[RISCV] Mark math functions as expanded for zvfhmin/zvfbfmin #112508
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
For regular floating point types we mark these as expanded on scalable vectors so they're not legal in the cost model, so this does the same for f16 w/ zvfhmin and bf16.
|
@llvm/pr-subscribers-llvm-analysis @llvm/pr-subscribers-backend-risc-v Author: Luke Lau (lukel97) ChangesFor regular floating point types we mark these as expanded on scalable vectors so they're not legal in the cost model, so this does the same for f16 w/ zvfhmin and bf16. Patch is 42.15 KiB, truncated to 20.00 KiB below, full version: https://github.com/llvm/llvm-project/pull/112508.diff 4 Files Affected:
diff --git a/llvm/lib/Target/RISCV/RISCVISelLowering.cpp b/llvm/lib/Target/RISCV/RISCVISelLowering.cpp
index bf333b7b790167..8db826f653156e 100644
--- a/llvm/lib/Target/RISCV/RISCVISelLowering.cpp
+++ b/llvm/lib/Target/RISCV/RISCVISelLowering.cpp
@@ -717,6 +717,10 @@ RISCVTargetLowering::RISCVTargetLowering(const TargetMachine &TM,
ISD::VECREDUCE_FADD, ISD::VECREDUCE_SEQ_FADD, ISD::VECREDUCE_FMIN,
ISD::VECREDUCE_FMAX, ISD::VECREDUCE_FMINIMUM, ISD::VECREDUCE_FMAXIMUM};
+ static const unsigned FloatingPointLibCallOps[] = {
+ ISD::FREM, ISD::FPOW, ISD::FCOS, ISD::FSIN, ISD::FSINCOS, ISD::FEXP,
+ ISD::FEXP2, ISD::FEXP10, ISD::FLOG, ISD::FLOG2, ISD::FLOG10};
+
if (!Subtarget.is64Bit()) {
// We must custom-lower certain vXi64 operations on RV32 due to the vector
// element type being illegal.
@@ -1002,17 +1006,7 @@ RISCVTargetLowering::RISCVTargetLowering(const TargetMachine &TM,
setOperationAction(FloatingPointVecReduceOps, VT, Custom);
// Expand FP operations that need libcalls.
- setOperationAction(ISD::FREM, VT, Expand);
- setOperationAction(ISD::FPOW, VT, Expand);
- setOperationAction(ISD::FCOS, VT, Expand);
- setOperationAction(ISD::FSIN, VT, Expand);
- setOperationAction(ISD::FSINCOS, VT, Expand);
- setOperationAction(ISD::FEXP, VT, Expand);
- setOperationAction(ISD::FEXP2, VT, Expand);
- setOperationAction(ISD::FEXP10, VT, Expand);
- setOperationAction(ISD::FLOG, VT, Expand);
- setOperationAction(ISD::FLOG2, VT, Expand);
- setOperationAction(ISD::FLOG10, VT, Expand);
+ setOperationAction(FloatingPointLibCallOps, VT, Expand);
setOperationAction(ISD::FCOPYSIGN, VT, Legal);
@@ -1097,6 +1091,9 @@ RISCVTargetLowering::RISCVTargetLowering(const TargetMachine &TM,
setOperationAction(ISD::FABS, VT, Expand);
setOperationAction(ISD::FCOPYSIGN, VT, Expand);
+ // Expand FP operations that need libcalls.
+ setOperationAction(FloatingPointLibCallOps, VT, Expand);
+
// Custom split nxv32[b]f16 since nxv32[b]f32 is not legal.
if (getLMUL(VT) == RISCVII::VLMUL::LMUL_8) {
setOperationAction(ZvfhminZvfbfminPromoteOps, VT, Custom);
diff --git a/llvm/test/Analysis/CostModel/RISCV/arith-fp.ll b/llvm/test/Analysis/CostModel/RISCV/arith-fp.ll
index 20d47001739ee1..b6b49982a73267 100644
--- a/llvm/test/Analysis/CostModel/RISCV/arith-fp.ll
+++ b/llvm/test/Analysis/CostModel/RISCV/arith-fp.ll
@@ -526,11 +526,11 @@ define void @frem() {
; CHECK-NEXT: Cost Model: Found an estimated cost of 12 for instruction: %V4BF16 = frem <4 x bfloat> undef, undef
; CHECK-NEXT: Cost Model: Found an estimated cost of 24 for instruction: %V8BF16 = frem <8 x bfloat> undef, undef
; CHECK-NEXT: Cost Model: Found an estimated cost of 63 for instruction: %V16BF16 = frem <16 x bfloat> undef, undef
-; CHECK-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %NXV1BF16 = frem <vscale x 1 x bfloat> undef, undef
-; CHECK-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %NXV2BF16 = frem <vscale x 2 x bfloat> undef, undef
-; CHECK-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %NXV4BF16 = frem <vscale x 4 x bfloat> undef, undef
-; CHECK-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %NXV8BF16 = frem <vscale x 8 x bfloat> undef, undef
-; CHECK-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %NXV16BF16 = frem <vscale x 16 x bfloat> undef, undef
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %NXV1BF16 = frem <vscale x 1 x bfloat> undef, undef
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %NXV2BF16 = frem <vscale x 2 x bfloat> undef, undef
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %NXV4BF16 = frem <vscale x 4 x bfloat> undef, undef
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %NXV8BF16 = frem <vscale x 8 x bfloat> undef, undef
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %NXV16BF16 = frem <vscale x 16 x bfloat> undef, undef
; CHECK-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V1F32 = frem <1 x float> undef, undef
; CHECK-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V2F32 = frem <2 x float> undef, undef
; CHECK-NEXT: Cost Model: Found an estimated cost of 12 for instruction: %V4F32 = frem <4 x float> undef, undef
@@ -593,37 +593,21 @@ define void @frem() {
}
define void @frem_f16() {
-; ZVFH-LABEL: 'frem_f16'
-; ZVFH-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %F16 = frem half undef, undef
-; ZVFH-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V1F16 = frem <1 x half> undef, undef
-; ZVFH-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V2F16 = frem <2 x half> undef, undef
-; ZVFH-NEXT: Cost Model: Found an estimated cost of 12 for instruction: %V4F16 = frem <4 x half> undef, undef
-; ZVFH-NEXT: Cost Model: Found an estimated cost of 24 for instruction: %V8F16 = frem <8 x half> undef, undef
-; ZVFH-NEXT: Cost Model: Found an estimated cost of 63 for instruction: %V16F16 = frem <16 x half> undef, undef
-; ZVFH-NEXT: Cost Model: Found an estimated cost of 127 for instruction: %V32F16 = frem <32 x half> undef, undef
-; ZVFH-NEXT: Cost Model: Invalid cost for instruction: %NXV1F16 = frem <vscale x 1 x half> undef, undef
-; ZVFH-NEXT: Cost Model: Invalid cost for instruction: %NXV2F16 = frem <vscale x 2 x half> undef, undef
-; ZVFH-NEXT: Cost Model: Invalid cost for instruction: %NXV4F16 = frem <vscale x 4 x half> undef, undef
-; ZVFH-NEXT: Cost Model: Invalid cost for instruction: %NXV8F16 = frem <vscale x 8 x half> undef, undef
-; ZVFH-NEXT: Cost Model: Invalid cost for instruction: %NXV16F16 = frem <vscale x 16 x half> undef, undef
-; ZVFH-NEXT: Cost Model: Invalid cost for instruction: %NXV32F16 = frem <vscale x 32 x half> undef, undef
-; ZVFH-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret void
-;
-; ZVFHMIN-LABEL: 'frem_f16'
-; ZVFHMIN-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %F16 = frem half undef, undef
-; ZVFHMIN-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V1F16 = frem <1 x half> undef, undef
-; ZVFHMIN-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V2F16 = frem <2 x half> undef, undef
-; ZVFHMIN-NEXT: Cost Model: Found an estimated cost of 12 for instruction: %V4F16 = frem <4 x half> undef, undef
-; ZVFHMIN-NEXT: Cost Model: Found an estimated cost of 24 for instruction: %V8F16 = frem <8 x half> undef, undef
-; ZVFHMIN-NEXT: Cost Model: Found an estimated cost of 63 for instruction: %V16F16 = frem <16 x half> undef, undef
-; ZVFHMIN-NEXT: Cost Model: Found an estimated cost of 127 for instruction: %V32F16 = frem <32 x half> undef, undef
-; ZVFHMIN-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %NXV1F16 = frem <vscale x 1 x half> undef, undef
-; ZVFHMIN-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %NXV2F16 = frem <vscale x 2 x half> undef, undef
-; ZVFHMIN-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %NXV4F16 = frem <vscale x 4 x half> undef, undef
-; ZVFHMIN-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %NXV8F16 = frem <vscale x 8 x half> undef, undef
-; ZVFHMIN-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %NXV16F16 = frem <vscale x 16 x half> undef, undef
-; ZVFHMIN-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %NXV32F16 = frem <vscale x 32 x half> undef, undef
-; ZVFHMIN-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret void
+; CHECK-LABEL: 'frem_f16'
+; CHECK-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %F16 = frem half undef, undef
+; CHECK-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V1F16 = frem <1 x half> undef, undef
+; CHECK-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V2F16 = frem <2 x half> undef, undef
+; CHECK-NEXT: Cost Model: Found an estimated cost of 12 for instruction: %V4F16 = frem <4 x half> undef, undef
+; CHECK-NEXT: Cost Model: Found an estimated cost of 24 for instruction: %V8F16 = frem <8 x half> undef, undef
+; CHECK-NEXT: Cost Model: Found an estimated cost of 63 for instruction: %V16F16 = frem <16 x half> undef, undef
+; CHECK-NEXT: Cost Model: Found an estimated cost of 127 for instruction: %V32F16 = frem <32 x half> undef, undef
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %NXV1F16 = frem <vscale x 1 x half> undef, undef
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %NXV2F16 = frem <vscale x 2 x half> undef, undef
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %NXV4F16 = frem <vscale x 4 x half> undef, undef
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %NXV8F16 = frem <vscale x 8 x half> undef, undef
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %NXV16F16 = frem <vscale x 16 x half> undef, undef
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %NXV32F16 = frem <vscale x 32 x half> undef, undef
+; CHECK-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret void
;
%F16 = frem half undef, undef
diff --git a/llvm/test/Analysis/CostModel/RISCV/fp-sqrt-pow.ll b/llvm/test/Analysis/CostModel/RISCV/fp-sqrt-pow.ll
index 1768222b8a92e5..efe17f2b76a70e 100644
--- a/llvm/test/Analysis/CostModel/RISCV/fp-sqrt-pow.ll
+++ b/llvm/test/Analysis/CostModel/RISCV/fp-sqrt-pow.ll
@@ -97,11 +97,11 @@ define void @pow() {
; CHECK-NEXT: Cost Model: Found an estimated cost of 44 for instruction: %3 = call <4 x bfloat> @llvm.pow.v4bf16(<4 x bfloat> undef, <4 x bfloat> undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 88 for instruction: %4 = call <8 x bfloat> @llvm.pow.v8bf16(<8 x bfloat> undef, <8 x bfloat> undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 191 for instruction: %5 = call <16 x bfloat> @llvm.pow.v16bf16(<16 x bfloat> undef, <16 x bfloat> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %6 = call <vscale x 1 x bfloat> @llvm.pow.nxv1bf16(<vscale x 1 x bfloat> undef, <vscale x 1 x bfloat> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %7 = call <vscale x 2 x bfloat> @llvm.pow.nxv2bf16(<vscale x 2 x bfloat> undef, <vscale x 2 x bfloat> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %8 = call <vscale x 4 x bfloat> @llvm.pow.nxv4bf16(<vscale x 4 x bfloat> undef, <vscale x 4 x bfloat> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %9 = call <vscale x 8 x bfloat> @llvm.pow.nxv8bf16(<vscale x 8 x bfloat> undef, <vscale x 8 x bfloat> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %10 = call <vscale x 16 x bfloat> @llvm.pow.nxv16bf16(<vscale x 16 x bfloat> undef, <vscale x 16 x bfloat> undef)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %6 = call <vscale x 1 x bfloat> @llvm.pow.nxv1bf16(<vscale x 1 x bfloat> undef, <vscale x 1 x bfloat> undef)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %7 = call <vscale x 2 x bfloat> @llvm.pow.nxv2bf16(<vscale x 2 x bfloat> undef, <vscale x 2 x bfloat> undef)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %8 = call <vscale x 4 x bfloat> @llvm.pow.nxv4bf16(<vscale x 4 x bfloat> undef, <vscale x 4 x bfloat> undef)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %9 = call <vscale x 8 x bfloat> @llvm.pow.nxv8bf16(<vscale x 8 x bfloat> undef, <vscale x 8 x bfloat> undef)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %10 = call <vscale x 16 x bfloat> @llvm.pow.nxv16bf16(<vscale x 16 x bfloat> undef, <vscale x 16 x bfloat> undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 10 for instruction: %11 = call float @llvm.pow.f32(float undef, float undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 22 for instruction: %12 = call <2 x float> @llvm.pow.v2f32(<2 x float> undef, <2 x float> undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 44 for instruction: %13 = call <4 x float> @llvm.pow.v4f32(<4 x float> undef, <4 x float> undef)
@@ -175,11 +175,11 @@ define void @pow_f16() {
; ZVFHMIN-NEXT: Cost Model: Found an estimated cost of 44 for instruction: %3 = call <4 x half> @llvm.pow.v4f16(<4 x half> undef, <4 x half> undef)
; ZVFHMIN-NEXT: Cost Model: Found an estimated cost of 88 for instruction: %4 = call <8 x half> @llvm.pow.v8f16(<8 x half> undef, <8 x half> undef)
; ZVFHMIN-NEXT: Cost Model: Found an estimated cost of 191 for instruction: %5 = call <16 x half> @llvm.pow.v16f16(<16 x half> undef, <16 x half> undef)
-; ZVFHMIN-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %6 = call <vscale x 1 x half> @llvm.pow.nxv1f16(<vscale x 1 x half> undef, <vscale x 1 x half> undef)
-; ZVFHMIN-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %7 = call <vscale x 2 x half> @llvm.pow.nxv2f16(<vscale x 2 x half> undef, <vscale x 2 x half> undef)
-; ZVFHMIN-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %8 = call <vscale x 4 x half> @llvm.pow.nxv4f16(<vscale x 4 x half> undef, <vscale x 4 x half> undef)
-; ZVFHMIN-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %9 = call <vscale x 8 x half> @llvm.pow.nxv8f16(<vscale x 8 x half> undef, <vscale x 8 x half> undef)
-; ZVFHMIN-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %10 = call <vscale x 16 x half> @llvm.pow.nxv16f16(<vscale x 16 x half> undef, <vscale x 16 x half> undef)
+; ZVFHMIN-NEXT: Cost Model: Invalid cost for instruction: %6 = call <vscale x 1 x half> @llvm.pow.nxv1f16(<vscale x 1 x half> undef, <vscale x 1 x half> undef)
+; ZVFHMIN-NEXT: Cost Model: Invalid cost for instruction: %7 = call <vscale x 2 x half> @llvm.pow.nxv2f16(<vscale x 2 x half> undef, <vscale x 2 x half> undef)
+; ZVFHMIN-NEXT: Cost Model: Invalid cost for instruction: %8 = call <vscale x 4 x half> @llvm.pow.nxv4f16(<vscale x 4 x half> undef, <vscale x 4 x half> undef)
+; ZVFHMIN-NEXT: Cost Model: Invalid cost for instruction: %9 = call <vscale x 8 x half> @llvm.pow.nxv8f16(<vscale x 8 x half> undef, <vscale x 8 x half> undef)
+; ZVFHMIN-NEXT: Cost Model: Invalid cost for instruction: %10 = call <vscale x 16 x half> @llvm.pow.nxv16f16(<vscale x 16 x half> undef, <vscale x 16 x half> undef)
; ZVFHMIN-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret void
;
call half @llvm.pow.f16(half undef, half undef)
diff --git a/llvm/test/Analysis/CostModel/RISCV/fp-trig-log-exp.ll b/llvm/test/Analysis/CostModel/RISCV/fp-trig-log-exp.ll
index d65fa43b89952e..34d6c93f4577a7 100644
--- a/llvm/test/Analysis/CostModel/RISCV/fp-trig-log-exp.ll
+++ b/llvm/test/Analysis/CostModel/RISCV/fp-trig-log-exp.ll
@@ -9,11 +9,11 @@ define void @sin() {
; CHECK-NEXT: Cost Model: Found an estimated cost of 44 for instruction: %3 = call <4 x bfloat> @llvm.sin.v4bf16(<4 x bfloat> undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 88 for instruction: %4 = call <8 x bfloat> @llvm.sin.v8bf16(<8 x bfloat> undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 191 for instruction: %5 = call <16 x bfloat> @llvm.sin.v16bf16(<16 x bfloat> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %6 = call <vscale x 1 x bfloat> @llvm.sin.nxv1bf16(<vscale x 1 x bfloat> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %7 = call <vscale x 2 x bfloat> @llvm.sin.nxv2bf16(<vscale x 2 x bfloat> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %8 = call <vscale x 4 x bfloat> @llvm.sin.nxv4bf16(<vscale x 4 x bfloat> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %9 = call <vscale x 8 x bfloat> @llvm.sin.nxv8bf16(<vscale x 8 x bfloat> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %10 = call <vscale x 16 x bfloat> @llvm.sin.nxv16bf16(<vscale x 16 x bfloat> undef)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %6 = call <vscale x 1 x bfloat> @llvm.sin.nxv1bf16(<vscale x 1 x bfloat> undef)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %7 = call <vscale x 2 x bfloat> @llvm.sin.nxv2bf16(<vscale x 2 x bfloat> undef)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %8 = call <vscale x 4 x bfloat> @llvm.sin.nxv4bf16(<vscale x 4 x bfloat> undef)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %9 = call <vscale x 8 x bfloat> @llvm.sin.nxv8bf16(<vscale x 8 x bfloat> undef)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %10 = call <vscale x 16 x bfloat> @llvm.sin.nxv16bf16(<vscale x 16 x bfloat> undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 10 for instruction: %11 = call float @llvm.sin.f32(float undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 22 for instruction: %12 = call <2 x float> @llvm.sin.v2f32(<2 x float> undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 44 for instruction: %13 = call <4 x float> @llvm.sin.v4f32(<4 x float> undef)
@@ -86,10 +86,10 @@ define void @sin_f16() {
; ZVFHMIN-NEXT: Cost Model: Found an estimated cost of 44 for instruction: %3 = call <4 x half> @llvm.sin.v4f16(<4 x half> undef)
; ZVFHMIN-NEXT: Cost Model: Found an estimated cost of 88 for instruction: %4 = call <8 x half> @llvm.sin.v8f16(<8 x half> undef)
; ZVFHMIN-NEXT: Cost Model: Found an estimated cost of 191 for instruction: %5 = call <16 x half> @llvm.sin.v16f16(<16 x half> undef)
-; ZVFHMIN-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %6 = call <vscale x 1 x half> @llvm.sin.nxv1f16(<vscale x 1 x half> undef)
-; ZVFHMIN-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %7 = call <vscale x 2 x half> @llvm.sin.nxv2f16(<vscale x 2 x half> undef)
-; ZVFHMIN-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %8 = call <vscale x 4 x half> @llvm.sin.nxv4f16(<vscale x 4 x half> undef)
-; ZVFHMIN-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %9 = call <vscale x 8 x half> @llvm.sin.nxv8f16(<vscale x 8 x half> undef)
+; ZVFHMIN-NEXT: Cost Model: Invalid cost for instruction: %6 = call <vscale x 1 x half> @llvm.sin.nxv1f16(<vscale x 1 x half> undef)
+; ZVFHMIN-NEXT: Cost Model: Invalid cost for instruction: %7 = call <vscale x 2 x half> @llvm.sin.nxv2f16(<vscale x 2 x half> undef)
+; ZVFHMIN-NEXT: Cost Model: Invalid cost for instruction: %8 = call <vscale x 4 x half> @llvm.sin.nxv4f16(<vscale x 4 x half> undef)
+; ZVFHMIN-NEXT: Cost Model: Invalid cost for instruction: %9 = call <vscale x 8 x half> @llvm.sin.nxv8f16(<vscale x 8 x half> undef)
; ZVFHMIN-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret void
;
call half @llvm.sin.f16(half undef)
@@ -111,11 +111,11 @@ define void @cos() {
; CHECK-NEXT: Cost Model: Found an estimated cost of 44 for instruction: %3 = call <4 x bfloat> @llvm.cos.v4bf16(<4 x bfloat> undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 88 for instruction: %4 = call <8 x bfloat> @llvm.cos.v8bf16(<8 x bfloat> undef)
; CHECK-NEXT: Cost Model: Found an estimated cost of 191 for instruction: %5 = call <16 x bfloat> @llvm.cos.v16bf16(<16 x bfloat> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %6 = call <vscale x 1 x bfloat> @llvm.cos.nxv1bf16(<vscale x 1 x bfloat> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %7 = call <vscale x 2 x bfloat> @llvm.cos.nxv2bf16(<vscale x 2 x bfloat> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %8 = call <vscale x 4 x bfloat> @llvm.cos.nxv4bf16(<vscale x 4 x bfloat> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %9 = call <vscale x 8 x bfloat> @llvm.cos.nxv8bf16(<vscale x 8 x bfloat> undef)
-; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %10 = call <vscale x 16 x bfloat> @llvm.cos.nxv16bf16(<vscale x 16 x bfloat> undef)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %6 = call <vscale x 1 x bfloat> @llvm.cos.nxv1bf16(<vscale x 1 x bfloat> undef)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %7 = call <vscale x 2 x bfloat> @llvm.cos.nxv2bf16(<vscale x 2 x bfloat> undef)
+; CHECK-NEXT: Cost Model: Invalid cost for instruction: %8 = call <vs...
[truncated]
|
wangpc-pp
reviewed
Oct 16, 2024
I thought the same thing as well, but I couldn't find any existing CodeGen tests to extend. I guess the original commit that set them to expand was mainly testing that the vectorizers wouldn't emit any of these. |
You can't really write a CodeGen test for them unless you use "not" and check for the error message. Without this patch they probably get a "cannot select" error from isel. With this patch they'll fail when LegalizeDAG tries to unroll them which can't be done for a scalable vector. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
For regular floating point types we mark these as expanded on scalable vectors so they're not legal in the cost model, so this does the same for f16 w/ zvfhmin and bf16.