[TTI][RISCV] Implement cost of some intrinsics with LMUL #117874
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Intrinsics include: sadd_sat/ssub_sat/uadd_sat/usub_sat/fabs/fsqrt/cttz/ctlz/ctpop
llvmbot
added
backend:RISC-V
llvm:analysis
Member
|
@llvm/pr-subscribers-backend-risc-v @llvm/pr-subscribers-llvm-analysis Author: LiqinWeng (LiqinWeng) ChangesIntrinsics include: sadd_sat/ssub_sat/uadd_sat/usub_sat/fabs/fsqrt/cttz/ctlz/ctpop Patch is 73.35 KiB, truncated to 20.00 KiB below, full version: https://github.com/llvm/llvm-project/pull/117874.diff 5 Files Affected:
diff --git a/llvm/lib/Target/RISCV/RISCVTargetTransformInfo.cpp b/llvm/lib/Target/RISCV/RISCVTargetTransformInfo.cpp
index 8f0ef69258b165..e2cd71fb5a9f40 100644
--- a/llvm/lib/Target/RISCV/RISCVTargetTransformInfo.cpp
+++ b/llvm/lib/Target/RISCV/RISCVTargetTransformInfo.cpp
@@ -1013,20 +1013,78 @@ RISCVTTIImpl::getIntrinsicInstrCost(const IntrinsicCostAttributes &ICA,
case Intrinsic::sadd_sat:
case Intrinsic::ssub_sat:
case Intrinsic::uadd_sat:
- case Intrinsic::usub_sat:
- case Intrinsic::fabs:
+ case Intrinsic::usub_sat: {
+ auto LT = getTypeLegalizationCost(RetTy);
+ if (ST->hasVInstructions() && LT.second.isVector()) {
+ unsigned Op;
+ switch (ICA.getID()) {
+ case Intrinsic::sadd_sat:
+ Op = RISCV::VSADD_VV;
+ break;
+ case Intrinsic::ssub_sat:
+ Op = RISCV::VSSUBU_VV;
+ break;
+ case Intrinsic::uadd_sat:
+ Op = RISCV::VSADDU_VV;
+ break;
+ case Intrinsic::usub_sat:
+ Op = RISCV::VSSUBU_VV;
+ break;
+ }
+ return LT.first * getRISCVInstructionCost(Op, LT.second, CostKind);
+ }
+ break;
+ }
+ case Intrinsic::fabs: {
+ auto LT = getTypeLegalizationCost(RetTy);
+ // FIXME: not get the correct cost about the llvm.sqrt.vxbf16
+ // LT.second is promote llvm::MVT::f32
+ if (ST->hasVInstructions() && LT.second.isVector()) {
+ // lui a0, 8
+ // addi a0, a0, -1
+ // vsetvli a1, zero, e16, m1, ta, ma
+ // vand.vx v8, v8, a0
+ if (LT.second.getVectorElementType() == MVT::bf16 ||
+ (LT.second.getVectorElementType() == MVT::f16 &&
+ !ST->hasVInstructionsF16()))
+ return LT.first * getRISCVInstructionCost(RISCV::VAND_VX, LT.second,
+ CostKind) +
+ 2;
+ else
+ return LT.first *
+ getRISCVInstructionCost(RISCV::VFSGNJX_VV, LT.second, CostKind);
+ }
+ break;
+ }
case Intrinsic::sqrt: {
auto LT = getTypeLegalizationCost(RetTy);
- if (ST->hasVInstructions() && LT.second.isVector())
- return LT.first;
+ // FIXME: not get the correct cost about the llvm.sqrt.vxbf16
+ // LT.second is promote llvm::MVT::f32
+ if (ST->hasVInstructions() && LT.second.isVector()) {
+ SmallVector<unsigned, 3> Opcodes;
+ if (LT.second.getVectorElementType() == MVT::bf16)
+ Opcodes = {RISCV::VFWCVTBF16_F_F_V, RISCV::VFSQRT_V,
+ RISCV::VFNCVTBF16_F_F_W};
+ else if (LT.second.getVectorElementType() == MVT::f16 &&
+ !ST->hasVInstructionsF16())
+ Opcodes = {RISCV::VFWCVT_F_F_V, RISCV::VFSQRT_V, RISCV::VFNCVT_F_F_W};
+ else
+ Opcodes = {RISCV::VFSQRT_V};
+ return LT.first *
+ getRISCVInstructionCost(RISCV::VFSQRT_V, LT.second, CostKind);
+ }
break;
}
case Intrinsic::cttz:
case Intrinsic::ctlz:
case Intrinsic::ctpop: {
auto LT = getTypeLegalizationCost(RetTy);
- if (ST->hasVInstructions() && ST->hasStdExtZvbb() && LT.second.isVector())
- return LT.first;
+ if (ST->hasVInstructions() && ST->hasStdExtZvbb() && LT.second.isVector()) {
+ unsigned Op = (Intrinsic::cttz) ? RISCV::VCTZ_V
+ : (Intrinsic::ctlz) ? RISCV::VCLZ_V
+ : RISCV::VCPOP_V;
+ return LT.first * getRISCVInstructionCost(Op, LT.second, CostKind);
+ }
break;
}
case Intrinsic::abs: {
diff --git a/llvm/test/Analysis/CostModel/RISCV/fp-min-max-abs.ll b/llvm/test/Analysis/CostModel/RISCV/fp-min-max-abs.ll
index 0b2c8da4438da2..6259c31a08bc1f 100644
--- a/llvm/test/Analysis/CostModel/RISCV/fp-min-max-abs.ll
+++ b/llvm/test/Analysis/CostModel/RISCV/fp-min-max-abs.ll
@@ -7,22 +7,22 @@ define void @fabs() {
; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %1 = call float @llvm.fabs.f32(float undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %2 = call <2 x float> @llvm.fabs.v2f32(<2 x float> undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %3 = call <4 x float> @llvm.fabs.v4f32(<4 x float> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %4 = call <8 x float> @llvm.fabs.v8f32(<8 x float> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %5 = call <16 x float> @llvm.fabs.v16f32(<16 x float> undef)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %4 = call <8 x float> @llvm.fabs.v8f32(<8 x float> undef)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %5 = call <16 x float> @llvm.fabs.v16f32(<16 x float> undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %6 = call <vscale x 1 x float> @llvm.fabs.nxv1f32(<vscale x 1 x float> undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %7 = call <vscale x 2 x float> @llvm.fabs.nxv2f32(<vscale x 2 x float> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %8 = call <vscale x 4 x float> @llvm.fabs.nxv4f32(<vscale x 4 x float> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %9 = call <vscale x 8 x float> @llvm.fabs.nxv8f32(<vscale x 8 x float> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %10 = call <vscale x 16 x float> @llvm.fabs.nxv16f32(<vscale x 16 x float> undef)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %8 = call <vscale x 4 x float> @llvm.fabs.nxv4f32(<vscale x 4 x float> undef)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %9 = call <vscale x 8 x float> @llvm.fabs.nxv8f32(<vscale x 8 x float> undef)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %10 = call <vscale x 16 x float> @llvm.fabs.nxv16f32(<vscale x 16 x float> undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %11 = call double @llvm.fabs.f64(double undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %12 = call <2 x double> @llvm.fabs.v2f64(<2 x double> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %13 = call <4 x double> @llvm.fabs.v4f64(<4 x double> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %14 = call <8 x double> @llvm.fabs.v8f64(<8 x double> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %15 = call <16 x double> @llvm.fabs.v16f64(<16 x double> undef)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %13 = call <4 x double> @llvm.fabs.v4f64(<4 x double> undef)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %14 = call <8 x double> @llvm.fabs.v8f64(<8 x double> undef)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %15 = call <16 x double> @llvm.fabs.v16f64(<16 x double> undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %16 = call <vscale x 1 x double> @llvm.fabs.nxv1f64(<vscale x 1 x double> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %17 = call <vscale x 2 x double> @llvm.fabs.nxv2f64(<vscale x 2 x double> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %18 = call <vscale x 4 x double> @llvm.fabs.nxv4f64(<vscale x 4 x double> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %19 = call <vscale x 8 x double> @llvm.fabs.nxv8f64(<vscale x 8 x double> undef)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %17 = call <vscale x 2 x double> @llvm.fabs.nxv2f64(<vscale x 2 x double> undef)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %18 = call <vscale x 4 x double> @llvm.fabs.nxv4f64(<vscale x 4 x double> undef)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %19 = call <vscale x 8 x double> @llvm.fabs.nxv8f64(<vscale x 8 x double> undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret void
;
call float @llvm.fabs.f32(float undef)
@@ -48,17 +48,29 @@ define void @fabs() {
}
define void @fabs_f16() {
-; CHECK-LABEL: 'fabs_f16'
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %1 = call half @llvm.fabs.f16(half undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %2 = call <2 x half> @llvm.fabs.v2f16(<2 x half> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %3 = call <4 x half> @llvm.fabs.v4f16(<4 x half> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %4 = call <8 x half> @llvm.fabs.v8f16(<8 x half> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %5 = call <16 x half> @llvm.fabs.v16f16(<16 x half> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %6 = call <vscale x 2 x half> @llvm.fabs.nxv2f16(<vscale x 2 x half> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %7 = call <vscale x 4 x half> @llvm.fabs.nxv4f16(<vscale x 4 x half> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %8 = call <vscale x 8 x half> @llvm.fabs.nxv8f16(<vscale x 8 x half> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %9 = call <vscale x 16 x half> @llvm.fabs.nxv16f16(<vscale x 16 x half> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret void
+; ZVFH-LABEL: 'fabs_f16'
+; ZVFH-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %1 = call half @llvm.fabs.f16(half undef)
+; ZVFH-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %2 = call <2 x half> @llvm.fabs.v2f16(<2 x half> undef)
+; ZVFH-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %3 = call <4 x half> @llvm.fabs.v4f16(<4 x half> undef)
+; ZVFH-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %4 = call <8 x half> @llvm.fabs.v8f16(<8 x half> undef)
+; ZVFH-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %5 = call <16 x half> @llvm.fabs.v16f16(<16 x half> undef)
+; ZVFH-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %6 = call <vscale x 2 x half> @llvm.fabs.nxv2f16(<vscale x 2 x half> undef)
+; ZVFH-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %7 = call <vscale x 4 x half> @llvm.fabs.nxv4f16(<vscale x 4 x half> undef)
+; ZVFH-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %8 = call <vscale x 8 x half> @llvm.fabs.nxv8f16(<vscale x 8 x half> undef)
+; ZVFH-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %9 = call <vscale x 16 x half> @llvm.fabs.nxv16f16(<vscale x 16 x half> undef)
+; ZVFH-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret void
+;
+; ZVFHMIN-LABEL: 'fabs_f16'
+; ZVFHMIN-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %1 = call half @llvm.fabs.f16(half undef)
+; ZVFHMIN-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %2 = call <2 x half> @llvm.fabs.v2f16(<2 x half> undef)
+; ZVFHMIN-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %3 = call <4 x half> @llvm.fabs.v4f16(<4 x half> undef)
+; ZVFHMIN-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %4 = call <8 x half> @llvm.fabs.v8f16(<8 x half> undef)
+; ZVFHMIN-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %5 = call <16 x half> @llvm.fabs.v16f16(<16 x half> undef)
+; ZVFHMIN-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %6 = call <vscale x 2 x half> @llvm.fabs.nxv2f16(<vscale x 2 x half> undef)
+; ZVFHMIN-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %7 = call <vscale x 4 x half> @llvm.fabs.nxv4f16(<vscale x 4 x half> undef)
+; ZVFHMIN-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %8 = call <vscale x 8 x half> @llvm.fabs.nxv8f16(<vscale x 8 x half> undef)
+; ZVFHMIN-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %9 = call <vscale x 16 x half> @llvm.fabs.nxv16f16(<vscale x 16 x half> undef)
+; ZVFHMIN-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret void
;
call half @llvm.fabs.f16(half undef)
call <2 x half> @llvm.fabs.v2f16(<2 x half> undef)
diff --git a/llvm/test/Analysis/CostModel/RISCV/fp-sqrt-pow.ll b/llvm/test/Analysis/CostModel/RISCV/fp-sqrt-pow.ll
index be9c19dc59a852..446627f6bf3c0e 100644
--- a/llvm/test/Analysis/CostModel/RISCV/fp-sqrt-pow.ll
+++ b/llvm/test/Analysis/CostModel/RISCV/fp-sqrt-pow.ll
@@ -8,30 +8,30 @@ define void @sqrt() {
; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %2 = call <2 x bfloat> @llvm.sqrt.v2bf16(<2 x bfloat> undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %3 = call <4 x bfloat> @llvm.sqrt.v4bf16(<4 x bfloat> undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %4 = call <8 x bfloat> @llvm.sqrt.v8bf16(<8 x bfloat> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %5 = call <16 x bfloat> @llvm.sqrt.v16bf16(<16 x bfloat> undef)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %5 = call <16 x bfloat> @llvm.sqrt.v16bf16(<16 x bfloat> undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %6 = call <vscale x 2 x bfloat> @llvm.sqrt.nxv2bf16(<vscale x 2 x bfloat> undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %7 = call <vscale x 4 x bfloat> @llvm.sqrt.nxv4bf16(<vscale x 4 x bfloat> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %8 = call <vscale x 8 x bfloat> @llvm.sqrt.nxv8bf16(<vscale x 8 x bfloat> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %9 = call <vscale x 16 x bfloat> @llvm.sqrt.nxv16bf16(<vscale x 16 x bfloat> undef)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %8 = call <vscale x 8 x bfloat> @llvm.sqrt.nxv8bf16(<vscale x 8 x bfloat> undef)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %9 = call <vscale x 16 x bfloat> @llvm.sqrt.nxv16bf16(<vscale x 16 x bfloat> undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %10 = call float @llvm.sqrt.f32(float undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %11 = call <2 x float> @llvm.sqrt.v2f32(<2 x float> undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %12 = call <4 x float> @llvm.sqrt.v4f32(<4 x float> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %13 = call <8 x float> @llvm.sqrt.v8f32(<8 x float> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %14 = call <16 x float> @llvm.sqrt.v16f32(<16 x float> undef)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %13 = call <8 x float> @llvm.sqrt.v8f32(<8 x float> undef)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %14 = call <16 x float> @llvm.sqrt.v16f32(<16 x float> undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %15 = call <vscale x 1 x float> @llvm.sqrt.nxv1f32(<vscale x 1 x float> undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %16 = call <vscale x 2 x float> @llvm.sqrt.nxv2f32(<vscale x 2 x float> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %17 = call <vscale x 4 x float> @llvm.sqrt.nxv4f32(<vscale x 4 x float> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %18 = call <vscale x 8 x float> @llvm.sqrt.nxv8f32(<vscale x 8 x float> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %19 = call <vscale x 16 x float> @llvm.sqrt.nxv16f32(<vscale x 16 x float> undef)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %17 = call <vscale x 4 x float> @llvm.sqrt.nxv4f32(<vscale x 4 x float> undef)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %18 = call <vscale x 8 x float> @llvm.sqrt.nxv8f32(<vscale x 8 x float> undef)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %19 = call <vscale x 16 x float> @llvm.sqrt.nxv16f32(<vscale x 16 x float> undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %20 = call double @llvm.sqrt.f64(double undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %21 = call <2 x double> @llvm.sqrt.v2f64(<2 x double> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %22 = call <4 x double> @llvm.sqrt.v4f64(<4 x double> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %23 = call <8 x double> @llvm.sqrt.v8f64(<8 x double> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %24 = call <16 x double> @llvm.sqrt.v16f64(<16 x double> undef)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %22 = call <4 x double> @llvm.sqrt.v4f64(<4 x double> undef)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %23 = call <8 x double> @llvm.sqrt.v8f64(<8 x double> undef)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %24 = call <16 x double> @llvm.sqrt.v16f64(<16 x double> undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %25 = call <vscale x 1 x double> @llvm.sqrt.nxv1f64(<vscale x 1 x double> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %26 = call <vscale x 2 x double> @llvm.sqrt.nxv2f64(<vscale x 2 x double> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %27 = call <vscale x 4 x double> @llvm.sqrt.nxv4f64(<vscale x 4 x double> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %28 = call <vscale x 8 x double> @llvm.sqrt.nxv8f64(<vscale x 8 x double> undef)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %26 = call <vscale x 2 x double> @llvm.sqrt.nxv2f64(<vscale x 2 x double> undef)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %27 = call <vscale x 4 x double> @llvm.sqrt.nxv4f64(<vscale x 4 x double> undef)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %28 = call <vscale x 8 x double> @llvm.sqrt.nxv8f64(<vscale x 8 x double> undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret void
;
call bfloat @llvm.sqrt.bf16(bfloat undef)
@@ -71,11 +71,11 @@ define void @sqrt_f16() {
; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %2 = call <2 x half> @llvm.sqrt.v2f16(<2 x half> undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %3 = call <4 x half> @llvm.sqrt.v4f16(<4 x half> undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %4 = call <8 x half> @llvm.sqrt.v8f16(<8 x half> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %5 = call <16 x half> @llvm.sqrt.v16f16(<16 x half> undef)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %5 = call <16 x half> @llvm.sqrt.v16f16(<16 x half> undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %6 = call <vscale x 2 x half> @llvm.sqrt.nxv2f16(<vscale x 2 x half> undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %7 = call <vscale x 4 x half> @llvm.sqrt.nxv4f16(<vscale x 4 x half> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %8 = call <vscale x 8 x half> @llvm.sqrt.nxv8f16(<vscale x 8 x half> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %9 = call <vscale x 16 x half> @llvm.sqrt.nxv16f16(<vscale x 16 x half> undef)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %8 = call <vscale x 8 x half> @llvm.sqrt.nxv8f16(<vscale x 8 x half> undef)
+; CHECK-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %9 = call <vscale x 16 x half> @llvm.sqrt.nxv16f16(<vscale x 16 x half> undef)
; CHECK-NEXT: Cost Model: Found an estimated cost...
[truncated]
|
lukel97
reviewed
Nov 28, 2024
| if (ST->hasVInstructions() && LT.second.isVector()) | ||
| return LT.first; | ||
| // FIXME: not get the correct cost about the llvm.sqrt.vxbf16 | ||
| // LT.second is promote llvm::MVT::f32 |
Contributor
There was a problem hiding this comment.
Not for this PR, but we probably want to handle the zvfhmin/zvfbfmin promotion for all intrinsics. We could pull this out of the switch statement later.
Contributor
Author
There was a problem hiding this comment.
I will create new pr deal with f16 && bf16
Comment on lines
1040
to
1041
| // FIXME: not get the correct cost about the llvm.sqrt.vxbf16 | ||
| // LT.second is promote llvm::MVT::f32 |
Contributor
There was a problem hiding this comment.
Is this comment still needed? The costs in the tests look correct
| !ST->hasVInstructionsF16()) | ||
| Opcodes = {RISCV::VFWCVT_F_F_V, RISCV::VFSQRT_V, RISCV::VFNCVT_F_F_W}; | ||
| else | ||
| Opcodes = {RISCV::VFSQRT_V}; |
lukel97
reviewed
Nov 30, 2024
| ; NOTE: Assertions have been autogenerated by utils/update_analyze_test_checks.py | ||
| ; RUN: opt < %s -passes="print<cost-model>" 2>&1 -disable-output -S -mtriple=riscv64 -mattr=+v,+f,+d,+zfh,+zvfh | FileCheck %s --check-prefixes=CHECK,ZVFH | ||
| ; RUN: opt < %s -passes="print<cost-model>" 2>&1 -disable-output -S -mtriple=riscv64 -mattr=+v,+f,+d,+zfh,+zvfhmin | FileCheck %s --check-prefixes=CHECK,ZVFHMIN | ||
| ; RUN: opt < %s -passes="print<cost-model>" 2>&1 -disable-output -S -mtriple=riscv64 -mattr=+v,+f,+d,+zfh,+zvfhmin,+zvfhbmin | FileCheck %s --check-prefixes=CHECK,ZVFHMIN |
| auto LT = getTypeLegalizationCost(RetTy); | ||
| if (ST->hasVInstructions() && LT.second.isVector()) | ||
| return LT.first; | ||
| // TODO: add f16/bf16, bf16 with zvfhbmin && f16 with zvfhmin |
Contributor
There was a problem hiding this comment.
Suggested change
| // TODO: add f16/bf16, bf16 with zvfhbmin && f16 with zvfhmin | |
| // TODO: add f16/bf16, bf16 with zvfbfmin && f16 with zvfhmin |
lukel97
reviewed
Dec 1, 2024
| @@ -1,28 +1,28 @@ | |||
| ; NOTE: Assertions have been autogenerated by utils/update_analyze_test_checks.py | |||
| ; RUN: opt < %s -passes="print<cost-model>" 2>&1 -disable-output -S -mtriple=riscv64 -mattr=+v,+f,+d,+zfh,+zvfh | FileCheck %s --check-prefixes=CHECK,ZVFH | |||
Contributor
There was a problem hiding this comment.
Can you add zvfbfmin to the ZVFH run line too? I think that matches the other tests with bfloat
lukel97
reviewed
Dec 2, 2024
Comment on lines
72
to
75
| call <2 x bfloat> @llvm.fabs.v2f16(<2 x bfloat> undef) | ||
| call <4 x bfloat> @llvm.fabs.v4f16(<4 x bfloat> undef) | ||
| call <8 x bfloat> @llvm.fabs.v8f16(<8 x bfloat> undef) | ||
| call <16 x bfloat> @llvm.fabs.v16f16(<16 x bfloat> undef) |
Contributor
There was a problem hiding this comment.
This should be @llvm.fabs.v2bf16?
| ; CHECK-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret void | ||
| ; | ||
| call half @llvm.fabs.f16(half undef) | ||
| call <2 x bfloat> @llvm.fabs.v2f16(<2 x bfloat> undef) |
Contributor
There was a problem hiding this comment.
Can you move the bf16 lines into the main @Fabs() test? The f16 tests were split out just to show the difference between zvfh and zvfhmin IIRC, which bf16 won't need
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Intrinsics include: sadd_sat/ssub_sat/uadd_sat/usub_sat/fabs/fsqrt/cttz/ctlz/ctpop