[llvm] 5b395bd - [CostModel][X86] Add costs for multiply-by-pow2 constants
Simon Pilgrim via llvm-commits
llvm-commits at lists.llvm.org
Wed Oct 20 05:11:45 PDT 2021
Author: Simon Pilgrim
Date: 2021-10-20T13:11:21+01:00
New Revision: 5b395bd633632b9f87f4d853095cb1a090a1efe6
URL: https://github.com/llvm/llvm-project/commit/5b395bd633632b9f87f4d853095cb1a090a1efe6
DIFF: https://github.com/llvm/llvm-project/commit/5b395bd633632b9f87f4d853095cb1a090a1efe6.diff
LOG: [CostModel][X86] Add costs for multiply-by-pow2 constants
These are folded to left shifts in the backend.
We should be able to extend this for multiply-by-negpow2 after D111968 has landed to resolve PR51436
Added:
Modified:
llvm/lib/Target/X86/X86TargetTransformInfo.cpp
llvm/test/Analysis/CostModel/X86/mul.ll
Removed:
################################################################################
diff --git a/llvm/lib/Target/X86/X86TargetTransformInfo.cpp b/llvm/lib/Target/X86/X86TargetTransformInfo.cpp
index 1a3828f78950..3488e3834990 100644
--- a/llvm/lib/Target/X86/X86TargetTransformInfo.cpp
+++ b/llvm/lib/Target/X86/X86TargetTransformInfo.cpp
@@ -236,11 +236,19 @@ InstructionCost X86TTIImpl::getArithmeticInstrCost(
}
}
- if ((ISD == ISD::SDIV || ISD == ISD::SREM || ISD == ISD::UDIV ||
- ISD == ISD::UREM) &&
+ if ((ISD == ISD::MUL || ISD == ISD::SDIV || ISD == ISD::SREM ||
+ ISD == ISD::UDIV || ISD == ISD::UREM) &&
(Op2Info == TargetTransformInfo::OK_UniformConstantValue ||
Op2Info == TargetTransformInfo::OK_NonUniformConstantValue) &&
Opd2PropInfo == TargetTransformInfo::OP_PowerOf2) {
+ // Vector multiply by pow2 will be simplified to shifts.
+ if (ISD == ISD::MUL) {
+ InstructionCost Cost = getArithmeticInstrCost(
+ Instruction::Shl, Ty, CostKind, Op1Info, Op2Info,
+ TargetTransformInfo::OP_None, TargetTransformInfo::OP_None);
+ return Cost;
+ }
+
if (ISD == ISD::SDIV || ISD == ISD::SREM) {
// On X86, vector signed division by constants power-of-two are
// normally expanded to the sequence SRA + SRL + ADD + SRA.
diff --git a/llvm/test/Analysis/CostModel/X86/mul.ll b/llvm/test/Analysis/CostModel/X86/mul.ll
index efa3cc2036d2..3213af05f6ae 100644
--- a/llvm/test/Analysis/CostModel/X86/mul.ll
+++ b/llvm/test/Analysis/CostModel/X86/mul.ll
@@ -13,10 +13,10 @@
define i32 @mul_constpow2() {
; SSE2-LABEL: 'mul_constpow2'
-; SSE2-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %I64 = mul i64 undef, 16
-; SSE2-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V2i64 = mul <2 x i64> undef, <i64 8, i64 16>
-; SSE2-NEXT: Cost Model: Found an estimated cost of 16 for instruction: %V4i64 = mul <4 x i64> undef, <i64 2, i64 4, i64 8, i64 16>
-; SSE2-NEXT: Cost Model: Found an estimated cost of 32 for instruction: %V8i64 = mul <8 x i64> undef, <i64 2, i64 4, i64 8, i64 16, i64 32, i64 64, i64 128, i64 256>
+; SSE2-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %I64 = mul i64 undef, 16
+; SSE2-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V2i64 = mul <2 x i64> undef, <i64 8, i64 16>
+; SSE2-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V4i64 = mul <4 x i64> undef, <i64 2, i64 4, i64 8, i64 16>
+; SSE2-NEXT: Cost Model: Found an estimated cost of 16 for instruction: %V8i64 = mul <8 x i64> undef, <i64 2, i64 4, i64 8, i64 16, i64 32, i64 64, i64 128, i64 256>
; SSE2-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %I32 = mul i32 undef, 16
; SSE2-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V4i32 = mul <4 x i32> undef, <i32 2, i32 4, i32 8, i32 16>
; SSE2-NEXT: Cost Model: Found an estimated cost of 12 for instruction: %V8i32 = mul <8 x i32> undef, <i32 2, i32 4, i32 8, i32 16, i32 32, i32 64, i32 128, i32 256>
@@ -32,10 +32,10 @@ define i32 @mul_constpow2() {
; SSE2-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
; SSSE3-LABEL: 'mul_constpow2'
-; SSSE3-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %I64 = mul i64 undef, 16
-; SSSE3-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V2i64 = mul <2 x i64> undef, <i64 8, i64 16>
-; SSSE3-NEXT: Cost Model: Found an estimated cost of 16 for instruction: %V4i64 = mul <4 x i64> undef, <i64 2, i64 4, i64 8, i64 16>
-; SSSE3-NEXT: Cost Model: Found an estimated cost of 32 for instruction: %V8i64 = mul <8 x i64> undef, <i64 2, i64 4, i64 8, i64 16, i64 32, i64 64, i64 128, i64 256>
+; SSSE3-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %I64 = mul i64 undef, 16
+; SSSE3-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V2i64 = mul <2 x i64> undef, <i64 8, i64 16>
+; SSSE3-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V4i64 = mul <4 x i64> undef, <i64 2, i64 4, i64 8, i64 16>
+; SSSE3-NEXT: Cost Model: Found an estimated cost of 16 for instruction: %V8i64 = mul <8 x i64> undef, <i64 2, i64 4, i64 8, i64 16, i64 32, i64 64, i64 128, i64 256>
; SSSE3-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %I32 = mul i32 undef, 16
; SSSE3-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V4i32 = mul <4 x i32> undef, <i32 2, i32 4, i32 8, i32 16>
; SSSE3-NEXT: Cost Model: Found an estimated cost of 12 for instruction: %V8i32 = mul <8 x i32> undef, <i32 2, i32 4, i32 8, i32 16, i32 32, i32 64, i32 128, i32 256>
@@ -51,10 +51,10 @@ define i32 @mul_constpow2() {
; SSSE3-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
; SSE42-LABEL: 'mul_constpow2'
-; SSE42-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %I64 = mul i64 undef, 16
-; SSE42-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V2i64 = mul <2 x i64> undef, <i64 8, i64 16>
-; SSE42-NEXT: Cost Model: Found an estimated cost of 12 for instruction: %V4i64 = mul <4 x i64> undef, <i64 2, i64 4, i64 8, i64 16>
-; SSE42-NEXT: Cost Model: Found an estimated cost of 24 for instruction: %V8i64 = mul <8 x i64> undef, <i64 2, i64 4, i64 8, i64 16, i64 32, i64 64, i64 128, i64 256>
+; SSE42-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %I64 = mul i64 undef, 16
+; SSE42-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V2i64 = mul <2 x i64> undef, <i64 8, i64 16>
+; SSE42-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V4i64 = mul <4 x i64> undef, <i64 2, i64 4, i64 8, i64 16>
+; SSE42-NEXT: Cost Model: Found an estimated cost of 16 for instruction: %V8i64 = mul <8 x i64> undef, <i64 2, i64 4, i64 8, i64 16, i64 32, i64 64, i64 128, i64 256>
; SSE42-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %I32 = mul i32 undef, 16
; SSE42-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V4i32 = mul <4 x i32> undef, <i32 2, i32 4, i32 8, i32 16>
; SSE42-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V8i32 = mul <8 x i32> undef, <i32 2, i32 4, i32 8, i32 16, i32 32, i32 64, i32 128, i32 256>
@@ -70,10 +70,10 @@ define i32 @mul_constpow2() {
; SSE42-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
; AVX1-LABEL: 'mul_constpow2'
-; AVX1-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %I64 = mul i64 undef, 16
-; AVX1-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V2i64 = mul <2 x i64> undef, <i64 8, i64 16>
-; AVX1-NEXT: Cost Model: Found an estimated cost of 12 for instruction: %V4i64 = mul <4 x i64> undef, <i64 2, i64 4, i64 8, i64 16>
-; AVX1-NEXT: Cost Model: Found an estimated cost of 24 for instruction: %V8i64 = mul <8 x i64> undef, <i64 2, i64 4, i64 8, i64 16, i64 32, i64 64, i64 128, i64 256>
+; AVX1-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %I64 = mul i64 undef, 16
+; AVX1-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V2i64 = mul <2 x i64> undef, <i64 8, i64 16>
+; AVX1-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V4i64 = mul <4 x i64> undef, <i64 2, i64 4, i64 8, i64 16>
+; AVX1-NEXT: Cost Model: Found an estimated cost of 12 for instruction: %V8i64 = mul <8 x i64> undef, <i64 2, i64 4, i64 8, i64 16, i64 32, i64 64, i64 128, i64 256>
; AVX1-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %I32 = mul i32 undef, 16
; AVX1-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V4i32 = mul <4 x i32> undef, <i32 2, i32 4, i32 8, i32 16>
; AVX1-NEXT: Cost Model: Found an estimated cost of 5 for instruction: %V8i32 = mul <8 x i32> undef, <i32 2, i32 4, i32 8, i32 16, i32 32, i32 64, i32 128, i32 256>
@@ -89,10 +89,10 @@ define i32 @mul_constpow2() {
; AVX1-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
; AVX2-LABEL: 'mul_constpow2'
-; AVX2-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %I64 = mul i64 undef, 16
-; AVX2-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V2i64 = mul <2 x i64> undef, <i64 8, i64 16>
-; AVX2-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V4i64 = mul <4 x i64> undef, <i64 2, i64 4, i64 8, i64 16>
-; AVX2-NEXT: Cost Model: Found an estimated cost of 12 for instruction: %V8i64 = mul <8 x i64> undef, <i64 2, i64 4, i64 8, i64 16, i64 32, i64 64, i64 128, i64 256>
+; AVX2-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %I64 = mul i64 undef, 16
+; AVX2-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2i64 = mul <2 x i64> undef, <i64 8, i64 16>
+; AVX2-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4i64 = mul <4 x i64> undef, <i64 2, i64 4, i64 8, i64 16>
+; AVX2-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V8i64 = mul <8 x i64> undef, <i64 2, i64 4, i64 8, i64 16, i64 32, i64 64, i64 128, i64 256>
; AVX2-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %I32 = mul i32 undef, 16
; AVX2-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V4i32 = mul <4 x i32> undef, <i32 2, i32 4, i32 8, i32 16>
; AVX2-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V8i32 = mul <8 x i32> undef, <i32 2, i32 4, i32 8, i32 16, i32 32, i32 64, i32 128, i32 256>
@@ -108,10 +108,10 @@ define i32 @mul_constpow2() {
; AVX2-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
; AVX512F-LABEL: 'mul_constpow2'
-; AVX512F-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %I64 = mul i64 undef, 16
-; AVX512F-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V2i64 = mul <2 x i64> undef, <i64 8, i64 16>
-; AVX512F-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V4i64 = mul <4 x i64> undef, <i64 2, i64 4, i64 8, i64 16>
-; AVX512F-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V8i64 = mul <8 x i64> undef, <i64 2, i64 4, i64 8, i64 16, i64 32, i64 64, i64 128, i64 256>
+; AVX512F-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %I64 = mul i64 undef, 16
+; AVX512F-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2i64 = mul <2 x i64> undef, <i64 8, i64 16>
+; AVX512F-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4i64 = mul <4 x i64> undef, <i64 2, i64 4, i64 8, i64 16>
+; AVX512F-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V8i64 = mul <8 x i64> undef, <i64 2, i64 4, i64 8, i64 16, i64 32, i64 64, i64 128, i64 256>
; AVX512F-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %I32 = mul i32 undef, 16
; AVX512F-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4i32 = mul <4 x i32> undef, <i32 2, i32 4, i32 8, i32 16>
; AVX512F-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V8i32 = mul <8 x i32> undef, <i32 2, i32 4, i32 8, i32 16, i32 32, i32 64, i32 128, i32 256>
@@ -127,10 +127,10 @@ define i32 @mul_constpow2() {
; AVX512F-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
; AVX512BW-LABEL: 'mul_constpow2'
-; AVX512BW-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %I64 = mul i64 undef, 16
-; AVX512BW-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V2i64 = mul <2 x i64> undef, <i64 8, i64 16>
-; AVX512BW-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V4i64 = mul <4 x i64> undef, <i64 2, i64 4, i64 8, i64 16>
-; AVX512BW-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V8i64 = mul <8 x i64> undef, <i64 2, i64 4, i64 8, i64 16, i64 32, i64 64, i64 128, i64 256>
+; AVX512BW-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %I64 = mul i64 undef, 16
+; AVX512BW-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2i64 = mul <2 x i64> undef, <i64 8, i64 16>
+; AVX512BW-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4i64 = mul <4 x i64> undef, <i64 2, i64 4, i64 8, i64 16>
+; AVX512BW-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V8i64 = mul <8 x i64> undef, <i64 2, i64 4, i64 8, i64 16, i64 32, i64 64, i64 128, i64 256>
; AVX512BW-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %I32 = mul i32 undef, 16
; AVX512BW-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4i32 = mul <4 x i32> undef, <i32 2, i32 4, i32 8, i32 16>
; AVX512BW-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V8i32 = mul <8 x i32> undef, <i32 2, i32 4, i32 8, i32 16, i32 32, i32 64, i32 128, i32 256>
@@ -146,22 +146,22 @@ define i32 @mul_constpow2() {
; AVX512BW-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
; SLM-LABEL: 'mul_constpow2'
-; SLM-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %I64 = mul i64 undef, 16
-; SLM-NEXT: Cost Model: Found an estimated cost of 17 for instruction: %V2i64 = mul <2 x i64> undef, <i64 8, i64 16>
-; SLM-NEXT: Cost Model: Found an estimated cost of 34 for instruction: %V4i64 = mul <4 x i64> undef, <i64 2, i64 4, i64 8, i64 16>
-; SLM-NEXT: Cost Model: Found an estimated cost of 68 for instruction: %V8i64 = mul <8 x i64> undef, <i64 2, i64 4, i64 8, i64 16, i64 32, i64 64, i64 128, i64 256>
+; SLM-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %I64 = mul i64 undef, 16
+; SLM-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V2i64 = mul <2 x i64> undef, <i64 8, i64 16>
+; SLM-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V4i64 = mul <4 x i64> undef, <i64 2, i64 4, i64 8, i64 16>
+; SLM-NEXT: Cost Model: Found an estimated cost of 16 for instruction: %V8i64 = mul <8 x i64> undef, <i64 2, i64 4, i64 8, i64 16, i64 32, i64 64, i64 128, i64 256>
; SLM-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %I32 = mul i32 undef, 16
-; SLM-NEXT: Cost Model: Found an estimated cost of 11 for instruction: %V4i32 = mul <4 x i32> undef, <i32 2, i32 4, i32 8, i32 16>
-; SLM-NEXT: Cost Model: Found an estimated cost of 22 for instruction: %V8i32 = mul <8 x i32> undef, <i32 2, i32 4, i32 8, i32 16, i32 32, i32 64, i32 128, i32 256>
-; SLM-NEXT: Cost Model: Found an estimated cost of 44 for instruction: %V16i32 = mul <16 x i32> undef, <i32 2, i32 4, i32 8, i32 16, i32 32, i32 64, i32 128, i32 256, i32 2, i32 4, i32 8, i32 16, i32 32, i32 64, i32 128, i32 256>
+; SLM-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V4i32 = mul <4 x i32> undef, <i32 2, i32 4, i32 8, i32 16>
+; SLM-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V8i32 = mul <8 x i32> undef, <i32 2, i32 4, i32 8, i32 16, i32 32, i32 64, i32 128, i32 256>
+; SLM-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V16i32 = mul <16 x i32> undef, <i32 2, i32 4, i32 8, i32 16, i32 32, i32 64, i32 128, i32 256, i32 2, i32 4, i32 8, i32 16, i32 32, i32 64, i32 128, i32 256>
; SLM-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %I16 = mul i16 undef, 16
-; SLM-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V8i16 = mul <8 x i16> undef, <i16 2, i16 4, i16 8, i16 16, i16 32, i16 64, i16 128, i16 256>
-; SLM-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V16i16 = mul <16 x i16> undef, <i16 2, i16 4, i16 8, i16 16, i16 32, i16 64, i16 128, i16 256, i16 2, i16 4, i16 8, i16 16, i16 32, i16 64, i16 128, i16 256>
-; SLM-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V32i16 = mul <32 x i16> undef, <i16 2, i16 4, i16 8, i16 16, i16 32, i16 64, i16 128, i16 256, i16 2, i16 4, i16 8, i16 16, i16 32, i16 64, i16 128, i16 256, i16 2, i16 4, i16 8, i16 16, i16 32, i16 64, i16 128, i16 256, i16 2, i16 4, i16 8, i16 16, i16 32, i16 64, i16 128, i16 256>
+; SLM-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V8i16 = mul <8 x i16> undef, <i16 2, i16 4, i16 8, i16 16, i16 32, i16 64, i16 128, i16 256>
+; SLM-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V16i16 = mul <16 x i16> undef, <i16 2, i16 4, i16 8, i16 16, i16 32, i16 64, i16 128, i16 256, i16 2, i16 4, i16 8, i16 16, i16 32, i16 64, i16 128, i16 256>
+; SLM-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V32i16 = mul <32 x i16> undef, <i16 2, i16 4, i16 8, i16 16, i16 32, i16 64, i16 128, i16 256, i16 2, i16 4, i16 8, i16 16, i16 32, i16 64, i16 128, i16 256, i16 2, i16 4, i16 8, i16 16, i16 32, i16 64, i16 128, i16 256, i16 2, i16 4, i16 8, i16 16, i16 32, i16 64, i16 128, i16 256>
; SLM-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %I8 = mul i8 undef, 16
-; SLM-NEXT: Cost Model: Found an estimated cost of 9 for instruction: %V16i8 = mul <16 x i8> undef, <i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16>
-; SLM-NEXT: Cost Model: Found an estimated cost of 20 for instruction: %V32i8 = mul <32 x i8> undef, <i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16>
-; SLM-NEXT: Cost Model: Found an estimated cost of 40 for instruction: %V64i8 = mul <64 x i8> undef, <i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16>
+; SLM-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %V16i8 = mul <16 x i8> undef, <i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16>
+; SLM-NEXT: Cost Model: Found an estimated cost of 16 for instruction: %V32i8 = mul <32 x i8> undef, <i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16>
+; SLM-NEXT: Cost Model: Found an estimated cost of 32 for instruction: %V64i8 = mul <64 x i8> undef, <i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16>
; SLM-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
%I64 = mul i64 undef, 16
@@ -189,14 +189,14 @@ define i32 @mul_constpow2() {
define i32 @mul_uniformconstpow2() {
; SSE2-LABEL: 'mul_uniformconstpow2'
-; SSE2-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %I64 = mul i64 undef, 16
-; SSE2-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V2i64 = mul <2 x i64> undef, <i64 16, i64 16>
-; SSE2-NEXT: Cost Model: Found an estimated cost of 16 for instruction: %V4i64 = mul <4 x i64> undef, <i64 16, i64 16, i64 16, i64 16>
-; SSE2-NEXT: Cost Model: Found an estimated cost of 32 for instruction: %V8i64 = mul <8 x i64> undef, <i64 16, i64 16, i64 16, i64 16, i64 16, i64 16, i64 16, i64 16>
+; SSE2-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %I64 = mul i64 undef, 16
+; SSE2-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2i64 = mul <2 x i64> undef, <i64 16, i64 16>
+; SSE2-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V4i64 = mul <4 x i64> undef, <i64 16, i64 16, i64 16, i64 16>
+; SSE2-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V8i64 = mul <8 x i64> undef, <i64 16, i64 16, i64 16, i64 16, i64 16, i64 16, i64 16, i64 16>
; SSE2-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %I32 = mul i32 undef, 16
-; SSE2-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V4i32 = mul <4 x i32> undef, <i32 16, i32 16, i32 16, i32 16>
-; SSE2-NEXT: Cost Model: Found an estimated cost of 12 for instruction: %V8i32 = mul <8 x i32> undef, <i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16>
-; SSE2-NEXT: Cost Model: Found an estimated cost of 24 for instruction: %V16i32 = mul <16 x i32> undef, <i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16>
+; SSE2-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4i32 = mul <4 x i32> undef, <i32 16, i32 16, i32 16, i32 16>
+; SSE2-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V8i32 = mul <8 x i32> undef, <i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16>
+; SSE2-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V16i32 = mul <16 x i32> undef, <i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16>
; SSE2-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %I16 = mul i16 undef, 16
; SSE2-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V8i16 = mul <8 x i16> undef, <i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16>
; SSE2-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V16i16 = mul <16 x i16> undef, <i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16>
@@ -208,14 +208,14 @@ define i32 @mul_uniformconstpow2() {
; SSE2-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
; SSSE3-LABEL: 'mul_uniformconstpow2'
-; SSSE3-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %I64 = mul i64 undef, 16
-; SSSE3-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V2i64 = mul <2 x i64> undef, <i64 16, i64 16>
-; SSSE3-NEXT: Cost Model: Found an estimated cost of 16 for instruction: %V4i64 = mul <4 x i64> undef, <i64 16, i64 16, i64 16, i64 16>
-; SSSE3-NEXT: Cost Model: Found an estimated cost of 32 for instruction: %V8i64 = mul <8 x i64> undef, <i64 16, i64 16, i64 16, i64 16, i64 16, i64 16, i64 16, i64 16>
+; SSSE3-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %I64 = mul i64 undef, 16
+; SSSE3-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2i64 = mul <2 x i64> undef, <i64 16, i64 16>
+; SSSE3-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V4i64 = mul <4 x i64> undef, <i64 16, i64 16, i64 16, i64 16>
+; SSSE3-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V8i64 = mul <8 x i64> undef, <i64 16, i64 16, i64 16, i64 16, i64 16, i64 16, i64 16, i64 16>
; SSSE3-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %I32 = mul i32 undef, 16
-; SSSE3-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V4i32 = mul <4 x i32> undef, <i32 16, i32 16, i32 16, i32 16>
-; SSSE3-NEXT: Cost Model: Found an estimated cost of 12 for instruction: %V8i32 = mul <8 x i32> undef, <i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16>
-; SSSE3-NEXT: Cost Model: Found an estimated cost of 24 for instruction: %V16i32 = mul <16 x i32> undef, <i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16>
+; SSSE3-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4i32 = mul <4 x i32> undef, <i32 16, i32 16, i32 16, i32 16>
+; SSSE3-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V8i32 = mul <8 x i32> undef, <i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16>
+; SSSE3-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V16i32 = mul <16 x i32> undef, <i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16>
; SSSE3-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %I16 = mul i16 undef, 16
; SSSE3-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V8i16 = mul <8 x i16> undef, <i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16>
; SSSE3-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V16i16 = mul <16 x i16> undef, <i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16>
@@ -227,14 +227,14 @@ define i32 @mul_uniformconstpow2() {
; SSSE3-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
; SSE42-LABEL: 'mul_uniformconstpow2'
-; SSE42-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %I64 = mul i64 undef, 16
-; SSE42-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V2i64 = mul <2 x i64> undef, <i64 16, i64 16>
-; SSE42-NEXT: Cost Model: Found an estimated cost of 12 for instruction: %V4i64 = mul <4 x i64> undef, <i64 16, i64 16, i64 16, i64 16>
-; SSE42-NEXT: Cost Model: Found an estimated cost of 24 for instruction: %V8i64 = mul <8 x i64> undef, <i64 16, i64 16, i64 16, i64 16, i64 16, i64 16, i64 16, i64 16>
+; SSE42-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %I64 = mul i64 undef, 16
+; SSE42-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2i64 = mul <2 x i64> undef, <i64 16, i64 16>
+; SSE42-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V4i64 = mul <4 x i64> undef, <i64 16, i64 16, i64 16, i64 16>
+; SSE42-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V8i64 = mul <8 x i64> undef, <i64 16, i64 16, i64 16, i64 16, i64 16, i64 16, i64 16, i64 16>
; SSE42-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %I32 = mul i32 undef, 16
-; SSE42-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V4i32 = mul <4 x i32> undef, <i32 16, i32 16, i32 16, i32 16>
-; SSE42-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V8i32 = mul <8 x i32> undef, <i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16>
-; SSE42-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V16i32 = mul <16 x i32> undef, <i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16>
+; SSE42-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4i32 = mul <4 x i32> undef, <i32 16, i32 16, i32 16, i32 16>
+; SSE42-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V8i32 = mul <8 x i32> undef, <i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16>
+; SSE42-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V16i32 = mul <16 x i32> undef, <i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16>
; SSE42-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %I16 = mul i16 undef, 16
; SSE42-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V8i16 = mul <8 x i16> undef, <i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16>
; SSE42-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V16i16 = mul <16 x i16> undef, <i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16>
@@ -246,14 +246,14 @@ define i32 @mul_uniformconstpow2() {
; SSE42-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
; AVX1-LABEL: 'mul_uniformconstpow2'
-; AVX1-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %I64 = mul i64 undef, 16
-; AVX1-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V2i64 = mul <2 x i64> undef, <i64 16, i64 16>
-; AVX1-NEXT: Cost Model: Found an estimated cost of 12 for instruction: %V4i64 = mul <4 x i64> undef, <i64 16, i64 16, i64 16, i64 16>
-; AVX1-NEXT: Cost Model: Found an estimated cost of 24 for instruction: %V8i64 = mul <8 x i64> undef, <i64 16, i64 16, i64 16, i64 16, i64 16, i64 16, i64 16, i64 16>
+; AVX1-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %I64 = mul i64 undef, 16
+; AVX1-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2i64 = mul <2 x i64> undef, <i64 16, i64 16>
+; AVX1-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V4i64 = mul <4 x i64> undef, <i64 16, i64 16, i64 16, i64 16>
+; AVX1-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V8i64 = mul <8 x i64> undef, <i64 16, i64 16, i64 16, i64 16, i64 16, i64 16, i64 16, i64 16>
; AVX1-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %I32 = mul i32 undef, 16
-; AVX1-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V4i32 = mul <4 x i32> undef, <i32 16, i32 16, i32 16, i32 16>
-; AVX1-NEXT: Cost Model: Found an estimated cost of 5 for instruction: %V8i32 = mul <8 x i32> undef, <i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16>
-; AVX1-NEXT: Cost Model: Found an estimated cost of 10 for instruction: %V16i32 = mul <16 x i32> undef, <i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16>
+; AVX1-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4i32 = mul <4 x i32> undef, <i32 16, i32 16, i32 16, i32 16>
+; AVX1-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V8i32 = mul <8 x i32> undef, <i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16>
+; AVX1-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V16i32 = mul <16 x i32> undef, <i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16>
; AVX1-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %I16 = mul i16 undef, 16
; AVX1-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V8i16 = mul <8 x i16> undef, <i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16>
; AVX1-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V16i16 = mul <16 x i16> undef, <i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16>
@@ -265,14 +265,14 @@ define i32 @mul_uniformconstpow2() {
; AVX1-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
; AVX2-LABEL: 'mul_uniformconstpow2'
-; AVX2-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %I64 = mul i64 undef, 16
-; AVX2-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V2i64 = mul <2 x i64> undef, <i64 16, i64 16>
-; AVX2-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V4i64 = mul <4 x i64> undef, <i64 16, i64 16, i64 16, i64 16>
-; AVX2-NEXT: Cost Model: Found an estimated cost of 12 for instruction: %V8i64 = mul <8 x i64> undef, <i64 16, i64 16, i64 16, i64 16, i64 16, i64 16, i64 16, i64 16>
+; AVX2-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %I64 = mul i64 undef, 16
+; AVX2-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2i64 = mul <2 x i64> undef, <i64 16, i64 16>
+; AVX2-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4i64 = mul <4 x i64> undef, <i64 16, i64 16, i64 16, i64 16>
+; AVX2-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V8i64 = mul <8 x i64> undef, <i64 16, i64 16, i64 16, i64 16, i64 16, i64 16, i64 16, i64 16>
; AVX2-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %I32 = mul i32 undef, 16
-; AVX2-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V4i32 = mul <4 x i32> undef, <i32 16, i32 16, i32 16, i32 16>
-; AVX2-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V8i32 = mul <8 x i32> undef, <i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16>
-; AVX2-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V16i32 = mul <16 x i32> undef, <i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16>
+; AVX2-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4i32 = mul <4 x i32> undef, <i32 16, i32 16, i32 16, i32 16>
+; AVX2-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V8i32 = mul <8 x i32> undef, <i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16>
+; AVX2-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V16i32 = mul <16 x i32> undef, <i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16>
; AVX2-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %I16 = mul i16 undef, 16
; AVX2-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V8i16 = mul <8 x i16> undef, <i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16>
; AVX2-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V16i16 = mul <16 x i16> undef, <i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16>
@@ -284,10 +284,10 @@ define i32 @mul_uniformconstpow2() {
; AVX2-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
; AVX512F-LABEL: 'mul_uniformconstpow2'
-; AVX512F-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %I64 = mul i64 undef, 16
-; AVX512F-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V2i64 = mul <2 x i64> undef, <i64 16, i64 16>
-; AVX512F-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V4i64 = mul <4 x i64> undef, <i64 16, i64 16, i64 16, i64 16>
-; AVX512F-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V8i64 = mul <8 x i64> undef, <i64 16, i64 16, i64 16, i64 16, i64 16, i64 16, i64 16, i64 16>
+; AVX512F-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %I64 = mul i64 undef, 16
+; AVX512F-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2i64 = mul <2 x i64> undef, <i64 16, i64 16>
+; AVX512F-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4i64 = mul <4 x i64> undef, <i64 16, i64 16, i64 16, i64 16>
+; AVX512F-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V8i64 = mul <8 x i64> undef, <i64 16, i64 16, i64 16, i64 16, i64 16, i64 16, i64 16, i64 16>
; AVX512F-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %I32 = mul i32 undef, 16
; AVX512F-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4i32 = mul <4 x i32> undef, <i32 16, i32 16, i32 16, i32 16>
; AVX512F-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V8i32 = mul <8 x i32> undef, <i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16>
@@ -303,10 +303,10 @@ define i32 @mul_uniformconstpow2() {
; AVX512F-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
; AVX512BW-LABEL: 'mul_uniformconstpow2'
-; AVX512BW-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %I64 = mul i64 undef, 16
-; AVX512BW-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V2i64 = mul <2 x i64> undef, <i64 16, i64 16>
-; AVX512BW-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V4i64 = mul <4 x i64> undef, <i64 16, i64 16, i64 16, i64 16>
-; AVX512BW-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V8i64 = mul <8 x i64> undef, <i64 16, i64 16, i64 16, i64 16, i64 16, i64 16, i64 16, i64 16>
+; AVX512BW-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %I64 = mul i64 undef, 16
+; AVX512BW-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2i64 = mul <2 x i64> undef, <i64 16, i64 16>
+; AVX512BW-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4i64 = mul <4 x i64> undef, <i64 16, i64 16, i64 16, i64 16>
+; AVX512BW-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V8i64 = mul <8 x i64> undef, <i64 16, i64 16, i64 16, i64 16, i64 16, i64 16, i64 16, i64 16>
; AVX512BW-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %I32 = mul i32 undef, 16
; AVX512BW-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4i32 = mul <4 x i32> undef, <i32 16, i32 16, i32 16, i32 16>
; AVX512BW-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V8i32 = mul <8 x i32> undef, <i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16>
@@ -322,22 +322,22 @@ define i32 @mul_uniformconstpow2() {
; AVX512BW-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
; SLM-LABEL: 'mul_uniformconstpow2'
-; SLM-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %I64 = mul i64 undef, 16
-; SLM-NEXT: Cost Model: Found an estimated cost of 17 for instruction: %V2i64 = mul <2 x i64> undef, <i64 16, i64 16>
-; SLM-NEXT: Cost Model: Found an estimated cost of 34 for instruction: %V4i64 = mul <4 x i64> undef, <i64 16, i64 16, i64 16, i64 16>
-; SLM-NEXT: Cost Model: Found an estimated cost of 68 for instruction: %V8i64 = mul <8 x i64> undef, <i64 16, i64 16, i64 16, i64 16, i64 16, i64 16, i64 16, i64 16>
+; SLM-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %I64 = mul i64 undef, 16
+; SLM-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2i64 = mul <2 x i64> undef, <i64 16, i64 16>
+; SLM-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V4i64 = mul <4 x i64> undef, <i64 16, i64 16, i64 16, i64 16>
+; SLM-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V8i64 = mul <8 x i64> undef, <i64 16, i64 16, i64 16, i64 16, i64 16, i64 16, i64 16, i64 16>
; SLM-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %I32 = mul i32 undef, 16
-; SLM-NEXT: Cost Model: Found an estimated cost of 11 for instruction: %V4i32 = mul <4 x i32> undef, <i32 16, i32 16, i32 16, i32 16>
-; SLM-NEXT: Cost Model: Found an estimated cost of 22 for instruction: %V8i32 = mul <8 x i32> undef, <i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16>
-; SLM-NEXT: Cost Model: Found an estimated cost of 44 for instruction: %V16i32 = mul <16 x i32> undef, <i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16>
+; SLM-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4i32 = mul <4 x i32> undef, <i32 16, i32 16, i32 16, i32 16>
+; SLM-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V8i32 = mul <8 x i32> undef, <i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16>
+; SLM-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V16i32 = mul <16 x i32> undef, <i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16>
; SLM-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %I16 = mul i16 undef, 16
-; SLM-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V8i16 = mul <8 x i16> undef, <i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16>
-; SLM-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V16i16 = mul <16 x i16> undef, <i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16>
-; SLM-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V32i16 = mul <32 x i16> undef, <i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16>
+; SLM-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V8i16 = mul <8 x i16> undef, <i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16>
+; SLM-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V16i16 = mul <16 x i16> undef, <i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16>
+; SLM-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V32i16 = mul <32 x i16> undef, <i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16>
; SLM-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %I8 = mul i8 undef, 16
-; SLM-NEXT: Cost Model: Found an estimated cost of 9 for instruction: %V16i8 = mul <16 x i8> undef, <i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16>
-; SLM-NEXT: Cost Model: Found an estimated cost of 20 for instruction: %V32i8 = mul <32 x i8> undef, <i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16>
-; SLM-NEXT: Cost Model: Found an estimated cost of 40 for instruction: %V64i8 = mul <64 x i8> undef, <i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16>
+; SLM-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %V16i8 = mul <16 x i8> undef, <i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16>
+; SLM-NEXT: Cost Model: Found an estimated cost of 16 for instruction: %V32i8 = mul <32 x i8> undef, <i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16>
+; SLM-NEXT: Cost Model: Found an estimated cost of 32 for instruction: %V64i8 = mul <64 x i8> undef, <i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16>
; SLM-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
%I64 = mul i64 undef, 16
More information about the llvm-commits
mailing list