[llvm] [RISCV] Remove codegen for vp_minnum, vp_maxnum (PR #189899)
Luke Lau via llvm-commits
llvm-commits at lists.llvm.org
Wed Apr 1 10:05:45 PDT 2026
https://github.com/lukel97 updated https://github.com/llvm/llvm-project/pull/189899
>From a2d36c3579775b1db34fae28073b889070674364 Mon Sep 17 00:00:00 2001
From: Luke Lau <luke at igalia.com>
Date: Wed, 1 Apr 2026 15:36:02 +0800
Subject: [PATCH] [RISCV] Remove codegen for vp_minnum, vp_maxnum
Part of the work to remove trivial VP intrinsics from the RISC-V backend, see https://discourse.llvm.org/t/rfc-remove-codegen-support-for-trivial-vp-intrinsics-in-the-risc-v-backend/87999
This splits off 2 intrinsics from #179622.
---
llvm/lib/Target/RISCV/RISCVISelLowering.cpp | 8 +-
.../Target/RISCV/RISCVTargetTransformInfo.h | 2 -
.../RISCV/rvv/fixed-vectors-vfmax-vp.ll | 183 +++-----
.../RISCV/rvv/fixed-vectors-vfmin-vp.ll | 183 +++-----
llvm/test/CodeGen/RISCV/rvv/vfmax-vp.ll | 398 ++++++++----------
llvm/test/CodeGen/RISCV/rvv/vfmin-vp.ll | 398 ++++++++----------
6 files changed, 471 insertions(+), 701 deletions(-)
diff --git a/llvm/lib/Target/RISCV/RISCVISelLowering.cpp b/llvm/lib/Target/RISCV/RISCVISelLowering.cpp
index cb6489fbeae09..8594cf1443279 100644
--- a/llvm/lib/Target/RISCV/RISCVISelLowering.cpp
+++ b/llvm/lib/Target/RISCV/RISCVISelLowering.cpp
@@ -889,7 +889,7 @@ RISCVTargetLowering::RISCVTargetLowering(const TargetMachine &TM,
ISD::VP_REDUCE_FMIN, ISD::VP_REDUCE_FMAX, ISD::VP_MERGE,
ISD::VP_SELECT, ISD::VP_SINT_TO_FP, ISD::VP_UINT_TO_FP,
ISD::VP_SETCC, ISD::VP_FP_ROUND, ISD::VP_FP_EXTEND,
- ISD::VP_SQRT, ISD::VP_FMINNUM, ISD::VP_FMAXNUM,
+ ISD::VP_SQRT,
ISD::VP_FCEIL, ISD::VP_FFLOOR, ISD::VP_FROUND,
ISD::VP_FROUNDEVEN, ISD::VP_FCOPYSIGN, ISD::VP_FROUNDTOZERO,
ISD::VP_FRINT, ISD::VP_FNEARBYINT, ISD::VP_IS_FPCLASS,
@@ -1222,8 +1222,6 @@ RISCVTargetLowering::RISCVTargetLowering(const TargetMachine &TM,
ISD::VP_REDUCE_FMIN,
ISD::VP_REDUCE_FMAX,
ISD::VP_SQRT,
- ISD::VP_FMINNUM,
- ISD::VP_FMAXNUM,
ISD::VP_FCEIL,
ISD::VP_FFLOOR,
ISD::VP_FROUND,
@@ -7705,11 +7703,9 @@ static unsigned getRISCVVLOp(SDValue Op) {
return RISCVISD::VFCVT_RTZ_XU_F_VL;
case ISD::FMINNUM:
case ISD::FMINIMUMNUM:
- case ISD::VP_FMINNUM:
return RISCVISD::VFMIN_VL;
case ISD::FMAXNUM:
case ISD::FMAXIMUMNUM:
- case ISD::VP_FMAXNUM:
return RISCVISD::VFMAX_VL;
case ISD::LRINT:
case ISD::LLRINT:
@@ -9062,8 +9058,6 @@ SDValue RISCVTargetLowering::LowerOperation(SDValue Op,
case ISD::VP_FABS:
case ISD::VP_SQRT:
case ISD::VP_FMA:
- case ISD::VP_FMINNUM:
- case ISD::VP_FMAXNUM:
case ISD::VP_FCOPYSIGN:
if (isPromotedOpNeedingSplit(Op, Subtarget))
return SplitVPOp(Op, DAG);
diff --git a/llvm/lib/Target/RISCV/RISCVTargetTransformInfo.h b/llvm/lib/Target/RISCV/RISCVTargetTransformInfo.h
index 18e0ab9b72608..5253999a58528 100644
--- a/llvm/lib/Target/RISCV/RISCVTargetTransformInfo.h
+++ b/llvm/lib/Target/RISCV/RISCVTargetTransformInfo.h
@@ -405,9 +405,7 @@ class RISCVTTIImpl final : public BasicTTIImplBase<RISCVTTIImpl> {
Intrinsic::vp_load,
Intrinsic::vp_load_ff,
Intrinsic::vp_lshr,
- Intrinsic::vp_maxnum,
Intrinsic::vp_merge,
- Intrinsic::vp_minnum,
Intrinsic::vp_mul,
Intrinsic::vp_nearbyint,
Intrinsic::vp_or,
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 403d0b8d57940..60465e948c411 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfmax-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfmax-vp.ll
@@ -11,19 +11,19 @@
define <2 x half> @vfmax_vv_v2f16(<2 x half> %va, <2 x half> %vb, <2 x i1> %m, i32 zeroext %evl) {
; ZVFH-LABEL: vfmax_vv_v2f16:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vsetvli zero, a0, e16, mf4, ta, ma
-; ZVFH-NEXT: vfmax.vv v8, v8, v9, v0.t
+; ZVFH-NEXT: vsetivli zero, 2, e16, mf4, ta, ma
+; ZVFH-NEXT: vfmax.vv v8, v8, v9
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmax_vv_v2f16:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: vsetvli zero, a0, e16, mf4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v10, v9, v0.t
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v9, v8, v0.t
+; ZVFHMIN-NEXT: vsetivli zero, 2, e16, mf4, ta, ma
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v10, v9
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v9, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, mf2, ta, ma
-; ZVFHMIN-NEXT: vfmax.vv v9, v9, v10, v0.t
+; ZVFHMIN-NEXT: vfmax.vv v9, v9, v10
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, mf4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v9, v0.t
+; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v9
; ZVFHMIN-NEXT: ret
%v = call <2 x half> @llvm.vp.maxnum.v2f16(<2 x half> %va, <2 x half> %vb, <2 x i1> %m, i32 %evl)
ret <2 x half> %v
@@ -32,13 +32,13 @@ define <2 x half> @vfmax_vv_v2f16(<2 x half> %va, <2 x half> %vb, <2 x i1> %m, i
define <2 x half> @vfmax_vv_v2f16_unmasked(<2 x half> %va, <2 x half> %vb, i32 zeroext %evl) {
; ZVFH-LABEL: vfmax_vv_v2f16_unmasked:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vsetvli zero, a0, e16, mf4, ta, ma
+; ZVFH-NEXT: vsetivli zero, 2, e16, mf4, ta, ma
; ZVFH-NEXT: vfmax.vv v8, v8, v9
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmax_vv_v2f16_unmasked:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: vsetvli zero, a0, e16, mf4, ta, ma
+; ZVFHMIN-NEXT: vsetivli zero, 2, e16, mf4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v10, v9
; ZVFHMIN-NEXT: vfwcvt.f.f.v v9, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, mf2, ta, ma
@@ -53,19 +53,19 @@ define <2 x half> @vfmax_vv_v2f16_unmasked(<2 x half> %va, <2 x half> %vb, i32 z
define <4 x half> @vfmax_vv_v4f16(<4 x half> %va, <4 x half> %vb, <4 x i1> %m, i32 zeroext %evl) {
; ZVFH-LABEL: vfmax_vv_v4f16:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vsetvli zero, a0, e16, mf2, ta, ma
-; ZVFH-NEXT: vfmax.vv v8, v8, v9, v0.t
+; ZVFH-NEXT: vsetivli zero, 4, e16, mf2, ta, ma
+; ZVFH-NEXT: vfmax.vv v8, v8, v9
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmax_vv_v4f16:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: vsetvli zero, a0, e16, mf2, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v10, v9, v0.t
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v9, v8, v0.t
+; ZVFHMIN-NEXT: vsetivli zero, 4, e16, mf2, ta, ma
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v10, v9
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v9, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m1, ta, ma
-; ZVFHMIN-NEXT: vfmax.vv v9, v9, v10, v0.t
+; ZVFHMIN-NEXT: vfmax.vv v9, v9, v10
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, mf2, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v9, v0.t
+; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v9
; ZVFHMIN-NEXT: ret
%v = call <4 x half> @llvm.vp.maxnum.v4f16(<4 x half> %va, <4 x half> %vb, <4 x i1> %m, i32 %evl)
ret <4 x half> %v
@@ -74,13 +74,13 @@ define <4 x half> @vfmax_vv_v4f16(<4 x half> %va, <4 x half> %vb, <4 x i1> %m, i
define <4 x half> @vfmax_vv_v4f16_unmasked(<4 x half> %va, <4 x half> %vb, i32 zeroext %evl) {
; ZVFH-LABEL: vfmax_vv_v4f16_unmasked:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vsetvli zero, a0, e16, mf2, ta, ma
+; ZVFH-NEXT: vsetivli zero, 4, e16, mf2, ta, ma
; ZVFH-NEXT: vfmax.vv v8, v8, v9
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmax_vv_v4f16_unmasked:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: vsetvli zero, a0, e16, mf2, ta, ma
+; ZVFHMIN-NEXT: vsetivli zero, 4, e16, mf2, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v10, v9
; ZVFHMIN-NEXT: vfwcvt.f.f.v v9, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m1, ta, ma
@@ -95,19 +95,19 @@ define <4 x half> @vfmax_vv_v4f16_unmasked(<4 x half> %va, <4 x half> %vb, i32 z
define <8 x half> @vfmax_vv_v8f16(<8 x half> %va, <8 x half> %vb, <8 x i1> %m, i32 zeroext %evl) {
; ZVFH-LABEL: vfmax_vv_v8f16:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vsetvli zero, a0, e16, m1, ta, ma
-; ZVFH-NEXT: vfmax.vv v8, v8, v9, v0.t
+; ZVFH-NEXT: vsetivli zero, 8, e16, m1, ta, ma
+; ZVFH-NEXT: vfmax.vv v8, v8, v9
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmax_vv_v8f16:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: vsetvli zero, a0, e16, m1, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v10, v9, v0.t
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v12, v8, v0.t
+; ZVFHMIN-NEXT: vsetivli zero, 8, e16, m1, ta, ma
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v10, v9
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v12, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m2, ta, ma
-; ZVFHMIN-NEXT: vfmax.vv v10, v12, v10, v0.t
+; ZVFHMIN-NEXT: vfmax.vv v10, v12, v10
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m1, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v10, v0.t
+; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v10
; ZVFHMIN-NEXT: ret
%v = call <8 x half> @llvm.vp.maxnum.v8f16(<8 x half> %va, <8 x half> %vb, <8 x i1> %m, i32 %evl)
ret <8 x half> %v
@@ -116,13 +116,13 @@ define <8 x half> @vfmax_vv_v8f16(<8 x half> %va, <8 x half> %vb, <8 x i1> %m, i
define <8 x half> @vfmax_vv_v8f16_unmasked(<8 x half> %va, <8 x half> %vb, i32 zeroext %evl) {
; ZVFH-LABEL: vfmax_vv_v8f16_unmasked:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vsetvli zero, a0, e16, m1, ta, ma
+; ZVFH-NEXT: vsetivli zero, 8, e16, m1, ta, ma
; ZVFH-NEXT: vfmax.vv v8, v8, v9
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmax_vv_v8f16_unmasked:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: vsetvli zero, a0, e16, m1, ta, ma
+; ZVFHMIN-NEXT: vsetivli zero, 8, e16, m1, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v10, v9
; ZVFHMIN-NEXT: vfwcvt.f.f.v v12, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m2, ta, ma
@@ -137,19 +137,19 @@ define <8 x half> @vfmax_vv_v8f16_unmasked(<8 x half> %va, <8 x half> %vb, i32 z
define <16 x half> @vfmax_vv_v16f16(<16 x half> %va, <16 x half> %vb, <16 x i1> %m, i32 zeroext %evl) {
; ZVFH-LABEL: vfmax_vv_v16f16:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vsetvli zero, a0, e16, m2, ta, ma
-; ZVFH-NEXT: vfmax.vv v8, v8, v10, v0.t
+; ZVFH-NEXT: vsetivli zero, 16, e16, m2, ta, ma
+; ZVFH-NEXT: vfmax.vv v8, v8, v10
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmax_vv_v16f16:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: vsetvli zero, a0, e16, m2, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v12, v10, v0.t
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8, v0.t
+; ZVFHMIN-NEXT: vsetivli zero, 16, e16, m2, ta, ma
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v12, v10
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m4, ta, ma
-; ZVFHMIN-NEXT: vfmax.vv v12, v16, v12, v0.t
+; ZVFHMIN-NEXT: vfmax.vv v12, v16, v12
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m2, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v12, v0.t
+; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v12
; ZVFHMIN-NEXT: ret
%v = call <16 x half> @llvm.vp.maxnum.v16f16(<16 x half> %va, <16 x half> %vb, <16 x i1> %m, i32 %evl)
ret <16 x half> %v
@@ -158,13 +158,13 @@ define <16 x half> @vfmax_vv_v16f16(<16 x half> %va, <16 x half> %vb, <16 x i1>
define <16 x half> @vfmax_vv_v16f16_unmasked(<16 x half> %va, <16 x half> %vb, i32 zeroext %evl) {
; ZVFH-LABEL: vfmax_vv_v16f16_unmasked:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vsetvli zero, a0, e16, m2, ta, ma
+; ZVFH-NEXT: vsetivli zero, 16, e16, m2, ta, ma
; ZVFH-NEXT: vfmax.vv v8, v8, v10
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmax_vv_v16f16_unmasked:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: vsetvli zero, a0, e16, m2, ta, ma
+; ZVFHMIN-NEXT: vsetivli zero, 16, e16, m2, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v12, v10
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m4, ta, ma
@@ -179,8 +179,8 @@ define <16 x half> @vfmax_vv_v16f16_unmasked(<16 x half> %va, <16 x half> %vb, i
define <2 x float> @vfmax_vv_v2f32(<2 x float> %va, <2 x float> %vb, <2 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_v2f32:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, ma
-; CHECK-NEXT: vfmax.vv v8, v8, v9, v0.t
+; CHECK-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
+; CHECK-NEXT: vfmax.vv v8, v8, v9
; CHECK-NEXT: ret
%v = call <2 x float> @llvm.vp.maxnum.v2f32(<2 x float> %va, <2 x float> %vb, <2 x i1> %m, i32 %evl)
ret <2 x float> %v
@@ -189,7 +189,7 @@ define <2 x float> @vfmax_vv_v2f32(<2 x float> %va, <2 x float> %vb, <2 x i1> %m
define <2 x float> @vfmax_vv_v2f32_unmasked(<2 x float> %va, <2 x float> %vb, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_v2f32_unmasked:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, ma
+; CHECK-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
; CHECK-NEXT: vfmax.vv v8, v8, v9
; CHECK-NEXT: ret
%v = call <2 x float> @llvm.vp.maxnum.v2f32(<2 x float> %va, <2 x float> %vb, <2 x i1> splat (i1 true), i32 %evl)
@@ -199,8 +199,8 @@ define <2 x float> @vfmax_vv_v2f32_unmasked(<2 x float> %va, <2 x float> %vb, i3
define <4 x float> @vfmax_vv_v4f32(<4 x float> %va, <4 x float> %vb, <4 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_v4f32:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, ma
-; CHECK-NEXT: vfmax.vv v8, v8, v9, v0.t
+; CHECK-NEXT: vsetivli zero, 4, e32, m1, ta, ma
+; CHECK-NEXT: vfmax.vv v8, v8, v9
; CHECK-NEXT: ret
%v = call <4 x float> @llvm.vp.maxnum.v4f32(<4 x float> %va, <4 x float> %vb, <4 x i1> %m, i32 %evl)
ret <4 x float> %v
@@ -209,7 +209,7 @@ define <4 x float> @vfmax_vv_v4f32(<4 x float> %va, <4 x float> %vb, <4 x i1> %m
define <4 x float> @vfmax_vv_v4f32_unmasked(<4 x float> %va, <4 x float> %vb, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_v4f32_unmasked:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, ma
+; CHECK-NEXT: vsetivli zero, 4, e32, m1, ta, ma
; CHECK-NEXT: vfmax.vv v8, v8, v9
; CHECK-NEXT: ret
%v = call <4 x float> @llvm.vp.maxnum.v4f32(<4 x float> %va, <4 x float> %vb, <4 x i1> splat (i1 true), i32 %evl)
@@ -219,8 +219,8 @@ define <4 x float> @vfmax_vv_v4f32_unmasked(<4 x float> %va, <4 x float> %vb, i3
define <8 x float> @vfmax_vv_v8f32(<8 x float> %va, <8 x float> %vb, <8 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_v8f32:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, ma
-; CHECK-NEXT: vfmax.vv v8, v8, v10, v0.t
+; CHECK-NEXT: vsetivli zero, 8, e32, m2, ta, ma
+; CHECK-NEXT: vfmax.vv v8, v8, v10
; CHECK-NEXT: ret
%v = call <8 x float> @llvm.vp.maxnum.v8f32(<8 x float> %va, <8 x float> %vb, <8 x i1> %m, i32 %evl)
ret <8 x float> %v
@@ -229,7 +229,7 @@ define <8 x float> @vfmax_vv_v8f32(<8 x float> %va, <8 x float> %vb, <8 x i1> %m
define <8 x float> @vfmax_vv_v8f32_unmasked(<8 x float> %va, <8 x float> %vb, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_v8f32_unmasked:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, ma
+; CHECK-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; CHECK-NEXT: vfmax.vv v8, v8, v10
; CHECK-NEXT: ret
%v = call <8 x float> @llvm.vp.maxnum.v8f32(<8 x float> %va, <8 x float> %vb, <8 x i1> splat (i1 true), i32 %evl)
@@ -239,8 +239,8 @@ define <8 x float> @vfmax_vv_v8f32_unmasked(<8 x float> %va, <8 x float> %vb, i3
define <16 x float> @vfmax_vv_v16f32(<16 x float> %va, <16 x float> %vb, <16 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_v16f32:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, ma
-; CHECK-NEXT: vfmax.vv v8, v8, v12, v0.t
+; CHECK-NEXT: vsetivli zero, 16, e32, m4, ta, ma
+; CHECK-NEXT: vfmax.vv v8, v8, v12
; CHECK-NEXT: ret
%v = call <16 x float> @llvm.vp.maxnum.v16f32(<16 x float> %va, <16 x float> %vb, <16 x i1> %m, i32 %evl)
ret <16 x float> %v
@@ -249,7 +249,7 @@ define <16 x float> @vfmax_vv_v16f32(<16 x float> %va, <16 x float> %vb, <16 x i
define <16 x float> @vfmax_vv_v16f32_unmasked(<16 x float> %va, <16 x float> %vb, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_v16f32_unmasked:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, ma
+; CHECK-NEXT: vsetivli zero, 16, e32, m4, ta, ma
; CHECK-NEXT: vfmax.vv v8, v8, v12
; CHECK-NEXT: ret
%v = call <16 x float> @llvm.vp.maxnum.v16f32(<16 x float> %va, <16 x float> %vb, <16 x i1> splat (i1 true), i32 %evl)
@@ -259,8 +259,8 @@ define <16 x float> @vfmax_vv_v16f32_unmasked(<16 x float> %va, <16 x float> %vb
define <2 x double> @vfmax_vv_v2f64(<2 x double> %va, <2 x double> %vb, <2 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_v2f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, ma
-; CHECK-NEXT: vfmax.vv v8, v8, v9, v0.t
+; CHECK-NEXT: vsetivli zero, 2, e64, m1, ta, ma
+; CHECK-NEXT: vfmax.vv v8, v8, v9
; CHECK-NEXT: ret
%v = call <2 x double> @llvm.vp.maxnum.v2f64(<2 x double> %va, <2 x double> %vb, <2 x i1> %m, i32 %evl)
ret <2 x double> %v
@@ -269,7 +269,7 @@ define <2 x double> @vfmax_vv_v2f64(<2 x double> %va, <2 x double> %vb, <2 x i1>
define <2 x double> @vfmax_vv_v2f64_unmasked(<2 x double> %va, <2 x double> %vb, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_v2f64_unmasked:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, ma
+; CHECK-NEXT: vsetivli zero, 2, e64, m1, ta, ma
; CHECK-NEXT: vfmax.vv v8, v8, v9
; CHECK-NEXT: ret
%v = call <2 x double> @llvm.vp.maxnum.v2f64(<2 x double> %va, <2 x double> %vb, <2 x i1> splat (i1 true), i32 %evl)
@@ -279,8 +279,8 @@ define <2 x double> @vfmax_vv_v2f64_unmasked(<2 x double> %va, <2 x double> %vb,
define <4 x double> @vfmax_vv_v4f64(<4 x double> %va, <4 x double> %vb, <4 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_v4f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, ma
-; CHECK-NEXT: vfmax.vv v8, v8, v10, v0.t
+; CHECK-NEXT: vsetivli zero, 4, e64, m2, ta, ma
+; CHECK-NEXT: vfmax.vv v8, v8, v10
; CHECK-NEXT: ret
%v = call <4 x double> @llvm.vp.maxnum.v4f64(<4 x double> %va, <4 x double> %vb, <4 x i1> %m, i32 %evl)
ret <4 x double> %v
@@ -289,7 +289,7 @@ define <4 x double> @vfmax_vv_v4f64(<4 x double> %va, <4 x double> %vb, <4 x i1>
define <4 x double> @vfmax_vv_v4f64_unmasked(<4 x double> %va, <4 x double> %vb, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_v4f64_unmasked:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, ma
+; CHECK-NEXT: vsetivli zero, 4, e64, m2, ta, ma
; CHECK-NEXT: vfmax.vv v8, v8, v10
; CHECK-NEXT: ret
%v = call <4 x double> @llvm.vp.maxnum.v4f64(<4 x double> %va, <4 x double> %vb, <4 x i1> splat (i1 true), i32 %evl)
@@ -299,8 +299,8 @@ define <4 x double> @vfmax_vv_v4f64_unmasked(<4 x double> %va, <4 x double> %vb,
define <8 x double> @vfmax_vv_v8f64(<8 x double> %va, <8 x double> %vb, <8 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_v8f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, ma
-; CHECK-NEXT: vfmax.vv v8, v8, v12, v0.t
+; CHECK-NEXT: vsetivli zero, 8, e64, m4, ta, ma
+; CHECK-NEXT: vfmax.vv v8, v8, v12
; CHECK-NEXT: ret
%v = call <8 x double> @llvm.vp.maxnum.v8f64(<8 x double> %va, <8 x double> %vb, <8 x i1> %m, i32 %evl)
ret <8 x double> %v
@@ -309,7 +309,7 @@ define <8 x double> @vfmax_vv_v8f64(<8 x double> %va, <8 x double> %vb, <8 x i1>
define <8 x double> @vfmax_vv_v8f64_unmasked(<8 x double> %va, <8 x double> %vb, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_v8f64_unmasked:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, ma
+; CHECK-NEXT: vsetivli zero, 8, e64, m4, ta, ma
; CHECK-NEXT: vfmax.vv v8, v8, v12
; CHECK-NEXT: ret
%v = call <8 x double> @llvm.vp.maxnum.v8f64(<8 x double> %va, <8 x double> %vb, <8 x i1> splat (i1 true), i32 %evl)
@@ -319,8 +319,8 @@ define <8 x double> @vfmax_vv_v8f64_unmasked(<8 x double> %va, <8 x double> %vb,
define <15 x double> @vfmax_vv_v15f64(<15 x double> %va, <15 x double> %vb, <15 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_v15f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: vfmax.vv v8, v8, v16, v0.t
+; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; CHECK-NEXT: vfmax.vv v8, v8, v16
; CHECK-NEXT: ret
%v = call <15 x double> @llvm.vp.maxnum.v15f64(<15 x double> %va, <15 x double> %vb, <15 x i1> %m, i32 %evl)
ret <15 x double> %v
@@ -329,7 +329,7 @@ define <15 x double> @vfmax_vv_v15f64(<15 x double> %va, <15 x double> %vb, <15
define <15 x double> @vfmax_vv_v15f64_unmasked(<15 x double> %va, <15 x double> %vb, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_v15f64_unmasked:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; CHECK-NEXT: vfmax.vv v8, v8, v16
; CHECK-NEXT: ret
%v = call <15 x double> @llvm.vp.maxnum.v15f64(<15 x double> %va, <15 x double> %vb, <15 x i1> splat (i1 true), i32 %evl)
@@ -339,8 +339,8 @@ define <15 x double> @vfmax_vv_v15f64_unmasked(<15 x double> %va, <15 x double>
define <16 x double> @vfmax_vv_v16f64(<16 x double> %va, <16 x double> %vb, <16 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_v16f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: vfmax.vv v8, v8, v16, v0.t
+; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; CHECK-NEXT: vfmax.vv v8, v8, v16
; CHECK-NEXT: ret
%v = call <16 x double> @llvm.vp.maxnum.v16f64(<16 x double> %va, <16 x double> %vb, <16 x i1> %m, i32 %evl)
ret <16 x double> %v
@@ -349,7 +349,7 @@ define <16 x double> @vfmax_vv_v16f64(<16 x double> %va, <16 x double> %vb, <16
define <16 x double> @vfmax_vv_v16f64_unmasked(<16 x double> %va, <16 x double> %vb, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_v16f64_unmasked:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; CHECK-NEXT: vfmax.vv v8, v8, v16
; CHECK-NEXT: ret
%v = call <16 x double> @llvm.vp.maxnum.v16f64(<16 x double> %va, <16 x double> %vb, <16 x i1> splat (i1 true), i32 %evl)
@@ -359,43 +359,12 @@ define <16 x double> @vfmax_vv_v16f64_unmasked(<16 x double> %va, <16 x double>
define <32 x double> @vfmax_vv_v32f64(<32 x double> %va, <32 x double> %vb, <32 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_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, sp, 16
-; CHECK-NEXT: vs8r.v v16, (a1) # vscale x 64-byte 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: 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 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, 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, v7
-; CHECK-NEXT: addi a1, sp, 16
-; CHECK-NEXT: vl8r.v v24, (a1) # vscale x 64-byte Folded Reload
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: vfmax.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: vle64.v v0, (a1)
+; CHECK-NEXT: vfmax.vv v8, v8, v24
+; CHECK-NEXT: vfmax.vv v16, v16, v0
; CHECK-NEXT: ret
%v = call <32 x double> @llvm.vp.maxnum.v32f64(<32 x double> %va, <32 x double> %vb, <32 x i1> %m, i32 %evl)
ret <32 x double> %v
@@ -406,22 +375,10 @@ define <32 x double> @vfmax_vv_v32f64_unmasked(<32 x double> %va, <32 x double>
; CHECK: # %bb.0:
; CHECK-NEXT: addi a1, a0, 128
; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; CHECK-NEXT: vle64.v v24, (a1)
-; CHECK-NEXT: vle64.v v0, (a0)
-; CHECK-NEXT: li a1, 16
-; CHECK-NEXT: mv a0, a2
-; CHECK-NEXT: bltu a2, a1, .LBB27_2
-; CHECK-NEXT: # %bb.1:
-; CHECK-NEXT: li a0, 16
-; CHECK-NEXT: .LBB27_2:
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: vfmax.vv v8, v8, v0
-; CHECK-NEXT: addi a0, a2, -16
-; CHECK-NEXT: sltu a1, a2, a0
-; CHECK-NEXT: addi a1, a1, -1
-; CHECK-NEXT: and a0, a1, a0
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: vfmax.vv v16, v16, v24
+; CHECK-NEXT: vle64.v v24, (a0)
+; CHECK-NEXT: vle64.v v0, (a1)
+; CHECK-NEXT: vfmax.vv v8, v8, v24
+; CHECK-NEXT: vfmax.vv v16, v16, v0
; CHECK-NEXT: ret
%v = call <32 x double> @llvm.vp.maxnum.v32f64(<32 x double> %va, <32 x double> %vb, <32 x i1> splat (i1 true), i32 %evl)
ret <32 x double> %v
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 56f7a8d48c5a1..6970d5ce2e987 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfmin-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfmin-vp.ll
@@ -11,19 +11,19 @@
define <2 x half> @vfmin_vv_v2f16(<2 x half> %va, <2 x half> %vb, <2 x i1> %m, i32 zeroext %evl) {
; ZVFH-LABEL: vfmin_vv_v2f16:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vsetvli zero, a0, e16, mf4, ta, ma
-; ZVFH-NEXT: vfmin.vv v8, v8, v9, v0.t
+; ZVFH-NEXT: vsetivli zero, 2, e16, mf4, ta, ma
+; ZVFH-NEXT: vfmin.vv v8, v8, v9
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmin_vv_v2f16:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: vsetvli zero, a0, e16, mf4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v10, v9, v0.t
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v9, v8, v0.t
+; ZVFHMIN-NEXT: vsetivli zero, 2, e16, mf4, ta, ma
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v10, v9
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v9, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, mf2, ta, ma
-; ZVFHMIN-NEXT: vfmin.vv v9, v9, v10, v0.t
+; ZVFHMIN-NEXT: vfmin.vv v9, v9, v10
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, mf4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v9, v0.t
+; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v9
; ZVFHMIN-NEXT: ret
%v = call <2 x half> @llvm.vp.minnum.v2f16(<2 x half> %va, <2 x half> %vb, <2 x i1> %m, i32 %evl)
ret <2 x half> %v
@@ -32,13 +32,13 @@ define <2 x half> @vfmin_vv_v2f16(<2 x half> %va, <2 x half> %vb, <2 x i1> %m, i
define <2 x half> @vfmin_vv_v2f16_unmasked(<2 x half> %va, <2 x half> %vb, i32 zeroext %evl) {
; ZVFH-LABEL: vfmin_vv_v2f16_unmasked:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vsetvli zero, a0, e16, mf4, ta, ma
+; ZVFH-NEXT: vsetivli zero, 2, e16, mf4, ta, ma
; ZVFH-NEXT: vfmin.vv v8, v8, v9
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmin_vv_v2f16_unmasked:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: vsetvli zero, a0, e16, mf4, ta, ma
+; ZVFHMIN-NEXT: vsetivli zero, 2, e16, mf4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v10, v9
; ZVFHMIN-NEXT: vfwcvt.f.f.v v9, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, mf2, ta, ma
@@ -53,19 +53,19 @@ define <2 x half> @vfmin_vv_v2f16_unmasked(<2 x half> %va, <2 x half> %vb, i32 z
define <4 x half> @vfmin_vv_v4f16(<4 x half> %va, <4 x half> %vb, <4 x i1> %m, i32 zeroext %evl) {
; ZVFH-LABEL: vfmin_vv_v4f16:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vsetvli zero, a0, e16, mf2, ta, ma
-; ZVFH-NEXT: vfmin.vv v8, v8, v9, v0.t
+; ZVFH-NEXT: vsetivli zero, 4, e16, mf2, ta, ma
+; ZVFH-NEXT: vfmin.vv v8, v8, v9
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmin_vv_v4f16:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: vsetvli zero, a0, e16, mf2, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v10, v9, v0.t
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v9, v8, v0.t
+; ZVFHMIN-NEXT: vsetivli zero, 4, e16, mf2, ta, ma
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v10, v9
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v9, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m1, ta, ma
-; ZVFHMIN-NEXT: vfmin.vv v9, v9, v10, v0.t
+; ZVFHMIN-NEXT: vfmin.vv v9, v9, v10
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, mf2, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v9, v0.t
+; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v9
; ZVFHMIN-NEXT: ret
%v = call <4 x half> @llvm.vp.minnum.v4f16(<4 x half> %va, <4 x half> %vb, <4 x i1> %m, i32 %evl)
ret <4 x half> %v
@@ -74,13 +74,13 @@ define <4 x half> @vfmin_vv_v4f16(<4 x half> %va, <4 x half> %vb, <4 x i1> %m, i
define <4 x half> @vfmin_vv_v4f16_unmasked(<4 x half> %va, <4 x half> %vb, i32 zeroext %evl) {
; ZVFH-LABEL: vfmin_vv_v4f16_unmasked:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vsetvli zero, a0, e16, mf2, ta, ma
+; ZVFH-NEXT: vsetivli zero, 4, e16, mf2, ta, ma
; ZVFH-NEXT: vfmin.vv v8, v8, v9
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmin_vv_v4f16_unmasked:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: vsetvli zero, a0, e16, mf2, ta, ma
+; ZVFHMIN-NEXT: vsetivli zero, 4, e16, mf2, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v10, v9
; ZVFHMIN-NEXT: vfwcvt.f.f.v v9, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m1, ta, ma
@@ -95,19 +95,19 @@ define <4 x half> @vfmin_vv_v4f16_unmasked(<4 x half> %va, <4 x half> %vb, i32 z
define <8 x half> @vfmin_vv_v8f16(<8 x half> %va, <8 x half> %vb, <8 x i1> %m, i32 zeroext %evl) {
; ZVFH-LABEL: vfmin_vv_v8f16:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vsetvli zero, a0, e16, m1, ta, ma
-; ZVFH-NEXT: vfmin.vv v8, v8, v9, v0.t
+; ZVFH-NEXT: vsetivli zero, 8, e16, m1, ta, ma
+; ZVFH-NEXT: vfmin.vv v8, v8, v9
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmin_vv_v8f16:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: vsetvli zero, a0, e16, m1, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v10, v9, v0.t
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v12, v8, v0.t
+; ZVFHMIN-NEXT: vsetivli zero, 8, e16, m1, ta, ma
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v10, v9
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v12, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m2, ta, ma
-; ZVFHMIN-NEXT: vfmin.vv v10, v12, v10, v0.t
+; ZVFHMIN-NEXT: vfmin.vv v10, v12, v10
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m1, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v10, v0.t
+; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v10
; ZVFHMIN-NEXT: ret
%v = call <8 x half> @llvm.vp.minnum.v8f16(<8 x half> %va, <8 x half> %vb, <8 x i1> %m, i32 %evl)
ret <8 x half> %v
@@ -116,13 +116,13 @@ define <8 x half> @vfmin_vv_v8f16(<8 x half> %va, <8 x half> %vb, <8 x i1> %m, i
define <8 x half> @vfmin_vv_v8f16_unmasked(<8 x half> %va, <8 x half> %vb, i32 zeroext %evl) {
; ZVFH-LABEL: vfmin_vv_v8f16_unmasked:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vsetvli zero, a0, e16, m1, ta, ma
+; ZVFH-NEXT: vsetivli zero, 8, e16, m1, ta, ma
; ZVFH-NEXT: vfmin.vv v8, v8, v9
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmin_vv_v8f16_unmasked:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: vsetvli zero, a0, e16, m1, ta, ma
+; ZVFHMIN-NEXT: vsetivli zero, 8, e16, m1, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v10, v9
; ZVFHMIN-NEXT: vfwcvt.f.f.v v12, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m2, ta, ma
@@ -137,19 +137,19 @@ define <8 x half> @vfmin_vv_v8f16_unmasked(<8 x half> %va, <8 x half> %vb, i32 z
define <16 x half> @vfmin_vv_v16f16(<16 x half> %va, <16 x half> %vb, <16 x i1> %m, i32 zeroext %evl) {
; ZVFH-LABEL: vfmin_vv_v16f16:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vsetvli zero, a0, e16, m2, ta, ma
-; ZVFH-NEXT: vfmin.vv v8, v8, v10, v0.t
+; ZVFH-NEXT: vsetivli zero, 16, e16, m2, ta, ma
+; ZVFH-NEXT: vfmin.vv v8, v8, v10
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmin_vv_v16f16:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: vsetvli zero, a0, e16, m2, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v12, v10, v0.t
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8, v0.t
+; ZVFHMIN-NEXT: vsetivli zero, 16, e16, m2, ta, ma
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v12, v10
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m4, ta, ma
-; ZVFHMIN-NEXT: vfmin.vv v12, v16, v12, v0.t
+; ZVFHMIN-NEXT: vfmin.vv v12, v16, v12
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m2, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v12, v0.t
+; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v12
; ZVFHMIN-NEXT: ret
%v = call <16 x half> @llvm.vp.minnum.v16f16(<16 x half> %va, <16 x half> %vb, <16 x i1> %m, i32 %evl)
ret <16 x half> %v
@@ -158,13 +158,13 @@ define <16 x half> @vfmin_vv_v16f16(<16 x half> %va, <16 x half> %vb, <16 x i1>
define <16 x half> @vfmin_vv_v16f16_unmasked(<16 x half> %va, <16 x half> %vb, i32 zeroext %evl) {
; ZVFH-LABEL: vfmin_vv_v16f16_unmasked:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vsetvli zero, a0, e16, m2, ta, ma
+; ZVFH-NEXT: vsetivli zero, 16, e16, m2, ta, ma
; ZVFH-NEXT: vfmin.vv v8, v8, v10
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmin_vv_v16f16_unmasked:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: vsetvli zero, a0, e16, m2, ta, ma
+; ZVFHMIN-NEXT: vsetivli zero, 16, e16, m2, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v12, v10
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m4, ta, ma
@@ -179,8 +179,8 @@ define <16 x half> @vfmin_vv_v16f16_unmasked(<16 x half> %va, <16 x half> %vb, i
define <2 x float> @vfmin_vv_v2f32(<2 x float> %va, <2 x float> %vb, <2 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_v2f32:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, ma
-; CHECK-NEXT: vfmin.vv v8, v8, v9, v0.t
+; CHECK-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
+; CHECK-NEXT: vfmin.vv v8, v8, v9
; CHECK-NEXT: ret
%v = call <2 x float> @llvm.vp.minnum.v2f32(<2 x float> %va, <2 x float> %vb, <2 x i1> %m, i32 %evl)
ret <2 x float> %v
@@ -189,7 +189,7 @@ define <2 x float> @vfmin_vv_v2f32(<2 x float> %va, <2 x float> %vb, <2 x i1> %m
define <2 x float> @vfmin_vv_v2f32_unmasked(<2 x float> %va, <2 x float> %vb, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_v2f32_unmasked:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, ma
+; CHECK-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
; CHECK-NEXT: vfmin.vv v8, v8, v9
; CHECK-NEXT: ret
%v = call <2 x float> @llvm.vp.minnum.v2f32(<2 x float> %va, <2 x float> %vb, <2 x i1> splat (i1 true), i32 %evl)
@@ -199,8 +199,8 @@ define <2 x float> @vfmin_vv_v2f32_unmasked(<2 x float> %va, <2 x float> %vb, i3
define <4 x float> @vfmin_vv_v4f32(<4 x float> %va, <4 x float> %vb, <4 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_v4f32:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, ma
-; CHECK-NEXT: vfmin.vv v8, v8, v9, v0.t
+; CHECK-NEXT: vsetivli zero, 4, e32, m1, ta, ma
+; CHECK-NEXT: vfmin.vv v8, v8, v9
; CHECK-NEXT: ret
%v = call <4 x float> @llvm.vp.minnum.v4f32(<4 x float> %va, <4 x float> %vb, <4 x i1> %m, i32 %evl)
ret <4 x float> %v
@@ -209,7 +209,7 @@ define <4 x float> @vfmin_vv_v4f32(<4 x float> %va, <4 x float> %vb, <4 x i1> %m
define <4 x float> @vfmin_vv_v4f32_unmasked(<4 x float> %va, <4 x float> %vb, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_v4f32_unmasked:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, ma
+; CHECK-NEXT: vsetivli zero, 4, e32, m1, ta, ma
; CHECK-NEXT: vfmin.vv v8, v8, v9
; CHECK-NEXT: ret
%v = call <4 x float> @llvm.vp.minnum.v4f32(<4 x float> %va, <4 x float> %vb, <4 x i1> splat (i1 true), i32 %evl)
@@ -219,8 +219,8 @@ define <4 x float> @vfmin_vv_v4f32_unmasked(<4 x float> %va, <4 x float> %vb, i3
define <8 x float> @vfmin_vv_v8f32(<8 x float> %va, <8 x float> %vb, <8 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_v8f32:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, ma
-; CHECK-NEXT: vfmin.vv v8, v8, v10, v0.t
+; CHECK-NEXT: vsetivli zero, 8, e32, m2, ta, ma
+; CHECK-NEXT: vfmin.vv v8, v8, v10
; CHECK-NEXT: ret
%v = call <8 x float> @llvm.vp.minnum.v8f32(<8 x float> %va, <8 x float> %vb, <8 x i1> %m, i32 %evl)
ret <8 x float> %v
@@ -229,7 +229,7 @@ define <8 x float> @vfmin_vv_v8f32(<8 x float> %va, <8 x float> %vb, <8 x i1> %m
define <8 x float> @vfmin_vv_v8f32_unmasked(<8 x float> %va, <8 x float> %vb, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_v8f32_unmasked:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, ma
+; CHECK-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; CHECK-NEXT: vfmin.vv v8, v8, v10
; CHECK-NEXT: ret
%v = call <8 x float> @llvm.vp.minnum.v8f32(<8 x float> %va, <8 x float> %vb, <8 x i1> splat (i1 true), i32 %evl)
@@ -239,8 +239,8 @@ define <8 x float> @vfmin_vv_v8f32_unmasked(<8 x float> %va, <8 x float> %vb, i3
define <16 x float> @vfmin_vv_v16f32(<16 x float> %va, <16 x float> %vb, <16 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_v16f32:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, ma
-; CHECK-NEXT: vfmin.vv v8, v8, v12, v0.t
+; CHECK-NEXT: vsetivli zero, 16, e32, m4, ta, ma
+; CHECK-NEXT: vfmin.vv v8, v8, v12
; CHECK-NEXT: ret
%v = call <16 x float> @llvm.vp.minnum.v16f32(<16 x float> %va, <16 x float> %vb, <16 x i1> %m, i32 %evl)
ret <16 x float> %v
@@ -249,7 +249,7 @@ define <16 x float> @vfmin_vv_v16f32(<16 x float> %va, <16 x float> %vb, <16 x i
define <16 x float> @vfmin_vv_v16f32_unmasked(<16 x float> %va, <16 x float> %vb, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_v16f32_unmasked:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, ma
+; CHECK-NEXT: vsetivli zero, 16, e32, m4, ta, ma
; CHECK-NEXT: vfmin.vv v8, v8, v12
; CHECK-NEXT: ret
%v = call <16 x float> @llvm.vp.minnum.v16f32(<16 x float> %va, <16 x float> %vb, <16 x i1> splat (i1 true), i32 %evl)
@@ -259,8 +259,8 @@ define <16 x float> @vfmin_vv_v16f32_unmasked(<16 x float> %va, <16 x float> %vb
define <2 x double> @vfmin_vv_v2f64(<2 x double> %va, <2 x double> %vb, <2 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_v2f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, ma
-; CHECK-NEXT: vfmin.vv v8, v8, v9, v0.t
+; CHECK-NEXT: vsetivli zero, 2, e64, m1, ta, ma
+; CHECK-NEXT: vfmin.vv v8, v8, v9
; CHECK-NEXT: ret
%v = call <2 x double> @llvm.vp.minnum.v2f64(<2 x double> %va, <2 x double> %vb, <2 x i1> %m, i32 %evl)
ret <2 x double> %v
@@ -269,7 +269,7 @@ define <2 x double> @vfmin_vv_v2f64(<2 x double> %va, <2 x double> %vb, <2 x i1>
define <2 x double> @vfmin_vv_v2f64_unmasked(<2 x double> %va, <2 x double> %vb, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_v2f64_unmasked:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, ma
+; CHECK-NEXT: vsetivli zero, 2, e64, m1, ta, ma
; CHECK-NEXT: vfmin.vv v8, v8, v9
; CHECK-NEXT: ret
%v = call <2 x double> @llvm.vp.minnum.v2f64(<2 x double> %va, <2 x double> %vb, <2 x i1> splat (i1 true), i32 %evl)
@@ -279,8 +279,8 @@ define <2 x double> @vfmin_vv_v2f64_unmasked(<2 x double> %va, <2 x double> %vb,
define <4 x double> @vfmin_vv_v4f64(<4 x double> %va, <4 x double> %vb, <4 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_v4f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, ma
-; CHECK-NEXT: vfmin.vv v8, v8, v10, v0.t
+; CHECK-NEXT: vsetivli zero, 4, e64, m2, ta, ma
+; CHECK-NEXT: vfmin.vv v8, v8, v10
; CHECK-NEXT: ret
%v = call <4 x double> @llvm.vp.minnum.v4f64(<4 x double> %va, <4 x double> %vb, <4 x i1> %m, i32 %evl)
ret <4 x double> %v
@@ -289,7 +289,7 @@ define <4 x double> @vfmin_vv_v4f64(<4 x double> %va, <4 x double> %vb, <4 x i1>
define <4 x double> @vfmin_vv_v4f64_unmasked(<4 x double> %va, <4 x double> %vb, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_v4f64_unmasked:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, ma
+; CHECK-NEXT: vsetivli zero, 4, e64, m2, ta, ma
; CHECK-NEXT: vfmin.vv v8, v8, v10
; CHECK-NEXT: ret
%v = call <4 x double> @llvm.vp.minnum.v4f64(<4 x double> %va, <4 x double> %vb, <4 x i1> splat (i1 true), i32 %evl)
@@ -299,8 +299,8 @@ define <4 x double> @vfmin_vv_v4f64_unmasked(<4 x double> %va, <4 x double> %vb,
define <8 x double> @vfmin_vv_v8f64(<8 x double> %va, <8 x double> %vb, <8 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_v8f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, ma
-; CHECK-NEXT: vfmin.vv v8, v8, v12, v0.t
+; CHECK-NEXT: vsetivli zero, 8, e64, m4, ta, ma
+; CHECK-NEXT: vfmin.vv v8, v8, v12
; CHECK-NEXT: ret
%v = call <8 x double> @llvm.vp.minnum.v8f64(<8 x double> %va, <8 x double> %vb, <8 x i1> %m, i32 %evl)
ret <8 x double> %v
@@ -309,7 +309,7 @@ define <8 x double> @vfmin_vv_v8f64(<8 x double> %va, <8 x double> %vb, <8 x i1>
define <8 x double> @vfmin_vv_v8f64_unmasked(<8 x double> %va, <8 x double> %vb, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_v8f64_unmasked:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, ma
+; CHECK-NEXT: vsetivli zero, 8, e64, m4, ta, ma
; CHECK-NEXT: vfmin.vv v8, v8, v12
; CHECK-NEXT: ret
%v = call <8 x double> @llvm.vp.minnum.v8f64(<8 x double> %va, <8 x double> %vb, <8 x i1> splat (i1 true), i32 %evl)
@@ -319,8 +319,8 @@ define <8 x double> @vfmin_vv_v8f64_unmasked(<8 x double> %va, <8 x double> %vb,
define <15 x double> @vfmin_vv_v15f64(<15 x double> %va, <15 x double> %vb, <15 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_v15f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: vfmin.vv v8, v8, v16, v0.t
+; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; CHECK-NEXT: vfmin.vv v8, v8, v16
; CHECK-NEXT: ret
%v = call <15 x double> @llvm.vp.minnum.v15f64(<15 x double> %va, <15 x double> %vb, <15 x i1> %m, i32 %evl)
ret <15 x double> %v
@@ -329,7 +329,7 @@ define <15 x double> @vfmin_vv_v15f64(<15 x double> %va, <15 x double> %vb, <15
define <15 x double> @vfmin_vv_v15f64_unmasked(<15 x double> %va, <15 x double> %vb, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_v15f64_unmasked:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; CHECK-NEXT: vfmin.vv v8, v8, v16
; CHECK-NEXT: ret
%v = call <15 x double> @llvm.vp.minnum.v15f64(<15 x double> %va, <15 x double> %vb, <15 x i1> splat (i1 true), i32 %evl)
@@ -339,8 +339,8 @@ define <15 x double> @vfmin_vv_v15f64_unmasked(<15 x double> %va, <15 x double>
define <16 x double> @vfmin_vv_v16f64(<16 x double> %va, <16 x double> %vb, <16 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_v16f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: vfmin.vv v8, v8, v16, v0.t
+; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
+; CHECK-NEXT: vfmin.vv v8, v8, v16
; CHECK-NEXT: ret
%v = call <16 x double> @llvm.vp.minnum.v16f64(<16 x double> %va, <16 x double> %vb, <16 x i1> %m, i32 %evl)
ret <16 x double> %v
@@ -349,7 +349,7 @@ define <16 x double> @vfmin_vv_v16f64(<16 x double> %va, <16 x double> %vb, <16
define <16 x double> @vfmin_vv_v16f64_unmasked(<16 x double> %va, <16 x double> %vb, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_v16f64_unmasked:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; CHECK-NEXT: vfmin.vv v8, v8, v16
; CHECK-NEXT: ret
%v = call <16 x double> @llvm.vp.minnum.v16f64(<16 x double> %va, <16 x double> %vb, <16 x i1> splat (i1 true), i32 %evl)
@@ -359,43 +359,12 @@ define <16 x double> @vfmin_vv_v16f64_unmasked(<16 x double> %va, <16 x double>
define <32 x double> @vfmin_vv_v32f64(<32 x double> %va, <32 x double> %vb, <32 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_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, sp, 16
-; CHECK-NEXT: vs8r.v v16, (a1) # vscale x 64-byte 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: 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 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, 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, v7
-; CHECK-NEXT: addi a1, sp, 16
-; CHECK-NEXT: vl8r.v v24, (a1) # vscale x 64-byte Folded Reload
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: vfmin.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: vle64.v v0, (a1)
+; CHECK-NEXT: vfmin.vv v8, v8, v24
+; CHECK-NEXT: vfmin.vv v16, v16, v0
; CHECK-NEXT: ret
%v = call <32 x double> @llvm.vp.minnum.v32f64(<32 x double> %va, <32 x double> %vb, <32 x i1> %m, i32 %evl)
ret <32 x double> %v
@@ -406,22 +375,10 @@ define <32 x double> @vfmin_vv_v32f64_unmasked(<32 x double> %va, <32 x double>
; CHECK: # %bb.0:
; CHECK-NEXT: addi a1, a0, 128
; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
-; CHECK-NEXT: vle64.v v24, (a1)
-; CHECK-NEXT: vle64.v v0, (a0)
-; CHECK-NEXT: li a1, 16
-; CHECK-NEXT: mv a0, a2
-; CHECK-NEXT: bltu a2, a1, .LBB27_2
-; CHECK-NEXT: # %bb.1:
-; CHECK-NEXT: li a0, 16
-; CHECK-NEXT: .LBB27_2:
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: vfmin.vv v8, v8, v0
-; CHECK-NEXT: addi a0, a2, -16
-; CHECK-NEXT: sltu a1, a2, a0
-; CHECK-NEXT: addi a1, a1, -1
-; CHECK-NEXT: and a0, a1, a0
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: vfmin.vv v16, v16, v24
+; CHECK-NEXT: vle64.v v24, (a0)
+; CHECK-NEXT: vle64.v v0, (a1)
+; CHECK-NEXT: vfmin.vv v8, v8, v24
+; CHECK-NEXT: vfmin.vv v16, v16, v0
; CHECK-NEXT: ret
%v = call <32 x double> @llvm.vp.minnum.v32f64(<32 x double> %va, <32 x double> %vb, <32 x i1> splat (i1 true), i32 %evl)
ret <32 x double> %v
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfmax-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vfmax-vp.ll
index 394887fee67fc..3f80199b00ff2 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfmax-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfmax-vp.ll
@@ -15,13 +15,13 @@
define <vscale x 1 x bfloat> @vfmax_vv_nxv1bf16(<vscale x 1 x bfloat> %va, <vscale x 1 x bfloat> %vb, <vscale x 1 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_nxv1bf16:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, ma
-; CHECK-NEXT: vfwcvtbf16.f.f.v v10, v9, v0.t
-; CHECK-NEXT: vfwcvtbf16.f.f.v v9, v8, v0.t
+; CHECK-NEXT: vsetvli a0, zero, e16, mf4, ta, ma
+; CHECK-NEXT: vfwcvtbf16.f.f.v v10, v9
+; CHECK-NEXT: vfwcvtbf16.f.f.v v9, v8
; CHECK-NEXT: vsetvli zero, zero, e32, mf2, ta, ma
-; CHECK-NEXT: vfmax.vv v9, v9, v10, v0.t
+; CHECK-NEXT: vfmax.vv v9, v9, v10
; CHECK-NEXT: vsetvli zero, zero, e16, mf4, ta, ma
-; CHECK-NEXT: vfncvtbf16.f.f.w v8, v9, v0.t
+; CHECK-NEXT: vfncvtbf16.f.f.w v8, v9
; CHECK-NEXT: ret
%v = call <vscale x 1 x bfloat> @llvm.vp.maxnum.nxv1bf16(<vscale x 1 x bfloat> %va, <vscale x 1 x bfloat> %vb, <vscale x 1 x i1> %m, i32 %evl)
ret <vscale x 1 x bfloat> %v
@@ -30,7 +30,7 @@ define <vscale x 1 x bfloat> @vfmax_vv_nxv1bf16(<vscale x 1 x bfloat> %va, <vsca
define <vscale x 1 x bfloat> @vfmax_vv_nxv1bf16_unmasked(<vscale x 1 x bfloat> %va, <vscale x 1 x bfloat> %vb, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_nxv1bf16_unmasked:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, ma
+; CHECK-NEXT: vsetvli a0, zero, e16, mf4, ta, ma
; CHECK-NEXT: vfwcvtbf16.f.f.v v10, v9
; CHECK-NEXT: vfwcvtbf16.f.f.v v9, v8
; CHECK-NEXT: vsetvli zero, zero, e32, mf2, ta, ma
@@ -45,13 +45,13 @@ define <vscale x 1 x bfloat> @vfmax_vv_nxv1bf16_unmasked(<vscale x 1 x bfloat> %
define <vscale x 2 x bfloat> @vfmax_vv_nxv2bf16(<vscale x 2 x bfloat> %va, <vscale x 2 x bfloat> %vb, <vscale x 2 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_nxv2bf16:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, ma
-; CHECK-NEXT: vfwcvtbf16.f.f.v v10, v9, v0.t
-; CHECK-NEXT: vfwcvtbf16.f.f.v v9, v8, v0.t
+; CHECK-NEXT: vsetvli a0, zero, e16, mf2, ta, ma
+; CHECK-NEXT: vfwcvtbf16.f.f.v v10, v9
+; CHECK-NEXT: vfwcvtbf16.f.f.v v9, v8
; CHECK-NEXT: vsetvli zero, zero, e32, m1, ta, ma
-; CHECK-NEXT: vfmax.vv v9, v9, v10, v0.t
+; CHECK-NEXT: vfmax.vv v9, v9, v10
; CHECK-NEXT: vsetvli zero, zero, e16, mf2, ta, ma
-; CHECK-NEXT: vfncvtbf16.f.f.w v8, v9, v0.t
+; CHECK-NEXT: vfncvtbf16.f.f.w v8, v9
; CHECK-NEXT: ret
%v = call <vscale x 2 x bfloat> @llvm.vp.maxnum.nxv2bf16(<vscale x 2 x bfloat> %va, <vscale x 2 x bfloat> %vb, <vscale x 2 x i1> %m, i32 %evl)
ret <vscale x 2 x bfloat> %v
@@ -60,7 +60,7 @@ define <vscale x 2 x bfloat> @vfmax_vv_nxv2bf16(<vscale x 2 x bfloat> %va, <vsca
define <vscale x 2 x bfloat> @vfmax_vv_nxv2bf16_unmasked(<vscale x 2 x bfloat> %va, <vscale x 2 x bfloat> %vb, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_nxv2bf16_unmasked:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, ma
+; CHECK-NEXT: vsetvli a0, zero, e16, mf2, ta, ma
; CHECK-NEXT: vfwcvtbf16.f.f.v v10, v9
; CHECK-NEXT: vfwcvtbf16.f.f.v v9, v8
; CHECK-NEXT: vsetvli zero, zero, e32, m1, ta, ma
@@ -75,13 +75,13 @@ define <vscale x 2 x bfloat> @vfmax_vv_nxv2bf16_unmasked(<vscale x 2 x bfloat> %
define <vscale x 4 x bfloat> @vfmax_vv_nxv4bf16(<vscale x 4 x bfloat> %va, <vscale x 4 x bfloat> %vb, <vscale x 4 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_nxv4bf16:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, ma
-; CHECK-NEXT: vfwcvtbf16.f.f.v v10, v9, v0.t
-; CHECK-NEXT: vfwcvtbf16.f.f.v v12, v8, v0.t
+; CHECK-NEXT: vsetvli a0, zero, e16, m1, ta, ma
+; CHECK-NEXT: vfwcvtbf16.f.f.v v10, v9
+; CHECK-NEXT: vfwcvtbf16.f.f.v v12, v8
; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, ma
-; CHECK-NEXT: vfmax.vv v10, v12, v10, v0.t
+; CHECK-NEXT: vfmax.vv v10, v12, v10
; CHECK-NEXT: vsetvli zero, zero, e16, m1, ta, ma
-; CHECK-NEXT: vfncvtbf16.f.f.w v8, v10, v0.t
+; CHECK-NEXT: vfncvtbf16.f.f.w v8, v10
; CHECK-NEXT: ret
%v = call <vscale x 4 x bfloat> @llvm.vp.maxnum.nxv4bf16(<vscale x 4 x bfloat> %va, <vscale x 4 x bfloat> %vb, <vscale x 4 x i1> %m, i32 %evl)
ret <vscale x 4 x bfloat> %v
@@ -90,7 +90,7 @@ define <vscale x 4 x bfloat> @vfmax_vv_nxv4bf16(<vscale x 4 x bfloat> %va, <vsca
define <vscale x 4 x bfloat> @vfmax_vv_nxv4bf16_unmasked(<vscale x 4 x bfloat> %va, <vscale x 4 x bfloat> %vb, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_nxv4bf16_unmasked:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, ma
+; CHECK-NEXT: vsetvli a0, zero, e16, m1, ta, ma
; CHECK-NEXT: vfwcvtbf16.f.f.v v10, v9
; CHECK-NEXT: vfwcvtbf16.f.f.v v12, v8
; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, ma
@@ -105,13 +105,13 @@ define <vscale x 4 x bfloat> @vfmax_vv_nxv4bf16_unmasked(<vscale x 4 x bfloat> %
define <vscale x 8 x bfloat> @vfmax_vv_nxv8bf16(<vscale x 8 x bfloat> %va, <vscale x 8 x bfloat> %vb, <vscale x 8 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_nxv8bf16:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, ma
-; CHECK-NEXT: vfwcvtbf16.f.f.v v12, v10, v0.t
-; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v8, v0.t
+; CHECK-NEXT: vsetvli a0, zero, e16, m2, ta, ma
+; CHECK-NEXT: vfwcvtbf16.f.f.v v12, v10
+; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v8
; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, ma
-; CHECK-NEXT: vfmax.vv v12, v16, v12, v0.t
+; CHECK-NEXT: vfmax.vv v12, v16, v12
; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, ma
-; CHECK-NEXT: vfncvtbf16.f.f.w v8, v12, v0.t
+; CHECK-NEXT: vfncvtbf16.f.f.w v8, v12
; CHECK-NEXT: ret
%v = call <vscale x 8 x bfloat> @llvm.vp.maxnum.nxv8bf16(<vscale x 8 x bfloat> %va, <vscale x 8 x bfloat> %vb, <vscale x 8 x i1> %m, i32 %evl)
ret <vscale x 8 x bfloat> %v
@@ -120,7 +120,7 @@ define <vscale x 8 x bfloat> @vfmax_vv_nxv8bf16(<vscale x 8 x bfloat> %va, <vsca
define <vscale x 8 x bfloat> @vfmax_vv_nxv8bf16_unmasked(<vscale x 8 x bfloat> %va, <vscale x 8 x bfloat> %vb, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_nxv8bf16_unmasked:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, ma
+; CHECK-NEXT: vsetvli a0, zero, e16, m2, ta, ma
; CHECK-NEXT: vfwcvtbf16.f.f.v v12, v10
; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v8
; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, ma
@@ -135,13 +135,13 @@ define <vscale x 8 x bfloat> @vfmax_vv_nxv8bf16_unmasked(<vscale x 8 x bfloat> %
define <vscale x 16 x bfloat> @vfmax_vv_nxv16bf16(<vscale x 16 x bfloat> %va, <vscale x 16 x bfloat> %vb, <vscale x 16 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_nxv16bf16:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, ma
-; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v12, v0.t
-; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v8, v0.t
+; 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
; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; CHECK-NEXT: vfmax.vv v16, v24, v16, v0.t
+; CHECK-NEXT: vfmax.vv v16, v24, v16
; CHECK-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; CHECK-NEXT: vfncvtbf16.f.f.w v8, v16, v0.t
+; CHECK-NEXT: vfncvtbf16.f.f.w v8, v16
; CHECK-NEXT: ret
%v = call <vscale x 16 x bfloat> @llvm.vp.maxnum.nxv16bf16(<vscale x 16 x bfloat> %va, <vscale x 16 x bfloat> %vb, <vscale x 16 x i1> %m, i32 %evl)
ret <vscale x 16 x bfloat> %v
@@ -150,7 +150,7 @@ 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: vsetvli zero, a0, e16, m4, ta, ma
+; 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
; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, ma
@@ -167,43 +167,26 @@ define <vscale x 32 x bfloat> @vfmax_vv_nxv32bf16(<vscale x 32 x bfloat> %va, <v
; 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: 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 a1, zero, e8, mf2, ta, ma
-; CHECK-NEXT: vmv1r.v v7, v0
-; CHECK-NEXT: csrr a2, vlenb
-; CHECK-NEXT: slli a1, a2, 1
-; CHECK-NEXT: srli a2, a2, 2
-; CHECK-NEXT: sub a3, a0, a1
-; CHECK-NEXT: vslidedown.vx v0, v0, a2
-; 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: vs8r.v v16, (a3) # vscale x 64-byte Folded Spill
-; CHECK-NEXT: vsetvli zero, a2, e16, m4, ta, ma
-; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v20, v0.t
-; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v12, v0.t
+; CHECK-NEXT: vsetvli a0, zero, e16, m4, ta, ma
+; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v16
+; CHECK-NEXT: addi a0, sp, 16
+; CHECK-NEXT: vs8r.v v24, (a0) # vscale x 64-byte Folded Spill
+; CHECK-NEXT: vfwcvtbf16.f.f.v v0, v8
+; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v20
+; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v12
+; CHECK-NEXT: vl8r.v v8, (a0) # vscale x 64-byte Folded Reload
; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; CHECK-NEXT: vfmax.vv v16, v16, v24, v0.t
+; CHECK-NEXT: vfmax.vv v0, v0, v8
; CHECK-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; CHECK-NEXT: vfncvtbf16.f.f.w v12, v16, v0.t
-; CHECK-NEXT: bltu a0, a1, .LBB10_2
-; CHECK-NEXT: # %bb.1:
-; CHECK-NEXT: mv a0, a1
-; CHECK-NEXT: .LBB10_2:
-; CHECK-NEXT: vmv1r.v v0, v7
-; CHECK-NEXT: addi a1, sp, 16
-; CHECK-NEXT: vl8r.v v24, (a1) # vscale x 64-byte Folded Reload
-; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, ma
-; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v24, v0.t
-; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v8, v0.t
+; CHECK-NEXT: vfncvtbf16.f.f.w v8, v0
; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; CHECK-NEXT: vfmax.vv v16, v24, v16, v0.t
+; CHECK-NEXT: vfmax.vv v16, v16, v24
; CHECK-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; CHECK-NEXT: vfncvtbf16.f.f.w v8, v16, v0.t
+; CHECK-NEXT: vfncvtbf16.f.f.w v12, v16
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add sp, sp, a0
@@ -220,43 +203,26 @@ define <vscale x 32 x bfloat> @vfmax_vv_nxv32bf16_unmasked(<vscale x 32 x bfloat
; 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: 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: csrr a2, vlenb
-; CHECK-NEXT: vsetvli a1, zero, e8, m4, ta, ma
-; CHECK-NEXT: vmset.m v24
-; 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, v24, a2
-; 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: vs8r.v v16, (a3) # vscale x 64-byte Folded Spill
-; CHECK-NEXT: vsetvli zero, a2, e16, m4, ta, ma
-; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v20, v0.t
-; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v12, v0.t
+; CHECK-NEXT: vsetvli a0, zero, e16, m4, ta, ma
+; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v16
+; CHECK-NEXT: addi a0, sp, 16
+; CHECK-NEXT: vs8r.v v24, (a0) # vscale x 64-byte Folded Spill
+; CHECK-NEXT: vfwcvtbf16.f.f.v v0, v8
+; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v20
+; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v12
+; CHECK-NEXT: vl8r.v v8, (a0) # vscale x 64-byte Folded Reload
; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; CHECK-NEXT: vfmax.vv v16, v16, v24, v0.t
+; CHECK-NEXT: vfmax.vv v0, v0, v8
; CHECK-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; CHECK-NEXT: vfncvtbf16.f.f.w v12, v16, v0.t
-; CHECK-NEXT: bltu a0, a1, .LBB11_2
-; CHECK-NEXT: # %bb.1:
-; CHECK-NEXT: mv a0, a1
-; CHECK-NEXT: .LBB11_2:
-; CHECK-NEXT: addi a1, sp, 16
-; CHECK-NEXT: vl8r.v v24, (a1) # vscale x 64-byte Folded Reload
-; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, ma
-; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v24
-; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v8
+; CHECK-NEXT: vfncvtbf16.f.f.w v8, v0
; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; CHECK-NEXT: vfmax.vv v16, v24, v16
+; CHECK-NEXT: vfmax.vv v16, v16, v24
; CHECK-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; CHECK-NEXT: vfncvtbf16.f.f.w v8, v16
+; CHECK-NEXT: vfncvtbf16.f.f.w v12, v16
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add sp, sp, a0
@@ -271,19 +237,19 @@ define <vscale x 32 x bfloat> @vfmax_vv_nxv32bf16_unmasked(<vscale x 32 x bfloat
define <vscale x 1 x half> @vfmax_vv_nxv1f16(<vscale x 1 x half> %va, <vscale x 1 x half> %vb, <vscale x 1 x i1> %m, i32 zeroext %evl) {
; ZVFH-LABEL: vfmax_vv_nxv1f16:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vsetvli zero, a0, e16, mf4, ta, ma
-; ZVFH-NEXT: vfmax.vv v8, v8, v9, v0.t
+; ZVFH-NEXT: vsetvli a0, zero, e16, mf4, ta, ma
+; ZVFH-NEXT: vfmax.vv v8, v8, v9
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmax_vv_nxv1f16:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: vsetvli zero, a0, e16, mf4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v10, v9, v0.t
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v9, v8, v0.t
+; ZVFHMIN-NEXT: vsetvli a0, zero, e16, mf4, ta, ma
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v10, v9
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v9, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, mf2, ta, ma
-; ZVFHMIN-NEXT: vfmax.vv v9, v9, v10, v0.t
+; ZVFHMIN-NEXT: vfmax.vv v9, v9, v10
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, mf4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v9, v0.t
+; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v9
; ZVFHMIN-NEXT: ret
%v = call <vscale x 1 x half> @llvm.vp.maxnum.nxv1f16(<vscale x 1 x half> %va, <vscale x 1 x half> %vb, <vscale x 1 x i1> %m, i32 %evl)
ret <vscale x 1 x half> %v
@@ -292,13 +258,13 @@ define <vscale x 1 x half> @vfmax_vv_nxv1f16(<vscale x 1 x half> %va, <vscale x
define <vscale x 1 x half> @vfmax_vv_nxv1f16_unmasked(<vscale x 1 x half> %va, <vscale x 1 x half> %vb, i32 zeroext %evl) {
; ZVFH-LABEL: vfmax_vv_nxv1f16_unmasked:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vsetvli zero, a0, e16, mf4, ta, ma
+; ZVFH-NEXT: vsetvli a0, zero, e16, mf4, ta, ma
; ZVFH-NEXT: vfmax.vv v8, v8, v9
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmax_vv_nxv1f16_unmasked:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: vsetvli zero, a0, e16, mf4, ta, ma
+; ZVFHMIN-NEXT: vsetvli a0, zero, e16, mf4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v10, v9
; ZVFHMIN-NEXT: vfwcvt.f.f.v v9, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, mf2, ta, ma
@@ -313,19 +279,19 @@ define <vscale x 1 x half> @vfmax_vv_nxv1f16_unmasked(<vscale x 1 x half> %va, <
define <vscale x 2 x half> @vfmax_vv_nxv2f16(<vscale x 2 x half> %va, <vscale x 2 x half> %vb, <vscale x 2 x i1> %m, i32 zeroext %evl) {
; ZVFH-LABEL: vfmax_vv_nxv2f16:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vsetvli zero, a0, e16, mf2, ta, ma
-; ZVFH-NEXT: vfmax.vv v8, v8, v9, v0.t
+; ZVFH-NEXT: vsetvli a0, zero, e16, mf2, ta, ma
+; ZVFH-NEXT: vfmax.vv v8, v8, v9
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmax_vv_nxv2f16:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: vsetvli zero, a0, e16, mf2, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v10, v9, v0.t
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v9, v8, v0.t
+; ZVFHMIN-NEXT: vsetvli a0, zero, e16, mf2, ta, ma
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v10, v9
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v9, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m1, ta, ma
-; ZVFHMIN-NEXT: vfmax.vv v9, v9, v10, v0.t
+; ZVFHMIN-NEXT: vfmax.vv v9, v9, v10
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, mf2, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v9, v0.t
+; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v9
; ZVFHMIN-NEXT: ret
%v = call <vscale x 2 x half> @llvm.vp.maxnum.nxv2f16(<vscale x 2 x half> %va, <vscale x 2 x half> %vb, <vscale x 2 x i1> %m, i32 %evl)
ret <vscale x 2 x half> %v
@@ -334,13 +300,13 @@ define <vscale x 2 x half> @vfmax_vv_nxv2f16(<vscale x 2 x half> %va, <vscale x
define <vscale x 2 x half> @vfmax_vv_nxv2f16_unmasked(<vscale x 2 x half> %va, <vscale x 2 x half> %vb, i32 zeroext %evl) {
; ZVFH-LABEL: vfmax_vv_nxv2f16_unmasked:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vsetvli zero, a0, e16, mf2, ta, ma
+; ZVFH-NEXT: vsetvli a0, zero, e16, mf2, ta, ma
; ZVFH-NEXT: vfmax.vv v8, v8, v9
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmax_vv_nxv2f16_unmasked:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: vsetvli zero, a0, e16, mf2, ta, ma
+; ZVFHMIN-NEXT: vsetvli a0, zero, e16, mf2, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v10, v9
; ZVFHMIN-NEXT: vfwcvt.f.f.v v9, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m1, ta, ma
@@ -355,19 +321,19 @@ define <vscale x 2 x half> @vfmax_vv_nxv2f16_unmasked(<vscale x 2 x half> %va, <
define <vscale x 4 x half> @vfmax_vv_nxv4f16(<vscale x 4 x half> %va, <vscale x 4 x half> %vb, <vscale x 4 x i1> %m, i32 zeroext %evl) {
; ZVFH-LABEL: vfmax_vv_nxv4f16:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vsetvli zero, a0, e16, m1, ta, ma
-; ZVFH-NEXT: vfmax.vv v8, v8, v9, v0.t
+; ZVFH-NEXT: vsetvli a0, zero, e16, m1, ta, ma
+; ZVFH-NEXT: vfmax.vv v8, v8, v9
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmax_vv_nxv4f16:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: vsetvli zero, a0, e16, m1, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v10, v9, v0.t
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v12, v8, v0.t
+; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m1, ta, ma
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v10, v9
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v12, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m2, ta, ma
-; ZVFHMIN-NEXT: vfmax.vv v10, v12, v10, v0.t
+; ZVFHMIN-NEXT: vfmax.vv v10, v12, v10
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m1, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v10, v0.t
+; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v10
; ZVFHMIN-NEXT: ret
%v = call <vscale x 4 x half> @llvm.vp.maxnum.nxv4f16(<vscale x 4 x half> %va, <vscale x 4 x half> %vb, <vscale x 4 x i1> %m, i32 %evl)
ret <vscale x 4 x half> %v
@@ -376,13 +342,13 @@ define <vscale x 4 x half> @vfmax_vv_nxv4f16(<vscale x 4 x half> %va, <vscale x
define <vscale x 4 x half> @vfmax_vv_nxv4f16_unmasked(<vscale x 4 x half> %va, <vscale x 4 x half> %vb, i32 zeroext %evl) {
; ZVFH-LABEL: vfmax_vv_nxv4f16_unmasked:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vsetvli zero, a0, e16, m1, ta, ma
+; ZVFH-NEXT: vsetvli a0, zero, e16, m1, ta, ma
; ZVFH-NEXT: vfmax.vv v8, v8, v9
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmax_vv_nxv4f16_unmasked:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: vsetvli zero, a0, e16, m1, ta, ma
+; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m1, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v10, v9
; ZVFHMIN-NEXT: vfwcvt.f.f.v v12, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m2, ta, ma
@@ -397,19 +363,19 @@ define <vscale x 4 x half> @vfmax_vv_nxv4f16_unmasked(<vscale x 4 x half> %va, <
define <vscale x 8 x half> @vfmax_vv_nxv8f16(<vscale x 8 x half> %va, <vscale x 8 x half> %vb, <vscale x 8 x i1> %m, i32 zeroext %evl) {
; ZVFH-LABEL: vfmax_vv_nxv8f16:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vsetvli zero, a0, e16, m2, ta, ma
-; ZVFH-NEXT: vfmax.vv v8, v8, v10, v0.t
+; ZVFH-NEXT: vsetvli a0, zero, e16, m2, ta, ma
+; ZVFH-NEXT: vfmax.vv v8, v8, v10
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmax_vv_nxv8f16:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: vsetvli zero, a0, e16, m2, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v12, v10, v0.t
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8, v0.t
+; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m2, ta, ma
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v12, v10
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m4, ta, ma
-; ZVFHMIN-NEXT: vfmax.vv v12, v16, v12, v0.t
+; ZVFHMIN-NEXT: vfmax.vv v12, v16, v12
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m2, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v12, v0.t
+; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v12
; ZVFHMIN-NEXT: ret
%v = call <vscale x 8 x half> @llvm.vp.maxnum.nxv8f16(<vscale x 8 x half> %va, <vscale x 8 x half> %vb, <vscale x 8 x i1> %m, i32 %evl)
ret <vscale x 8 x half> %v
@@ -418,13 +384,13 @@ define <vscale x 8 x half> @vfmax_vv_nxv8f16(<vscale x 8 x half> %va, <vscale x
define <vscale x 8 x half> @vfmax_vv_nxv8f16_unmasked(<vscale x 8 x half> %va, <vscale x 8 x half> %vb, i32 zeroext %evl) {
; ZVFH-LABEL: vfmax_vv_nxv8f16_unmasked:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vsetvli zero, a0, e16, m2, ta, ma
+; ZVFH-NEXT: vsetvli a0, zero, e16, m2, ta, ma
; ZVFH-NEXT: vfmax.vv v8, v8, v10
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmax_vv_nxv8f16_unmasked:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: vsetvli zero, a0, e16, m2, ta, ma
+; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m2, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v12, v10
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m4, ta, ma
@@ -439,19 +405,19 @@ define <vscale x 8 x half> @vfmax_vv_nxv8f16_unmasked(<vscale x 8 x half> %va, <
define <vscale x 16 x half> @vfmax_vv_nxv16f16(<vscale x 16 x half> %va, <vscale x 16 x half> %vb, <vscale x 16 x i1> %m, i32 zeroext %evl) {
; ZVFH-LABEL: vfmax_vv_nxv16f16:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vsetvli zero, a0, e16, m4, ta, ma
-; ZVFH-NEXT: vfmax.vv v8, v8, v12, v0.t
+; ZVFH-NEXT: vsetvli a0, zero, e16, m4, ta, ma
+; ZVFH-NEXT: vfmax.vv v8, v8, v12
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmax_vv_nxv16f16:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: vsetvli zero, a0, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12, v0.t
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v8, v0.t
+; 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
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfmax.vv v16, v24, v16, v0.t
+; ZVFHMIN-NEXT: vfmax.vv v16, v24, v16
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v16, v0.t
+; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v16
; ZVFHMIN-NEXT: ret
%v = call <vscale x 16 x half> @llvm.vp.maxnum.nxv16f16(<vscale x 16 x half> %va, <vscale x 16 x half> %vb, <vscale x 16 x i1> %m, i32 %evl)
ret <vscale x 16 x half> %v
@@ -460,13 +426,13 @@ define <vscale x 16 x half> @vfmax_vv_nxv16f16(<vscale x 16 x half> %va, <vscale
define <vscale x 16 x half> @vfmax_vv_nxv16f16_unmasked(<vscale x 16 x half> %va, <vscale x 16 x half> %vb, i32 zeroext %evl) {
; ZVFH-LABEL: vfmax_vv_nxv16f16_unmasked:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vsetvli zero, a0, e16, m4, ta, ma
+; ZVFH-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; ZVFH-NEXT: vfmax.vv v8, v8, v12
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmax_vv_nxv16f16_unmasked:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: vsetvli zero, a0, e16, m4, ta, ma
+; 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
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
@@ -481,51 +447,34 @@ define <vscale x 16 x half> @vfmax_vv_nxv16f16_unmasked(<vscale x 16 x half> %va
define <vscale x 32 x half> @vfmax_vv_nxv32f16(<vscale x 32 x half> %va, <vscale x 32 x half> %vb, <vscale x 32 x i1> %m, i32 zeroext %evl) {
; ZVFH-LABEL: vfmax_vv_nxv32f16:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vsetvli zero, a0, e16, m8, ta, ma
-; ZVFH-NEXT: vfmax.vv v8, v8, v16, v0.t
+; ZVFH-NEXT: vsetvli a0, zero, e16, m8, ta, ma
+; ZVFH-NEXT: vfmax.vv v8, v8, v16
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmax_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: 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 a1, zero, e8, mf2, ta, ma
-; ZVFHMIN-NEXT: vmv1r.v v7, v0
-; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a1, a2, 1
-; ZVFHMIN-NEXT: srli a2, a2, 2
-; ZVFHMIN-NEXT: sub a3, a0, a1
-; ZVFHMIN-NEXT: vslidedown.vx v0, v0, a2
-; ZVFHMIN-NEXT: sltu a2, a0, a3
-; ZVFHMIN-NEXT: addi a2, a2, -1
-; ZVFHMIN-NEXT: and a2, a2, a3
-; ZVFHMIN-NEXT: addi a3, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v16, (a3) # vscale x 64-byte Folded Spill
-; ZVFHMIN-NEXT: vsetvli zero, a2, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20, v0.t
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12, v0.t
+; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v16
+; ZVFHMIN-NEXT: addi a0, sp, 16
+; ZVFHMIN-NEXT: vs8r.v v24, (a0) # vscale x 64-byte Folded Spill
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v8
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
+; ZVFHMIN-NEXT: vl8r.v v8, (a0) # vscale x 64-byte Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfmax.vv v16, v16, v24, v0.t
+; ZVFHMIN-NEXT: vfmax.vv v0, v0, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v12, v16, v0.t
-; ZVFHMIN-NEXT: bltu a0, a1, .LBB22_2
-; ZVFHMIN-NEXT: # %bb.1:
-; ZVFHMIN-NEXT: mv a0, a1
-; ZVFHMIN-NEXT: .LBB22_2:
-; ZVFHMIN-NEXT: vmv1r.v v0, v7
-; ZVFHMIN-NEXT: addi a1, sp, 16
-; ZVFHMIN-NEXT: vl8r.v v24, (a1) # vscale x 64-byte Folded Reload
-; ZVFHMIN-NEXT: vsetvli zero, a0, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v24, v0.t
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v8, v0.t
+; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v0
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfmax.vv v16, v24, v16, v0.t
+; ZVFHMIN-NEXT: vfmax.vv v16, v16, v24
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v16, v0.t
+; ZVFHMIN-NEXT: vfncvt.f.f.w v12, v16
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: add sp, sp, a0
@@ -540,7 +489,7 @@ define <vscale x 32 x half> @vfmax_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
define <vscale x 32 x half> @vfmax_vv_nxv32f16_unmasked(<vscale x 32 x half> %va, <vscale x 32 x half> %vb, i32 zeroext %evl) {
; ZVFH-LABEL: vfmax_vv_nxv32f16_unmasked:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vsetvli zero, a0, e16, m8, ta, ma
+; ZVFH-NEXT: vsetvli a0, zero, e16, m8, ta, ma
; ZVFH-NEXT: vfmax.vv v8, v8, v16
; ZVFH-NEXT: ret
;
@@ -548,43 +497,26 @@ define <vscale x 32 x half> @vfmax_vv_nxv32f16_unmasked(<vscale x 32 x half> %va
; 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: 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: csrr a2, vlenb
-; ZVFHMIN-NEXT: vsetvli a1, zero, e8, m4, ta, ma
-; ZVFHMIN-NEXT: vmset.m v24
-; 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, v24, a2
-; ZVFHMIN-NEXT: sltu a2, a0, a3
-; ZVFHMIN-NEXT: addi a2, a2, -1
-; ZVFHMIN-NEXT: and a2, a2, a3
-; ZVFHMIN-NEXT: addi a3, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v16, (a3) # vscale x 64-byte Folded Spill
-; ZVFHMIN-NEXT: vsetvli zero, a2, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20, v0.t
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12, v0.t
+; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v16
+; ZVFHMIN-NEXT: addi a0, sp, 16
+; ZVFHMIN-NEXT: vs8r.v v24, (a0) # vscale x 64-byte Folded Spill
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v8
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
+; ZVFHMIN-NEXT: vl8r.v v8, (a0) # vscale x 64-byte Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfmax.vv v16, v16, v24, v0.t
+; ZVFHMIN-NEXT: vfmax.vv v0, v0, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v12, v16, v0.t
-; ZVFHMIN-NEXT: bltu a0, a1, .LBB23_2
-; ZVFHMIN-NEXT: # %bb.1:
-; ZVFHMIN-NEXT: mv a0, a1
-; ZVFHMIN-NEXT: .LBB23_2:
-; ZVFHMIN-NEXT: addi a1, sp, 16
-; ZVFHMIN-NEXT: vl8r.v v24, (a1) # vscale x 64-byte Folded Reload
-; ZVFHMIN-NEXT: vsetvli zero, a0, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v24
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v8
+; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v0
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfmax.vv v16, v24, v16
+; ZVFHMIN-NEXT: vfmax.vv v16, v16, v24
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v16
+; ZVFHMIN-NEXT: vfncvt.f.f.w v12, v16
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: add sp, sp, a0
@@ -599,8 +531,8 @@ define <vscale x 32 x half> @vfmax_vv_nxv32f16_unmasked(<vscale x 32 x half> %va
define <vscale x 1 x float> @vfmax_vv_nxv1f32(<vscale x 1 x float> %va, <vscale x 1 x float> %vb, <vscale x 1 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_nxv1f32:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, ma
-; CHECK-NEXT: vfmax.vv v8, v8, v9, v0.t
+; CHECK-NEXT: vsetvli a0, zero, e32, mf2, ta, ma
+; CHECK-NEXT: vfmax.vv v8, v8, v9
; CHECK-NEXT: ret
%v = call <vscale x 1 x float> @llvm.vp.maxnum.nxv1f32(<vscale x 1 x float> %va, <vscale x 1 x float> %vb, <vscale x 1 x i1> %m, i32 %evl)
ret <vscale x 1 x float> %v
@@ -609,7 +541,7 @@ define <vscale x 1 x float> @vfmax_vv_nxv1f32(<vscale x 1 x float> %va, <vscale
define <vscale x 1 x float> @vfmax_vv_nxv1f32_unmasked(<vscale x 1 x float> %va, <vscale x 1 x float> %vb, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_nxv1f32_unmasked:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, ma
+; CHECK-NEXT: vsetvli a0, zero, e32, mf2, ta, ma
; CHECK-NEXT: vfmax.vv v8, v8, v9
; CHECK-NEXT: ret
%v = call <vscale x 1 x float> @llvm.vp.maxnum.nxv1f32(<vscale x 1 x float> %va, <vscale x 1 x float> %vb, <vscale x 1 x i1> splat (i1 true), i32 %evl)
@@ -619,8 +551,8 @@ define <vscale x 1 x float> @vfmax_vv_nxv1f32_unmasked(<vscale x 1 x float> %va,
define <vscale x 2 x float> @vfmax_vv_nxv2f32(<vscale x 2 x float> %va, <vscale x 2 x float> %vb, <vscale x 2 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_nxv2f32:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, ma
-; CHECK-NEXT: vfmax.vv v8, v8, v9, v0.t
+; CHECK-NEXT: vsetvli a0, zero, e32, m1, ta, ma
+; CHECK-NEXT: vfmax.vv v8, v8, v9
; CHECK-NEXT: ret
%v = call <vscale x 2 x float> @llvm.vp.maxnum.nxv2f32(<vscale x 2 x float> %va, <vscale x 2 x float> %vb, <vscale x 2 x i1> %m, i32 %evl)
ret <vscale x 2 x float> %v
@@ -629,7 +561,7 @@ define <vscale x 2 x float> @vfmax_vv_nxv2f32(<vscale x 2 x float> %va, <vscale
define <vscale x 2 x float> @vfmax_vv_nxv2f32_unmasked(<vscale x 2 x float> %va, <vscale x 2 x float> %vb, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_nxv2f32_unmasked:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, ma
+; CHECK-NEXT: vsetvli a0, zero, e32, m1, ta, ma
; CHECK-NEXT: vfmax.vv v8, v8, v9
; CHECK-NEXT: ret
%v = call <vscale x 2 x float> @llvm.vp.maxnum.nxv2f32(<vscale x 2 x float> %va, <vscale x 2 x float> %vb, <vscale x 2 x i1> splat (i1 true), i32 %evl)
@@ -639,8 +571,8 @@ define <vscale x 2 x float> @vfmax_vv_nxv2f32_unmasked(<vscale x 2 x float> %va,
define <vscale x 4 x float> @vfmax_vv_nxv4f32(<vscale x 4 x float> %va, <vscale x 4 x float> %vb, <vscale x 4 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_nxv4f32:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, ma
-; CHECK-NEXT: vfmax.vv v8, v8, v10, v0.t
+; CHECK-NEXT: vsetvli a0, zero, e32, m2, ta, ma
+; CHECK-NEXT: vfmax.vv v8, v8, v10
; CHECK-NEXT: ret
%v = call <vscale x 4 x float> @llvm.vp.maxnum.nxv4f32(<vscale x 4 x float> %va, <vscale x 4 x float> %vb, <vscale x 4 x i1> %m, i32 %evl)
ret <vscale x 4 x float> %v
@@ -649,7 +581,7 @@ define <vscale x 4 x float> @vfmax_vv_nxv4f32(<vscale x 4 x float> %va, <vscale
define <vscale x 4 x float> @vfmax_vv_nxv4f32_unmasked(<vscale x 4 x float> %va, <vscale x 4 x float> %vb, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_nxv4f32_unmasked:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, ma
+; CHECK-NEXT: vsetvli a0, zero, e32, m2, ta, ma
; CHECK-NEXT: vfmax.vv v8, v8, v10
; CHECK-NEXT: ret
%v = call <vscale x 4 x float> @llvm.vp.maxnum.nxv4f32(<vscale x 4 x float> %va, <vscale x 4 x float> %vb, <vscale x 4 x i1> splat (i1 true), i32 %evl)
@@ -659,8 +591,8 @@ define <vscale x 4 x float> @vfmax_vv_nxv4f32_unmasked(<vscale x 4 x float> %va,
define <vscale x 8 x float> @vfmax_vv_nxv8f32(<vscale x 8 x float> %va, <vscale x 8 x float> %vb, <vscale x 8 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_nxv8f32:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, ma
-; CHECK-NEXT: vfmax.vv v8, v8, v12, v0.t
+; CHECK-NEXT: vsetvli a0, zero, e32, m4, ta, ma
+; CHECK-NEXT: vfmax.vv v8, v8, v12
; CHECK-NEXT: ret
%v = call <vscale x 8 x float> @llvm.vp.maxnum.nxv8f32(<vscale x 8 x float> %va, <vscale x 8 x float> %vb, <vscale x 8 x i1> %m, i32 %evl)
ret <vscale x 8 x float> %v
@@ -669,7 +601,7 @@ define <vscale x 8 x float> @vfmax_vv_nxv8f32(<vscale x 8 x float> %va, <vscale
define <vscale x 8 x float> @vfmax_vv_nxv8f32_unmasked(<vscale x 8 x float> %va, <vscale x 8 x float> %vb, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_nxv8f32_unmasked:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, ma
+; CHECK-NEXT: vsetvli a0, zero, e32, m4, ta, ma
; CHECK-NEXT: vfmax.vv v8, v8, v12
; CHECK-NEXT: ret
%v = call <vscale x 8 x float> @llvm.vp.maxnum.nxv8f32(<vscale x 8 x float> %va, <vscale x 8 x float> %vb, <vscale x 8 x i1> splat (i1 true), i32 %evl)
@@ -679,8 +611,8 @@ define <vscale x 8 x float> @vfmax_vv_nxv8f32_unmasked(<vscale x 8 x float> %va,
define <vscale x 16 x float> @vfmax_vv_nxv16f32(<vscale x 16 x float> %va, <vscale x 16 x float> %vb, <vscale x 16 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_nxv16f32:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
-; CHECK-NEXT: vfmax.vv v8, v8, v16, v0.t
+; CHECK-NEXT: vsetvli a0, zero, e32, m8, ta, ma
+; CHECK-NEXT: vfmax.vv v8, v8, v16
; CHECK-NEXT: ret
%v = call <vscale x 16 x float> @llvm.vp.maxnum.nxv16f32(<vscale x 16 x float> %va, <vscale x 16 x float> %vb, <vscale x 16 x i1> %m, i32 %evl)
ret <vscale x 16 x float> %v
@@ -689,7 +621,7 @@ define <vscale x 16 x float> @vfmax_vv_nxv16f32(<vscale x 16 x float> %va, <vsca
define <vscale x 16 x float> @vfmax_vv_nxv16f32_unmasked(<vscale x 16 x float> %va, <vscale x 16 x float> %vb, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_nxv16f32_unmasked:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
+; CHECK-NEXT: vsetvli a0, zero, e32, m8, ta, ma
; CHECK-NEXT: vfmax.vv v8, v8, v16
; CHECK-NEXT: ret
%v = call <vscale x 16 x float> @llvm.vp.maxnum.nxv16f32(<vscale x 16 x float> %va, <vscale x 16 x float> %vb, <vscale x 16 x i1> splat (i1 true), i32 %evl)
@@ -699,8 +631,8 @@ define <vscale x 16 x float> @vfmax_vv_nxv16f32_unmasked(<vscale x 16 x float> %
define <vscale x 1 x double> @vfmax_vv_nxv1f64(<vscale x 1 x double> %va, <vscale x 1 x double> %vb, <vscale x 1 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_nxv1f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, ma
-; CHECK-NEXT: vfmax.vv v8, v8, v9, v0.t
+; CHECK-NEXT: vsetvli a0, zero, e64, m1, ta, ma
+; CHECK-NEXT: vfmax.vv v8, v8, v9
; CHECK-NEXT: ret
%v = call <vscale x 1 x double> @llvm.vp.maxnum.nxv1f64(<vscale x 1 x double> %va, <vscale x 1 x double> %vb, <vscale x 1 x i1> %m, i32 %evl)
ret <vscale x 1 x double> %v
@@ -709,7 +641,7 @@ define <vscale x 1 x double> @vfmax_vv_nxv1f64(<vscale x 1 x double> %va, <vscal
define <vscale x 1 x double> @vfmax_vv_nxv1f64_unmasked(<vscale x 1 x double> %va, <vscale x 1 x double> %vb, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_nxv1f64_unmasked:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, ma
+; CHECK-NEXT: vsetvli a0, zero, e64, m1, ta, ma
; CHECK-NEXT: vfmax.vv v8, v8, v9
; CHECK-NEXT: ret
%v = call <vscale x 1 x double> @llvm.vp.maxnum.nxv1f64(<vscale x 1 x double> %va, <vscale x 1 x double> %vb, <vscale x 1 x i1> splat (i1 true), i32 %evl)
@@ -719,8 +651,8 @@ define <vscale x 1 x double> @vfmax_vv_nxv1f64_unmasked(<vscale x 1 x double> %v
define <vscale x 2 x double> @vfmax_vv_nxv2f64(<vscale x 2 x double> %va, <vscale x 2 x double> %vb, <vscale x 2 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_nxv2f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, ma
-; CHECK-NEXT: vfmax.vv v8, v8, v10, v0.t
+; CHECK-NEXT: vsetvli a0, zero, e64, m2, ta, ma
+; CHECK-NEXT: vfmax.vv v8, v8, v10
; CHECK-NEXT: ret
%v = call <vscale x 2 x double> @llvm.vp.maxnum.nxv2f64(<vscale x 2 x double> %va, <vscale x 2 x double> %vb, <vscale x 2 x i1> %m, i32 %evl)
ret <vscale x 2 x double> %v
@@ -729,7 +661,7 @@ define <vscale x 2 x double> @vfmax_vv_nxv2f64(<vscale x 2 x double> %va, <vscal
define <vscale x 2 x double> @vfmax_vv_nxv2f64_unmasked(<vscale x 2 x double> %va, <vscale x 2 x double> %vb, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_nxv2f64_unmasked:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, ma
+; CHECK-NEXT: vsetvli a0, zero, e64, m2, ta, ma
; CHECK-NEXT: vfmax.vv v8, v8, v10
; CHECK-NEXT: ret
%v = call <vscale x 2 x double> @llvm.vp.maxnum.nxv2f64(<vscale x 2 x double> %va, <vscale x 2 x double> %vb, <vscale x 2 x i1> splat (i1 true), i32 %evl)
@@ -739,8 +671,8 @@ define <vscale x 2 x double> @vfmax_vv_nxv2f64_unmasked(<vscale x 2 x double> %v
define <vscale x 4 x double> @vfmax_vv_nxv4f64(<vscale x 4 x double> %va, <vscale x 4 x double> %vb, <vscale x 4 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_nxv4f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, ma
-; CHECK-NEXT: vfmax.vv v8, v8, v12, v0.t
+; CHECK-NEXT: vsetvli a0, zero, e64, m4, ta, ma
+; CHECK-NEXT: vfmax.vv v8, v8, v12
; CHECK-NEXT: ret
%v = call <vscale x 4 x double> @llvm.vp.maxnum.nxv4f64(<vscale x 4 x double> %va, <vscale x 4 x double> %vb, <vscale x 4 x i1> %m, i32 %evl)
ret <vscale x 4 x double> %v
@@ -749,7 +681,7 @@ define <vscale x 4 x double> @vfmax_vv_nxv4f64(<vscale x 4 x double> %va, <vscal
define <vscale x 4 x double> @vfmax_vv_nxv4f64_unmasked(<vscale x 4 x double> %va, <vscale x 4 x double> %vb, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_nxv4f64_unmasked:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, ma
+; CHECK-NEXT: vsetvli a0, zero, e64, m4, ta, ma
; CHECK-NEXT: vfmax.vv v8, v8, v12
; CHECK-NEXT: ret
%v = call <vscale x 4 x double> @llvm.vp.maxnum.nxv4f64(<vscale x 4 x double> %va, <vscale x 4 x double> %vb, <vscale x 4 x i1> splat (i1 true), i32 %evl)
@@ -759,8 +691,8 @@ define <vscale x 4 x double> @vfmax_vv_nxv4f64_unmasked(<vscale x 4 x double> %v
define <vscale x 8 x double> @vfmax_vv_nxv8f64(<vscale x 8 x double> %va, <vscale x 8 x double> %vb, <vscale x 8 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_nxv8f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: vfmax.vv v8, v8, v16, v0.t
+; CHECK-NEXT: vsetvli a0, zero, e64, m8, ta, ma
+; CHECK-NEXT: vfmax.vv v8, v8, v16
; CHECK-NEXT: ret
%v = call <vscale x 8 x double> @llvm.vp.maxnum.nxv8f64(<vscale x 8 x double> %va, <vscale x 8 x double> %vb, <vscale x 8 x i1> %m, i32 %evl)
ret <vscale x 8 x double> %v
@@ -769,7 +701,7 @@ define <vscale x 8 x double> @vfmax_vv_nxv8f64(<vscale x 8 x double> %va, <vscal
define <vscale x 8 x double> @vfmax_vv_nxv8f64_unmasked(<vscale x 8 x double> %va, <vscale x 8 x double> %vb, i32 zeroext %evl) {
; CHECK-LABEL: vfmax_vv_nxv8f64_unmasked:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT: vsetvli a0, zero, e64, m8, ta, ma
; CHECK-NEXT: vfmax.vv v8, v8, v16
; CHECK-NEXT: ret
%v = call <vscale x 8 x double> @llvm.vp.maxnum.nxv8f64(<vscale x 8 x double> %va, <vscale x 8 x double> %vb, <vscale x 8 x i1> splat (i1 true), i32 %evl)
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfmin-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vfmin-vp.ll
index 5c5542619b6ef..2940617d4b17c 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfmin-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfmin-vp.ll
@@ -15,13 +15,13 @@
define <vscale x 1 x bfloat> @vfmin_vv_nxv1bf16(<vscale x 1 x bfloat> %va, <vscale x 1 x bfloat> %vb, <vscale x 1 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_nxv1bf16:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, ma
-; CHECK-NEXT: vfwcvtbf16.f.f.v v10, v9, v0.t
-; CHECK-NEXT: vfwcvtbf16.f.f.v v9, v8, v0.t
+; CHECK-NEXT: vsetvli a0, zero, e16, mf4, ta, ma
+; CHECK-NEXT: vfwcvtbf16.f.f.v v10, v9
+; CHECK-NEXT: vfwcvtbf16.f.f.v v9, v8
; CHECK-NEXT: vsetvli zero, zero, e32, mf2, ta, ma
-; CHECK-NEXT: vfmin.vv v9, v9, v10, v0.t
+; CHECK-NEXT: vfmin.vv v9, v9, v10
; CHECK-NEXT: vsetvli zero, zero, e16, mf4, ta, ma
-; CHECK-NEXT: vfncvtbf16.f.f.w v8, v9, v0.t
+; CHECK-NEXT: vfncvtbf16.f.f.w v8, v9
; CHECK-NEXT: ret
%v = call <vscale x 1 x bfloat> @llvm.vp.minnum.nxv1bf16(<vscale x 1 x bfloat> %va, <vscale x 1 x bfloat> %vb, <vscale x 1 x i1> %m, i32 %evl)
ret <vscale x 1 x bfloat> %v
@@ -30,7 +30,7 @@ define <vscale x 1 x bfloat> @vfmin_vv_nxv1bf16(<vscale x 1 x bfloat> %va, <vsca
define <vscale x 1 x bfloat> @vfmin_vv_nxv1bf16_unmasked(<vscale x 1 x bfloat> %va, <vscale x 1 x bfloat> %vb, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_nxv1bf16_unmasked:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, ma
+; CHECK-NEXT: vsetvli a0, zero, e16, mf4, ta, ma
; CHECK-NEXT: vfwcvtbf16.f.f.v v10, v9
; CHECK-NEXT: vfwcvtbf16.f.f.v v9, v8
; CHECK-NEXT: vsetvli zero, zero, e32, mf2, ta, ma
@@ -45,13 +45,13 @@ define <vscale x 1 x bfloat> @vfmin_vv_nxv1bf16_unmasked(<vscale x 1 x bfloat> %
define <vscale x 2 x bfloat> @vfmin_vv_nxv2bf16(<vscale x 2 x bfloat> %va, <vscale x 2 x bfloat> %vb, <vscale x 2 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_nxv2bf16:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, ma
-; CHECK-NEXT: vfwcvtbf16.f.f.v v10, v9, v0.t
-; CHECK-NEXT: vfwcvtbf16.f.f.v v9, v8, v0.t
+; CHECK-NEXT: vsetvli a0, zero, e16, mf2, ta, ma
+; CHECK-NEXT: vfwcvtbf16.f.f.v v10, v9
+; CHECK-NEXT: vfwcvtbf16.f.f.v v9, v8
; CHECK-NEXT: vsetvli zero, zero, e32, m1, ta, ma
-; CHECK-NEXT: vfmin.vv v9, v9, v10, v0.t
+; CHECK-NEXT: vfmin.vv v9, v9, v10
; CHECK-NEXT: vsetvli zero, zero, e16, mf2, ta, ma
-; CHECK-NEXT: vfncvtbf16.f.f.w v8, v9, v0.t
+; CHECK-NEXT: vfncvtbf16.f.f.w v8, v9
; CHECK-NEXT: ret
%v = call <vscale x 2 x bfloat> @llvm.vp.minnum.nxv2bf16(<vscale x 2 x bfloat> %va, <vscale x 2 x bfloat> %vb, <vscale x 2 x i1> %m, i32 %evl)
ret <vscale x 2 x bfloat> %v
@@ -60,7 +60,7 @@ define <vscale x 2 x bfloat> @vfmin_vv_nxv2bf16(<vscale x 2 x bfloat> %va, <vsca
define <vscale x 2 x bfloat> @vfmin_vv_nxv2bf16_unmasked(<vscale x 2 x bfloat> %va, <vscale x 2 x bfloat> %vb, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_nxv2bf16_unmasked:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, ma
+; CHECK-NEXT: vsetvli a0, zero, e16, mf2, ta, ma
; CHECK-NEXT: vfwcvtbf16.f.f.v v10, v9
; CHECK-NEXT: vfwcvtbf16.f.f.v v9, v8
; CHECK-NEXT: vsetvli zero, zero, e32, m1, ta, ma
@@ -75,13 +75,13 @@ define <vscale x 2 x bfloat> @vfmin_vv_nxv2bf16_unmasked(<vscale x 2 x bfloat> %
define <vscale x 4 x bfloat> @vfmin_vv_nxv4bf16(<vscale x 4 x bfloat> %va, <vscale x 4 x bfloat> %vb, <vscale x 4 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_nxv4bf16:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, ma
-; CHECK-NEXT: vfwcvtbf16.f.f.v v10, v9, v0.t
-; CHECK-NEXT: vfwcvtbf16.f.f.v v12, v8, v0.t
+; CHECK-NEXT: vsetvli a0, zero, e16, m1, ta, ma
+; CHECK-NEXT: vfwcvtbf16.f.f.v v10, v9
+; CHECK-NEXT: vfwcvtbf16.f.f.v v12, v8
; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, ma
-; CHECK-NEXT: vfmin.vv v10, v12, v10, v0.t
+; CHECK-NEXT: vfmin.vv v10, v12, v10
; CHECK-NEXT: vsetvli zero, zero, e16, m1, ta, ma
-; CHECK-NEXT: vfncvtbf16.f.f.w v8, v10, v0.t
+; CHECK-NEXT: vfncvtbf16.f.f.w v8, v10
; CHECK-NEXT: ret
%v = call <vscale x 4 x bfloat> @llvm.vp.minnum.nxv4bf16(<vscale x 4 x bfloat> %va, <vscale x 4 x bfloat> %vb, <vscale x 4 x i1> %m, i32 %evl)
ret <vscale x 4 x bfloat> %v
@@ -90,7 +90,7 @@ define <vscale x 4 x bfloat> @vfmin_vv_nxv4bf16(<vscale x 4 x bfloat> %va, <vsca
define <vscale x 4 x bfloat> @vfmin_vv_nxv4bf16_unmasked(<vscale x 4 x bfloat> %va, <vscale x 4 x bfloat> %vb, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_nxv4bf16_unmasked:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, ma
+; CHECK-NEXT: vsetvli a0, zero, e16, m1, ta, ma
; CHECK-NEXT: vfwcvtbf16.f.f.v v10, v9
; CHECK-NEXT: vfwcvtbf16.f.f.v v12, v8
; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, ma
@@ -105,13 +105,13 @@ define <vscale x 4 x bfloat> @vfmin_vv_nxv4bf16_unmasked(<vscale x 4 x bfloat> %
define <vscale x 8 x bfloat> @vfmin_vv_nxv8bf16(<vscale x 8 x bfloat> %va, <vscale x 8 x bfloat> %vb, <vscale x 8 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_nxv8bf16:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, ma
-; CHECK-NEXT: vfwcvtbf16.f.f.v v12, v10, v0.t
-; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v8, v0.t
+; CHECK-NEXT: vsetvli a0, zero, e16, m2, ta, ma
+; CHECK-NEXT: vfwcvtbf16.f.f.v v12, v10
+; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v8
; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, ma
-; CHECK-NEXT: vfmin.vv v12, v16, v12, v0.t
+; CHECK-NEXT: vfmin.vv v12, v16, v12
; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, ma
-; CHECK-NEXT: vfncvtbf16.f.f.w v8, v12, v0.t
+; CHECK-NEXT: vfncvtbf16.f.f.w v8, v12
; CHECK-NEXT: ret
%v = call <vscale x 8 x bfloat> @llvm.vp.minnum.nxv8bf16(<vscale x 8 x bfloat> %va, <vscale x 8 x bfloat> %vb, <vscale x 8 x i1> %m, i32 %evl)
ret <vscale x 8 x bfloat> %v
@@ -120,7 +120,7 @@ define <vscale x 8 x bfloat> @vfmin_vv_nxv8bf16(<vscale x 8 x bfloat> %va, <vsca
define <vscale x 8 x bfloat> @vfmin_vv_nxv8bf16_unmasked(<vscale x 8 x bfloat> %va, <vscale x 8 x bfloat> %vb, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_nxv8bf16_unmasked:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, ma
+; CHECK-NEXT: vsetvli a0, zero, e16, m2, ta, ma
; CHECK-NEXT: vfwcvtbf16.f.f.v v12, v10
; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v8
; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, ma
@@ -135,13 +135,13 @@ define <vscale x 8 x bfloat> @vfmin_vv_nxv8bf16_unmasked(<vscale x 8 x bfloat> %
define <vscale x 16 x bfloat> @vfmin_vv_nxv16bf16(<vscale x 16 x bfloat> %va, <vscale x 16 x bfloat> %vb, <vscale x 16 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_nxv16bf16:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, ma
-; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v12, v0.t
-; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v8, v0.t
+; 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
; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; CHECK-NEXT: vfmin.vv v16, v24, v16, v0.t
+; CHECK-NEXT: vfmin.vv v16, v24, v16
; CHECK-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; CHECK-NEXT: vfncvtbf16.f.f.w v8, v16, v0.t
+; CHECK-NEXT: vfncvtbf16.f.f.w v8, v16
; CHECK-NEXT: ret
%v = call <vscale x 16 x bfloat> @llvm.vp.minnum.nxv16bf16(<vscale x 16 x bfloat> %va, <vscale x 16 x bfloat> %vb, <vscale x 16 x i1> %m, i32 %evl)
ret <vscale x 16 x bfloat> %v
@@ -150,7 +150,7 @@ 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: vsetvli zero, a0, e16, m4, ta, ma
+; 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
; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, ma
@@ -167,43 +167,26 @@ define <vscale x 32 x bfloat> @vfmin_vv_nxv32bf16(<vscale x 32 x bfloat> %va, <v
; 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: 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 a1, zero, e8, mf2, ta, ma
-; CHECK-NEXT: vmv1r.v v7, v0
-; CHECK-NEXT: csrr a2, vlenb
-; CHECK-NEXT: slli a1, a2, 1
-; CHECK-NEXT: srli a2, a2, 2
-; CHECK-NEXT: sub a3, a0, a1
-; CHECK-NEXT: vslidedown.vx v0, v0, a2
-; 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: vs8r.v v16, (a3) # vscale x 64-byte Folded Spill
-; CHECK-NEXT: vsetvli zero, a2, e16, m4, ta, ma
-; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v20, v0.t
-; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v12, v0.t
+; CHECK-NEXT: vsetvli a0, zero, e16, m4, ta, ma
+; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v16
+; CHECK-NEXT: addi a0, sp, 16
+; CHECK-NEXT: vs8r.v v24, (a0) # vscale x 64-byte Folded Spill
+; CHECK-NEXT: vfwcvtbf16.f.f.v v0, v8
+; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v20
+; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v12
+; CHECK-NEXT: vl8r.v v8, (a0) # vscale x 64-byte Folded Reload
; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; CHECK-NEXT: vfmin.vv v16, v16, v24, v0.t
+; CHECK-NEXT: vfmin.vv v0, v0, v8
; CHECK-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; CHECK-NEXT: vfncvtbf16.f.f.w v12, v16, v0.t
-; CHECK-NEXT: bltu a0, a1, .LBB10_2
-; CHECK-NEXT: # %bb.1:
-; CHECK-NEXT: mv a0, a1
-; CHECK-NEXT: .LBB10_2:
-; CHECK-NEXT: vmv1r.v v0, v7
-; CHECK-NEXT: addi a1, sp, 16
-; CHECK-NEXT: vl8r.v v24, (a1) # vscale x 64-byte Folded Reload
-; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, ma
-; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v24, v0.t
-; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v8, v0.t
+; CHECK-NEXT: vfncvtbf16.f.f.w v8, v0
; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; CHECK-NEXT: vfmin.vv v16, v24, v16, v0.t
+; CHECK-NEXT: vfmin.vv v16, v16, v24
; CHECK-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; CHECK-NEXT: vfncvtbf16.f.f.w v8, v16, v0.t
+; CHECK-NEXT: vfncvtbf16.f.f.w v12, v16
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add sp, sp, a0
@@ -220,43 +203,26 @@ define <vscale x 32 x bfloat> @vfmin_vv_nxv32bf16_unmasked(<vscale x 32 x bfloat
; 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: 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: csrr a2, vlenb
-; CHECK-NEXT: vsetvli a1, zero, e8, m4, ta, ma
-; CHECK-NEXT: vmset.m v24
-; 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, v24, a2
-; 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: vs8r.v v16, (a3) # vscale x 64-byte Folded Spill
-; CHECK-NEXT: vsetvli zero, a2, e16, m4, ta, ma
-; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v20, v0.t
-; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v12, v0.t
+; CHECK-NEXT: vsetvli a0, zero, e16, m4, ta, ma
+; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v16
+; CHECK-NEXT: addi a0, sp, 16
+; CHECK-NEXT: vs8r.v v24, (a0) # vscale x 64-byte Folded Spill
+; CHECK-NEXT: vfwcvtbf16.f.f.v v0, v8
+; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v20
+; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v12
+; CHECK-NEXT: vl8r.v v8, (a0) # vscale x 64-byte Folded Reload
; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; CHECK-NEXT: vfmin.vv v16, v16, v24, v0.t
+; CHECK-NEXT: vfmin.vv v0, v0, v8
; CHECK-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; CHECK-NEXT: vfncvtbf16.f.f.w v12, v16, v0.t
-; CHECK-NEXT: bltu a0, a1, .LBB11_2
-; CHECK-NEXT: # %bb.1:
-; CHECK-NEXT: mv a0, a1
-; CHECK-NEXT: .LBB11_2:
-; CHECK-NEXT: addi a1, sp, 16
-; CHECK-NEXT: vl8r.v v24, (a1) # vscale x 64-byte Folded Reload
-; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, ma
-; CHECK-NEXT: vfwcvtbf16.f.f.v v16, v24
-; CHECK-NEXT: vfwcvtbf16.f.f.v v24, v8
+; CHECK-NEXT: vfncvtbf16.f.f.w v8, v0
; CHECK-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; CHECK-NEXT: vfmin.vv v16, v24, v16
+; CHECK-NEXT: vfmin.vv v16, v16, v24
; CHECK-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; CHECK-NEXT: vfncvtbf16.f.f.w v8, v16
+; CHECK-NEXT: vfncvtbf16.f.f.w v12, v16
; CHECK-NEXT: csrr a0, vlenb
; CHECK-NEXT: slli a0, a0, 3
; CHECK-NEXT: add sp, sp, a0
@@ -271,19 +237,19 @@ define <vscale x 32 x bfloat> @vfmin_vv_nxv32bf16_unmasked(<vscale x 32 x bfloat
define <vscale x 1 x half> @vfmin_vv_nxv1f16(<vscale x 1 x half> %va, <vscale x 1 x half> %vb, <vscale x 1 x i1> %m, i32 zeroext %evl) {
; ZVFH-LABEL: vfmin_vv_nxv1f16:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vsetvli zero, a0, e16, mf4, ta, ma
-; ZVFH-NEXT: vfmin.vv v8, v8, v9, v0.t
+; ZVFH-NEXT: vsetvli a0, zero, e16, mf4, ta, ma
+; ZVFH-NEXT: vfmin.vv v8, v8, v9
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmin_vv_nxv1f16:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: vsetvli zero, a0, e16, mf4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v10, v9, v0.t
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v9, v8, v0.t
+; ZVFHMIN-NEXT: vsetvli a0, zero, e16, mf4, ta, ma
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v10, v9
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v9, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, mf2, ta, ma
-; ZVFHMIN-NEXT: vfmin.vv v9, v9, v10, v0.t
+; ZVFHMIN-NEXT: vfmin.vv v9, v9, v10
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, mf4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v9, v0.t
+; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v9
; ZVFHMIN-NEXT: ret
%v = call <vscale x 1 x half> @llvm.vp.minnum.nxv1f16(<vscale x 1 x half> %va, <vscale x 1 x half> %vb, <vscale x 1 x i1> %m, i32 %evl)
ret <vscale x 1 x half> %v
@@ -292,13 +258,13 @@ define <vscale x 1 x half> @vfmin_vv_nxv1f16(<vscale x 1 x half> %va, <vscale x
define <vscale x 1 x half> @vfmin_vv_nxv1f16_unmasked(<vscale x 1 x half> %va, <vscale x 1 x half> %vb, i32 zeroext %evl) {
; ZVFH-LABEL: vfmin_vv_nxv1f16_unmasked:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vsetvli zero, a0, e16, mf4, ta, ma
+; ZVFH-NEXT: vsetvli a0, zero, e16, mf4, ta, ma
; ZVFH-NEXT: vfmin.vv v8, v8, v9
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmin_vv_nxv1f16_unmasked:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: vsetvli zero, a0, e16, mf4, ta, ma
+; ZVFHMIN-NEXT: vsetvli a0, zero, e16, mf4, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v10, v9
; ZVFHMIN-NEXT: vfwcvt.f.f.v v9, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, mf2, ta, ma
@@ -313,19 +279,19 @@ define <vscale x 1 x half> @vfmin_vv_nxv1f16_unmasked(<vscale x 1 x half> %va, <
define <vscale x 2 x half> @vfmin_vv_nxv2f16(<vscale x 2 x half> %va, <vscale x 2 x half> %vb, <vscale x 2 x i1> %m, i32 zeroext %evl) {
; ZVFH-LABEL: vfmin_vv_nxv2f16:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vsetvli zero, a0, e16, mf2, ta, ma
-; ZVFH-NEXT: vfmin.vv v8, v8, v9, v0.t
+; ZVFH-NEXT: vsetvli a0, zero, e16, mf2, ta, ma
+; ZVFH-NEXT: vfmin.vv v8, v8, v9
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmin_vv_nxv2f16:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: vsetvli zero, a0, e16, mf2, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v10, v9, v0.t
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v9, v8, v0.t
+; ZVFHMIN-NEXT: vsetvli a0, zero, e16, mf2, ta, ma
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v10, v9
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v9, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m1, ta, ma
-; ZVFHMIN-NEXT: vfmin.vv v9, v9, v10, v0.t
+; ZVFHMIN-NEXT: vfmin.vv v9, v9, v10
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, mf2, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v9, v0.t
+; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v9
; ZVFHMIN-NEXT: ret
%v = call <vscale x 2 x half> @llvm.vp.minnum.nxv2f16(<vscale x 2 x half> %va, <vscale x 2 x half> %vb, <vscale x 2 x i1> %m, i32 %evl)
ret <vscale x 2 x half> %v
@@ -334,13 +300,13 @@ define <vscale x 2 x half> @vfmin_vv_nxv2f16(<vscale x 2 x half> %va, <vscale x
define <vscale x 2 x half> @vfmin_vv_nxv2f16_unmasked(<vscale x 2 x half> %va, <vscale x 2 x half> %vb, i32 zeroext %evl) {
; ZVFH-LABEL: vfmin_vv_nxv2f16_unmasked:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vsetvli zero, a0, e16, mf2, ta, ma
+; ZVFH-NEXT: vsetvli a0, zero, e16, mf2, ta, ma
; ZVFH-NEXT: vfmin.vv v8, v8, v9
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmin_vv_nxv2f16_unmasked:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: vsetvli zero, a0, e16, mf2, ta, ma
+; ZVFHMIN-NEXT: vsetvli a0, zero, e16, mf2, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v10, v9
; ZVFHMIN-NEXT: vfwcvt.f.f.v v9, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m1, ta, ma
@@ -355,19 +321,19 @@ define <vscale x 2 x half> @vfmin_vv_nxv2f16_unmasked(<vscale x 2 x half> %va, <
define <vscale x 4 x half> @vfmin_vv_nxv4f16(<vscale x 4 x half> %va, <vscale x 4 x half> %vb, <vscale x 4 x i1> %m, i32 zeroext %evl) {
; ZVFH-LABEL: vfmin_vv_nxv4f16:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vsetvli zero, a0, e16, m1, ta, ma
-; ZVFH-NEXT: vfmin.vv v8, v8, v9, v0.t
+; ZVFH-NEXT: vsetvli a0, zero, e16, m1, ta, ma
+; ZVFH-NEXT: vfmin.vv v8, v8, v9
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmin_vv_nxv4f16:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: vsetvli zero, a0, e16, m1, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v10, v9, v0.t
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v12, v8, v0.t
+; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m1, ta, ma
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v10, v9
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v12, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m2, ta, ma
-; ZVFHMIN-NEXT: vfmin.vv v10, v12, v10, v0.t
+; ZVFHMIN-NEXT: vfmin.vv v10, v12, v10
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m1, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v10, v0.t
+; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v10
; ZVFHMIN-NEXT: ret
%v = call <vscale x 4 x half> @llvm.vp.minnum.nxv4f16(<vscale x 4 x half> %va, <vscale x 4 x half> %vb, <vscale x 4 x i1> %m, i32 %evl)
ret <vscale x 4 x half> %v
@@ -376,13 +342,13 @@ define <vscale x 4 x half> @vfmin_vv_nxv4f16(<vscale x 4 x half> %va, <vscale x
define <vscale x 4 x half> @vfmin_vv_nxv4f16_unmasked(<vscale x 4 x half> %va, <vscale x 4 x half> %vb, i32 zeroext %evl) {
; ZVFH-LABEL: vfmin_vv_nxv4f16_unmasked:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vsetvli zero, a0, e16, m1, ta, ma
+; ZVFH-NEXT: vsetvli a0, zero, e16, m1, ta, ma
; ZVFH-NEXT: vfmin.vv v8, v8, v9
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmin_vv_nxv4f16_unmasked:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: vsetvli zero, a0, e16, m1, ta, ma
+; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m1, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v10, v9
; ZVFHMIN-NEXT: vfwcvt.f.f.v v12, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m2, ta, ma
@@ -397,19 +363,19 @@ define <vscale x 4 x half> @vfmin_vv_nxv4f16_unmasked(<vscale x 4 x half> %va, <
define <vscale x 8 x half> @vfmin_vv_nxv8f16(<vscale x 8 x half> %va, <vscale x 8 x half> %vb, <vscale x 8 x i1> %m, i32 zeroext %evl) {
; ZVFH-LABEL: vfmin_vv_nxv8f16:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vsetvli zero, a0, e16, m2, ta, ma
-; ZVFH-NEXT: vfmin.vv v8, v8, v10, v0.t
+; ZVFH-NEXT: vsetvli a0, zero, e16, m2, ta, ma
+; ZVFH-NEXT: vfmin.vv v8, v8, v10
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmin_vv_nxv8f16:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: vsetvli zero, a0, e16, m2, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v12, v10, v0.t
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8, v0.t
+; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m2, ta, ma
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v12, v10
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m4, ta, ma
-; ZVFHMIN-NEXT: vfmin.vv v12, v16, v12, v0.t
+; ZVFHMIN-NEXT: vfmin.vv v12, v16, v12
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m2, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v12, v0.t
+; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v12
; ZVFHMIN-NEXT: ret
%v = call <vscale x 8 x half> @llvm.vp.minnum.nxv8f16(<vscale x 8 x half> %va, <vscale x 8 x half> %vb, <vscale x 8 x i1> %m, i32 %evl)
ret <vscale x 8 x half> %v
@@ -418,13 +384,13 @@ define <vscale x 8 x half> @vfmin_vv_nxv8f16(<vscale x 8 x half> %va, <vscale x
define <vscale x 8 x half> @vfmin_vv_nxv8f16_unmasked(<vscale x 8 x half> %va, <vscale x 8 x half> %vb, i32 zeroext %evl) {
; ZVFH-LABEL: vfmin_vv_nxv8f16_unmasked:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vsetvli zero, a0, e16, m2, ta, ma
+; ZVFH-NEXT: vsetvli a0, zero, e16, m2, ta, ma
; ZVFH-NEXT: vfmin.vv v8, v8, v10
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmin_vv_nxv8f16_unmasked:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: vsetvli zero, a0, e16, m2, ta, ma
+; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m2, ta, ma
; ZVFHMIN-NEXT: vfwcvt.f.f.v v12, v10
; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m4, ta, ma
@@ -439,19 +405,19 @@ define <vscale x 8 x half> @vfmin_vv_nxv8f16_unmasked(<vscale x 8 x half> %va, <
define <vscale x 16 x half> @vfmin_vv_nxv16f16(<vscale x 16 x half> %va, <vscale x 16 x half> %vb, <vscale x 16 x i1> %m, i32 zeroext %evl) {
; ZVFH-LABEL: vfmin_vv_nxv16f16:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vsetvli zero, a0, e16, m4, ta, ma
-; ZVFH-NEXT: vfmin.vv v8, v8, v12, v0.t
+; ZVFH-NEXT: vsetvli a0, zero, e16, m4, ta, ma
+; ZVFH-NEXT: vfmin.vv v8, v8, v12
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmin_vv_nxv16f16:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: vsetvli zero, a0, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12, v0.t
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v8, v0.t
+; 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
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfmin.vv v16, v24, v16, v0.t
+; ZVFHMIN-NEXT: vfmin.vv v16, v24, v16
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v16, v0.t
+; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v16
; ZVFHMIN-NEXT: ret
%v = call <vscale x 16 x half> @llvm.vp.minnum.nxv16f16(<vscale x 16 x half> %va, <vscale x 16 x half> %vb, <vscale x 16 x i1> %m, i32 %evl)
ret <vscale x 16 x half> %v
@@ -460,13 +426,13 @@ define <vscale x 16 x half> @vfmin_vv_nxv16f16(<vscale x 16 x half> %va, <vscale
define <vscale x 16 x half> @vfmin_vv_nxv16f16_unmasked(<vscale x 16 x half> %va, <vscale x 16 x half> %vb, i32 zeroext %evl) {
; ZVFH-LABEL: vfmin_vv_nxv16f16_unmasked:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vsetvli zero, a0, e16, m4, ta, ma
+; ZVFH-NEXT: vsetvli a0, zero, e16, m4, ta, ma
; ZVFH-NEXT: vfmin.vv v8, v8, v12
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmin_vv_nxv16f16_unmasked:
; ZVFHMIN: # %bb.0:
-; ZVFHMIN-NEXT: vsetvli zero, a0, e16, m4, ta, ma
+; 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
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
@@ -481,51 +447,34 @@ define <vscale x 16 x half> @vfmin_vv_nxv16f16_unmasked(<vscale x 16 x half> %va
define <vscale x 32 x half> @vfmin_vv_nxv32f16(<vscale x 32 x half> %va, <vscale x 32 x half> %vb, <vscale x 32 x i1> %m, i32 zeroext %evl) {
; ZVFH-LABEL: vfmin_vv_nxv32f16:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vsetvli zero, a0, e16, m8, ta, ma
-; ZVFH-NEXT: vfmin.vv v8, v8, v16, v0.t
+; ZVFH-NEXT: vsetvli a0, zero, e16, m8, ta, ma
+; ZVFH-NEXT: vfmin.vv v8, v8, v16
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vfmin_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: 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 a1, zero, e8, mf2, ta, ma
-; ZVFHMIN-NEXT: vmv1r.v v7, v0
-; ZVFHMIN-NEXT: csrr a2, vlenb
-; ZVFHMIN-NEXT: slli a1, a2, 1
-; ZVFHMIN-NEXT: srli a2, a2, 2
-; ZVFHMIN-NEXT: sub a3, a0, a1
-; ZVFHMIN-NEXT: vslidedown.vx v0, v0, a2
-; ZVFHMIN-NEXT: sltu a2, a0, a3
-; ZVFHMIN-NEXT: addi a2, a2, -1
-; ZVFHMIN-NEXT: and a2, a2, a3
-; ZVFHMIN-NEXT: addi a3, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v16, (a3) # vscale x 64-byte Folded Spill
-; ZVFHMIN-NEXT: vsetvli zero, a2, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20, v0.t
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12, v0.t
+; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v16
+; ZVFHMIN-NEXT: addi a0, sp, 16
+; ZVFHMIN-NEXT: vs8r.v v24, (a0) # vscale x 64-byte Folded Spill
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v8
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
+; ZVFHMIN-NEXT: vl8r.v v8, (a0) # vscale x 64-byte Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfmin.vv v16, v16, v24, v0.t
+; ZVFHMIN-NEXT: vfmin.vv v0, v0, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v12, v16, v0.t
-; ZVFHMIN-NEXT: bltu a0, a1, .LBB22_2
-; ZVFHMIN-NEXT: # %bb.1:
-; ZVFHMIN-NEXT: mv a0, a1
-; ZVFHMIN-NEXT: .LBB22_2:
-; ZVFHMIN-NEXT: vmv1r.v v0, v7
-; ZVFHMIN-NEXT: addi a1, sp, 16
-; ZVFHMIN-NEXT: vl8r.v v24, (a1) # vscale x 64-byte Folded Reload
-; ZVFHMIN-NEXT: vsetvli zero, a0, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v24, v0.t
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v8, v0.t
+; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v0
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfmin.vv v16, v24, v16, v0.t
+; ZVFHMIN-NEXT: vfmin.vv v16, v16, v24
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v16, v0.t
+; ZVFHMIN-NEXT: vfncvt.f.f.w v12, v16
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: add sp, sp, a0
@@ -540,7 +489,7 @@ define <vscale x 32 x half> @vfmin_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
define <vscale x 32 x half> @vfmin_vv_nxv32f16_unmasked(<vscale x 32 x half> %va, <vscale x 32 x half> %vb, i32 zeroext %evl) {
; ZVFH-LABEL: vfmin_vv_nxv32f16_unmasked:
; ZVFH: # %bb.0:
-; ZVFH-NEXT: vsetvli zero, a0, e16, m8, ta, ma
+; ZVFH-NEXT: vsetvli a0, zero, e16, m8, ta, ma
; ZVFH-NEXT: vfmin.vv v8, v8, v16
; ZVFH-NEXT: ret
;
@@ -548,43 +497,26 @@ define <vscale x 32 x half> @vfmin_vv_nxv32f16_unmasked(<vscale x 32 x half> %va
; 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: 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: csrr a2, vlenb
-; ZVFHMIN-NEXT: vsetvli a1, zero, e8, m4, ta, ma
-; ZVFHMIN-NEXT: vmset.m v24
-; 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, v24, a2
-; ZVFHMIN-NEXT: sltu a2, a0, a3
-; ZVFHMIN-NEXT: addi a2, a2, -1
-; ZVFHMIN-NEXT: and a2, a2, a3
-; ZVFHMIN-NEXT: addi a3, sp, 16
-; ZVFHMIN-NEXT: vs8r.v v16, (a3) # vscale x 64-byte Folded Spill
-; ZVFHMIN-NEXT: vsetvli zero, a2, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20, v0.t
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12, v0.t
+; ZVFHMIN-NEXT: vsetvli a0, zero, e16, m4, ta, ma
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v16
+; ZVFHMIN-NEXT: addi a0, sp, 16
+; ZVFHMIN-NEXT: vs8r.v v24, (a0) # vscale x 64-byte Folded Spill
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v0, v8
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v20
+; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v12
+; ZVFHMIN-NEXT: vl8r.v v8, (a0) # vscale x 64-byte Folded Reload
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfmin.vv v16, v16, v24, v0.t
+; ZVFHMIN-NEXT: vfmin.vv v0, v0, v8
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v12, v16, v0.t
-; ZVFHMIN-NEXT: bltu a0, a1, .LBB23_2
-; ZVFHMIN-NEXT: # %bb.1:
-; ZVFHMIN-NEXT: mv a0, a1
-; ZVFHMIN-NEXT: .LBB23_2:
-; ZVFHMIN-NEXT: addi a1, sp, 16
-; ZVFHMIN-NEXT: vl8r.v v24, (a1) # vscale x 64-byte Folded Reload
-; ZVFHMIN-NEXT: vsetvli zero, a0, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v16, v24
-; ZVFHMIN-NEXT: vfwcvt.f.f.v v24, v8
+; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v0
; ZVFHMIN-NEXT: vsetvli zero, zero, e32, m8, ta, ma
-; ZVFHMIN-NEXT: vfmin.vv v16, v24, v16
+; ZVFHMIN-NEXT: vfmin.vv v16, v16, v24
; ZVFHMIN-NEXT: vsetvli zero, zero, e16, m4, ta, ma
-; ZVFHMIN-NEXT: vfncvt.f.f.w v8, v16
+; ZVFHMIN-NEXT: vfncvt.f.f.w v12, v16
; ZVFHMIN-NEXT: csrr a0, vlenb
; ZVFHMIN-NEXT: slli a0, a0, 3
; ZVFHMIN-NEXT: add sp, sp, a0
@@ -599,8 +531,8 @@ define <vscale x 32 x half> @vfmin_vv_nxv32f16_unmasked(<vscale x 32 x half> %va
define <vscale x 1 x float> @vfmin_vv_nxv1f32(<vscale x 1 x float> %va, <vscale x 1 x float> %vb, <vscale x 1 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_nxv1f32:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, ma
-; CHECK-NEXT: vfmin.vv v8, v8, v9, v0.t
+; CHECK-NEXT: vsetvli a0, zero, e32, mf2, ta, ma
+; CHECK-NEXT: vfmin.vv v8, v8, v9
; CHECK-NEXT: ret
%v = call <vscale x 1 x float> @llvm.vp.minnum.nxv1f32(<vscale x 1 x float> %va, <vscale x 1 x float> %vb, <vscale x 1 x i1> %m, i32 %evl)
ret <vscale x 1 x float> %v
@@ -609,7 +541,7 @@ define <vscale x 1 x float> @vfmin_vv_nxv1f32(<vscale x 1 x float> %va, <vscale
define <vscale x 1 x float> @vfmin_vv_nxv1f32_unmasked(<vscale x 1 x float> %va, <vscale x 1 x float> %vb, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_nxv1f32_unmasked:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, ma
+; CHECK-NEXT: vsetvli a0, zero, e32, mf2, ta, ma
; CHECK-NEXT: vfmin.vv v8, v8, v9
; CHECK-NEXT: ret
%v = call <vscale x 1 x float> @llvm.vp.minnum.nxv1f32(<vscale x 1 x float> %va, <vscale x 1 x float> %vb, <vscale x 1 x i1> splat (i1 true), i32 %evl)
@@ -619,8 +551,8 @@ define <vscale x 1 x float> @vfmin_vv_nxv1f32_unmasked(<vscale x 1 x float> %va,
define <vscale x 2 x float> @vfmin_vv_nxv2f32(<vscale x 2 x float> %va, <vscale x 2 x float> %vb, <vscale x 2 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_nxv2f32:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, ma
-; CHECK-NEXT: vfmin.vv v8, v8, v9, v0.t
+; CHECK-NEXT: vsetvli a0, zero, e32, m1, ta, ma
+; CHECK-NEXT: vfmin.vv v8, v8, v9
; CHECK-NEXT: ret
%v = call <vscale x 2 x float> @llvm.vp.minnum.nxv2f32(<vscale x 2 x float> %va, <vscale x 2 x float> %vb, <vscale x 2 x i1> %m, i32 %evl)
ret <vscale x 2 x float> %v
@@ -629,7 +561,7 @@ define <vscale x 2 x float> @vfmin_vv_nxv2f32(<vscale x 2 x float> %va, <vscale
define <vscale x 2 x float> @vfmin_vv_nxv2f32_unmasked(<vscale x 2 x float> %va, <vscale x 2 x float> %vb, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_nxv2f32_unmasked:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, ma
+; CHECK-NEXT: vsetvli a0, zero, e32, m1, ta, ma
; CHECK-NEXT: vfmin.vv v8, v8, v9
; CHECK-NEXT: ret
%v = call <vscale x 2 x float> @llvm.vp.minnum.nxv2f32(<vscale x 2 x float> %va, <vscale x 2 x float> %vb, <vscale x 2 x i1> splat (i1 true), i32 %evl)
@@ -639,8 +571,8 @@ define <vscale x 2 x float> @vfmin_vv_nxv2f32_unmasked(<vscale x 2 x float> %va,
define <vscale x 4 x float> @vfmin_vv_nxv4f32(<vscale x 4 x float> %va, <vscale x 4 x float> %vb, <vscale x 4 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_nxv4f32:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, ma
-; CHECK-NEXT: vfmin.vv v8, v8, v10, v0.t
+; CHECK-NEXT: vsetvli a0, zero, e32, m2, ta, ma
+; CHECK-NEXT: vfmin.vv v8, v8, v10
; CHECK-NEXT: ret
%v = call <vscale x 4 x float> @llvm.vp.minnum.nxv4f32(<vscale x 4 x float> %va, <vscale x 4 x float> %vb, <vscale x 4 x i1> %m, i32 %evl)
ret <vscale x 4 x float> %v
@@ -649,7 +581,7 @@ define <vscale x 4 x float> @vfmin_vv_nxv4f32(<vscale x 4 x float> %va, <vscale
define <vscale x 4 x float> @vfmin_vv_nxv4f32_unmasked(<vscale x 4 x float> %va, <vscale x 4 x float> %vb, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_nxv4f32_unmasked:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, ma
+; CHECK-NEXT: vsetvli a0, zero, e32, m2, ta, ma
; CHECK-NEXT: vfmin.vv v8, v8, v10
; CHECK-NEXT: ret
%v = call <vscale x 4 x float> @llvm.vp.minnum.nxv4f32(<vscale x 4 x float> %va, <vscale x 4 x float> %vb, <vscale x 4 x i1> splat (i1 true), i32 %evl)
@@ -659,8 +591,8 @@ define <vscale x 4 x float> @vfmin_vv_nxv4f32_unmasked(<vscale x 4 x float> %va,
define <vscale x 8 x float> @vfmin_vv_nxv8f32(<vscale x 8 x float> %va, <vscale x 8 x float> %vb, <vscale x 8 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_nxv8f32:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, ma
-; CHECK-NEXT: vfmin.vv v8, v8, v12, v0.t
+; CHECK-NEXT: vsetvli a0, zero, e32, m4, ta, ma
+; CHECK-NEXT: vfmin.vv v8, v8, v12
; CHECK-NEXT: ret
%v = call <vscale x 8 x float> @llvm.vp.minnum.nxv8f32(<vscale x 8 x float> %va, <vscale x 8 x float> %vb, <vscale x 8 x i1> %m, i32 %evl)
ret <vscale x 8 x float> %v
@@ -669,7 +601,7 @@ define <vscale x 8 x float> @vfmin_vv_nxv8f32(<vscale x 8 x float> %va, <vscale
define <vscale x 8 x float> @vfmin_vv_nxv8f32_unmasked(<vscale x 8 x float> %va, <vscale x 8 x float> %vb, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_nxv8f32_unmasked:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, ma
+; CHECK-NEXT: vsetvli a0, zero, e32, m4, ta, ma
; CHECK-NEXT: vfmin.vv v8, v8, v12
; CHECK-NEXT: ret
%v = call <vscale x 8 x float> @llvm.vp.minnum.nxv8f32(<vscale x 8 x float> %va, <vscale x 8 x float> %vb, <vscale x 8 x i1> splat (i1 true), i32 %evl)
@@ -679,8 +611,8 @@ define <vscale x 8 x float> @vfmin_vv_nxv8f32_unmasked(<vscale x 8 x float> %va,
define <vscale x 16 x float> @vfmin_vv_nxv16f32(<vscale x 16 x float> %va, <vscale x 16 x float> %vb, <vscale x 16 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_nxv16f32:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
-; CHECK-NEXT: vfmin.vv v8, v8, v16, v0.t
+; CHECK-NEXT: vsetvli a0, zero, e32, m8, ta, ma
+; CHECK-NEXT: vfmin.vv v8, v8, v16
; CHECK-NEXT: ret
%v = call <vscale x 16 x float> @llvm.vp.minnum.nxv16f32(<vscale x 16 x float> %va, <vscale x 16 x float> %vb, <vscale x 16 x i1> %m, i32 %evl)
ret <vscale x 16 x float> %v
@@ -689,7 +621,7 @@ define <vscale x 16 x float> @vfmin_vv_nxv16f32(<vscale x 16 x float> %va, <vsca
define <vscale x 16 x float> @vfmin_vv_nxv16f32_unmasked(<vscale x 16 x float> %va, <vscale x 16 x float> %vb, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_nxv16f32_unmasked:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e32, m8, ta, ma
+; CHECK-NEXT: vsetvli a0, zero, e32, m8, ta, ma
; CHECK-NEXT: vfmin.vv v8, v8, v16
; CHECK-NEXT: ret
%v = call <vscale x 16 x float> @llvm.vp.minnum.nxv16f32(<vscale x 16 x float> %va, <vscale x 16 x float> %vb, <vscale x 16 x i1> splat (i1 true), i32 %evl)
@@ -699,8 +631,8 @@ define <vscale x 16 x float> @vfmin_vv_nxv16f32_unmasked(<vscale x 16 x float> %
define <vscale x 1 x double> @vfmin_vv_nxv1f64(<vscale x 1 x double> %va, <vscale x 1 x double> %vb, <vscale x 1 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_nxv1f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, ma
-; CHECK-NEXT: vfmin.vv v8, v8, v9, v0.t
+; CHECK-NEXT: vsetvli a0, zero, e64, m1, ta, ma
+; CHECK-NEXT: vfmin.vv v8, v8, v9
; CHECK-NEXT: ret
%v = call <vscale x 1 x double> @llvm.vp.minnum.nxv1f64(<vscale x 1 x double> %va, <vscale x 1 x double> %vb, <vscale x 1 x i1> %m, i32 %evl)
ret <vscale x 1 x double> %v
@@ -709,7 +641,7 @@ define <vscale x 1 x double> @vfmin_vv_nxv1f64(<vscale x 1 x double> %va, <vscal
define <vscale x 1 x double> @vfmin_vv_nxv1f64_unmasked(<vscale x 1 x double> %va, <vscale x 1 x double> %vb, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_nxv1f64_unmasked:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, ma
+; CHECK-NEXT: vsetvli a0, zero, e64, m1, ta, ma
; CHECK-NEXT: vfmin.vv v8, v8, v9
; CHECK-NEXT: ret
%v = call <vscale x 1 x double> @llvm.vp.minnum.nxv1f64(<vscale x 1 x double> %va, <vscale x 1 x double> %vb, <vscale x 1 x i1> splat (i1 true), i32 %evl)
@@ -719,8 +651,8 @@ define <vscale x 1 x double> @vfmin_vv_nxv1f64_unmasked(<vscale x 1 x double> %v
define <vscale x 2 x double> @vfmin_vv_nxv2f64(<vscale x 2 x double> %va, <vscale x 2 x double> %vb, <vscale x 2 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_nxv2f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, ma
-; CHECK-NEXT: vfmin.vv v8, v8, v10, v0.t
+; CHECK-NEXT: vsetvli a0, zero, e64, m2, ta, ma
+; CHECK-NEXT: vfmin.vv v8, v8, v10
; CHECK-NEXT: ret
%v = call <vscale x 2 x double> @llvm.vp.minnum.nxv2f64(<vscale x 2 x double> %va, <vscale x 2 x double> %vb, <vscale x 2 x i1> %m, i32 %evl)
ret <vscale x 2 x double> %v
@@ -729,7 +661,7 @@ define <vscale x 2 x double> @vfmin_vv_nxv2f64(<vscale x 2 x double> %va, <vscal
define <vscale x 2 x double> @vfmin_vv_nxv2f64_unmasked(<vscale x 2 x double> %va, <vscale x 2 x double> %vb, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_nxv2f64_unmasked:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, ma
+; CHECK-NEXT: vsetvli a0, zero, e64, m2, ta, ma
; CHECK-NEXT: vfmin.vv v8, v8, v10
; CHECK-NEXT: ret
%v = call <vscale x 2 x double> @llvm.vp.minnum.nxv2f64(<vscale x 2 x double> %va, <vscale x 2 x double> %vb, <vscale x 2 x i1> splat (i1 true), i32 %evl)
@@ -739,8 +671,8 @@ define <vscale x 2 x double> @vfmin_vv_nxv2f64_unmasked(<vscale x 2 x double> %v
define <vscale x 4 x double> @vfmin_vv_nxv4f64(<vscale x 4 x double> %va, <vscale x 4 x double> %vb, <vscale x 4 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_nxv4f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, ma
-; CHECK-NEXT: vfmin.vv v8, v8, v12, v0.t
+; CHECK-NEXT: vsetvli a0, zero, e64, m4, ta, ma
+; CHECK-NEXT: vfmin.vv v8, v8, v12
; CHECK-NEXT: ret
%v = call <vscale x 4 x double> @llvm.vp.minnum.nxv4f64(<vscale x 4 x double> %va, <vscale x 4 x double> %vb, <vscale x 4 x i1> %m, i32 %evl)
ret <vscale x 4 x double> %v
@@ -749,7 +681,7 @@ define <vscale x 4 x double> @vfmin_vv_nxv4f64(<vscale x 4 x double> %va, <vscal
define <vscale x 4 x double> @vfmin_vv_nxv4f64_unmasked(<vscale x 4 x double> %va, <vscale x 4 x double> %vb, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_nxv4f64_unmasked:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, ma
+; CHECK-NEXT: vsetvli a0, zero, e64, m4, ta, ma
; CHECK-NEXT: vfmin.vv v8, v8, v12
; CHECK-NEXT: ret
%v = call <vscale x 4 x double> @llvm.vp.minnum.nxv4f64(<vscale x 4 x double> %va, <vscale x 4 x double> %vb, <vscale x 4 x i1> splat (i1 true), i32 %evl)
@@ -759,8 +691,8 @@ define <vscale x 4 x double> @vfmin_vv_nxv4f64_unmasked(<vscale x 4 x double> %v
define <vscale x 8 x double> @vfmin_vv_nxv8f64(<vscale x 8 x double> %va, <vscale x 8 x double> %vb, <vscale x 8 x i1> %m, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_nxv8f64:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
-; CHECK-NEXT: vfmin.vv v8, v8, v16, v0.t
+; CHECK-NEXT: vsetvli a0, zero, e64, m8, ta, ma
+; CHECK-NEXT: vfmin.vv v8, v8, v16
; CHECK-NEXT: ret
%v = call <vscale x 8 x double> @llvm.vp.minnum.nxv8f64(<vscale x 8 x double> %va, <vscale x 8 x double> %vb, <vscale x 8 x i1> %m, i32 %evl)
ret <vscale x 8 x double> %v
@@ -769,7 +701,7 @@ define <vscale x 8 x double> @vfmin_vv_nxv8f64(<vscale x 8 x double> %va, <vscal
define <vscale x 8 x double> @vfmin_vv_nxv8f64_unmasked(<vscale x 8 x double> %va, <vscale x 8 x double> %vb, i32 zeroext %evl) {
; CHECK-LABEL: vfmin_vv_nxv8f64_unmasked:
; CHECK: # %bb.0:
-; CHECK-NEXT: vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT: vsetvli a0, zero, e64, m8, ta, ma
; CHECK-NEXT: vfmin.vv v8, v8, v16
; CHECK-NEXT: ret
%v = call <vscale x 8 x double> @llvm.vp.minnum.nxv8f64(<vscale x 8 x double> %va, <vscale x 8 x double> %vb, <vscale x 8 x i1> splat (i1 true), i32 %evl)
More information about the llvm-commits
mailing list