,7 +10114,8 @@ define <vscale x 32 x half> @vfnmadd_vf_nxv32f16(<vscale x 32 x half> %va, half
; ZVFHMIN-NEXT: add a4, sp, a4
; ZVFHMIN-NEXT: addi a4, a4, 16
; ZVFHMIN-NEXT: vs8r.v v24, (a4) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vmv4r.v v4, v12
+; ZVFHMIN-NEXT: addi a4, sp, 16
+; ZVFHMIN-NEXT: vs8r.v v8, (a4) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8
; ZVFHMIN-NEXT: csrr a4, vlenb
; ZVFHMIN-NEXT: slli a4, a4, 4
@@ -10223,10 +10152,8 @@ define <vscale x 32 x half> @vfnmadd_vf_nxv32f16(<vscale x 32 x half> %va, half
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
; ZVFHMIN-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: addi a3, sp, 16
-; ZVFHMIN-NEXT: vl8r.v v16, (a3) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v20
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v4
; ZVFHMIN-NEXT: csrr a3, vlenb
; ZVFHMIN-NEXT: slli a3, a3, 3
; ZVFHMIN-NEXT: mv a4, a3
@@ -10235,7 +10162,9 @@ define <vscale x 32 x half> @vfnmadd_vf_nxv32f16(<vscale x 32 x half> %va, half
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v4
+; ZVFHMIN-NEXT: addi a3, sp, 16
+; ZVFHMIN-NEXT: vl8r.v v16, (a3) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v20
; ZVFHMIN-NEXT: csrr a3, vlenb
; ZVFHMIN-NEXT: slli a3, a3, 4
; ZVFHMIN-NEXT: add a3, sp, a3
@@ -10456,17 +10385,9 @@ define <vscale x 32 x half> @vfnmadd_vf_nxv32f16_unmasked(<vscale x 32 x half> %
; ZVFHMIN-NEXT: fmv.x.h a1, fa0
; ZVFHMIN-NEXT: lui a2, 8
; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m8, ta, ma
-; ZVFHMIN-NEXT: vmset.m v7
+; ZVFHMIN-NEXT: vmset.m v24
; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: vmv.v.x v24, a1
-; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a1, a1, 3
-; ZVFHMIN-NEXT: mv a4, a1
-; ZVFHMIN-NEXT: slli a1, a1, 1
-; ZVFHMIN-NEXT: add a1, a1, a4
-; ZVFHMIN-NEXT: add a1, sp, a1
-; ZVFHMIN-NEXT: addi a1, a1, 16
-; ZVFHMIN-NEXT: vs8r.v v24, (a1) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vmv.v.x v0, a1
; ZVFHMIN-NEXT: vsetvli zero, a0, e16, m8, ta, ma
; ZVFHMIN-NEXT: vxor.vx v8, v8, a2
; ZVFHMIN-NEXT: csrr a1, vlenb
@@ -10481,17 +10402,17 @@ define <vscale x 32 x half> @vfnmadd_vf_nxv32f16_unmasked(<vscale x 32 x half> %
; ZVFHMIN-NEXT: srli a3, a3, 2
; ZVFHMIN-NEXT: sub a2, a0, a1
; ZVFHMIN-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
-; ZVFHMIN-NEXT: vslidedown.vx v0, v7, a3
+; ZVFHMIN-NEXT: vslidedown.vx v8, v24, a3
+; ZVFHMIN-NEXT: csrr a3, vlenb
+; ZVFHMIN-NEXT: slli a3, a3, 4
+; ZVFHMIN-NEXT: add a3, sp, a3
+; ZVFHMIN-NEXT: addi a3, a3, 16
+; ZVFHMIN-NEXT: vs1r.v v8, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: sltu a3, a0, a2
; ZVFHMIN-NEXT: addi a3, a3, -1
; ZVFHMIN-NEXT: and a2, a3, a2
; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20
-; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: slli a3, a3, 4
-; ZVFHMIN-NEXT: add a3, sp, a3
-; ZVFHMIN-NEXT: addi a3, a3, 16
-; ZVFHMIN-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
; ZVFHMIN-NEXT: csrr a3, vlenb
; ZVFHMIN-NEXT: slli a3, a3, 3
@@ -10500,13 +10421,13 @@ define <vscale x 32 x half> @vfnmadd_vf_nxv32f16_unmasked(<vscale x 32 x half> %
; ZVFHMIN-NEXT: add a3, a3, a4
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
-; ZVFHMIN-NEXT: vl8r.v v24, (a3) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v28
+; ZVFHMIN-NEXT: vs8r.v v0, (a3) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v4
; ZVFHMIN-NEXT: csrr a3, vlenb
; ZVFHMIN-NEXT: slli a3, a3, 4
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
-; ZVFHMIN-NEXT: vl8r.v v24, (a3) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vl1r.v v0, (a3) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, a2, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfmadd.vv v8, v16, v24, v0.t
; ZVFHMIN-NEXT: addi a2, sp, 16
@@ -10581,17 +10502,17 @@ define <vscale x 32 x half> @vfnmadd_vf_nxv32f16_unmasked_commute(<vscale x 32 x
; ZVFHMIN-NEXT: fmv.x.h a1, fa0
; ZVFHMIN-NEXT: lui a2, 8
; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m8, ta, ma
-; ZVFHMIN-NEXT: vmset.m v7
+; ZVFHMIN-NEXT: vmset.m v24
; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: vmv.v.x v24, a1
-; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a1, a1, 3
-; ZVFHMIN-NEXT: mv a4, a1
-; ZVFHMIN-NEXT: slli a1, a1, 1
-; ZVFHMIN-NEXT: add a1, a1, a4
-; ZVFHMIN-NEXT: add a1, sp, a1
-; ZVFHMIN-NEXT: addi a1, a1, 16
-; ZVFHMIN-NEXT: vs8r.v v24, (a1) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: slli a3, a3, 3
+; ZVFHMIN-NEXT: mv a4, a3
+; ZVFHMIN-NEXT: slli a3, a3, 1
+; ZVFHMIN-NEXT: add a3, a3, a4
+; ZVFHMIN-NEXT: add a3, sp, a3
+; ZVFHMIN-NEXT: addi a3, a3, 16
+; ZVFHMIN-NEXT: vs1r.v v24, (a3) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: csrr a3, vlenb
+; ZVFHMIN-NEXT: vmv.v.x v0, a1
; ZVFHMIN-NEXT: vsetvli zero, a0, e16, m8, ta, ma
; ZVFHMIN-NEXT: vxor.vx v8, v8, a2
; ZVFHMIN-NEXT: csrr a1, vlenb
@@ -10599,24 +10520,32 @@ define <vscale x 32 x half> @vfnmadd_vf_nxv32f16_unmasked_commute(<vscale x 32 x
; ZVFHMIN-NEXT: add a1, sp, a1
; ZVFHMIN-NEXT: addi a1, a1, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vxor.vx v16, v16, a2
+; ZVFHMIN-NEXT: vxor.vx v24, v16, a2
; ZVFHMIN-NEXT: addi a1, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vs8r.v v24, (a1) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: slli a1, a3, 1
; ZVFHMIN-NEXT: srli a3, a3, 2
; ZVFHMIN-NEXT: sub a2, a0, a1
+; ZVFHMIN-NEXT: csrr a4, vlenb
+; ZVFHMIN-NEXT: slli a4, a4, 3
+; ZVFHMIN-NEXT: mv a5, a4
+; ZVFHMIN-NEXT: slli a4, a4, 1
+; ZVFHMIN-NEXT: add a4, a4, a5
+; ZVFHMIN-NEXT: add a4, sp, a4
+; ZVFHMIN-NEXT: addi a4, a4, 16
+; ZVFHMIN-NEXT: vl1r.v v16, (a4) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
-; ZVFHMIN-NEXT: vslidedown.vx v0, v7, a3
-; ZVFHMIN-NEXT: sltu a3, a0, a2
-; ZVFHMIN-NEXT: addi a3, a3, -1
-; ZVFHMIN-NEXT: and a2, a3, a2
-; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20
+; ZVFHMIN-NEXT: vslidedown.vx v8, v16, a3
; ZVFHMIN-NEXT: csrr a3, vlenb
; ZVFHMIN-NEXT: slli a3, a3, 4
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
-; ZVFHMIN-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vs1r.v v8, (a3) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: sltu a3, a0, a2
+; ZVFHMIN-NEXT: addi a3, a3, -1
+; ZVFHMIN-NEXT: and a2, a3, a2
+; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m4, ta, ma
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v28
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
; ZVFHMIN-NEXT: csrr a3, vlenb
; ZVFHMIN-NEXT: slli a3, a3, 3
@@ -10625,13 +10554,13 @@ define <vscale x 32 x half> @vfnmadd_vf_nxv32f16_unmasked_commute(<vscale x 32 x
; ZVFHMIN-NEXT: add a3, a3, a4
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
-; ZVFHMIN-NEXT: vl8r.v v16, (a3) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v20
+; ZVFHMIN-NEXT: vs8r.v v0, (a3) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v4
; ZVFHMIN-NEXT: csrr a3, vlenb
; ZVFHMIN-NEXT: slli a3, a3, 4
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
-; ZVFHMIN-NEXT: vl8r.v v16, (a3) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vl1r.v v0, (a3) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, a2, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfmadd.vv v24, v8, v16, v0.t
; ZVFHMIN-NEXT: addi a2, sp, 16
@@ -11000,28 +10929,39 @@ define <vscale x 32 x half> @vfnmadd_vf_nxv32f16_neg_splat_unmasked(<vscale x 32
; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m8, ta, ma
; ZVFHMIN-NEXT: vmset.m v24
; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: vmv.v.x v0, a1
+; ZVFHMIN-NEXT: slli a4, a3, 5
+; ZVFHMIN-NEXT: add a3, a4, a3
+; ZVFHMIN-NEXT: add a3, sp, a3
+; ZVFHMIN-NEXT: addi a3, a3, 16
+; ZVFHMIN-NEXT: vs1r.v v24, (a3) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: csrr a3, vlenb
+; ZVFHMIN-NEXT: vmv.v.x v24, a1
; ZVFHMIN-NEXT: vsetvli zero, a0, e16, m8, ta, ma
; ZVFHMIN-NEXT: vxor.vx v16, v16, a2
-; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a4, a1, 5
-; ZVFHMIN-NEXT: add a1, a4, a1
-; ZVFHMIN-NEXT: add a1, sp, a1
-; ZVFHMIN-NEXT: addi a1, a1, 16
+; ZVFHMIN-NEXT: addi a1, sp, 16
; ZVFHMIN-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: slli a1, a3, 1
; ZVFHMIN-NEXT: srli a3, a3, 2
-; ZVFHMIN-NEXT: vxor.vx v0, v0, a2
+; ZVFHMIN-NEXT: vxor.vx v0, v24, a2
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a2, a2, 3
+; ZVFHMIN-NEXT: slli a2, a2, 4
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
; ZVFHMIN-NEXT: vs8r.v v0, (a2) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: sub a2, a0, a1
+; ZVFHMIN-NEXT: csrr a4, vlenb
+; ZVFHMIN-NEXT: slli a5, a4, 5
+; ZVFHMIN-NEXT: add a4, a5, a4
+; ZVFHMIN-NEXT: add a4, sp, a4
+; ZVFHMIN-NEXT: addi a4, a4, 16
+; ZVFHMIN-NEXT: vl1r.v v24, (a4) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
; ZVFHMIN-NEXT: vslidedown.vx v24, v24, a3
; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: slli a3, a3, 4
+; ZVFHMIN-NEXT: slli a3, a3, 3
+; ZVFHMIN-NEXT: mv a4, a3
+; ZVFHMIN-NEXT: slli a3, a3, 1
+; ZVFHMIN-NEXT: add a3, a3, a4
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
; ZVFHMIN-NEXT: vs1r.v v24, (a3) # Unknown-size Folded Spill
@@ -11029,64 +10969,69 @@ define <vscale x 32 x half> @vfnmadd_vf_nxv32f16_neg_splat_unmasked(<vscale x 32
; ZVFHMIN-NEXT: addi a3, a3, -1
; ZVFHMIN-NEXT: and a2, a3, a2
; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: mv a4, a3
-; ZVFHMIN-NEXT: slli a3, a3, 3
-; ZVFHMIN-NEXT: add a4, a4, a3
-; ZVFHMIN-NEXT: slli a3, a3, 1
-; ZVFHMIN-NEXT: add a3, a3, a4
+; ZVFHMIN-NEXT: slli a4, a3, 5
+; ZVFHMIN-NEXT: add a3, a4, a3
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: slli a4, a3, 4
-; ZVFHMIN-NEXT: add a3, a4, a3
+; ZVFHMIN-NEXT: mv a4, a3
+; ZVFHMIN-NEXT: slli a3, a3, 3
+; ZVFHMIN-NEXT: add a4, a4, a3
+; ZVFHMIN-NEXT: slli a3, a3, 1
+; ZVFHMIN-NEXT: add a3, a3, a4
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
; ZVFHMIN-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v20
; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: slli a4, a3, 5
-; ZVFHMIN-NEXT: add a3, a4, a3
+; ZVFHMIN-NEXT: slli a3, a3, 3
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
-; ZVFHMIN-NEXT: vl8r.v v16, (a3) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v20
-; ZVFHMIN-NEXT: addi a3, sp, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v4
; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: slli a3, a3, 4
+; ZVFHMIN-NEXT: slli a3, a3, 3
+; ZVFHMIN-NEXT: mv a4, a3
+; ZVFHMIN-NEXT: slli a3, a3, 1
+; ZVFHMIN-NEXT: add a3, a3, a4
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
; ZVFHMIN-NEXT: vl1r.v v0, (a3) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: slli a4, a3, 4
-; ZVFHMIN-NEXT: add a3, a4, a3
+; ZVFHMIN-NEXT: mv a4, a3
+; ZVFHMIN-NEXT: slli a3, a3, 3
+; ZVFHMIN-NEXT: add a4, a4, a3
+; ZVFHMIN-NEXT: slli a3, a3, 1
+; ZVFHMIN-NEXT: add a3, a3, a4
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
; ZVFHMIN-NEXT: vl8r.v v24, (a3) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: addi a3, sp, 16
+; ZVFHMIN-NEXT: csrr a3, vlenb
+; ZVFHMIN-NEXT: slli a3, a3, 3
+; ZVFHMIN-NEXT: add a3, sp, a3
+; ZVFHMIN-NEXT: addi a3, a3, 16
; ZVFHMIN-NEXT: vl8r.v v8, (a3) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, a2, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfmadd.vv v16, v24, v8, v0.t
-; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a3, a2, 5
-; ZVFHMIN-NEXT: add a2, a3, a2
-; ZVFHMIN-NEXT: add a2, sp, a2
-; ZVFHMIN-NEXT: addi a2, a2, 16
+; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vl8r.v v8, (a2) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v8
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a3, a2, 4
-; ZVFHMIN-NEXT: add a2, a3, a2
+; ZVFHMIN-NEXT: mv a3, a2
+; ZVFHMIN-NEXT: slli a2, a2, 3
+; ZVFHMIN-NEXT: add a3, a3, a2
+; ZVFHMIN-NEXT: slli a2, a2, 1
+; ZVFHMIN-NEXT: add a2, a2, a3
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
; ZVFHMIN-NEXT: vs8r.v v0, (a2) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vfncvt.f.f.w v28, v16
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a2, a2, 3
+; ZVFHMIN-NEXT: slli a2, a2, 4
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
; ZVFHMIN-NEXT: vl8r.v v8, (a2) # Unknown-size Folded Reload
@@ -11096,18 +11041,18 @@ define <vscale x 32 x half> @vfnmadd_vf_nxv32f16_neg_splat_unmasked(<vscale x 32
; ZVFHMIN-NEXT: mv a0, a1
; ZVFHMIN-NEXT: .LBB296_2:
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: mv a2, a1
-; ZVFHMIN-NEXT: slli a1, a1, 3
-; ZVFHMIN-NEXT: add a2, a2, a1
-; ZVFHMIN-NEXT: slli a1, a1, 1
-; ZVFHMIN-NEXT: add a1, a1, a2
+; ZVFHMIN-NEXT: slli a2, a1, 5
+; ZVFHMIN-NEXT: add a1, a2, a1
; ZVFHMIN-NEXT: add a1, sp, a1
; ZVFHMIN-NEXT: addi a1, a1, 16
; ZVFHMIN-NEXT: vl8r.v v0, (a1) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v0
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a2, a1, 4
-; ZVFHMIN-NEXT: add a1, a2, a1
+; ZVFHMIN-NEXT: mv a2, a1
+; ZVFHMIN-NEXT: slli a1, a1, 3
+; ZVFHMIN-NEXT: add a2, a2, a1
+; ZVFHMIN-NEXT: slli a1, a1, 1
+; ZVFHMIN-NEXT: add 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
@@ -11147,15 +11092,13 @@ define <vscale x 32 x half> @vfnmadd_vf_nxv32f16_neg_splat_unmasked_commute(<vsc
; ZVFHMIN-NEXT: addi sp, sp, -16
; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
; ZVFHMIN-NEXT: csrr a1, vlenb
+; ZVFHMIN-NEXT: slli a1, a1, 4
; ZVFHMIN-NEXT: mv a2, a1
-; ZVFHMIN-NEXT: slli a1, a1, 3
-; ZVFHMIN-NEXT: add a2, a2, a1
-; ZVFHMIN-NEXT: slli a1, a1, 2
+; ZVFHMIN-NEXT: slli a1, a1, 1
; ZVFHMIN-NEXT: add a1, a1, a2
; ZVFHMIN-NEXT: sub sp, sp, a1
-; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x29, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 41 * vlenb
-; ZVFHMIN-NEXT: vmv8r.v v0, v16
-; ZVFHMIN-NEXT: vmv8r.v v16, v8
+; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x30, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 48 * vlenb
+; ZVFHMIN-NEXT: vmv8r.v v0, v8
; ZVFHMIN-NEXT: fmv.x.h a1, fa0
; ZVFHMIN-NEXT: lui a2, 8
; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m8, ta, ma
@@ -11163,26 +11106,25 @@ define <vscale x 32 x half> @vfnmadd_vf_nxv32f16_neg_splat_unmasked_commute(<vsc
; ZVFHMIN-NEXT: csrr a3, vlenb
; ZVFHMIN-NEXT: vmv.v.x v24, a1
; ZVFHMIN-NEXT: vsetvli zero, a0, e16, m8, ta, ma
-; ZVFHMIN-NEXT: vxor.vx v0, v0, a2
+; ZVFHMIN-NEXT: vxor.vx v16, v16, a2
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a4, a1, 5
-; ZVFHMIN-NEXT: add a1, a4, a1
+; ZVFHMIN-NEXT: slli a1, a1, 3
+; ZVFHMIN-NEXT: mv a4, a1
+; ZVFHMIN-NEXT: slli a1, a1, 1
+; ZVFHMIN-NEXT: add a1, a1, a4
; ZVFHMIN-NEXT: add a1, sp, a1
; ZVFHMIN-NEXT: addi a1, a1, 16
-; ZVFHMIN-NEXT: vs8r.v v0, (a1) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: slli a1, a3, 1
; ZVFHMIN-NEXT: srli a3, a3, 2
-; ZVFHMIN-NEXT: vxor.vx v0, v24, a2
-; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a2, a2, 3
-; ZVFHMIN-NEXT: add a2, sp, a2
-; ZVFHMIN-NEXT: addi a2, a2, 16
-; ZVFHMIN-NEXT: vs8r.v v0, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vxor.vx v24, v24, a2
+; ZVFHMIN-NEXT: addi a2, sp, 16
+; ZVFHMIN-NEXT: vs8r.v v24, (a2) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: sub a2, a0, a1
; ZVFHMIN-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
; ZVFHMIN-NEXT: vslidedown.vx v8, v8, a3
; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: slli a3, a3, 4
+; ZVFHMIN-NEXT: slli a3, a3, 5
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
; ZVFHMIN-NEXT: vs1r.v v8, (a3) # Unknown-size Folded Spill
@@ -11190,66 +11132,61 @@ define <vscale x 32 x half> @vfnmadd_vf_nxv32f16_neg_splat_unmasked_commute(<vsc
; ZVFHMIN-NEXT: addi a3, a3, -1
; ZVFHMIN-NEXT: and a2, a3, a2
; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: mv a4, a3
; ZVFHMIN-NEXT: slli a3, a3, 3
-; ZVFHMIN-NEXT: add a4, a4, a3
-; ZVFHMIN-NEXT: slli a3, a3, 1
+; ZVFHMIN-NEXT: mv a4, a3
+; ZVFHMIN-NEXT: slli a3, a3, 2
; ZVFHMIN-NEXT: add a3, a3, a4
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
-; ZVFHMIN-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vs8r.v v0, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v20
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v4
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v20
; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: slli a4, a3, 5
-; ZVFHMIN-NEXT: add a3, a4, a3
+; ZVFHMIN-NEXT: slli a3, a3, 4
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
-; ZVFHMIN-NEXT: vl8r.v v16, (a3) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20
+; ZVFHMIN-NEXT: vs8r.v v0, (a3) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v28
; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: slli a4, a3, 4
-; ZVFHMIN-NEXT: add a3, a4, a3
+; ZVFHMIN-NEXT: slli a3, a3, 3
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
-; ZVFHMIN-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v4
-; ZVFHMIN-NEXT: addi a3, sp, 16
; ZVFHMIN-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: slli a3, a3, 4
+; ZVFHMIN-NEXT: slli a3, a3, 5
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
; ZVFHMIN-NEXT: vl1r.v v0, (a3) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: slli a4, a3, 4
-; ZVFHMIN-NEXT: add a3, a4, a3
+; ZVFHMIN-NEXT: slli a3, a3, 4
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
-; ZVFHMIN-NEXT: vl8r.v v24, (a3) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: addi a3, sp, 16
; ZVFHMIN-NEXT: vl8r.v v16, (a3) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: csrr a3, vlenb
+; ZVFHMIN-NEXT: slli a3, a3, 3
+; ZVFHMIN-NEXT: add a3, sp, a3
+; ZVFHMIN-NEXT: addi a3, a3, 16
+; ZVFHMIN-NEXT: vl8r.v v24, (a3) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, a2, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfmadd.vv v8, v16, v24, v0.t
+; ZVFHMIN-NEXT: vfmadd.vv v8, v24, v16, v0.t
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a3, a2, 5
-; ZVFHMIN-NEXT: add a2, a3, a2
+; ZVFHMIN-NEXT: slli a2, a2, 3
+; ZVFHMIN-NEXT: mv a3, a2
+; ZVFHMIN-NEXT: slli a2, a2, 1
+; ZVFHMIN-NEXT: add a2, a2, a3
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
-; ZVFHMIN-NEXT: vl8r.v v16, (a2) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vl8r.v v0, (a2) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v16
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v0
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a3, a2, 4
-; ZVFHMIN-NEXT: add a2, a3, a2
+; ZVFHMIN-NEXT: slli a2, a2, 5
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
-; ZVFHMIN-NEXT: vs8r.v v0, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vfncvt.f.f.w v28, v8
-; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a2, a2, 3
-; ZVFHMIN-NEXT: add a2, sp, a2
-; ZVFHMIN-NEXT: addi a2, a2, 16
+; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vl8r.v v8, (a2) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8
; ZVFHMIN-NEXT: bltu a0, a1, .LBB297_2
@@ -11257,18 +11194,16 @@ define <vscale x 32 x half> @vfnmadd_vf_nxv32f16_neg_splat_unmasked_commute(<vsc
; ZVFHMIN-NEXT: mv a0, a1
; ZVFHMIN-NEXT: .LBB297_2:
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: mv a2, a1
; ZVFHMIN-NEXT: slli a1, a1, 3
-; ZVFHMIN-NEXT: add a2, a2, a1
-; ZVFHMIN-NEXT: slli a1, a1, 1
+; ZVFHMIN-NEXT: mv a2, a1
+; ZVFHMIN-NEXT: slli a1, a1, 2
; ZVFHMIN-NEXT: add 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: vfwcvt.f.f.v v8, v0
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a2, a1, 4
-; ZVFHMIN-NEXT: add a1, a2, a1
+; ZVFHMIN-NEXT: slli a1, a1, 5
; ZVFHMIN-NEXT: add a1, sp, a1
; ZVFHMIN-NEXT: addi a1, a1, 16
; ZVFHMIN-NEXT: vl8r.v v0, (a1) # Unknown-size Folded Reload
@@ -11278,10 +11213,9 @@ define <vscale x 32 x half> @vfnmadd_vf_nxv32f16_neg_splat_unmasked_commute(<vsc
; ZVFHMIN-NEXT: vfncvt.f.f.w v24, v8
; ZVFHMIN-NEXT: vmv8r.v v8, v24
; ZVFHMIN-NEXT: csrr a0, vlenb
+; ZVFHMIN-NEXT: slli a0, a0, 4
; ZVFHMIN-NEXT: mv a1, a0
-; ZVFHMIN-NEXT: slli a0, a0, 3
-; ZVFHMIN-NEXT: add a1, a1, a0
-; ZVFHMIN-NEXT: slli a0, a0, 2
+; ZVFHMIN-NEXT: slli a0, a0, 1
; ZVFHMIN-NEXT: add a0, a0, a1
; ZVFHMIN-NEXT: add sp, sp, a0
; ZVFHMIN-NEXT: .cfi_def_cfa sp, 16
@@ -11584,124 +11518,115 @@ define <vscale x 32 x half> @vfnmsub_vv_nxv32f16_unmasked(<vscale x 32 x half> %
; ZVFHMIN-NEXT: addi sp, sp, -16
; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
; ZVFHMIN-NEXT: csrr a2, vlenb
+; ZVFHMIN-NEXT: slli a2, a2, 4
; ZVFHMIN-NEXT: mv a3, a2
-; ZVFHMIN-NEXT: slli a2, a2, 3
-; ZVFHMIN-NEXT: add a3, a3, a2
-; ZVFHMIN-NEXT: slli a2, a2, 2
+; ZVFHMIN-NEXT: slli a2, a2, 1
; ZVFHMIN-NEXT: add a2, a2, a3
; ZVFHMIN-NEXT: sub sp, sp, a2
-; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x29, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 41 * vlenb
+; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x30, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 48 * vlenb
; ZVFHMIN-NEXT: vl8re16.v v24, (a0)
+; ZVFHMIN-NEXT: lui a2, 8
+; ZVFHMIN-NEXT: vsetvli a0, zero, e8, m4, ta, ma
+; ZVFHMIN-NEXT: vmset.m v7
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: mv a2, a0
; ZVFHMIN-NEXT: slli a0, a0, 3
-; ZVFHMIN-NEXT: add a2, a2, a0
-; ZVFHMIN-NEXT: slli a0, a0, 1
-; ZVFHMIN-NEXT: add a0, a0, a2
+; ZVFHMIN-NEXT: mv a3, a0
+; ZVFHMIN-NEXT: slli a0, a0, 2
+; ZVFHMIN-NEXT: add a0, a0, a3
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: lui a2, 8
-; ZVFHMIN-NEXT: vsetvli a0, zero, e8, m4, ta, ma
-; ZVFHMIN-NEXT: vmset.m v7
+; ZVFHMIN-NEXT: vs1r.v v7, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a3, vlenb
; ZVFHMIN-NEXT: vsetvli zero, a1, e16, m8, ta, ma
-; ZVFHMIN-NEXT: vxor.vx v16, v16, a2
+; ZVFHMIN-NEXT: vxor.vx v0, v16, a2
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a4, a0, 5
-; ZVFHMIN-NEXT: add a0, a4, a0
+; ZVFHMIN-NEXT: slli a0, a0, 3
+; ZVFHMIN-NEXT: mv a4, a0
+; ZVFHMIN-NEXT: slli a0, a0, 1
+; ZVFHMIN-NEXT: add a0, a0, a4
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vs8r.v v0, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: slli a0, a3, 1
; ZVFHMIN-NEXT: srli a3, a3, 2
; ZVFHMIN-NEXT: sub a4, a1, a0
-; ZVFHMIN-NEXT: vsetvli a5, zero, e8, mf2, ta, ma
-; ZVFHMIN-NEXT: vslidedown.vx v16, v7, a3
-; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: slli a3, a3, 4
-; ZVFHMIN-NEXT: add a3, sp, a3
-; ZVFHMIN-NEXT: addi a3, a3, 16
-; ZVFHMIN-NEXT: vs1r.v v16, (a3) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: sltu a3, a1, a4
; ZVFHMIN-NEXT: csrr a5, vlenb
-; ZVFHMIN-NEXT: mv a6, a5
; ZVFHMIN-NEXT: slli a5, a5, 3
-; ZVFHMIN-NEXT: add a6, a6, a5
-; ZVFHMIN-NEXT: slli a5, a5, 1
+; ZVFHMIN-NEXT: mv a6, a5
+; ZVFHMIN-NEXT: slli a5, a5, 2
; ZVFHMIN-NEXT: add 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: vl1r.v v16, (a5) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vsetvli a5, zero, e8, mf2, ta, ma
+; ZVFHMIN-NEXT: vslidedown.vx v16, v16, a3
+; ZVFHMIN-NEXT: csrr a3, vlenb
+; ZVFHMIN-NEXT: slli a3, a3, 5
+; ZVFHMIN-NEXT: add a3, sp, a3
+; ZVFHMIN-NEXT: addi a3, a3, 16
+; ZVFHMIN-NEXT: vs1r.v v16, (a3) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: sltu a3, a1, a4
; ZVFHMIN-NEXT: vsetvli zero, a1, e16, m8, ta, ma
-; ZVFHMIN-NEXT: vxor.vx v0, v16, a2
-; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a2, a2, 3
-; ZVFHMIN-NEXT: add a2, sp, a2
-; ZVFHMIN-NEXT: addi a2, a2, 16
-; ZVFHMIN-NEXT: vs8r.v v0, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vxor.vx v16, v24, a2
+; ZVFHMIN-NEXT: addi a2, sp, 16
+; ZVFHMIN-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: addi a3, a3, -1
; ZVFHMIN-NEXT: and a3, a3, a4
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: mv a4, a2
; ZVFHMIN-NEXT: slli a2, a2, 3
-; ZVFHMIN-NEXT: add a4, a4, a2
-; ZVFHMIN-NEXT: slli a2, a2, 1
+; ZVFHMIN-NEXT: mv a4, a2
+; ZVFHMIN-NEXT: slli a2, a2, 2
; ZVFHMIN-NEXT: add a2, a2, a4
; 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 v16, v12
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a4, a2, 4
-; ZVFHMIN-NEXT: add a2, a4, a2
+; ZVFHMIN-NEXT: slli a2, a2, 4
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
-; ZVFHMIN-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vs8r.v v24, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v4
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v20
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a4, a2, 5
-; ZVFHMIN-NEXT: add a2, a4, a2
+; ZVFHMIN-NEXT: slli a2, a2, 3
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
-; ZVFHMIN-NEXT: vl8r.v v8, (a2) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v4
-; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a2, a2, 4
+; ZVFHMIN-NEXT: slli a2, a2, 5
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
; ZVFHMIN-NEXT: vl1r.v v0, (a2) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a4, a2, 4
-; ZVFHMIN-NEXT: add a2, a4, a2
+; ZVFHMIN-NEXT: slli a2, a2, 4
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
-; ZVFHMIN-NEXT: vl8r.v v16, (a2) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vl8r.v v8, (a2) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: csrr a2, vlenb
+; ZVFHMIN-NEXT: slli a2, a2, 3
+; ZVFHMIN-NEXT: add a2, sp, a2
+; ZVFHMIN-NEXT: addi a2, a2, 16
+; ZVFHMIN-NEXT: vl8r.v v16, (a2) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfmadd.vv v24, v16, v8, v0.t
+; ZVFHMIN-NEXT: vfmadd.vv v24, v8, v16, v0.t
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a3, a2, 5
-; ZVFHMIN-NEXT: add a2, a3, a2
+; ZVFHMIN-NEXT: slli a2, a2, 3
+; ZVFHMIN-NEXT: mv a3, a2
+; ZVFHMIN-NEXT: slli a2, a2, 1
+; ZVFHMIN-NEXT: add a2, a2, a3
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
; ZVFHMIN-NEXT: vl8r.v v8, (a2) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v8
; ZVFHMIN-NEXT: vfncvt.f.f.w v12, v24
-; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a2, a2, 3
-; ZVFHMIN-NEXT: add a2, sp, a2
-; ZVFHMIN-NEXT: addi a2, a2, 16
+; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vl8r.v v16, (a2) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v16
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a3, a2, 5
-; ZVFHMIN-NEXT: add a2, a3, a2
+; ZVFHMIN-NEXT: slli a2, a2, 5
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
; ZVFHMIN-NEXT: vs8r.v v24, (a2) # Unknown-size Folded Spill
@@ -11710,18 +11635,16 @@ define <vscale x 32 x half> @vfnmsub_vv_nxv32f16_unmasked(<vscale x 32 x half> %
; ZVFHMIN-NEXT: mv a1, a0
; ZVFHMIN-NEXT: .LBB300_2:
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: mv a2, a0
; ZVFHMIN-NEXT: slli a0, a0, 3
-; ZVFHMIN-NEXT: add a2, a2, a0
-; ZVFHMIN-NEXT: slli a0, a0, 1
+; ZVFHMIN-NEXT: mv a2, a0
+; ZVFHMIN-NEXT: slli a0, a0, 2
; ZVFHMIN-NEXT: add a0, a0, a2
; 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: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a2, a0, 5
-; ZVFHMIN-NEXT: add a0, a2, a0
+; ZVFHMIN-NEXT: slli a0, a0, 5
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
@@ -11730,10 +11653,9 @@ define <vscale x 32 x half> @vfnmsub_vv_nxv32f16_unmasked(<vscale x 32 x half> %
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v0
; ZVFHMIN-NEXT: csrr a0, vlenb
+; ZVFHMIN-NEXT: slli a0, a0, 4
; ZVFHMIN-NEXT: mv a1, a0
-; ZVFHMIN-NEXT: slli a0, a0, 3
-; ZVFHMIN-NEXT: add a1, a1, a0
-; ZVFHMIN-NEXT: slli a0, a0, 2
+; ZVFHMIN-NEXT: slli a0, a0, 1
; ZVFHMIN-NEXT: add a0, a0, a1
; ZVFHMIN-NEXT: add sp, sp, a0
; ZVFHMIN-NEXT: .cfi_def_cfa sp, 16
@@ -11759,109 +11681,101 @@ define <vscale x 32 x half> @vfnmsub_vv_nxv32f16_unmasked_commuted(<vscale x 32
; ZVFHMIN-NEXT: addi sp, sp, -16
; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
; ZVFHMIN-NEXT: csrr a2, vlenb
+; ZVFHMIN-NEXT: slli a2, a2, 4
; ZVFHMIN-NEXT: mv a3, a2
-; ZVFHMIN-NEXT: slli a2, a2, 3
-; ZVFHMIN-NEXT: add a3, a3, a2
-; ZVFHMIN-NEXT: slli a2, a2, 2
+; ZVFHMIN-NEXT: slli a2, a2, 1
; ZVFHMIN-NEXT: add a2, a2, a3
; ZVFHMIN-NEXT: sub sp, sp, a2
-; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x29, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 41 * vlenb
-; ZVFHMIN-NEXT: vmv8r.v v0, v16
-; ZVFHMIN-NEXT: vmv8r.v v16, v8
+; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x30, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 48 * vlenb
+; ZVFHMIN-NEXT: vmv8r.v v0, v8
; ZVFHMIN-NEXT: vl8re16.v v24, (a0)
; ZVFHMIN-NEXT: lui a2, 8
; ZVFHMIN-NEXT: vsetvli a0, zero, e8, m4, ta, ma
; ZVFHMIN-NEXT: vmset.m v8
; ZVFHMIN-NEXT: csrr a3, vlenb
; ZVFHMIN-NEXT: vsetvli zero, a1, e16, m8, ta, ma
-; ZVFHMIN-NEXT: vxor.vx v0, v0, a2
+; ZVFHMIN-NEXT: vxor.vx v16, v16, a2
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a4, a0, 5
-; ZVFHMIN-NEXT: add a0, a4, a0
+; ZVFHMIN-NEXT: slli a0, a0, 3
+; ZVFHMIN-NEXT: mv a4, a0
+; ZVFHMIN-NEXT: slli a0, a0, 1
+; ZVFHMIN-NEXT: add a0, a0, a4
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vs8r.v v0, (a0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: slli a0, a3, 1
; ZVFHMIN-NEXT: srli a3, a3, 2
; ZVFHMIN-NEXT: sub a4, a1, a0
; ZVFHMIN-NEXT: vsetvli a5, zero, e8, mf2, ta, ma
; ZVFHMIN-NEXT: vslidedown.vx v8, v8, a3
; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: slli a3, a3, 4
+; ZVFHMIN-NEXT: slli a3, a3, 5
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
; ZVFHMIN-NEXT: vs1r.v v8, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: sltu a3, a1, a4
; ZVFHMIN-NEXT: vsetvli zero, a1, e16, m8, ta, ma
-; ZVFHMIN-NEXT: vxor.vx v0, v24, a2
-; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a2, a2, 3
-; ZVFHMIN-NEXT: add a2, sp, a2
-; ZVFHMIN-NEXT: addi a2, a2, 16
-; ZVFHMIN-NEXT: vs8r.v v0, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vxor.vx v24, v24, a2
+; ZVFHMIN-NEXT: addi a2, sp, 16
+; ZVFHMIN-NEXT: vs8r.v v24, (a2) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: addi a3, a3, -1
; ZVFHMIN-NEXT: and a3, a3, a4
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: mv a4, a2
; ZVFHMIN-NEXT: slli a2, a2, 3
-; ZVFHMIN-NEXT: add a4, a4, a2
-; ZVFHMIN-NEXT: slli a2, a2, 1
+; ZVFHMIN-NEXT: mv a4, a2
+; ZVFHMIN-NEXT: slli a2, a2, 2
; ZVFHMIN-NEXT: add a2, a2, a4
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
-; ZVFHMIN-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
+; 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 v8, v20
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v4
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v20
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a4, a2, 5
-; ZVFHMIN-NEXT: add a2, a4, a2
+; ZVFHMIN-NEXT: slli a2, a2, 4
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
-; ZVFHMIN-NEXT: vl8r.v v16, (a2) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20
+; ZVFHMIN-NEXT: vs8r.v v0, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v28
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a4, a2, 4
-; ZVFHMIN-NEXT: add a2, a4, a2
+; ZVFHMIN-NEXT: slli a2, a2, 3
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
-; ZVFHMIN-NEXT: vs8r.v v24, (a2) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v4
-; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a2, a2, 4
+; ZVFHMIN-NEXT: slli a2, a2, 5
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
; ZVFHMIN-NEXT: vl1r.v v0, (a2) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a4, a2, 4
-; ZVFHMIN-NEXT: add a2, a4, a2
+; ZVFHMIN-NEXT: slli a2, a2, 4
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
-; ZVFHMIN-NEXT: vl8r.v v24, (a2) # Unknown-size Folded Reload
-; 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: vfmadd.vv v8, v24, v16, v0.t
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a3, a2, 5
-; ZVFHMIN-NEXT: add a2, a3, a2
+; ZVFHMIN-NEXT: slli a2, a2, 3
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
-; ZVFHMIN-NEXT: vl8r.v v16, (a2) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v16
+; ZVFHMIN-NEXT: vl8r.v v24, (a2) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
+; ZVFHMIN-NEXT: vfmadd.vv v8, v16, v24, v0.t
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a3, a2, 4
-; ZVFHMIN-NEXT: add a2, a3, a2
+; ZVFHMIN-NEXT: slli a2, a2, 3
+; ZVFHMIN-NEXT: mv a3, a2
+; ZVFHMIN-NEXT: slli a2, a2, 1
+; ZVFHMIN-NEXT: add a2, a2, a3
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
-; ZVFHMIN-NEXT: vs8r.v v0, (a2) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vfncvt.f.f.w v28, v8
+; ZVFHMIN-NEXT: vl8r.v v0, (a2) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v0
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a2, a2, 3
+; ZVFHMIN-NEXT: slli a2, a2, 5
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
+; ZVFHMIN-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vfncvt.f.f.w v28, v8
+; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vl8r.v v8, (a2) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8
; ZVFHMIN-NEXT: bltu a1, a0, .LBB301_2
@@ -11869,18 +11783,16 @@ define <vscale x 32 x half> @vfnmsub_vv_nxv32f16_unmasked_commuted(<vscale x 32
; ZVFHMIN-NEXT: mv a1, a0
; ZVFHMIN-NEXT: .LBB301_2:
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: mv a2, a0
; ZVFHMIN-NEXT: slli a0, a0, 3
-; ZVFHMIN-NEXT: add a2, a2, a0
-; ZVFHMIN-NEXT: slli a0, a0, 1
+; ZVFHMIN-NEXT: mv a2, a0
+; ZVFHMIN-NEXT: slli a0, a0, 2
; ZVFHMIN-NEXT: add a0, a0, a2
; 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 v8, v0
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a2, a0, 4
-; ZVFHMIN-NEXT: add a0, a2, a0
+; ZVFHMIN-NEXT: slli a0, a0, 5
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vl8r.v v0, (a0) # Unknown-size Folded Reload
@@ -11890,10 +11802,9 @@ define <vscale x 32 x half> @vfnmsub_vv_nxv32f16_unmasked_commuted(<vscale x 32
; ZVFHMIN-NEXT: vfncvt.f.f.w v24, v8
; ZVFHMIN-NEXT: vmv8r.v v8, v24
; ZVFHMIN-NEXT: csrr a0, vlenb
+; ZVFHMIN-NEXT: slli a0, a0, 4
; ZVFHMIN-NEXT: mv a1, a0
-; ZVFHMIN-NEXT: slli a0, a0, 3
-; ZVFHMIN-NEXT: add a1, a1, a0
-; ZVFHMIN-NEXT: slli a0, a0, 2
+; ZVFHMIN-NEXT: slli a0, a0, 1
; ZVFHMIN-NEXT: add a0, a0, a1
; ZVFHMIN-NEXT: add sp, sp, a0
; ZVFHMIN-NEXT: .cfi_def_cfa sp, 16
@@ -11918,12 +11829,12 @@ define <vscale x 32 x half> @vfnmsub_vf_nxv32f16(<vscale x 32 x half> %va, half
; ZVFHMIN-NEXT: addi sp, sp, -16
; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a1, a1, 3
+; ZVFHMIN-NEXT: slli a1, a1, 4
; ZVFHMIN-NEXT: mv a2, a1
-; ZVFHMIN-NEXT: slli a1, a1, 2
+; ZVFHMIN-NEXT: slli a1, a1, 1
; ZVFHMIN-NEXT: add a1, a1, a2
; 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: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x30, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 48 * vlenb
; ZVFHMIN-NEXT: vmv8r.v v24, v16
; ZVFHMIN-NEXT: fmv.x.h a2, fa0
; ZVFHMIN-NEXT: lui a3, 8
@@ -11947,7 +11858,7 @@ define <vscale x 32 x half> @vfnmsub_vf_nxv32f16(<vscale x 32 x half> %va, half
; ZVFHMIN-NEXT: csrr a4, vlenb
; ZVFHMIN-NEXT: slli a4, a4, 3
; ZVFHMIN-NEXT: mv a5, a4
-; ZVFHMIN-NEXT: slli a4, a4, 1
+; ZVFHMIN-NEXT: slli a4, a4, 2
; ZVFHMIN-NEXT: add a4, a4, a5
; ZVFHMIN-NEXT: add a4, sp, a4
; ZVFHMIN-NEXT: addi a4, a4, 16
@@ -11962,17 +11873,25 @@ define <vscale x 32 x half> @vfnmsub_vf_nxv32f16(<vscale x 32 x half> %va, half
; ZVFHMIN-NEXT: add a4, sp, a4
; ZVFHMIN-NEXT: addi a4, a4, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a4) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vmv4r.v v4, v28
+; ZVFHMIN-NEXT: csrr a4, vlenb
+; ZVFHMIN-NEXT: slli a4, a4, 3
+; ZVFHMIN-NEXT: mv a5, a4
+; ZVFHMIN-NEXT: slli a4, a4, 1
+; ZVFHMIN-NEXT: add a4, a4, a5
+; ZVFHMIN-NEXT: add a4, sp, a4
+; ZVFHMIN-NEXT: addi a4, a4, 16
+; ZVFHMIN-NEXT: vs8r.v v24, (a4) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a4, vlenb
; ZVFHMIN-NEXT: slli a4, a4, 5
; ZVFHMIN-NEXT: add a4, sp, a4
; 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: vmv4r.v v4, v12
; ZVFHMIN-NEXT: csrr a4, vlenb
; ZVFHMIN-NEXT: slli a4, a4, 3
; ZVFHMIN-NEXT: mv a5, a4
-; ZVFHMIN-NEXT: slli a4, a4, 1
+; ZVFHMIN-NEXT: slli a4, a4, 2
; ZVFHMIN-NEXT: add a4, a4, a5
; ZVFHMIN-NEXT: add a4, sp, a4
; ZVFHMIN-NEXT: addi a4, a4, 16
@@ -11996,25 +11915,28 @@ define <vscale x 32 x half> @vfnmsub_vf_nxv32f16(<vscale x 32 x half> %va, half
; ZVFHMIN-NEXT: csrr a3, vlenb
; ZVFHMIN-NEXT: slli a3, a3, 3
; ZVFHMIN-NEXT: mv a4, a3
-; ZVFHMIN-NEXT: slli a3, a3, 1
+; ZVFHMIN-NEXT: slli a3, a3, 2
; ZVFHMIN-NEXT: add a3, a3, a4
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: sub a2, a0, a2
; ZVFHMIN-NEXT: srli a1, a1, 2
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v4
; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: slli a3, a3, 4
+; ZVFHMIN-NEXT: slli a3, a3, 3
+; ZVFHMIN-NEXT: mv a4, a3
+; ZVFHMIN-NEXT: slli a3, a3, 1
+; ZVFHMIN-NEXT: add a3, a3, a4
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
-; ZVFHMIN-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vl8r.v v8, (a3) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: slli a3, a3, 5
+; ZVFHMIN-NEXT: slli a3, a3, 4
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
-; ZVFHMIN-NEXT: vl8r.v v8, (a3) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
+; ZVFHMIN-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v4
; ZVFHMIN-NEXT: sltu a0, a0, a2
; ZVFHMIN-NEXT: vsetvli a3, zero, e8, mf2, ta, ma
; ZVFHMIN-NEXT: vslidedown.vx v0, v0, a1
@@ -12023,7 +11945,7 @@ define <vscale x 32 x half> @vfnmsub_vf_nxv32f16(<vscale x 32 x half> %va, half
; ZVFHMIN-NEXT: csrr a1, vlenb
; ZVFHMIN-NEXT: slli a1, a1, 3
; ZVFHMIN-NEXT: mv a2, a1
-; ZVFHMIN-NEXT: slli a1, a1, 1
+; ZVFHMIN-NEXT: slli a1, a1, 2
; ZVFHMIN-NEXT: add a1, a1, a2
; ZVFHMIN-NEXT: add a1, sp, a1
; ZVFHMIN-NEXT: addi a1, a1, 16
@@ -12044,9 +11966,9 @@ define <vscale x 32 x half> @vfnmsub_vf_nxv32f16(<vscale x 32 x half> %va, half
; 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, 3
+; ZVFHMIN-NEXT: slli a0, a0, 4
; ZVFHMIN-NEXT: mv a1, a0
-; ZVFHMIN-NEXT: slli a0, a0, 2
+; ZVFHMIN-NEXT: slli a0, a0, 1
; ZVFHMIN-NEXT: add a0, a0, a1
; ZVFHMIN-NEXT: add sp, sp, a0
; ZVFHMIN-NEXT: .cfi_def_cfa sp, 16
@@ -12093,11 +12015,7 @@ define <vscale x 32 x half> @vfnmsub_vf_nxv32f16_commute(<vscale x 32 x half> %v
; ZVFHMIN-NEXT: vxor.vx v8, v8, a3, v0.t
; ZVFHMIN-NEXT: slli a2, a1, 1
; ZVFHMIN-NEXT: mv a3, a0
-; ZVFHMIN-NEXT: csrr a4, vlenb
-; ZVFHMIN-NEXT: slli a4, a4, 4
-; ZVFHMIN-NEXT: add a4, sp, a4
-; ZVFHMIN-NEXT: addi a4, a4, 16
-; ZVFHMIN-NEXT: vs8r.v v8, (a4) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vmv4r.v v4, v12
; ZVFHMIN-NEXT: vsetvli a4, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8
; ZVFHMIN-NEXT: bltu a0, a2, .LBB303_2
@@ -12110,7 +12028,11 @@ define <vscale x 32 x half> @vfnmsub_vf_nxv32f16_commute(<vscale x 32 x half> %v
; ZVFHMIN-NEXT: add a4, sp, a4
; ZVFHMIN-NEXT: addi a4, a4, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a4) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vmv4r.v v4, v28
+; ZVFHMIN-NEXT: csrr a4, vlenb
+; ZVFHMIN-NEXT: slli a4, a4, 4
+; ZVFHMIN-NEXT: add a4, sp, a4
+; ZVFHMIN-NEXT: addi a4, a4, 16
+; ZVFHMIN-NEXT: vs8r.v v24, (a4) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a4, vlenb
; ZVFHMIN-NEXT: slli a4, a4, 3
; ZVFHMIN-NEXT: mv a5, a4
@@ -12129,21 +12051,21 @@ define <vscale x 32 x half> @vfnmsub_vf_nxv32f16_commute(<vscale x 32 x half> %v
; ZVFHMIN-NEXT: vfmadd.vv v16, v24, v8, v0.t
; ZVFHMIN-NEXT: addi a3, sp, 16
; ZVFHMIN-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m4, ta, ma
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v4
+; ZVFHMIN-NEXT: sub a2, a0, a2
+; ZVFHMIN-NEXT: srli a1, a1, 2
; ZVFHMIN-NEXT: csrr a3, vlenb
; ZVFHMIN-NEXT: slli a3, a3, 4
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
; ZVFHMIN-NEXT: vl8r.v v8, (a3) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
-; ZVFHMIN-NEXT: sub a2, a0, a2
-; ZVFHMIN-NEXT: srli a1, a1, 2
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v4
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: slli a3, a3, 4
+; ZVFHMIN-NEXT: slli a3, a3, 3
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
-; ZVFHMIN-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a3, vlenb
; ZVFHMIN-NEXT: slli a3, a3, 3
; ZVFHMIN-NEXT: mv a4, a3
@@ -12153,26 +12075,16 @@ define <vscale x 32 x half> @vfnmsub_vf_nxv32f16_commute(<vscale x 32 x half> %v
; ZVFHMIN-NEXT: addi a3, a3, 16
; ZVFHMIN-NEXT: vl8r.v v8, (a3) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
-; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: slli a3, a3, 3
-; ZVFHMIN-NEXT: add a3, sp, a3
-; ZVFHMIN-NEXT: addi a3, a3, 16
-; ZVFHMIN-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: sltu a0, a0, a2
; ZVFHMIN-NEXT: vsetvli a3, zero, e8, mf2, ta, ma
; ZVFHMIN-NEXT: vslidedown.vx v0, v0, a1
; ZVFHMIN-NEXT: addi a0, a0, -1
; ZVFHMIN-NEXT: and a0, a0, a2
; 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: vl8r.v v8, (a1) # Unknown-size Folded Reload
-; 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: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfmadd.vv v16, v24, v8, v0.t
; ZVFHMIN-NEXT: addi a0, sp, 16
@@ -12237,12 +12149,12 @@ define <vscale x 32 x half> @vfnmsub_vf_nxv32f16_unmasked(<vscale x 32 x half> %
; ZVFHMIN-NEXT: srli a3, a3, 2
; ZVFHMIN-NEXT: sub a4, a0, a1
; ZVFHMIN-NEXT: vsetvli a5, zero, e8, mf2, ta, ma
-; ZVFHMIN-NEXT: vslidedown.vx v0, v7, a3
+; ZVFHMIN-NEXT: vslidedown.vx v16, v7, a3
; ZVFHMIN-NEXT: sltu a3, a0, a4
; ZVFHMIN-NEXT: addi a3, a3, -1
; ZVFHMIN-NEXT: and a3, a3, a4
; ZVFHMIN-NEXT: vsetvli a4, zero, e16, m8, ta, ma
-; ZVFHMIN-NEXT: vmv.v.x v24, a2
+; ZVFHMIN-NEXT: vmv.v.x v0, a2
; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v20
; ZVFHMIN-NEXT: csrr a2, vlenb
@@ -12250,7 +12162,7 @@ define <vscale x 32 x half> @vfnmsub_vf_nxv32f16_unmasked(<vscale x 32 x half> %
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vmv4r.v v8, v24
+; ZVFHMIN-NEXT: vmv4r.v v8, v0
; ZVFHMIN-NEXT: csrr a2, vlenb
; ZVFHMIN-NEXT: slli a2, a2, 3
; ZVFHMIN-NEXT: mv a4, a2
@@ -12259,7 +12171,8 @@ define <vscale x 32 x half> @vfnmsub_vf_nxv32f16_unmasked(<vscale x 32 x half> %
; 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 v8, v28
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v4
+; ZVFHMIN-NEXT: vmv1r.v v0, v16
; ZVFHMIN-NEXT: csrr a2, vlenb
; ZVFHMIN-NEXT: slli a2, a2, 4
; ZVFHMIN-NEXT: add a2, sp, a2
@@ -12364,22 +12277,22 @@ define <vscale x 32 x half> @vfnmsub_vf_nxv32f16_unmasked_commute(<vscale x 32 x
; ZVFHMIN-NEXT: addi a4, a4, 16
; ZVFHMIN-NEXT: vs8r.v v24, (a4) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vsetvli zero, a0, e16, m8, ta, ma
-; ZVFHMIN-NEXT: vxor.vx v16, v8, a1
+; ZVFHMIN-NEXT: vxor.vx v24, v8, a1
; ZVFHMIN-NEXT: addi a1, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vs8r.v v24, (a1) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: slli a1, a3, 1
; ZVFHMIN-NEXT: srli a3, a3, 2
; ZVFHMIN-NEXT: sub a4, a0, a1
; ZVFHMIN-NEXT: vsetvli a5, zero, e8, mf2, ta, ma
-; ZVFHMIN-NEXT: vslidedown.vx v0, v7, a3
+; ZVFHMIN-NEXT: vslidedown.vx v20, v7, a3
; ZVFHMIN-NEXT: sltu a3, a0, a4
; ZVFHMIN-NEXT: addi a3, a3, -1
; ZVFHMIN-NEXT: and a3, a3, a4
; ZVFHMIN-NEXT: vsetvli a4, zero, e16, m8, ta, ma
-; ZVFHMIN-NEXT: vmv.v.x v24, a2
+; ZVFHMIN-NEXT: vmv.v.x v0, a2
; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v20
-; ZVFHMIN-NEXT: vmv4r.v v16, v24
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v28
+; ZVFHMIN-NEXT: vmv4r.v v16, v0
; ZVFHMIN-NEXT: csrr a2, vlenb
; ZVFHMIN-NEXT: slli a2, a2, 3
; ZVFHMIN-NEXT: mv a4, a2
@@ -12388,12 +12301,13 @@ define <vscale x 32 x half> @vfnmsub_vf_nxv32f16_unmasked_commute(<vscale x 32 x
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
; ZVFHMIN-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v28
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v4
; ZVFHMIN-NEXT: csrr a2, vlenb
; ZVFHMIN-NEXT: slli a2, a2, 3
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
-; ZVFHMIN-NEXT: vs8r.v v16, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vs8r.v v24, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vmv1r.v v0, v20
; ZVFHMIN-NEXT: csrr a2, vlenb
; ZVFHMIN-NEXT: slli a2, a2, 4
; ZVFHMIN-NEXT: add a2, sp, a2
@@ -12481,12 +12395,15 @@ define <vscale x 32 x half> @vfnmsub_vf_nxv32f16_neg_splat(<vscale x 32 x half>
; 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, 4
+; ZVFHMIN-NEXT: slli a1, a1, 5
; ZVFHMIN-NEXT: add a1, sp, a1
; ZVFHMIN-NEXT: addi a1, a1, 16
; ZVFHMIN-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a1, a1, 5
+; ZVFHMIN-NEXT: slli a1, a1, 3
+; ZVFHMIN-NEXT: mv a2, a1
+; ZVFHMIN-NEXT: slli a1, a1, 1
+; ZVFHMIN-NEXT: add 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
@@ -12499,15 +12416,7 @@ define <vscale x 32 x half> @vfnmsub_vf_nxv32f16_neg_splat(<vscale x 32 x half>
; ZVFHMIN-NEXT: vxor.vx v24, v16, a3, v0.t
; ZVFHMIN-NEXT: slli a2, a1, 1
; ZVFHMIN-NEXT: mv a3, a0
-; ZVFHMIN-NEXT: vmv4r.v v20, v28
-; ZVFHMIN-NEXT: csrr a4, vlenb
-; ZVFHMIN-NEXT: slli a4, a4, 3
-; ZVFHMIN-NEXT: mv a5, a4
-; ZVFHMIN-NEXT: slli a4, a4, 1
-; ZVFHMIN-NEXT: add a4, a4, a5
-; ZVFHMIN-NEXT: add a4, sp, a4
-; ZVFHMIN-NEXT: addi a4, a4, 16
-; ZVFHMIN-NEXT: vs8r.v v16, (a4) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vmv4r.v v4, v28
; ZVFHMIN-NEXT: vsetvli a4, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v24
; ZVFHMIN-NEXT: bltu a0, a2, .LBB306_2
@@ -12515,56 +12424,54 @@ define <vscale x 32 x half> @vfnmsub_vf_nxv32f16_neg_splat(<vscale x 32 x half>
; ZVFHMIN-NEXT: mv a3, a2
; ZVFHMIN-NEXT: .LBB306_2:
; ZVFHMIN-NEXT: csrr a4, vlenb
-; ZVFHMIN-NEXT: slli a4, a4, 4
+; ZVFHMIN-NEXT: slli a4, a4, 5
; ZVFHMIN-NEXT: add a4, sp, a4
; 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: csrr a4, vlenb
-; ZVFHMIN-NEXT: slli a4, a4, 3
+; ZVFHMIN-NEXT: slli a4, a4, 4
; ZVFHMIN-NEXT: add a4, sp, a4
; ZVFHMIN-NEXT: addi a4, a4, 16
; ZVFHMIN-NEXT: vs8r.v v24, (a4) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vmv4r.v v4, v12
; ZVFHMIN-NEXT: csrr a4, vlenb
-; ZVFHMIN-NEXT: slli a4, a4, 5
+; ZVFHMIN-NEXT: slli a4, a4, 3
+; ZVFHMIN-NEXT: mv a5, a4
+; ZVFHMIN-NEXT: slli a4, a4, 1
+; ZVFHMIN-NEXT: add a4, a4, a5
; ZVFHMIN-NEXT: add a4, sp, a4
; 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: csrr a4, vlenb
-; ZVFHMIN-NEXT: slli a4, a4, 3
+; ZVFHMIN-NEXT: slli a4, a4, 4
; ZVFHMIN-NEXT: add a4, sp, a4
; ZVFHMIN-NEXT: addi a4, a4, 16
; ZVFHMIN-NEXT: vl8r.v v8, (a4) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfmadd.vv v16, v24, v8, v0.t
; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: slli a3, a3, 3
+; ZVFHMIN-NEXT: slli a3, a3, 4
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
; ZVFHMIN-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m4, ta, ma
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v4
+; ZVFHMIN-NEXT: addi a3, sp, 16
+; ZVFHMIN-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: sub a2, a0, a2
+; ZVFHMIN-NEXT: srli a1, a1, 2
; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: slli a3, a3, 3
-; ZVFHMIN-NEXT: mv a4, a3
-; ZVFHMIN-NEXT: slli a3, a3, 1
-; ZVFHMIN-NEXT: add a3, a3, a4
+; ZVFHMIN-NEXT: slli a3, a3, 5
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
; ZVFHMIN-NEXT: vl8r.v v24, (a3) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v28
-; ZVFHMIN-NEXT: sub a2, a0, a2
-; ZVFHMIN-NEXT: srli a1, a1, 2
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v4
-; ZVFHMIN-NEXT: addi a3, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: slli a3, a3, 5
+; ZVFHMIN-NEXT: slli a3, a3, 3
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
-; ZVFHMIN-NEXT: vl8r.v v8, (a3) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
+; ZVFHMIN-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a3, vlenb
; ZVFHMIN-NEXT: slli a3, a3, 3
; ZVFHMIN-NEXT: mv a4, a3
@@ -12572,7 +12479,8 @@ define <vscale x 32 x half> @vfnmsub_vf_nxv32f16_neg_splat(<vscale x 32 x half>
; ZVFHMIN-NEXT: add a3, a3, a4
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
-; ZVFHMIN-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vl8r.v v8, (a3) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
; ZVFHMIN-NEXT: sltu a0, a0, a2
; ZVFHMIN-NEXT: vsetvli a3, zero, e8, mf2, ta, ma
; ZVFHMIN-NEXT: vslidedown.vx v0, v0, a1
@@ -12580,18 +12488,15 @@ define <vscale x 32 x half> @vfnmsub_vf_nxv32f16_neg_splat(<vscale x 32 x half>
; ZVFHMIN-NEXT: and a0, a0, a2
; ZVFHMIN-NEXT: csrr a1, vlenb
; ZVFHMIN-NEXT: slli a1, a1, 3
-; ZVFHMIN-NEXT: mv a2, a1
-; ZVFHMIN-NEXT: slli a1, a1, 1
-; ZVFHMIN-NEXT: add a1, a1, a2
; ZVFHMIN-NEXT: add a1, sp, a1
; ZVFHMIN-NEXT: addi a1, a1, 16
; ZVFHMIN-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: addi a1, sp, 16
-; ZVFHMIN-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vl8r.v v16, (a1) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfmadd.vv v16, v8, v24, v0.t
+; ZVFHMIN-NEXT: vfmadd.vv v16, v24, v8, v0.t
; 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
@@ -12627,14 +12532,14 @@ define <vscale x 32 x half> @vfnmsub_vf_nxv32f16_neg_splat_commute(<vscale x 32
; ZVFHMIN-NEXT: addi sp, sp, -16
; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a1, a1, 3
-; ZVFHMIN-NEXT: mv a2, a1
-; ZVFHMIN-NEXT: slli a1, a1, 2
-; ZVFHMIN-NEXT: add 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: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x20, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 32 * vlenb
; ZVFHMIN-NEXT: csrr a1, vlenb
; ZVFHMIN-NEXT: slli a1, a1, 3
+; ZVFHMIN-NEXT: mv a2, a1
+; ZVFHMIN-NEXT: slli a1, a1, 1
+; ZVFHMIN-NEXT: add 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
@@ -12647,15 +12552,11 @@ define <vscale x 32 x half> @vfnmsub_vf_nxv32f16_neg_splat_commute(<vscale x 32
; ZVFHMIN-NEXT: vxor.vx v16, v16, a3, v0.t
; ZVFHMIN-NEXT: slli a2, a1, 1
; ZVFHMIN-NEXT: mv a3, a0
-; ZVFHMIN-NEXT: csrr a4, vlenb
-; ZVFHMIN-NEXT: slli a4, a4, 4
-; ZVFHMIN-NEXT: add a4, sp, a4
-; ZVFHMIN-NEXT: addi a4, a4, 16
-; ZVFHMIN-NEXT: vs8r.v v16, (a4) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vmv4r.v v4, v20
; ZVFHMIN-NEXT: vsetvli a4, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v16
; ZVFHMIN-NEXT: csrr a4, vlenb
-; ZVFHMIN-NEXT: slli a4, a4, 5
+; ZVFHMIN-NEXT: slli a4, a4, 4
; ZVFHMIN-NEXT: add a4, sp, a4
; ZVFHMIN-NEXT: addi a4, a4, 16
; ZVFHMIN-NEXT: vs8r.v v24, (a4) # Unknown-size Folded Spill
@@ -12665,32 +12566,28 @@ define <vscale x 32 x half> @vfnmsub_vf_nxv32f16_neg_splat_commute(<vscale x 32
; ZVFHMIN-NEXT: .LBB307_2:
; ZVFHMIN-NEXT: csrr a4, vlenb
; ZVFHMIN-NEXT: slli a4, a4, 3
+; ZVFHMIN-NEXT: mv a5, a4
+; ZVFHMIN-NEXT: slli a4, a4, 1
+; ZVFHMIN-NEXT: add a4, a4, a5
; ZVFHMIN-NEXT: add a4, sp, a4
; ZVFHMIN-NEXT: addi a4, a4, 16
; ZVFHMIN-NEXT: vl8r.v v16, (a4) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v16
; ZVFHMIN-NEXT: csrr a4, vlenb
; ZVFHMIN-NEXT: slli a4, a4, 3
-; ZVFHMIN-NEXT: mv a5, a4
-; ZVFHMIN-NEXT: slli a4, a4, 1
-; ZVFHMIN-NEXT: add a4, a4, a5
; ZVFHMIN-NEXT: add a4, sp, a4
; ZVFHMIN-NEXT: addi a4, a4, 16
; ZVFHMIN-NEXT: vs8r.v v24, (a4) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vmv4r.v v4, v20
; ZVFHMIN-NEXT: addi a4, sp, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a4) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v8
; ZVFHMIN-NEXT: csrr a4, vlenb
-; ZVFHMIN-NEXT: slli a4, a4, 5
+; ZVFHMIN-NEXT: slli a4, a4, 4
; ZVFHMIN-NEXT: add a4, sp, a4
; ZVFHMIN-NEXT: addi a4, a4, 16
; ZVFHMIN-NEXT: vl8r.v v16, (a4) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: csrr a4, vlenb
; ZVFHMIN-NEXT: slli a4, a4, 3
-; ZVFHMIN-NEXT: mv a5, a4
-; ZVFHMIN-NEXT: slli a4, a4, 1
-; ZVFHMIN-NEXT: add a4, a4, a5
; ZVFHMIN-NEXT: add a4, sp, a4
; ZVFHMIN-NEXT: addi a4, a4, 16
; ZVFHMIN-NEXT: vl8r.v v8, (a4) # Unknown-size Folded Reload
@@ -12698,27 +12595,28 @@ define <vscale x 32 x half> @vfnmsub_vf_nxv32f16_neg_splat_commute(<vscale x 32
; ZVFHMIN-NEXT: vfmadd.vv v24, v16, v8, v0.t
; ZVFHMIN-NEXT: csrr a3, vlenb
; ZVFHMIN-NEXT: slli a3, a3, 3
-; ZVFHMIN-NEXT: mv a4, a3
-; ZVFHMIN-NEXT: slli a3, a3, 1
-; ZVFHMIN-NEXT: add a3, a3, a4
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
; ZVFHMIN-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: slli a3, a3, 4
-; ZVFHMIN-NEXT: add a3, sp, a3
-; ZVFHMIN-NEXT: addi a3, a3, 16
-; ZVFHMIN-NEXT: vl8r.v v24, (a3) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vmv4r.v v28, v4
; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v28
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v4
; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: slli a3, a3, 5
+; ZVFHMIN-NEXT: slli a3, a3, 4
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
; ZVFHMIN-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: sub a2, a0, a2
; ZVFHMIN-NEXT: srli a1, a1, 2
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v4
+; ZVFHMIN-NEXT: csrr a3, vlenb
+; ZVFHMIN-NEXT: slli a3, a3, 3
+; ZVFHMIN-NEXT: mv a4, a3
+; ZVFHMIN-NEXT: slli a3, a3, 1
+; ZVFHMIN-NEXT: add a3, a3, a4
+; ZVFHMIN-NEXT: add a3, sp, a3
+; ZVFHMIN-NEXT: addi a3, a3, 16
+; ZVFHMIN-NEXT: vl8r.v v16, (a3) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20
; ZVFHMIN-NEXT: addi a3, sp, 16
; ZVFHMIN-NEXT: vl8r.v v8, (a3) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
@@ -12728,7 +12626,7 @@ define <vscale x 32 x half> @vfnmsub_vf_nxv32f16_neg_splat_commute(<vscale x 32
; ZVFHMIN-NEXT: addi a0, a0, -1
; ZVFHMIN-NEXT: and a0, a0, a2
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a1, a1, 5
+; ZVFHMIN-NEXT: slli a1, a1, 4
; ZVFHMIN-NEXT: add a1, sp, a1
; ZVFHMIN-NEXT: addi a1, a1, 16
; ZVFHMIN-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
@@ -12736,9 +12634,6 @@ define <vscale x 32 x half> @vfnmsub_vf_nxv32f16_neg_splat_commute(<vscale x 32
; ZVFHMIN-NEXT: vfmadd.vv v16, v8, v24, v0.t
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 3
-; ZVFHMIN-NEXT: mv a1, a0
-; ZVFHMIN-NEXT: slli a0, a0, 1
-; ZVFHMIN-NEXT: add 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
@@ -12746,10 +12641,7 @@ define <vscale x 32 x half> @vfnmsub_vf_nxv32f16_neg_splat_commute(<vscale x 32
; 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, 3
-; ZVFHMIN-NEXT: mv a1, a0
-; ZVFHMIN-NEXT: slli a0, a0, 2
-; ZVFHMIN-NEXT: add a0, a0, a1
+; ZVFHMIN-NEXT: slli a0, a0, 5
; ZVFHMIN-NEXT: add sp, sp, a0
; ZVFHMIN-NEXT: .cfi_def_cfa sp, 16
; ZVFHMIN-NEXT: addi sp, sp, 16
@@ -12774,77 +12666,99 @@ define <vscale x 32 x half> @vfnmsub_vf_nxv32f16_neg_splat_unmasked(<vscale x 32
; ZVFHMIN-NEXT: addi sp, sp, -16
; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a1, a1, 3
; ZVFHMIN-NEXT: mv a2, a1
+; ZVFHMIN-NEXT: slli a1, a1, 3
+; ZVFHMIN-NEXT: add a2, a2, a1
; ZVFHMIN-NEXT: slli a1, a1, 2
; ZVFHMIN-NEXT: add a1, a1, a2
; 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: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x29, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 41 * vlenb
; ZVFHMIN-NEXT: fmv.x.h a1, fa0
; ZVFHMIN-NEXT: lui a2, 8
; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m8, ta, ma
-; ZVFHMIN-NEXT: vmset.m v7
+; ZVFHMIN-NEXT: vmset.m v24
+; ZVFHMIN-NEXT: csrr a3, vlenb
+; ZVFHMIN-NEXT: slli a4, a3, 5
+; ZVFHMIN-NEXT: add a3, a4, a3
+; ZVFHMIN-NEXT: add a3, sp, a3
+; ZVFHMIN-NEXT: addi a3, a3, 16
+; ZVFHMIN-NEXT: vs1r.v v24, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a3, vlenb
; ZVFHMIN-NEXT: vmv.v.x v24, a1
; ZVFHMIN-NEXT: slli a1, a3, 1
; ZVFHMIN-NEXT: srli a3, a3, 2
; ZVFHMIN-NEXT: vsetvli zero, a0, e16, m8, ta, ma
-; ZVFHMIN-NEXT: vxor.vx v24, v24, a2
-; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a2, a2, 5
-; ZVFHMIN-NEXT: add a2, sp, a2
-; ZVFHMIN-NEXT: addi a2, a2, 16
-; ZVFHMIN-NEXT: vs8r.v v24, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vxor.vx v0, v24, a2
+; ZVFHMIN-NEXT: addi a2, sp, 16
+; ZVFHMIN-NEXT: vs8r.v v0, (a2) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: sub a2, a0, a1
+; ZVFHMIN-NEXT: csrr a4, vlenb
+; ZVFHMIN-NEXT: slli a5, a4, 5
+; ZVFHMIN-NEXT: add a4, a5, a4
+; ZVFHMIN-NEXT: add a4, sp, a4
+; ZVFHMIN-NEXT: addi a4, a4, 16
+; ZVFHMIN-NEXT: vl1r.v v24, (a4) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
-; ZVFHMIN-NEXT: vslidedown.vx v0, v7, a3
+; ZVFHMIN-NEXT: vslidedown.vx v24, v24, a3
+; ZVFHMIN-NEXT: csrr a3, vlenb
+; ZVFHMIN-NEXT: slli a3, a3, 4
+; ZVFHMIN-NEXT: add a3, sp, a3
+; ZVFHMIN-NEXT: addi a3, a3, 16
+; ZVFHMIN-NEXT: vs1r.v v24, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: sltu a3, a0, a2
; ZVFHMIN-NEXT: addi a3, a3, -1
; ZVFHMIN-NEXT: and a2, a3, a2
; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: slli a3, a3, 3
-; ZVFHMIN-NEXT: mv a4, a3
-; ZVFHMIN-NEXT: slli a3, a3, 1
-; ZVFHMIN-NEXT: add a3, a3, a4
+; ZVFHMIN-NEXT: slli a4, a3, 5
+; ZVFHMIN-NEXT: add a3, a4, a3
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
; ZVFHMIN-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20
; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: slli a3, a3, 3
+; ZVFHMIN-NEXT: slli a4, a3, 4
+; ZVFHMIN-NEXT: add a3, a4, a3
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
; ZVFHMIN-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: slli a3, a3, 4
+; ZVFHMIN-NEXT: mv a4, a3
+; ZVFHMIN-NEXT: slli a3, a3, 3
+; ZVFHMIN-NEXT: add a4, a4, a3
+; ZVFHMIN-NEXT: slli a3, a3, 1
+; ZVFHMIN-NEXT: add a3, a3, a4
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
-; ZVFHMIN-NEXT: addi a3, sp, 16
+; ZVFHMIN-NEXT: csrr a3, vlenb
+; ZVFHMIN-NEXT: slli a3, a3, 3
+; ZVFHMIN-NEXT: add a3, sp, a3
+; ZVFHMIN-NEXT: addi a3, a3, 16
; ZVFHMIN-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v4
; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: slli a3, a3, 5
+; ZVFHMIN-NEXT: slli a3, a3, 4
+; ZVFHMIN-NEXT: add a3, sp, a3
+; ZVFHMIN-NEXT: addi a3, a3, 16
+; ZVFHMIN-NEXT: vl1r.v v0, (a3) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: csrr a3, vlenb
+; ZVFHMIN-NEXT: slli a4, a3, 4
+; ZVFHMIN-NEXT: add a3, a4, a3
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
; ZVFHMIN-NEXT: vl8r.v v16, (a3) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v20
; ZVFHMIN-NEXT: csrr a3, vlenb
; ZVFHMIN-NEXT: slli a3, a3, 3
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
-; ZVFHMIN-NEXT: vl8r.v v16, (a3) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: addi a3, sp, 16
; ZVFHMIN-NEXT: vl8r.v v24, (a3) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, a2, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfmadd.vv v8, v24, v16, v0.t
; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v20, v8
-; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a2, a2, 5
-; ZVFHMIN-NEXT: add a2, sp, a2
-; ZVFHMIN-NEXT: addi a2, a2, 16
+; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vl8r.v v8, (a2) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v8
; ZVFHMIN-NEXT: bltu a0, a1, .LBB308_2
@@ -12852,27 +12766,31 @@ define <vscale x 32 x half> @vfnmsub_vf_nxv32f16_neg_splat_unmasked(<vscale x 32
; ZVFHMIN-NEXT: mv a0, a1
; ZVFHMIN-NEXT: .LBB308_2:
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a1, a1, 3
-; ZVFHMIN-NEXT: mv a2, a1
-; ZVFHMIN-NEXT: slli a1, a1, 1
-; ZVFHMIN-NEXT: add a1, a1, a2
+; ZVFHMIN-NEXT: slli a2, a1, 5
+; ZVFHMIN-NEXT: add a1, a2, a1
; ZVFHMIN-NEXT: add a1, sp, a1
; ZVFHMIN-NEXT: addi a1, a1, 16
; ZVFHMIN-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v24
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a1, a1, 5
+; ZVFHMIN-NEXT: slli a2, a1, 4
+; ZVFHMIN-NEXT: add a1, a2, a1
; ZVFHMIN-NEXT: add a1, sp, a1
; ZVFHMIN-NEXT: addi a1, a1, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a1, a1, 4
+; ZVFHMIN-NEXT: mv a2, a1
+; ZVFHMIN-NEXT: slli a1, a1, 3
+; ZVFHMIN-NEXT: add a2, a2, a1
+; ZVFHMIN-NEXT: slli a1, a1, 1
+; ZVFHMIN-NEXT: add a1, a1, a2
; ZVFHMIN-NEXT: add a1, sp, a1
; ZVFHMIN-NEXT: addi a1, a1, 16
; ZVFHMIN-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v8
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a1, a1, 5
+; ZVFHMIN-NEXT: slli a2, a1, 4
+; ZVFHMIN-NEXT: add a1, a2, a1
; ZVFHMIN-NEXT: add a1, sp, a1
; ZVFHMIN-NEXT: addi a1, a1, 16
; ZVFHMIN-NEXT: vl8r.v v8, (a1) # Unknown-size Folded Reload
@@ -12882,8 +12800,9 @@ define <vscale x 32 x half> @vfnmsub_vf_nxv32f16_neg_splat_unmasked(<vscale x 32
; ZVFHMIN-NEXT: vfncvt.f.f.w v16, v0
; ZVFHMIN-NEXT: vmv8r.v v8, v16
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: mv a1, a0
+; ZVFHMIN-NEXT: slli a0, a0, 3
+; ZVFHMIN-NEXT: add a1, a1, a0
; ZVFHMIN-NEXT: slli a0, a0, 2
; ZVFHMIN-NEXT: add a0, a0, a1
; ZVFHMIN-NEXT: add sp, sp, a0
@@ -12910,82 +12829,96 @@ define <vscale x 32 x half> @vfnmsub_vf_nxv32f16_neg_splat_unmasked_commute(<vsc
; ZVFHMIN-NEXT: addi sp, sp, -16
; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a1, a1, 3
+; ZVFHMIN-NEXT: slli a1, a1, 4
; ZVFHMIN-NEXT: mv a2, a1
-; ZVFHMIN-NEXT: slli a1, a1, 2
+; ZVFHMIN-NEXT: slli a1, a1, 1
; ZVFHMIN-NEXT: add a1, a1, a2
; 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: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x30, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 48 * vlenb
+; ZVFHMIN-NEXT: vmv8r.v v0, v8
; ZVFHMIN-NEXT: fmv.x.h a1, fa0
; ZVFHMIN-NEXT: lui a2, 8
; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m8, ta, ma
-; ZVFHMIN-NEXT: vmset.m v7
+; ZVFHMIN-NEXT: vmset.m v8
; ZVFHMIN-NEXT: csrr a3, vlenb
; ZVFHMIN-NEXT: vmv.v.x v24, a1
; ZVFHMIN-NEXT: slli a1, a3, 1
; ZVFHMIN-NEXT: srli a3, a3, 2
; ZVFHMIN-NEXT: vsetvli zero, a0, e16, m8, ta, ma
; ZVFHMIN-NEXT: vxor.vx v24, v24, a2
-; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a2, a2, 5
-; ZVFHMIN-NEXT: add a2, sp, a2
-; ZVFHMIN-NEXT: addi a2, a2, 16
+; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vs8r.v v24, (a2) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: sub a2, a0, a1
; ZVFHMIN-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
-; ZVFHMIN-NEXT: vslidedown.vx v0, v7, a3
+; ZVFHMIN-NEXT: vslidedown.vx v8, v8, a3
+; ZVFHMIN-NEXT: csrr a3, vlenb
+; ZVFHMIN-NEXT: slli a3, a3, 3
+; ZVFHMIN-NEXT: mv a4, a3
+; ZVFHMIN-NEXT: slli a3, a3, 1
+; ZVFHMIN-NEXT: add a3, a3, a4
+; ZVFHMIN-NEXT: add a3, sp, a3
+; ZVFHMIN-NEXT: addi a3, a3, 16
+; ZVFHMIN-NEXT: vs1r.v v8, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: sltu a3, a0, a2
; ZVFHMIN-NEXT: addi a3, a3, -1
; ZVFHMIN-NEXT: and a2, a3, a2
; ZVFHMIN-NEXT: csrr a3, vlenb
; ZVFHMIN-NEXT: slli a3, a3, 3
; ZVFHMIN-NEXT: mv a4, a3
-; ZVFHMIN-NEXT: slli a3, a3, 1
+; ZVFHMIN-NEXT: slli a3, a3, 2
; ZVFHMIN-NEXT: add a3, a3, a4
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
; ZVFHMIN-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v20
+; ZVFHMIN-NEXT: csrr a3, vlenb
+; ZVFHMIN-NEXT: slli a3, a3, 4
+; ZVFHMIN-NEXT: add a3, sp, a3
+; ZVFHMIN-NEXT: addi a3, a3, 16
+; ZVFHMIN-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: csrr a3, vlenb
+; ZVFHMIN-NEXT: slli a3, a3, 5
+; ZVFHMIN-NEXT: add a3, sp, a3
+; ZVFHMIN-NEXT: addi a3, a3, 16
+; ZVFHMIN-NEXT: vs8r.v v0, (a3) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v4
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v28
; ZVFHMIN-NEXT: csrr a3, vlenb
; ZVFHMIN-NEXT: slli a3, a3, 3
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
-; ZVFHMIN-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vmv8r.v v16, v8
+; ZVFHMIN-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: slli a3, a3, 4
+; ZVFHMIN-NEXT: slli a3, a3, 3
+; ZVFHMIN-NEXT: mv a4, a3
+; ZVFHMIN-NEXT: slli a3, a3, 1
+; ZVFHMIN-NEXT: add a3, a3, a4
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
-; ZVFHMIN-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v20
+; ZVFHMIN-NEXT: vl1r.v v0, (a3) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: slli a3, a3, 5
+; ZVFHMIN-NEXT: slli a3, a3, 4
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
; ZVFHMIN-NEXT: vl8r.v v16, (a3) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20
-; ZVFHMIN-NEXT: addi a3, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a3, vlenb
; ZVFHMIN-NEXT: slli a3, a3, 3
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
-; ZVFHMIN-NEXT: vl8r.v v16, (a3) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: addi a3, sp, 16
; ZVFHMIN-NEXT: vl8r.v v24, (a3) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, a2, e32, m8, ta, ma
; ZVFHMIN-NEXT: vfmadd.vv v8, v24, v16, v0.t
; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v20, v8
-; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a2, a2, 5
-; ZVFHMIN-NEXT: add a2, sp, a2
-; ZVFHMIN-NEXT: addi a2, a2, 16
+; ZVFHMIN-NEXT: addi a2, sp, 16
; ZVFHMIN-NEXT: vl8r.v v24, (a2) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v24
; ZVFHMIN-NEXT: csrr a2, vlenb
; ZVFHMIN-NEXT: slli a2, a2, 3
+; ZVFHMIN-NEXT: mv a3, a2
+; ZVFHMIN-NEXT: slli a2, a2, 1
+; ZVFHMIN-NEXT: add a2, a2, a3
; ZVFHMIN-NEXT: add a2, sp, a2
; ZVFHMIN-NEXT: addi a2, a2, 16
; ZVFHMIN-NEXT: vs8r.v v8, (a2) # Unknown-size Folded Spill
@@ -12996,20 +12929,23 @@ define <vscale x 32 x half> @vfnmsub_vf_nxv32f16_neg_splat_unmasked_commute(<vsc
; ZVFHMIN-NEXT: csrr a1, vlenb
; ZVFHMIN-NEXT: slli a1, a1, 3
; ZVFHMIN-NEXT: mv a2, a1
-; ZVFHMIN-NEXT: slli a1, a1, 1
+; ZVFHMIN-NEXT: slli a1, a1, 2
; ZVFHMIN-NEXT: add 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: vfwcvt.f.f.v v8, v24
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a1, a1, 4
+; ZVFHMIN-NEXT: slli a1, a1, 5
; ZVFHMIN-NEXT: add a1, sp, a1
; ZVFHMIN-NEXT: addi a1, a1, 16
; ZVFHMIN-NEXT: vl8r.v v0, (a1) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v0
; ZVFHMIN-NEXT: csrr a1, vlenb
; ZVFHMIN-NEXT: slli a1, a1, 3
+; ZVFHMIN-NEXT: mv a2, a1
+; ZVFHMIN-NEXT: slli a1, a1, 1
+; ZVFHMIN-NEXT: add 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
@@ -13019,9 +12955,9 @@ define <vscale x 32 x half> @vfnmsub_vf_nxv32f16_neg_splat_unmasked_commute(<vsc
; ZVFHMIN-NEXT: vfncvt.f.f.w v16, v24
; ZVFHMIN-NEXT: vmv8r.v v8, v16
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 3
+; ZVFHMIN-NEXT: slli a0, a0, 4
; ZVFHMIN-NEXT: mv a1, a0
-; ZVFHMIN-NEXT: slli a0, a0, 2
+; ZVFHMIN-NEXT: slli a0, a0, 1
; ZVFHMIN-NEXT: add a0, a0, a1
; ZVFHMIN-NEXT: add sp, sp, a0
; ZVFHMIN-NEXT: .cfi_def_cfa sp, 16
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfmadd-constrained-sdnode.ll b/llvm/test/CodeGen/RISCV/rvv/vfmadd-constrained-sdnode.ll
index 1d471ab2404b17..24ca0469f3073f 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfmadd-constrained-sdnode.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfmadd-constrained-sdnode.ll
@@ -228,66 +228,69 @@ define <vscale x 32 x bfloat> @vfmadd_vv_nxv32bf16(<vscale x 32 x bfloat> %va, <
; 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: vmv8r.v v0, v16
-; 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 v16, (a1) # Unknown-size Folded Spill
-; CHECK-NEXT: vmv8r.v v16, v8
-; CHECK-NEXT: vl8re16.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 v8, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: vsetvli a0, zero, e16, m4, ta, ma
-; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v16
+; CHECK-NEXT: vl8re16.v v16, (a0)
; 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: vfwcvtbf16.f.f.v v24, v0
+; CHECK-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vsetvli a0, zero, e16, m4, ta, ma
+; CHECK-NEXT: vfwcvtbf16.f.f.v v24, 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: vs8r.v v24, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT: vfwcvtbf16.f.f.v v0, v8
+; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v0
+; CHECK-NEXT: addi a0, sp, 16
+; CHECK-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
; 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 v24, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vfwcvtbf16.f.f.v v0, v16
; 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 v8, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vl8r.v v24, (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, zero, e32, m8, ta, ma
-; CHECK-NEXT: vfmadd.vv v0, v8, v24
+; CHECK-NEXT: vfmadd.vv v0, v16, v24
; CHECK-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v20
+; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v12
; 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: 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: vfwcvtbf16.f.f.v v8, v20
; 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 v24, (a0) # Unknown-size Folded Reload
-; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v28
+; CHECK-NEXT: vfwcvtbf16.f.f.v v8, v28
; 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 v24, (a0) # Unknown-size Folded Reload
+; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v28
+; 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 v24, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; CHECK-NEXT: vfmadd.vv v16, v8, v24
; CHECK-NEXT: vsetvli zero, zero, e16, m4, ta, ma
@@ -665,66 +668,69 @@ define <vscale x 32 x half> @vfmadd_vv_nxv32f16(<vscale x 32 x half> %va, <vscal
; ZVFHMIN-NEXT: sub sp, sp, a1
; 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: addi a1, sp, 16
+; 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 v16, v8
-; ZVFHMIN-NEXT: vl8re16.v v8, (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, v16
+; ZVFHMIN-NEXT: vl8re16.v v16, (a0)
; 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 v24, (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 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 v0, v8
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v0
+; ZVFHMIN-NEXT: addi a0, sp, 16
+; ZVFHMIN-NEXT: vs8r.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: 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: vfwcvt.f.f.v v0, v16
; 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 v8, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: addi a0, sp, 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, v8, v24
+; ZVFHMIN-NEXT: vfmadd.vv v0, v16, v24
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: li a1, 24
-; 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: vs8r.v v24, (a0) # Unknown-size Folded Spill
-; 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: 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: vl8r.v v24, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v28
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v28
; 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 v16, 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: vl8r.v v24, (a0) # Unknown-size Folded Reload
; 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
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfmadd-sdnode.ll b/llvm/test/CodeGen/RISCV/rvv/vfmadd-sdnode.ll
index 88fd81a5a2f7bc..b78c4aad2fc4d5 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfmadd-sdnode.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfmadd-sdnode.ll
@@ -227,17 +227,12 @@ define <vscale x 32 x bfloat> @vfmadd_vv_nxv32bf16(<vscale x 32 x bfloat> %va, <
; ZVFH-NEXT: sub sp, sp, a1
; ZVFH-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x20, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 32 * vlenb
; ZVFH-NEXT: vmv8r.v v0, v16
-; ZVFH-NEXT: addi a1, sp, 16
+; 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: vs8r.v v16, (a1) # Unknown-size Folded Spill
-; ZVFH-NEXT: vmv8r.v v16, v8
-; ZVFH-NEXT: vl8re16.v v8, (a0)
-; ZVFH-NEXT: csrr a0, vlenb
-; ZVFH-NEXT: slli a0, a0, 4
-; ZVFH-NEXT: add a0, sp, a0
-; ZVFH-NEXT: addi a0, a0, 16
-; ZVFH-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
-; ZVFH-NEXT: vsetvli a0, zero, e16, m4, ta, ma
-; ZVFH-NEXT: vfwcvtbf16.f.f.v v24, v16
+; ZVFH-NEXT: vl8re16.v v16, (a0)
; ZVFH-NEXT: csrr a0, vlenb
; ZVFH-NEXT: slli a0, a0, 3
; ZVFH-NEXT: mv a1, a0
@@ -245,14 +240,17 @@ define <vscale x 32 x bfloat> @vfmadd_vv_nxv32bf16(<vscale x 32 x bfloat> %va, <
; ZVFH-NEXT: add a0, a0, a1
; ZVFH-NEXT: add a0, sp, a0
; ZVFH-NEXT: addi a0, a0, 16
-; ZVFH-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
-; ZVFH-NEXT: vfwcvtbf16.f.f.v v24, v0
+; ZVFH-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
+; ZVFH-NEXT: vsetvli a0, zero, e16, m4, ta, ma
+; ZVFH-NEXT: vfwcvtbf16.f.f.v v24, v8
; ZVFH-NEXT: csrr a0, vlenb
-; ZVFH-NEXT: slli a0, a0, 3
+; ZVFH-NEXT: slli a0, a0, 4
; ZVFH-NEXT: add a0, sp, a0
; ZVFH-NEXT: addi a0, a0, 16
; ZVFH-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
-; ZVFH-NEXT: vfwcvtbf16.f.f.v v0, v8
+; ZVFH-NEXT: vfwcvtbf16.f.f.v v16, v0
+; ZVFH-NEXT: addi a0, sp, 16
+; ZVFH-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
; ZVFH-NEXT: csrr a0, vlenb
; ZVFH-NEXT: slli a0, a0, 3
; ZVFH-NEXT: mv a1, a0
@@ -260,33 +258,30 @@ define <vscale x 32 x bfloat> @vfmadd_vv_nxv32bf16(<vscale x 32 x bfloat> %va, <
; ZVFH-NEXT: add a0, a0, a1
; ZVFH-NEXT: add a0, sp, a0
; ZVFH-NEXT: addi a0, a0, 16
-; ZVFH-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; ZVFH-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
+; ZVFH-NEXT: vfwcvtbf16.f.f.v v0, v16
; ZVFH-NEXT: csrr a0, vlenb
-; ZVFH-NEXT: slli a0, a0, 3
+; ZVFH-NEXT: slli a0, a0, 4
; ZVFH-NEXT: add a0, sp, a0
; ZVFH-NEXT: addi a0, a0, 16
-; ZVFH-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; ZVFH-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; ZVFH-NEXT: addi a0, sp, 16
+; ZVFH-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; ZVFH-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFH-NEXT: vfmadd.vv v0, v8, v24
+; ZVFH-NEXT: vfmadd.vv v0, v16, v24
; ZVFH-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFH-NEXT: vfwcvtbf16.f.f.v v24, v20
+; ZVFH-NEXT: vfwcvtbf16.f.f.v v24, v12
; ZVFH-NEXT: csrr a0, vlenb
-; ZVFH-NEXT: slli a0, a0, 3
-; ZVFH-NEXT: mv a1, a0
-; ZVFH-NEXT: slli a0, a0, 1
-; ZVFH-NEXT: add a0, a0, a1
+; ZVFH-NEXT: slli a0, a0, 4
; ZVFH-NEXT: add a0, sp, a0
; ZVFH-NEXT: addi a0, a0, 16
; ZVFH-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
-; ZVFH-NEXT: addi a0, sp, 16
-; ZVFH-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; ZVFH-NEXT: vfwcvtbf16.f.f.v v8, v20
; ZVFH-NEXT: csrr a0, vlenb
-; ZVFH-NEXT: slli a0, a0, 4
+; ZVFH-NEXT: slli a0, a0, 3
; ZVFH-NEXT: add a0, sp, a0
; ZVFH-NEXT: addi a0, a0, 16
; ZVFH-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
-; ZVFH-NEXT: vfwcvtbf16.f.f.v v16, v28
+; ZVFH-NEXT: vfwcvtbf16.f.f.v v8, v28
; ZVFH-NEXT: csrr a0, vlenb
; ZVFH-NEXT: slli a0, a0, 3
; ZVFH-NEXT: mv a1, a0
@@ -295,6 +290,12 @@ define <vscale x 32 x bfloat> @vfmadd_vv_nxv32bf16(<vscale x 32 x bfloat> %va, <
; ZVFH-NEXT: add a0, sp, a0
; ZVFH-NEXT: addi a0, a0, 16
; ZVFH-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; ZVFH-NEXT: vfwcvtbf16.f.f.v v16, v28
+; ZVFH-NEXT: csrr a0, vlenb
+; ZVFH-NEXT: slli a0, a0, 4
+; ZVFH-NEXT: add a0, sp, a0
+; ZVFH-NEXT: addi a0, a0, 16
+; ZVFH-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
; ZVFH-NEXT: vsetvli zero, zero, e32, m8, ta, ma
; ZVFH-NEXT: vfmadd.vv v16, v8, v24
; ZVFH-NEXT: vsetvli zero, zero, e16, m4, ta, ma
@@ -317,66 +318,69 @@ define <vscale x 32 x bfloat> @vfmadd_vv_nxv32bf16(<vscale x 32 x bfloat> %va, <
; ZVFHMIN-NEXT: sub sp, sp, a1
; 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: addi a1, sp, 16
+; 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 v16, v8
-; ZVFHMIN-NEXT: vl8re16.v v8, (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: vfwcvtbf16.f.f.v v24, v16
+; ZVFHMIN-NEXT: vl8re16.v v16, (a0)
; 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 v24, (a0) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vfwcvtbf16.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: vfwcvtbf16.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: vfwcvtbf16.f.f.v v0, v8
+; ZVFHMIN-NEXT: vfwcvtbf16.f.f.v v16, v0
+; ZVFHMIN-NEXT: addi a0, sp, 16
+; ZVFHMIN-NEXT: vs8r.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: 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: vfwcvtbf16.f.f.v v0, v16
; 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 v8, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: addi a0, sp, 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, v8, v24
+; ZVFHMIN-NEXT: vfmadd.vv v0, v16, v24
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvtbf16.f.f.v v24, v20
+; ZVFHMIN-NEXT: vfwcvtbf16.f.f.v v24, v12
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: li a1, 24
-; 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: vs8r.v v24, (a0) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: addi a0, sp, 16
-; ZVFHMIN-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvtbf16.f.f.v v8, v20
; 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: vl8r.v v24, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvtbf16.f.f.v v16, v28
+; ZVFHMIN-NEXT: vfwcvtbf16.f.f.v v8, v28
; 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: vfwcvtbf16.f.f.v v16, 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: vl8r.v v24, (a0) # Unknown-size Folded Reload
; 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
@@ -399,22 +403,23 @@ define <vscale x 32 x bfloat> @vfmadd_vf_nxv32bf16(<vscale x 32 x bfloat> %va, <
; ZVFH-NEXT: addi sp, sp, -16
; ZVFH-NEXT: .cfi_def_cfa_offset 16
; ZVFH-NEXT: csrr a0, vlenb
-; ZVFH-NEXT: slli a0, a0, 5
+; ZVFH-NEXT: slli a0, a0, 3
+; ZVFH-NEXT: mv a1, a0
+; ZVFH-NEXT: slli a0, a0, 2
+; ZVFH-NEXT: add a0, a0, a1
; ZVFH-NEXT: sub sp, sp, a0
-; ZVFH-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x20, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 32 * vlenb
-; ZVFH-NEXT: vmv8r.v v0, v16
+; ZVFH-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x28, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 40 * vlenb
; ZVFH-NEXT: addi a0, sp, 16
; ZVFH-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
-; ZVFH-NEXT: vmv8r.v v16, v8
; ZVFH-NEXT: fmv.x.h a0, fa0
; ZVFH-NEXT: vsetvli a1, zero, e16, m4, ta, ma
-; ZVFH-NEXT: vfwcvtbf16.f.f.v v24, v16
+; ZVFH-NEXT: vfwcvtbf16.f.f.v v24, v8
; ZVFH-NEXT: csrr a1, vlenb
; ZVFH-NEXT: slli a1, a1, 4
; ZVFH-NEXT: add a1, sp, a1
; ZVFH-NEXT: addi a1, a1, 16
; ZVFH-NEXT: vs8r.v v24, (a1) # Unknown-size Folded Spill
-; ZVFH-NEXT: vfwcvtbf16.f.f.v v24, v0
+; ZVFH-NEXT: vfwcvtbf16.f.f.v v24, v16
; ZVFH-NEXT: csrr a1, vlenb
; ZVFH-NEXT: slli a1, a1, 3
; ZVFH-NEXT: add a1, sp, a1
@@ -439,7 +444,12 @@ define <vscale x 32 x bfloat> @vfmadd_vf_nxv32bf16(<vscale x 32 x bfloat> %va, <
; ZVFH-NEXT: addi a0, a0, 16
; ZVFH-NEXT: vl8r.v v0, (a0) # Unknown-size Folded Reload
; ZVFH-NEXT: vsetvli a0, zero, e16, m4, ta, ma
-; ZVFH-NEXT: vfwcvtbf16.f.f.v v8, v0
+; ZVFH-NEXT: vfwcvtbf16.f.f.v v16, v0
+; ZVFH-NEXT: csrr a0, vlenb
+; ZVFH-NEXT: slli a0, a0, 5
+; ZVFH-NEXT: add a0, sp, a0
+; ZVFH-NEXT: addi a0, a0, 16
+; ZVFH-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
; ZVFH-NEXT: csrr a0, vlenb
; ZVFH-NEXT: slli a0, a0, 4
; ZVFH-NEXT: add a0, sp, a0
@@ -450,16 +460,20 @@ define <vscale x 32 x bfloat> @vfmadd_vf_nxv32bf16(<vscale x 32 x bfloat> %va, <
; ZVFH-NEXT: add a0, sp, a0
; ZVFH-NEXT: addi a0, a0, 16
; ZVFH-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; ZVFH-NEXT: csrr a0, vlenb
+; ZVFH-NEXT: slli a0, a0, 5
+; ZVFH-NEXT: add a0, sp, a0
+; ZVFH-NEXT: addi a0, a0, 16
+; ZVFH-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; ZVFH-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFH-NEXT: vfmadd.vv v8, v24, v0
-; ZVFH-NEXT: vmv.v.v v24, v8
-; ZVFH-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFH-NEXT: vfwcvtbf16.f.f.v v8, v20
+; ZVFH-NEXT: vfmadd.vv v16, v24, v0
; ZVFH-NEXT: csrr a0, vlenb
-; ZVFH-NEXT: slli a0, a0, 4
+; ZVFH-NEXT: slli a0, a0, 5
; ZVFH-NEXT: add a0, sp, a0
; ZVFH-NEXT: addi a0, a0, 16
-; ZVFH-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; ZVFH-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
+; ZVFH-NEXT: vsetvli zero, zero, e16, m4, ta, ma
+; ZVFH-NEXT: vfwcvtbf16.f.f.v v24, v12
; ZVFH-NEXT: addi a0, sp, 16
; ZVFH-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; ZVFH-NEXT: vfwcvtbf16.f.f.v v8, v20
@@ -472,18 +486,21 @@ define <vscale x 32 x bfloat> @vfmadd_vf_nxv32bf16(<vscale x 32 x bfloat> %va, <
; ZVFH-NEXT: addi a0, a0, 16
; ZVFH-NEXT: vl8r.v v0, (a0) # Unknown-size Folded Reload
; ZVFH-NEXT: vfwcvtbf16.f.f.v v16, v4
+; ZVFH-NEXT: vsetvli zero, zero, e32, m8, ta, ma
+; ZVFH-NEXT: vfmadd.vv v16, v8, v24
; ZVFH-NEXT: csrr a0, vlenb
-; ZVFH-NEXT: slli a0, a0, 4
+; ZVFH-NEXT: slli a0, a0, 5
; ZVFH-NEXT: add a0, sp, a0
; ZVFH-NEXT: addi a0, a0, 16
-; ZVFH-NEXT: vl8r.v v0, (a0) # Unknown-size Folded Reload
-; ZVFH-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFH-NEXT: vfmadd.vv v16, v8, v0
+; ZVFH-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
; ZVFH-NEXT: vsetvli zero, zero, e16, m4, ta, ma
; ZVFH-NEXT: vfncvtbf16.f.f.w v8, v24
; ZVFH-NEXT: vfncvtbf16.f.f.w v12, v16
; ZVFH-NEXT: csrr a0, vlenb
-; ZVFH-NEXT: slli a0, a0, 5
+; ZVFH-NEXT: slli a0, a0, 3
+; ZVFH-NEXT: mv a1, a0
+; ZVFH-NEXT: slli a0, a0, 2
+; ZVFH-NEXT: add a0, a0, a1
; ZVFH-NEXT: add sp, sp, a0
; ZVFH-NEXT: .cfi_def_cfa sp, 16
; ZVFH-NEXT: addi sp, sp, 16
@@ -495,22 +512,21 @@ define <vscale x 32 x bfloat> @vfmadd_vf_nxv32bf16(<vscale x 32 x bfloat> %va, <
; 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: li a1, 40
+; ZVFHMIN-NEXT: mul a0, a0, a1
; 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: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x28, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 40 * vlenb
; ZVFHMIN-NEXT: addi a0, sp, 16
; ZVFHMIN-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vmv8r.v v16, v8
; ZVFHMIN-NEXT: fmv.x.h a0, fa0
; ZVFHMIN-NEXT: vsetvli a1, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvtbf16.f.f.v v24, v16
+; ZVFHMIN-NEXT: vfwcvtbf16.f.f.v v24, v8
; 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: vs8r.v v24, (a1) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vfwcvtbf16.f.f.v v24, v0
+; ZVFHMIN-NEXT: vfwcvtbf16.f.f.v v24, v16
; ZVFHMIN-NEXT: csrr a1, vlenb
; ZVFHMIN-NEXT: slli a1, a1, 3
; ZVFHMIN-NEXT: add a1, sp, a1
@@ -531,7 +547,12 @@ define <vscale x 32 x bfloat> @vfmadd_vf_nxv32bf16(<vscale x 32 x bfloat> %va, <
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vl8r.v v0, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvtbf16.f.f.v v8, v0
+; ZVFHMIN-NEXT: vfwcvtbf16.f.f.v v16, v0
+; 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: vs8r.v v16, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 4
; ZVFHMIN-NEXT: add a0, sp, a0
@@ -542,16 +563,20 @@ define <vscale x 32 x bfloat> @vfmadd_vf_nxv32bf16(<vscale x 32 x bfloat> %va, <
; 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, 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: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfmadd.vv v8, v24, v0
-; ZVFHMIN-NEXT: vmv.v.v v24, v8
-; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvtbf16.f.f.v v8, v20
+; ZVFHMIN-NEXT: vfmadd.vv v16, v24, v0
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 4
+; ZVFHMIN-NEXT: slli a0, a0, 5
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
+; ZVFHMIN-NEXT: vfwcvtbf16.f.f.v v24, v12
; ZVFHMIN-NEXT: addi a0, sp, 16
; ZVFHMIN-NEXT: vl8r.v v16, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfwcvtbf16.f.f.v v8, v20
@@ -562,18 +587,19 @@ define <vscale x 32 x bfloat> @vfmadd_vf_nxv32bf16(<vscale x 32 x bfloat> %va, <
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vl8r.v v0, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfwcvtbf16.f.f.v v16, v4
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
+; ZVFHMIN-NEXT: vfmadd.vv v16, v8, v24
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 4
+; ZVFHMIN-NEXT: slli a0, a0, 5
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vl8r.v v0, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfmadd.vv v16, v8, v0
+; ZVFHMIN-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfncvtbf16.f.f.w v8, v24
; ZVFHMIN-NEXT: vfncvtbf16.f.f.w v12, v16
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 5
+; ZVFHMIN-NEXT: li a1, 40
+; ZVFHMIN-NEXT: mul a0, a0, a1
; ZVFHMIN-NEXT: add sp, sp, a0
; ZVFHMIN-NEXT: .cfi_def_cfa sp, 16
; ZVFHMIN-NEXT: addi sp, sp, 16
@@ -876,66 +902,69 @@ define <vscale x 32 x half> @vfmadd_vv_nxv32f16(<vscale x 32 x half> %va, <vscal
; ZVFHMIN-NEXT: sub sp, sp, a1
; 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: addi a1, sp, 16
+; 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 v16, v8
-; ZVFHMIN-NEXT: vl8re16.v v8, (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, v16
+; ZVFHMIN-NEXT: vl8re16.v v16, (a0)
; 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 v24, (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 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 v0, v8
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v0
+; ZVFHMIN-NEXT: addi a0, sp, 16
+; ZVFHMIN-NEXT: vs8r.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: 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: vfwcvt.f.f.v v0, v16
; 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 v8, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: addi a0, sp, 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, v8, v24
+; ZVFHMIN-NEXT: vfmadd.vv v0, v16, v24
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: li a1, 24
-; 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: vs8r.v v24, (a0) # Unknown-size Folded Spill
-; 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: 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: vl8r.v v24, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v28
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v28
; 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 v16, 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: vl8r.v v24, (a0) # Unknown-size Folded Reload
; 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
@@ -964,22 +993,21 @@ 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: li a1, 40
+; ZVFHMIN-NEXT: mul a0, a0, a1
; 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: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x28, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 40 * vlenb
; ZVFHMIN-NEXT: addi a0, sp, 16
; ZVFHMIN-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vmv8r.v v16, v8
; ZVFHMIN-NEXT: fmv.x.h a0, fa0
; ZVFHMIN-NEXT: vsetvli a1, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v16
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v8
; 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: vs8r.v v24, (a1) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v0
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v16
; ZVFHMIN-NEXT: csrr a1, vlenb
; ZVFHMIN-NEXT: slli a1, a1, 3
; ZVFHMIN-NEXT: add a1, sp, a1
@@ -1000,7 +1028,12 @@ define <vscale x 32 x half> @vfmadd_vf_nxv32f16(<vscale x 32 x half> %va, <vscal
; ZVFHMIN-NEXT: addi a0, a0, 16
; ZVFHMIN-NEXT: vl8r.v v0, (a0) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v0
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v0
+; 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: vs8r.v v16, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 4
; ZVFHMIN-NEXT: add a0, sp, a0
@@ -1011,16 +1044,20 @@ define <vscale x 32 x half> @vfmadd_vf_nxv32f16(<vscale x 32 x half> %va, <vscal
; 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, 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: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfmadd.vv v8, v24, v0
-; ZVFHMIN-NEXT: vmv.v.v v24, v8
-; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v20
+; ZVFHMIN-NEXT: vfmadd.vv v16, v24, v0
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 4
+; ZVFHMIN-NEXT: slli a0, a0, 5
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vs8r.v v8, (a0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vs8r.v v16, (a0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
+; 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
@@ -1031,18 +1068,19 @@ define <vscale x 32 x half> @vfmadd_vf_nxv32f16(<vscale x 32 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, v4
+; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
+; ZVFHMIN-NEXT: vfmadd.vv v16, v8, v24
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 4
+; ZVFHMIN-NEXT: slli a0, a0, 5
; ZVFHMIN-NEXT: add a0, sp, a0
; ZVFHMIN-NEXT: addi a0, a0, 16
-; ZVFHMIN-NEXT: vl8r.v v0, (a0) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfmadd.vv v16, v8, v0
+; 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: vfncvt.f.f.w v12, v16
; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 5
+; ZVFHMIN-NEXT: li a1, 40
+; ZVFHMIN-NEXT: mul a0, a0, a1
; ZVFHMIN-NEXT: add sp, sp, a0
; ZVFHMIN-NEXT: .cfi_def_cfa sp, 16
; ZVFHMIN-NEXT: addi sp, sp, 16
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfmsub-constrained-sdnode.ll b/llvm/test/CodeGen/RISCV/rvv/vfmsub-constrained-sdnode.ll
index 978347fa4fc10a..f6769601f488ef 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfmsub-constrained-sdnode.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfmsub-constrained-sdnode.ll
@@ -343,32 +343,37 @@ define <vscale x 32 x half> @vfmsub_vv_nxv32f16(<vscale x 32 x half> %va, <vscal
; 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: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a1, a1, 4
+; 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: vl8re16.v v24, (a0)
; 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: lui a0, 8
; ZVFHMIN-NEXT: vsetvli a1, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v16
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v16
; ZVFHMIN-NEXT: addi a1, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v0, (a1) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vs8r.v v24, (a1) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vsetvli a1, zero, e16, m8, ta, ma
; ZVFHMIN-NEXT: vxor.vx v8, v8, a0
; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, 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 v0, v24
; 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 v0, v24, v16
; 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: vl8r.v v16, (a0) # Unknown-size Folded Reload
@@ -378,7 +383,7 @@ define <vscale x 32 x half> @vfmsub_vv_nxv32f16(<vscale x 32 x half> %va, <vscal
; ZVFHMIN-NEXT: vs8r.v v24, (a0) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, 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: vl8r.v v8, (a0) # Unknown-size Folded Reload
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfmul-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vfmul-vp.ll
index f4a236df4c9e4f..1f5a17496b067a 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfmul-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfmul-vp.ll
@@ -610,75 +610,57 @@ define <vscale x 32 x half> @vfmul_vf_nxv32f16(<vscale x 32 x half> %va, half %b
; ZVFHMIN-NEXT: addi sp, sp, -16
; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a2, a1, 4
+; ZVFHMIN-NEXT: slli a2, a1, 3
; ZVFHMIN-NEXT: add a1, a2, a1
; ZVFHMIN-NEXT: sub sp, sp, a1
-; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x11, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 17 * vlenb
-; ZVFHMIN-NEXT: vmv8r.v v16, v8
+; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x09, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 9 * vlenb
+; ZVFHMIN-NEXT: vmv1r.v v24, v0
; ZVFHMIN-NEXT: fmv.x.h a1, fa0
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: addi a3, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v20
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m8, ta, ma
-; ZVFHMIN-NEXT: vmv.v.x v16, a1
-; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a3, a1, 3
-; ZVFHMIN-NEXT: add a1, a3, a1
-; ZVFHMIN-NEXT: add a1, sp, a1
-; ZVFHMIN-NEXT: addi a1, a1, 16
-; ZVFHMIN-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vmv.v.x v0, a1
; ZVFHMIN-NEXT: slli a1, a2, 1
; ZVFHMIN-NEXT: srli a2, a2, 2
; ZVFHMIN-NEXT: sub a3, a0, a1
-; ZVFHMIN-NEXT: csrr a4, vlenb
-; ZVFHMIN-NEXT: slli a4, a4, 3
-; ZVFHMIN-NEXT: add a4, sp, a4
-; ZVFHMIN-NEXT: addi a4, a4, 16
-; ZVFHMIN-NEXT: vs1r.v v0, (a4) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: addi a4, sp, 16
+; ZVFHMIN-NEXT: vs1r.v v24, (a4) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
-; ZVFHMIN-NEXT: vslidedown.vx v0, v0, a2
+; ZVFHMIN-NEXT: vslidedown.vx v12, v24, a2
; ZVFHMIN-NEXT: sltu a2, a0, a3
; ZVFHMIN-NEXT: addi a2, a2, -1
; ZVFHMIN-NEXT: and a2, a2, a3
+; ZVFHMIN-NEXT: vmv4r.v v24, v0
; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: slli a4, a3, 3
-; ZVFHMIN-NEXT: add a3, a4, a3
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
-; ZVFHMIN-NEXT: vl8r.v v24, (a3) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v28
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v4
+; ZVFHMIN-NEXT: vmv1r.v v0, v12
; ZVFHMIN-NEXT: vsetvli zero, a2, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfmul.vv v16, v8, v16, v0.t
+; ZVFHMIN-NEXT: vfmul.vv v16, v16, v24, v0.t
; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v12, v16
; ZVFHMIN-NEXT: bltu a0, a1, .LBB22_2
; ZVFHMIN-NEXT: # %bb.1:
; ZVFHMIN-NEXT: mv a0, a1
; ZVFHMIN-NEXT: .LBB22_2:
-; ZVFHMIN-NEXT: addi a1, sp, 16
-; ZVFHMIN-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v24
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a2, a1, 3
-; ZVFHMIN-NEXT: add a1, a2, a1
; ZVFHMIN-NEXT: add a1, sp, a1
; ZVFHMIN-NEXT: addi a1, a1, 16
; ZVFHMIN-NEXT: vl8r.v v0, (a1) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v0
-; 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: addi a1, sp, 16
; ZVFHMIN-NEXT: vl1r.v v0, (a1) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; 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 a1, a0, 4
+; ZVFHMIN-NEXT: slli a1, a0, 3
; ZVFHMIN-NEXT: add a0, a1, a0
; ZVFHMIN-NEXT: add sp, sp, a0
; ZVFHMIN-NEXT: .cfi_def_cfa sp, 16
@@ -703,55 +685,41 @@ define <vscale x 32 x half> @vfmul_vf_nxv32f16_unmasked(<vscale x 32 x half> %va
; ZVFHMIN-NEXT: addi sp, sp, -16
; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a1, a1, 4
+; ZVFHMIN-NEXT: slli a1, a1, 3
; ZVFHMIN-NEXT: sub sp, sp, a1
-; 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 v16, v8
+; 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: fmv.x.h a1, fa0
; ZVFHMIN-NEXT: csrr a2, vlenb
; ZVFHMIN-NEXT: vsetvli a3, zero, e8, m4, ta, ma
-; ZVFHMIN-NEXT: vmset.m v7
-; ZVFHMIN-NEXT: addi a3, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vmset.m v16
; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v20
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m8, ta, ma
-; ZVFHMIN-NEXT: vmv.v.x v16, a1
-; 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: vmv.v.x v0, a1
; ZVFHMIN-NEXT: slli a1, a2, 1
; ZVFHMIN-NEXT: srli a2, a2, 2
; ZVFHMIN-NEXT: sub a3, a0, a1
; ZVFHMIN-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
-; ZVFHMIN-NEXT: vslidedown.vx v0, v7, a2
+; ZVFHMIN-NEXT: vslidedown.vx v12, v16, a2
; ZVFHMIN-NEXT: sltu a2, a0, a3
; ZVFHMIN-NEXT: addi a2, a2, -1
; ZVFHMIN-NEXT: and a2, a2, a3
-; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: slli a3, a3, 3
-; ZVFHMIN-NEXT: add a3, sp, a3
-; ZVFHMIN-NEXT: addi a3, a3, 16
-; ZVFHMIN-NEXT: vl8r.v v24, (a3) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vmv4r.v v16, v0
+; ZVFHMIN-NEXT: addi a3, sp, 16
+; ZVFHMIN-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v28
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v4
+; ZVFHMIN-NEXT: vmv1r.v v0, v12
; ZVFHMIN-NEXT: vsetvli zero, a2, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfmul.vv v16, v8, v16, v0.t
+; ZVFHMIN-NEXT: vfmul.vv v16, v24, v16, v0.t
; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v12, v16
; ZVFHMIN-NEXT: bltu a0, a1, .LBB23_2
; ZVFHMIN-NEXT: # %bb.1:
; ZVFHMIN-NEXT: mv a0, a1
; ZVFHMIN-NEXT: .LBB23_2:
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8
; ZVFHMIN-NEXT: addi a1, sp, 16
-; ZVFHMIN-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v24
-; 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: vl8r.v v0, (a1) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v0
; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
@@ -759,7 +727,7 @@ define <vscale x 32 x half> @vfmul_vf_nxv32f16_unmasked(<vscale x 32 x half> %va
; 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, 4
+; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: add sp, sp, a0
; ZVFHMIN-NEXT: .cfi_def_cfa sp, 16
; ZVFHMIN-NEXT: addi sp, sp, 16
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfptrunc-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vfptrunc-vp.ll
index 874813f0575953..29f85344787aa5 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfptrunc-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfptrunc-vp.ll
@@ -96,15 +96,8 @@ declare <vscale x 16 x float> @llvm.vp.fptrunc.nxv16f64.nxv16f32(<vscale x 16 x
define <vscale x 16 x float> @vfptrunc_nxv16f32_nxv16f64(<vscale x 16 x double> %a, <vscale x 16 x i1> %m, i32 zeroext %vl) {
; CHECK-LABEL: vfptrunc_nxv16f32_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, 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: vmv8r.v v24, v16
; CHECK-NEXT: csrr a1, vlenb
; CHECK-NEXT: srli a2, a1, 3
; CHECK-NEXT: sub a3, a0, a1
@@ -113,8 +106,6 @@ define <vscale x 16 x float> @vfptrunc_nxv16f32_nxv16f64(<vscale x 16 x double>
; CHECK-NEXT: sltu a2, a0, a3
; CHECK-NEXT: addi a2, a2, -1
; CHECK-NEXT: and a2, a2, a3
-; 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: vfncvt.f.f.w v20, v24, v0.t
; CHECK-NEXT: bltu a0, a1, .LBB7_2
@@ -125,12 +116,6 @@ define <vscale x 16 x float> @vfptrunc_nxv16f32_nxv16f64(<vscale x 16 x double>
; 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
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add sp, sp, a0
-; CHECK-NEXT: .cfi_def_cfa sp, 16
-; CHECK-NEXT: addi sp, sp, 16
-; CHECK-NEXT: .cfi_def_cfa_offset 0
; CHECK-NEXT: ret
%v = call <vscale x 16 x float> @llvm.vp.fptrunc.nxv16f64.nxv16f32(<vscale x 16 x double> %a, <vscale x 16 x i1> %m, i32 %vl)
ret <vscale x 16 x float> %v
@@ -196,9 +181,9 @@ define <vscale x 32 x float> @vfptrunc_nxv32f32_nxv32f64(<vscale x 32 x double>
; CHECK-NEXT: and a0, a3, a0
; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: addi a3, sp, 16
-; CHECK-NEXT: vl8r.v v8, (a3) # Unknown-size Folded Reload
+; CHECK-NEXT: vl8r.v v24, (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: vfncvt.f.f.w v12, v24, v0.t
; CHECK-NEXT: bltu a2, a1, .LBB8_6
; CHECK-NEXT: # %bb.5:
; CHECK-NEXT: mv a2, a1
@@ -208,10 +193,9 @@ define <vscale x 32 x float> @vfptrunc_nxv32f32_nxv32f64(<vscale x 32 x double>
; 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: vl8r.v v24, (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: vfncvt.f.f.w v8, v24, v0.t
; 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/vfsub-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vfsub-vp.ll
index 25a80e66c4a527..9bff9a13056864 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfsub-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfsub-vp.ll
@@ -476,75 +476,57 @@ define <vscale x 32 x bfloat> @vfsub_vf_nxv32bf16(<vscale x 32 x bfloat> %va, bf
; CHECK-NEXT: addi sp, sp, -16
; CHECK-NEXT: .cfi_def_cfa_offset 16
; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a2, a1, 4
+; CHECK-NEXT: slli a2, a1, 3
; CHECK-NEXT: add a1, a2, a1
; CHECK-NEXT: sub sp, sp, a1
-; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x11, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 17 * vlenb
-; CHECK-NEXT: vmv8r.v v16, v8
+; CHECK-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x09, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 9 * vlenb
+; CHECK-NEXT: vmv1r.v v24, v0
; CHECK-NEXT: fmv.x.h a1, fa0
; CHECK-NEXT: csrr a2, vlenb
-; CHECK-NEXT: addi a3, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
; CHECK-NEXT: vsetvli a3, zero, e16, m4, ta, ma
-; CHECK-NEXT: vfwcvtbf16.f.f.v v8, v20
+; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v12
; CHECK-NEXT: vsetvli a3, zero, e16, m8, ta, ma
-; CHECK-NEXT: vmv.v.x v16, a1
-; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a3, a1, 3
-; CHECK-NEXT: add a1, a3, a1
-; CHECK-NEXT: add a1, sp, a1
-; CHECK-NEXT: addi a1, a1, 16
-; CHECK-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
+; CHECK-NEXT: vmv.v.x v0, a1
; CHECK-NEXT: slli a1, a2, 1
; CHECK-NEXT: srli a2, a2, 2
; CHECK-NEXT: sub a3, a0, a1
-; 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: vs1r.v v0, (a4) # Unknown-size Folded Spill
+; CHECK-NEXT: addi a4, sp, 16
+; CHECK-NEXT: vs1r.v v24, (a4) # Unknown-size Folded Spill
; CHECK-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
-; CHECK-NEXT: vslidedown.vx v0, v0, a2
+; CHECK-NEXT: vslidedown.vx v12, v24, a2
; CHECK-NEXT: sltu a2, a0, a3
; CHECK-NEXT: addi a2, a2, -1
; CHECK-NEXT: and a2, a2, a3
+; CHECK-NEXT: vmv4r.v v24, v0
; CHECK-NEXT: csrr a3, vlenb
-; CHECK-NEXT: slli a4, a3, 3
-; CHECK-NEXT: add a3, a4, a3
; CHECK-NEXT: add a3, sp, a3
; CHECK-NEXT: addi a3, a3, 16
-; CHECK-NEXT: vl8r.v v24, (a3) # Unknown-size Folded Reload
+; CHECK-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
; CHECK-NEXT: vsetvli a3, zero, e16, m4, ta, ma
-; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v28
+; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v4
+; CHECK-NEXT: vmv1r.v v0, v12
; CHECK-NEXT: vsetvli zero, a2, e32, m8, ta, ma
-; CHECK-NEXT: vfsub.vv v16, v8, v16, v0.t
+; CHECK-NEXT: vfsub.vv v16, v16, v24, v0.t
; CHECK-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; CHECK-NEXT: vfncvtbf16.f.f.w v12, v16
; CHECK-NEXT: bltu a0, a1, .LBB22_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB22_2:
-; CHECK-NEXT: addi a1, sp, 16
-; CHECK-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
-; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v24
+; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v8
; CHECK-NEXT: csrr a1, vlenb
-; CHECK-NEXT: slli a2, a1, 3
-; CHECK-NEXT: add a1, a2, a1
; CHECK-NEXT: add a1, sp, a1
; CHECK-NEXT: addi a1, a1, 16
; CHECK-NEXT: vl8r.v v0, (a1) # Unknown-size Folded Reload
; CHECK-NEXT: vfwcvtbf16.f.f.v v24, 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: addi a1, sp, 16
; CHECK-NEXT: vl1r.v v0, (a1) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; CHECK-NEXT: vfsub.vv v16, v16, v24, v0.t
; CHECK-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; CHECK-NEXT: vfncvtbf16.f.f.w v8, v16
; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a1, a0, 4
+; CHECK-NEXT: slli a1, a0, 3
; CHECK-NEXT: add a0, a1, a0
; CHECK-NEXT: add sp, sp, a0
; CHECK-NEXT: .cfi_def_cfa sp, 16
@@ -563,55 +545,41 @@ define <vscale x 32 x bfloat> @vfsub_vf_nxv32bf16_unmasked(<vscale x 32 x bfloat
; 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: vmv8r.v v16, v8
+; 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: fmv.x.h a1, fa0
; CHECK-NEXT: csrr a2, vlenb
; CHECK-NEXT: vsetvli a3, zero, e8, m4, ta, ma
-; CHECK-NEXT: vmset.m v7
-; CHECK-NEXT: addi a3, sp, 16
-; CHECK-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
+; CHECK-NEXT: vmset.m v16
; CHECK-NEXT: vsetvli a3, zero, e16, m4, ta, ma
-; CHECK-NEXT: vfwcvtbf16.f.f.v v8, v20
+; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v12
; CHECK-NEXT: vsetvli a3, zero, e16, m8, ta, ma
-; CHECK-NEXT: vmv.v.x v16, 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 v16, (a1) # Unknown-size Folded Spill
+; CHECK-NEXT: vmv.v.x v0, a1
; CHECK-NEXT: slli a1, a2, 1
; CHECK-NEXT: srli a2, a2, 2
; CHECK-NEXT: sub a3, a0, a1
; CHECK-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
-; CHECK-NEXT: vslidedown.vx v0, v7, a2
+; CHECK-NEXT: vslidedown.vx v12, v16, a2
; CHECK-NEXT: sltu a2, a0, a3
; CHECK-NEXT: addi a2, a2, -1
; CHECK-NEXT: and a2, a2, a3
-; CHECK-NEXT: csrr a3, vlenb
-; CHECK-NEXT: slli a3, a3, 3
-; CHECK-NEXT: add a3, sp, a3
-; CHECK-NEXT: addi a3, a3, 16
-; CHECK-NEXT: vl8r.v v24, (a3) # Unknown-size Folded Reload
+; CHECK-NEXT: vmv4r.v v16, v0
+; CHECK-NEXT: addi a3, sp, 16
+; CHECK-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
; CHECK-NEXT: vsetvli a3, zero, e16, m4, ta, ma
-; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v28
+; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v4
+; CHECK-NEXT: vmv1r.v v0, v12
; CHECK-NEXT: vsetvli zero, a2, e32, m8, ta, ma
-; CHECK-NEXT: vfsub.vv v16, v8, v16, v0.t
+; CHECK-NEXT: vfsub.vv v16, v24, v16, v0.t
; CHECK-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; CHECK-NEXT: vfncvtbf16.f.f.w v12, v16
; CHECK-NEXT: bltu a0, a1, .LBB23_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a0, a1
; CHECK-NEXT: .LBB23_2:
+; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v8
; CHECK-NEXT: addi a1, sp, 16
-; CHECK-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
-; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v24
-; 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 v0, (a1) # Unknown-size Folded Reload
; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v0
; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
@@ -619,7 +587,7 @@ define <vscale x 32 x bfloat> @vfsub_vf_nxv32bf16_unmasked(<vscale x 32 x bfloat
; CHECK-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; CHECK-NEXT: vfncvtbf16.f.f.w v8, v16
; 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: .cfi_def_cfa sp, 16
; CHECK-NEXT: addi sp, sp, 16
@@ -1232,75 +1200,57 @@ define <vscale x 32 x half> @vfsub_vf_nxv32f16(<vscale x 32 x half> %va, half %b
; ZVFHMIN-NEXT: addi sp, sp, -16
; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a2, a1, 4
+; ZVFHMIN-NEXT: slli a2, a1, 3
; ZVFHMIN-NEXT: add a1, a2, a1
; ZVFHMIN-NEXT: sub sp, sp, a1
-; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x11, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 17 * vlenb
-; ZVFHMIN-NEXT: vmv8r.v v16, v8
+; ZVFHMIN-NEXT: .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x09, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 9 * vlenb
+; ZVFHMIN-NEXT: vmv1r.v v24, v0
; ZVFHMIN-NEXT: fmv.x.h a1, fa0
; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: addi a3, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v20
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m8, ta, ma
-; ZVFHMIN-NEXT: vmv.v.x v16, a1
-; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a3, a1, 3
-; ZVFHMIN-NEXT: add a1, a3, a1
-; ZVFHMIN-NEXT: add a1, sp, a1
-; ZVFHMIN-NEXT: addi a1, a1, 16
-; ZVFHMIN-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vmv.v.x v0, a1
; ZVFHMIN-NEXT: slli a1, a2, 1
; ZVFHMIN-NEXT: srli a2, a2, 2
; ZVFHMIN-NEXT: sub a3, a0, a1
-; ZVFHMIN-NEXT: csrr a4, vlenb
-; ZVFHMIN-NEXT: slli a4, a4, 3
-; ZVFHMIN-NEXT: add a4, sp, a4
-; ZVFHMIN-NEXT: addi a4, a4, 16
-; ZVFHMIN-NEXT: vs1r.v v0, (a4) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: addi a4, sp, 16
+; ZVFHMIN-NEXT: vs1r.v v24, (a4) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
-; ZVFHMIN-NEXT: vslidedown.vx v0, v0, a2
+; ZVFHMIN-NEXT: vslidedown.vx v12, v24, a2
; ZVFHMIN-NEXT: sltu a2, a0, a3
; ZVFHMIN-NEXT: addi a2, a2, -1
; ZVFHMIN-NEXT: and a2, a2, a3
+; ZVFHMIN-NEXT: vmv4r.v v24, v0
; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: slli a4, a3, 3
-; ZVFHMIN-NEXT: add a3, a4, a3
; ZVFHMIN-NEXT: add a3, sp, a3
; ZVFHMIN-NEXT: addi a3, a3, 16
-; ZVFHMIN-NEXT: vl8r.v v24, (a3) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vs8r.v v24, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v28
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v4
+; ZVFHMIN-NEXT: vmv1r.v v0, v12
; ZVFHMIN-NEXT: vsetvli zero, a2, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfsub.vv v16, v8, v16, v0.t
+; ZVFHMIN-NEXT: vfsub.vv v16, v16, v24, v0.t
; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v12, v16
; ZVFHMIN-NEXT: bltu a0, a1, .LBB46_2
; ZVFHMIN-NEXT: # %bb.1:
; ZVFHMIN-NEXT: mv a0, a1
; ZVFHMIN-NEXT: .LBB46_2:
-; ZVFHMIN-NEXT: addi a1, sp, 16
-; ZVFHMIN-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v24
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a2, a1, 3
-; ZVFHMIN-NEXT: add a1, a2, a1
; ZVFHMIN-NEXT: add a1, sp, a1
; ZVFHMIN-NEXT: addi a1, a1, 16
; ZVFHMIN-NEXT: vl8r.v v0, (a1) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v0
-; 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: addi a1, sp, 16
; ZVFHMIN-NEXT: vl1r.v v0, (a1) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
; 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 a1, a0, 4
+; ZVFHMIN-NEXT: slli a1, a0, 3
; ZVFHMIN-NEXT: add a0, a1, a0
; ZVFHMIN-NEXT: add sp, sp, a0
; ZVFHMIN-NEXT: .cfi_def_cfa sp, 16
@@ -1325,55 +1275,41 @@ define <vscale x 32 x half> @vfsub_vf_nxv32f16_unmasked(<vscale x 32 x half> %va
; ZVFHMIN-NEXT: addi sp, sp, -16
; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
; ZVFHMIN-NEXT: csrr a1, vlenb
-; ZVFHMIN-NEXT: slli a1, a1, 4
+; ZVFHMIN-NEXT: slli a1, a1, 3
; ZVFHMIN-NEXT: sub sp, sp, a1
-; 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 v16, v8
+; 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: fmv.x.h a1, fa0
; ZVFHMIN-NEXT: csrr a2, vlenb
; ZVFHMIN-NEXT: vsetvli a3, zero, e8, m4, ta, ma
-; ZVFHMIN-NEXT: vmset.m v7
-; ZVFHMIN-NEXT: addi a3, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v8, (a3) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vmset.m v16
; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v20
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v12
; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m8, ta, ma
-; ZVFHMIN-NEXT: vmv.v.x v16, a1
-; 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: vmv.v.x v0, a1
; ZVFHMIN-NEXT: slli a1, a2, 1
; ZVFHMIN-NEXT: srli a2, a2, 2
; ZVFHMIN-NEXT: sub a3, a0, a1
; ZVFHMIN-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
-; ZVFHMIN-NEXT: vslidedown.vx v0, v7, a2
+; ZVFHMIN-NEXT: vslidedown.vx v12, v16, a2
; ZVFHMIN-NEXT: sltu a2, a0, a3
; ZVFHMIN-NEXT: addi a2, a2, -1
; ZVFHMIN-NEXT: and a2, a2, a3
-; ZVFHMIN-NEXT: csrr a3, vlenb
-; ZVFHMIN-NEXT: slli a3, a3, 3
-; ZVFHMIN-NEXT: add a3, sp, a3
-; ZVFHMIN-NEXT: addi a3, a3, 16
-; ZVFHMIN-NEXT: vl8r.v v24, (a3) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vmv4r.v v16, v0
+; ZVFHMIN-NEXT: addi a3, sp, 16
+; ZVFHMIN-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
; ZVFHMIN-NEXT: vsetvli a3, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v28
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v4
+; ZVFHMIN-NEXT: vmv1r.v v0, v12
; ZVFHMIN-NEXT: vsetvli zero, a2, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfsub.vv v16, v8, v16, v0.t
+; ZVFHMIN-NEXT: vfsub.vv v16, v24, v16, v0.t
; ZVFHMIN-NEXT: vsetvli a2, zero, e16, m4, ta, ma
; ZVFHMIN-NEXT: vfncvt.f.f.w v12, v16
; ZVFHMIN-NEXT: bltu a0, a1, .LBB47_2
; ZVFHMIN-NEXT: # %bb.1:
; ZVFHMIN-NEXT: mv a0, a1
; ZVFHMIN-NEXT: .LBB47_2:
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8
; ZVFHMIN-NEXT: addi a1, sp, 16
-; ZVFHMIN-NEXT: vl8r.v v24, (a1) # Unknown-size Folded Reload
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v24
-; 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: vl8r.v v0, (a1) # Unknown-size Folded Reload
; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v0
; ZVFHMIN-NEXT: vsetvli zero, a0, e32, m8, ta, ma
@@ -1381,7 +1317,7 @@ define <vscale x 32 x half> @vfsub_vf_nxv32f16_unmasked(<vscale x 32 x half> %va
; 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, 4
+; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: add sp, sp, a0
; ZVFHMIN-NEXT: .cfi_def_cfa sp, 16
; ZVFHMIN-NEXT: addi sp, sp, 16
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfwmacc-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vfwmacc-vp.ll
index 6cd3884f029fdd..1b343788e03f8a 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfwmacc-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfwmacc-vp.ll
@@ -633,28 +633,14 @@ define <vscale x 16 x float> @vfmacc_vv_nxv16f32(<vscale x 16 x half> %a, <vscal
;
; ZVFHMIN-LABEL: vfmacc_vv_nxv16f32:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: addi sp, sp, -16
-; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
-; ZVFHMIN-NEXT: csrr a1, vlenb
-; 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: addi a1, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vmv4r.v v4, v12
+; ZVFHMIN-NEXT: vmv4r.v v24, v8
; 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: addi a0, sp, 16
-; ZVFHMIN-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v24, v0.t
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v4, v0.t
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfmadd.vv v24, v16, v8, v0.t
+; ZVFHMIN-NEXT: vfmadd.vv v24, v8, v16, v0.t
; ZVFHMIN-NEXT: vmv.v.v v8, v24
-; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 3
-; ZVFHMIN-NEXT: add sp, sp, a0
-; ZVFHMIN-NEXT: .cfi_def_cfa sp, 16
-; ZVFHMIN-NEXT: addi sp, sp, 16
-; ZVFHMIN-NEXT: .cfi_def_cfa_offset 0
; ZVFHMIN-NEXT: ret
%aext = call <vscale x 16 x float> @llvm.vp.fpext.nxv16f32.nxv16f16(<vscale x 16 x half> %a, <vscale x 16 x i1> %m, i32 %evl)
%bext = call <vscale x 16 x float> @llvm.vp.fpext.nxv16f32.nxv16f16(<vscale x 16 x half> %b, <vscale x 16 x i1> %m, i32 %evl)
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfwnmacc-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vfwnmacc-vp.ll
index 0a0bc6696a9f96..9e5da5a4effaf5 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfwnmacc-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfwnmacc-vp.ll
@@ -590,28 +590,14 @@ define <vscale x 16 x float> @vfnmacc_vv_nxv16f32(<vscale x 16 x half> %a, <vsca
;
; ZVFHMIN-LABEL: vfnmacc_vv_nxv16f32:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: addi sp, sp, -16
-; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
-; ZVFHMIN-NEXT: csrr a1, vlenb
-; 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: addi a1, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vmv4r.v v4, v12
+; ZVFHMIN-NEXT: vmv4r.v v24, v8
; 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: addi a0, sp, 16
-; ZVFHMIN-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v24, v0.t
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v4, v0.t
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfnmadd.vv v24, v16, v8, v0.t
+; ZVFHMIN-NEXT: vfnmadd.vv v24, v8, v16, v0.t
; ZVFHMIN-NEXT: vmv.v.v v8, v24
-; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 3
-; ZVFHMIN-NEXT: add sp, sp, a0
-; ZVFHMIN-NEXT: .cfi_def_cfa sp, 16
-; ZVFHMIN-NEXT: addi sp, sp, 16
-; ZVFHMIN-NEXT: .cfi_def_cfa_offset 0
; ZVFHMIN-NEXT: ret
%aext = call <vscale x 16 x float> @llvm.vp.fpext.nxv16f32.nxv16f16(<vscale x 16 x half> %a, <vscale x 16 x i1> %m, i32 %evl)
%bext = call <vscale x 16 x float> @llvm.vp.fpext.nxv16f32.nxv16f16(<vscale x 16 x half> %b, <vscale x 16 x i1> %m, i32 %evl)
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfwnmsac-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vfwnmsac-vp.ll
index b5f7ef3380869f..c44fc1aa269621 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfwnmsac-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfwnmsac-vp.ll
@@ -566,28 +566,14 @@ define <vscale x 16 x float> @vfnmsac_vv_nxv16f32(<vscale x 16 x half> %a, <vsca
;
; ZVFHMIN-LABEL: vfnmsac_vv_nxv16f32:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: addi sp, sp, -16
-; ZVFHMIN-NEXT: .cfi_def_cfa_offset 16
-; ZVFHMIN-NEXT: csrr a1, vlenb
-; 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: addi a1, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v16, (a1) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT: vmv4r.v v4, v12
+; ZVFHMIN-NEXT: vmv4r.v v24, v8
; 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: addi a0, sp, 16
-; ZVFHMIN-NEXT: vl8r.v v8, (a0) # Unknown-size Folded Reload
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v8, v24, v0.t
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v4, v0.t
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfnmsub.vv v24, v16, v8, v0.t
+; ZVFHMIN-NEXT: vfnmsub.vv v24, v8, v16, v0.t
; ZVFHMIN-NEXT: vmv.v.v v8, v24
-; ZVFHMIN-NEXT: csrr a0, vlenb
-; ZVFHMIN-NEXT: slli a0, a0, 3
-; ZVFHMIN-NEXT: add sp, sp, a0
-; ZVFHMIN-NEXT: .cfi_def_cfa sp, 16
-; ZVFHMIN-NEXT: addi sp, sp, 16
-; ZVFHMIN-NEXT: .cfi_def_cfa_offset 0
; ZVFHMIN-NEXT: ret
%aext = call <vscale x 16 x float> @llvm.vp.fpext.nxv16f32.nxv16f16(<vscale x 16 x half> %a, <vscale x 16 x i1> %m, i32 %evl)
%bext = call <vscale x 16 x float> @llvm.vp.fpext.nxv16f32.nxv16f16(<vscale x 16 x half> %b, <vscale x 16 x i1> %m, i32 %evl)
diff --git a/llvm/test/CodeGen/RISCV/rvv/vpstore.ll b/llvm/test/CodeGen/RISCV/rvv/vpstore.ll
index 8978dc268d4e52..0a2da7a3780444 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vpstore.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vpstore.ll
@@ -462,15 +462,7 @@ declare void @llvm.vp.store.nxv17f64.p0(<vscale x 17 x double>, ptr, <vscale x 1
define void @vpstore_nxv17f64(<vscale x 17 x double> %val, ptr %ptr, <vscale x 17 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vpstore_nxv17f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: addi sp, sp, -16
-; CHECK-NEXT: .cfi_def_cfa_offset 16
-; CHECK-NEXT: csrr a3, vlenb
-; CHECK-NEXT: slli a3, a3, 3
-; CHECK-NEXT: sub sp, sp, a3
-; 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 v24, v0
-; CHECK-NEXT: addi a3, sp, 16
-; CHECK-NEXT: vs8r.v v16, (a3) # Unknown-size Folded Spill
+; CHECK-NEXT: vmv1r.v v7, v0
; CHECK-NEXT: csrr a3, vlenb
; CHECK-NEXT: slli a4, a3, 1
; CHECK-NEXT: mv a5, a2
@@ -483,14 +475,14 @@ define void @vpstore_nxv17f64(<vscale x 17 x double> %val, ptr %ptr, <vscale x 1
; CHECK-NEXT: # %bb.3:
; CHECK-NEXT: mv a6, a3
; CHECK-NEXT: .LBB36_4:
-; CHECK-NEXT: vmv1r.v v0, v24
-; CHECK-NEXT: vl8re64.v v16, (a0)
+; CHECK-NEXT: vmv1r.v v0, v7
+; CHECK-NEXT: vl8re64.v v24, (a0)
; 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: srli a6, a3, 3
; CHECK-NEXT: vsetvli a7, zero, e8, mf4, ta, ma
-; CHECK-NEXT: vslidedown.vx v0, v24, a6
+; CHECK-NEXT: vslidedown.vx v0, v7, a6
; CHECK-NEXT: slli a6, a3, 3
; CHECK-NEXT: sub a4, a2, a4
; CHECK-NEXT: sltu a5, a5, a0
@@ -500,10 +492,8 @@ define void @vpstore_nxv17f64(<vscale x 17 x double> %val, ptr %ptr, <vscale x 1
; CHECK-NEXT: addi a2, a2, -1
; CHECK-NEXT: and a5, a5, a0
; CHECK-NEXT: and a0, a2, a4
-; CHECK-NEXT: addi a2, sp, 16
-; CHECK-NEXT: vl8r.v v8, (a2) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, a5, e64, m8, ta, ma
-; CHECK-NEXT: vse64.v v8, (a6), v0.t
+; CHECK-NEXT: vse64.v v16, (a6), v0.t
; CHECK-NEXT: bltu a0, a3, .LBB36_6
; CHECK-NEXT: # %bb.5:
; CHECK-NEXT: mv a0, a3
@@ -511,16 +501,10 @@ define void @vpstore_nxv17f64(<vscale x 17 x double> %val, ptr %ptr, <vscale x 1
; CHECK-NEXT: slli a2, a3, 4
; CHECK-NEXT: srli a3, a3, 2
; CHECK-NEXT: vsetvli a4, zero, e8, mf2, ta, ma
-; CHECK-NEXT: vslidedown.vx v0, v24, a3
+; CHECK-NEXT: vslidedown.vx v0, v7, a3
; CHECK-NEXT: add a1, a1, a2
; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: vse64.v v16, (a1), v0.t
-; CHECK-NEXT: csrr a0, vlenb
-; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: add sp, sp, a0
-; CHECK-NEXT: .cfi_def_cfa sp, 16
-; CHECK-NEXT: addi sp, sp, 16
-; CHECK-NEXT: .cfi_def_cfa_offset 0
+; CHECK-NEXT: vse64.v v24, (a1), v0.t
; CHECK-NEXT: ret
call void @llvm.vp.store.nxv17f64.p0(<vscale x 17 x double> %val, ptr %ptr, <vscale x 17 x i1> %m, i32 %evl)
ret void
diff --git a/llvm/test/CodeGen/RISCV/rvv/vselect-fp.ll b/llvm/test/CodeGen/RISCV/rvv/vselect-fp.ll
index 1fc33dc73a27dc..6ef6d5ff37b52c 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vselect-fp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vselect-fp.ll
@@ -514,19 +514,20 @@ define void @vselect_legalize_regression(<vscale x 16 x double> %a, <vscale x 16
; CHECK-LABEL: vselect_legalize_regression:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetvli a2, zero, e8, m2, ta, ma
-; CHECK-NEXT: vlm.v v24, (a0)
+; CHECK-NEXT: vlm.v v7, (a0)
; CHECK-NEXT: csrr a0, vlenb
+; CHECK-NEXT: vsetvli a2, zero, e64, m8, ta, ma
+; CHECK-NEXT: vmv.v.i v24, 0
; CHECK-NEXT: srli a2, a0, 3
; CHECK-NEXT: slli a0, a0, 3
-; CHECK-NEXT: vmand.mm v7, v0, v24
+; CHECK-NEXT: vsetvli a3, zero, e8, m2, ta, ma
+; CHECK-NEXT: vmand.mm v7, v0, v7
; CHECK-NEXT: vsetvli a3, zero, e8, mf4, ta, ma
; CHECK-NEXT: vslidedown.vx v0, v7, a2
; CHECK-NEXT: add a0, a1, a0
; CHECK-NEXT: vsetvli a2, zero, e64, m8, ta, ma
-; CHECK-NEXT: vmv.v.i v24, 0
; CHECK-NEXT: vmerge.vvm v16, v24, v16, v0
; CHECK-NEXT: vmv1r.v v0, v7
-; CHECK-NEXT: vmv.v.i v24, 0
; CHECK-NEXT: vmerge.vvm v8, v24, v8, v0
; CHECK-NEXT: vs8r.v v8, (a1)
; CHECK-NEXT: vs8r.v v16, (a0)
diff --git a/llvm/test/CodeGen/RISCV/rvv/vselect-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vselect-vp.ll
index bb51f0592dc17a..107817cefe3e9a 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vselect-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vselect-vp.ll
@@ -354,48 +354,38 @@ 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: 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: csrr a1, vlenb
; CHECK-NEXT: slli a1, a1, 3
-; CHECK-NEXT: add a1, sp, a1
-; CHECK-NEXT: addi a1, a1, 16
+; 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: addi a1, sp, 16
; CHECK-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
-; CHECK-NEXT: vmv1r.v v24, v0
+; CHECK-NEXT: vmv1r.v v7, v0
; CHECK-NEXT: csrr a3, vlenb
; CHECK-NEXT: slli a4, a3, 3
; CHECK-NEXT: slli a1, a3, 1
; CHECK-NEXT: srli a3, a3, 2
; CHECK-NEXT: add a4, a0, a4
; CHECK-NEXT: sub a5, a2, a1
-; CHECK-NEXT: vl8re32.v v8, (a4)
+; CHECK-NEXT: vl8re32.v v24, (a4)
; CHECK-NEXT: sltu a4, a2, a5
; CHECK-NEXT: addi a4, a4, -1
-; CHECK-NEXT: vl8re32.v v0, (a0)
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vs8r.v v0, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vl8re32.v v8, (a0)
; 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, a4, a5
; CHECK-NEXT: vsetvli zero, a4, e32, m8, ta, ma
-; CHECK-NEXT: vmerge.vvm v16, v8, v16, v0
+; CHECK-NEXT: vmerge.vvm v16, v24, v16, v0
; CHECK-NEXT: bltu a2, a1, .LBB27_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a2, a1
; CHECK-NEXT: .LBB27_2:
-; 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: vmv1r.v v0, v7
; 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: vmerge.vvm v8, v24, v8, v0
+; CHECK-NEXT: vmerge.vvm v8, v8, v24, v0
; 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: .cfi_def_cfa sp, 16
; CHECK-NEXT: addi sp, sp, 16
@@ -413,48 +403,38 @@ 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: 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: csrr a1, vlenb
; CHECK-NEXT: slli a1, a1, 3
-; CHECK-NEXT: add a1, sp, a1
-; CHECK-NEXT: addi a1, a1, 16
+; 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: addi a1, sp, 16
; CHECK-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
-; CHECK-NEXT: vmv1r.v v24, v0
+; CHECK-NEXT: vmv1r.v v7, v0
; CHECK-NEXT: csrr a1, vlenb
; CHECK-NEXT: slli a3, a1, 3
; CHECK-NEXT: slli a2, a1, 1
; CHECK-NEXT: srli a4, a1, 2
; CHECK-NEXT: add a3, a0, a3
; CHECK-NEXT: sub a5, a1, a2
-; CHECK-NEXT: vl8re32.v v8, (a3)
+; CHECK-NEXT: vl8re32.v v24, (a3)
; CHECK-NEXT: sltu a3, a1, a5
; CHECK-NEXT: addi a3, a3, -1
-; CHECK-NEXT: vl8re32.v v0, (a0)
-; CHECK-NEXT: addi a0, sp, 16
-; CHECK-NEXT: vs8r.v v0, (a0) # Unknown-size Folded Spill
+; CHECK-NEXT: vl8re32.v v8, (a0)
; CHECK-NEXT: vsetvli a0, zero, e8, mf2, ta, ma
-; CHECK-NEXT: vslidedown.vx v0, v24, a4
+; CHECK-NEXT: vslidedown.vx v0, v0, a4
; CHECK-NEXT: and a3, a3, a5
; CHECK-NEXT: vsetvli zero, a3, e32, m8, ta, ma
-; CHECK-NEXT: vmerge.vvm v16, v8, v16, v0
+; CHECK-NEXT: vmerge.vvm v16, v24, v16, v0
; CHECK-NEXT: bltu a1, a2, .LBB28_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a1, a2
; CHECK-NEXT: .LBB28_2:
-; 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: vmv1r.v v0, v7
; CHECK-NEXT: addi a0, sp, 16
; CHECK-NEXT: vl8r.v v24, (a0) # Unknown-size Folded Reload
; CHECK-NEXT: vsetvli zero, a1, e32, m8, ta, ma
-; CHECK-NEXT: vmerge.vvm v8, v24, v8, v0
+; CHECK-NEXT: vmerge.vvm v8, v8, v24, v0
; 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: .cfi_def_cfa sp, 16
; CHECK-NEXT: addi sp, sp, 16
@@ -702,47 +682,37 @@ define <vscale x 16 x double> @select_nxv16f64(<vscale x 16 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: 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: csrr a1, vlenb
; CHECK-NEXT: slli a1, a1, 3
-; CHECK-NEXT: add a1, sp, a1
-; CHECK-NEXT: addi a1, a1, 16
+; 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: addi a1, sp, 16
; CHECK-NEXT: vs8r.v v8, (a1) # Unknown-size Folded Spill
-; CHECK-NEXT: vmv1r.v v24, v0
+; CHECK-NEXT: vmv1r.v v7, v0
; CHECK-NEXT: csrr a1, vlenb
; CHECK-NEXT: slli a3, a1, 3
; CHECK-NEXT: sub a4, a2, a1
; CHECK-NEXT: add a3, a0, a3
; CHECK-NEXT: sltu a5, a2, a4
-; CHECK-NEXT: vl8re64.v v8, (a3)
+; CHECK-NEXT: vl8re64.v v24, (a3)
; CHECK-NEXT: addi a5, a5, -1
; CHECK-NEXT: srli a3, 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: vl8re64.v v8, (a0)
; CHECK-NEXT: vsetvli a0, zero, e8, mf4, ta, ma
-; CHECK-NEXT: vslidedown.vx v0, v24, a3
+; CHECK-NEXT: vslidedown.vx v0, v0, a3
; CHECK-NEXT: and a4, a5, a4
; CHECK-NEXT: vsetvli zero, a4, e64, m8, ta, ma
-; CHECK-NEXT: vmerge.vvm v16, v8, v16, v0
+; CHECK-NEXT: vmerge.vvm v16, v24, v16, v0
; CHECK-NEXT: bltu a2, a1, .LBB48_2
; CHECK-NEXT: # %bb.1:
; CHECK-NEXT: mv a2, a1
; CHECK-NEXT: .LBB48_2:
-; 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: vmv1r.v v0, v7
; 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: vmerge.vvm v8, v8, v24, v0
; 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: .cfi_def_cfa sp, 16
; CHECK-NEXT: addi sp, sp, 16
diff --git a/llvm/test/CodeGen/RISCV/rvv/vtrunc-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vtrunc-vp.ll
index e62b7a00396388..525788d3e2fe06 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vtrunc-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vtrunc-vp.ll
@@ -337,9 +337,9 @@ define <vscale x 32 x i32> @vtrunc_nxv32i64_nxv32i32(<vscale x 32 x i64> %a, <vs
; CHECK-NEXT: and a0, a3, a0
; CHECK-NEXT: vmv1r.v v0, v6
; CHECK-NEXT: addi a3, sp, 16
-; CHECK-NEXT: vl8r.v v8, (a3) # Unknown-size Folded Reload
+; CHECK-NEXT: vl8r.v v24, (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: vnsrl.wi v12, v24, 0, v0.t
; CHECK-NEXT: bltu a2, a1, .LBB17_6
; CHECK-NEXT: # %bb.5:
; CHECK-NEXT: mv a2, a1
@@ -349,10 +349,9 @@ define <vscale x 32 x i32> @vtrunc_nxv32i64_nxv32i32(<vscale x 32 x i64> %a, <vs
; 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: vl8r.v v24, (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: vnsrl.wi v8, v24, 0, v0.t
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 4
; CHECK-NEXT: add sp, sp, a0
More information about the llvm-commits
mailing list