Skip to content

[SLP]Improve/fix extracts calculations for non-power-of-2 elements. #93213

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
May 24, 2024
Merged

[SLP]Improve/fix extracts calculations for non-power-of-2 elements. #93213

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
147 changes: 95 additions & 52 deletions llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp
View file
Open in desktop

Large diffs are not rendered by default.

23 changes: 5 additions & 18 deletions llvm/test/Transforms/SLPVectorizer/AMDGPU/reduction.ll
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -56,25 +56,12 @@ define half @reduction_half16(<16 x half> %vec16) {
;
; VI-LABEL: @reduction_half16(
; VI-NEXT: entry:
; VI-NEXT: [[ELT8:%.*]] = extractelement <16 x half> [[VEC16:%.*]], i64 8
; VI-NEXT: [[ELT9:%.*]] = extractelement <16 x half> [[VEC16]], i64 9
; VI-NEXT: [[ELT10:%.*]] = extractelement <16 x half> [[VEC16]], i64 10
; VI-NEXT: [[ELT11:%.*]] = extractelement <16 x half> [[VEC16]], i64 11
; VI-NEXT: [[ELT12:%.*]] = extractelement <16 x half> [[VEC16]], i64 12
; VI-NEXT: [[ELT13:%.*]] = extractelement <16 x half> [[VEC16]], i64 13
; VI-NEXT: [[ELT14:%.*]] = extractelement <16 x half> [[VEC16]], i64 14
; VI-NEXT: [[ELT15:%.*]] = extractelement <16 x half> [[VEC16]], i64 15
; VI-NEXT: [[TMP0:%.*]] = shufflevector <16 x half> [[VEC16]], <16 x half> poison, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
; VI-NEXT: [[TMP0:%.*]] = shufflevector <16 x half> [[VEC16:%.*]], <16 x half> poison, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
; VI-NEXT: [[TMP1:%.*]] = call fast half @llvm.vector.reduce.fadd.v8f16(half 0xH8000, <8 x half> [[TMP0]])
; VI-NEXT: [[OP_RDX:%.*]] = fadd fast half [[TMP1]], [[ELT8]]
; VI-NEXT: [[OP_RDX1:%.*]] = fadd fast half [[ELT9]], [[ELT10]]
; VI-NEXT: [[OP_RDX2:%.*]] = fadd fast half [[ELT11]], [[ELT12]]
; VI-NEXT: [[OP_RDX3:%.*]] = fadd fast half [[ELT13]], [[ELT14]]
; VI-NEXT: [[OP_RDX4:%.*]] = fadd fast half [[OP_RDX]], [[OP_RDX1]]
; VI-NEXT: [[OP_RDX5:%.*]] = fadd fast half [[OP_RDX2]], [[OP_RDX3]]
; VI-NEXT: [[OP_RDX6:%.*]] = fadd fast half [[OP_RDX4]], [[OP_RDX5]]
; VI-NEXT: [[OP_RDX7:%.*]] = fadd fast half [[OP_RDX6]], [[ELT15]]
; VI-NEXT: ret half [[OP_RDX7]]
; VI-NEXT: [[TMP2:%.*]] = shufflevector <16 x half> [[VEC16]], <16 x half> poison, <8 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
; VI-NEXT: [[TMP3:%.*]] = call fast half @llvm.vector.reduce.fadd.v8f16(half 0xH8000, <8 x half> [[TMP2]])
; VI-NEXT: [[OP_RDX:%.*]] = fadd fast half [[TMP1]], [[TMP3]]
; VI-NEXT: ret half [[OP_RDX]]
;
entry:
%elt0 = extractelement <16 x half> %vec16, i64 0
Expand Down
144 changes: 60 additions & 84 deletions llvm/test/Transforms/SLPVectorizer/RISCV/math-function.ll
View file
Open in desktop

Large diffs are not rendered by default.

28 changes: 13 additions & 15 deletions llvm/test/Transforms/SLPVectorizer/X86/alternate-calls-inseltpoison.ll
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -51,25 +51,23 @@ define <8 x float> @ceil_floor(<8 x float> %a) {
;
; AVX-LABEL: @ceil_floor(
; AVX-NEXT: [[A0:%.*]] = extractelement <8 x float> [[A:%.*]], i64 0
; AVX-NEXT: [[A1:%.*]] = extractelement <8 x float> [[A]], i64 1
; AVX-NEXT: [[A2:%.*]] = extractelement <8 x float> [[A]], i64 2
; AVX-NEXT: [[A3:%.*]] = extractelement <8 x float> [[A]], i64 3
; AVX-NEXT: [[AB0:%.*]] = call float @llvm.ceil.f32(float [[A0]])
; AVX-NEXT: [[AB1:%.*]] = call float @llvm.floor.f32(float [[A1]])
; AVX-NEXT: [[AB2:%.*]] = call float @llvm.floor.f32(float [[A2]])
; AVX-NEXT: [[TMP1:%.*]] = shufflevector <8 x float> [[A]], <8 x float> poison, <2 x i32> <i32 1, i32 2>
; AVX-NEXT: [[TMP2:%.*]] = call <2 x float> @llvm.floor.v2f32(<2 x float> [[TMP1]])
; AVX-NEXT: [[AB3:%.*]] = call float @llvm.ceil.f32(float [[A3]])
; AVX-NEXT: [[TMP1:%.*]] = shufflevector <8 x float> [[A]], <8 x float> poison, <2 x i32> <i32 4, i32 5>
; AVX-NEXT: [[TMP2:%.*]] = call <2 x float> @llvm.ceil.v2f32(<2 x float> [[TMP1]])
; AVX-NEXT: [[TMP3:%.*]] = shufflevector <8 x float> [[A]], <8 x float> poison, <2 x i32> <i32 6, i32 7>
; AVX-NEXT: [[TMP4:%.*]] = call <2 x float> @llvm.floor.v2f32(<2 x float> [[TMP3]])
; AVX-NEXT: [[TMP3:%.*]] = shufflevector <8 x float> [[A]], <8 x float> poison, <2 x i32> <i32 4, i32 5>
; AVX-NEXT: [[TMP4:%.*]] = call <2 x float> @llvm.ceil.v2f32(<2 x float> [[TMP3]])
; AVX-NEXT: [[TMP5:%.*]] = shufflevector <8 x float> [[A]], <8 x float> poison, <2 x i32> <i32 6, i32 7>
; AVX-NEXT: [[TMP6:%.*]] = call <2 x float> @llvm.floor.v2f32(<2 x float> [[TMP5]])
; AVX-NEXT: [[R0:%.*]] = insertelement <8 x float> poison, float [[AB0]], i64 0
; AVX-NEXT: [[R1:%.*]] = insertelement <8 x float> [[R0]], float [[AB1]], i64 1
; AVX-NEXT: [[R2:%.*]] = insertelement <8 x float> [[R1]], float [[AB2]], i64 2
; AVX-NEXT: [[R3:%.*]] = insertelement <8 x float> [[R2]], float [[AB3]], i64 3
; AVX-NEXT: [[TMP5:%.*]] = shufflevector <2 x float> [[TMP2]], <2 x float> poison, <8 x i32> <i32 0, i32 1, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
; AVX-NEXT: [[R52:%.*]] = shufflevector <8 x float> [[R3]], <8 x float> [[TMP5]], <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 8, i32 9, i32 poison, i32 poison>
; AVX-NEXT: [[TMP6:%.*]] = shufflevector <2 x float> [[TMP4]], <2 x float> poison, <8 x i32> <i32 0, i32 1, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
; AVX-NEXT: [[R71:%.*]] = shufflevector <8 x float> [[R52]], <8 x float> [[TMP6]], <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 8, i32 9>
; AVX-NEXT: [[TMP7:%.*]] = shufflevector <2 x float> [[TMP2]], <2 x float> poison, <8 x i32> <i32 0, i32 1, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
; AVX-NEXT: [[R23:%.*]] = shufflevector <8 x float> [[R0]], <8 x float> [[TMP7]], <8 x i32> <i32 0, i32 8, i32 9, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
; AVX-NEXT: [[R3:%.*]] = insertelement <8 x float> [[R23]], float [[AB3]], i64 3
; AVX-NEXT: [[TMP8:%.*]] = shufflevector <2 x float> [[TMP4]], <2 x float> poison, <8 x i32> <i32 0, i32 1, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
; AVX-NEXT: [[R52:%.*]] = shufflevector <8 x float> [[R3]], <8 x float> [[TMP8]], <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 8, i32 9, i32 poison, i32 poison>
; AVX-NEXT: [[TMP9:%.*]] = shufflevector <2 x float> [[TMP6]], <2 x float> poison, <8 x i32> <i32 0, i32 1, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
; AVX-NEXT: [[R71:%.*]] = shufflevector <8 x float> [[R52]], <8 x float> [[TMP9]], <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 8, i32 9>
; AVX-NEXT: ret <8 x float> [[R71]]
;
; AVX2-LABEL: @ceil_floor(
Expand Down
28 changes: 13 additions & 15 deletions llvm/test/Transforms/SLPVectorizer/X86/alternate-calls.ll
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -51,25 +51,23 @@ define <8 x float> @ceil_floor(<8 x float> %a) {
;
; AVX-LABEL: @ceil_floor(
; AVX-NEXT: [[A0:%.*]] = extractelement <8 x float> [[A:%.*]], i64 0
; AVX-NEXT: [[A1:%.*]] = extractelement <8 x float> [[A]], i64 1
; AVX-NEXT: [[A2:%.*]] = extractelement <8 x float> [[A]], i64 2
; AVX-NEXT: [[A3:%.*]] = extractelement <8 x float> [[A]], i64 3
; AVX-NEXT: [[AB0:%.*]] = call float @llvm.ceil.f32(float [[A0]])
; AVX-NEXT: [[AB1:%.*]] = call float @llvm.floor.f32(float [[A1]])
; AVX-NEXT: [[AB2:%.*]] = call float @llvm.floor.f32(float [[A2]])
; AVX-NEXT: [[TMP1:%.*]] = shufflevector <8 x float> [[A]], <8 x float> poison, <2 x i32> <i32 1, i32 2>
; AVX-NEXT: [[TMP2:%.*]] = call <2 x float> @llvm.floor.v2f32(<2 x float> [[TMP1]])
; AVX-NEXT: [[AB3:%.*]] = call float @llvm.ceil.f32(float [[A3]])
; AVX-NEXT: [[TMP1:%.*]] = shufflevector <8 x float> [[A]], <8 x float> poison, <2 x i32> <i32 4, i32 5>
; AVX-NEXT: [[TMP2:%.*]] = call <2 x float> @llvm.ceil.v2f32(<2 x float> [[TMP1]])
; AVX-NEXT: [[TMP3:%.*]] = shufflevector <8 x float> [[A]], <8 x float> poison, <2 x i32> <i32 6, i32 7>
; AVX-NEXT: [[TMP4:%.*]] = call <2 x float> @llvm.floor.v2f32(<2 x float> [[TMP3]])
; AVX-NEXT: [[TMP3:%.*]] = shufflevector <8 x float> [[A]], <8 x float> poison, <2 x i32> <i32 4, i32 5>
; AVX-NEXT: [[TMP4:%.*]] = call <2 x float> @llvm.ceil.v2f32(<2 x float> [[TMP3]])
; AVX-NEXT: [[TMP5:%.*]] = shufflevector <8 x float> [[A]], <8 x float> poison, <2 x i32> <i32 6, i32 7>
; AVX-NEXT: [[TMP6:%.*]] = call <2 x float> @llvm.floor.v2f32(<2 x float> [[TMP5]])
; AVX-NEXT: [[R0:%.*]] = insertelement <8 x float> poison, float [[AB0]], i64 0
; AVX-NEXT: [[R1:%.*]] = insertelement <8 x float> [[R0]], float [[AB1]], i64 1
; AVX-NEXT: [[R2:%.*]] = insertelement <8 x float> [[R1]], float [[AB2]], i64 2
; AVX-NEXT: [[R3:%.*]] = insertelement <8 x float> [[R2]], float [[AB3]], i64 3
; AVX-NEXT: [[TMP5:%.*]] = shufflevector <2 x float> [[TMP2]], <2 x float> poison, <8 x i32> <i32 0, i32 1, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
; AVX-NEXT: [[R52:%.*]] = shufflevector <8 x float> [[R3]], <8 x float> [[TMP5]], <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 8, i32 9, i32 poison, i32 poison>
; AVX-NEXT: [[TMP6:%.*]] = shufflevector <2 x float> [[TMP4]], <2 x float> poison, <8 x i32> <i32 0, i32 1, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
; AVX-NEXT: [[R71:%.*]] = shufflevector <8 x float> [[R52]], <8 x float> [[TMP6]], <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 8, i32 9>
; AVX-NEXT: [[TMP7:%.*]] = shufflevector <2 x float> [[TMP2]], <2 x float> poison, <8 x i32> <i32 0, i32 1, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
; AVX-NEXT: [[R23:%.*]] = shufflevector <8 x float> [[R0]], <8 x float> [[TMP7]], <8 x i32> <i32 0, i32 8, i32 9, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
; AVX-NEXT: [[R3:%.*]] = insertelement <8 x float> [[R23]], float [[AB3]], i64 3
; AVX-NEXT: [[TMP8:%.*]] = shufflevector <2 x float> [[TMP4]], <2 x float> poison, <8 x i32> <i32 0, i32 1, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
; AVX-NEXT: [[R52:%.*]] = shufflevector <8 x float> [[R3]], <8 x float> [[TMP8]], <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 8, i32 9, i32 poison, i32 poison>
; AVX-NEXT: [[TMP9:%.*]] = shufflevector <2 x float> [[TMP6]], <2 x float> poison, <8 x i32> <i32 0, i32 1, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
; AVX-NEXT: [[R71:%.*]] = shufflevector <8 x float> [[R52]], <8 x float> [[TMP9]], <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 8, i32 9>
; AVX-NEXT: ret <8 x float> [[R71]]
;
; AVX2-LABEL: @ceil_floor(
Expand Down
74 changes: 14 additions & 60 deletions llvm/test/Transforms/SLPVectorizer/X86/hadd-inseltpoison.ll
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,10 +1,10 @@
; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
; RUN: opt < %s -mtriple=x86_64-unknown -passes=slp-vectorizer,instcombine -S | FileCheck %s --check-prefixes=CHECK,SSE
; RUN: opt < %s -mtriple=x86_64-unknown -mcpu=slm -passes=slp-vectorizer,instcombine -S | FileCheck %s --check-prefixes=CHECK,SLM
; RUN: opt < %s -mtriple=x86_64-unknown -mcpu=corei7-avx -passes=slp-vectorizer,instcombine -S | FileCheck %s --check-prefixes=CHECK,AVX,AVX1
; RUN: opt < %s -mtriple=x86_64-unknown -mcpu=core-avx2 -passes=slp-vectorizer,instcombine -S | FileCheck %s --check-prefixes=CHECK,AVX,AVX2
; RUN: opt < %s -mtriple=x86_64-unknown -mcpu=knl -passes=slp-vectorizer,instcombine -S | FileCheck %s --check-prefixes=CHECK,AVX,AVX512
; RUN: opt < %s -mtriple=x86_64-unknown -mcpu=skx -passes=slp-vectorizer,instcombine -S | FileCheck %s --check-prefixes=CHECK,AVX,AVX512
; RUN: opt < %s -mtriple=x86_64-unknown -mcpu=corei7-avx -passes=slp-vectorizer,instcombine -S | FileCheck %s --check-prefixes=CHECK,AVX
; RUN: opt < %s -mtriple=x86_64-unknown -mcpu=core-avx2 -passes=slp-vectorizer,instcombine -S | FileCheck %s --check-prefixes=CHECK,AVX
; RUN: opt < %s -mtriple=x86_64-unknown -mcpu=knl -passes=slp-vectorizer,instcombine -S | FileCheck %s --check-prefixes=CHECK,AVX
; RUN: opt < %s -mtriple=x86_64-unknown -mcpu=skx -passes=slp-vectorizer,instcombine -S | FileCheck %s --check-prefixes=CHECK,AVX

;
; 128-bit vectors
Expand Down Expand Up @@ -213,62 +213,16 @@ define <4 x double> @test_v4f64(<4 x double> %a, <4 x double> %b) {

; PR50392
define <4 x double> @test_v4f64_partial_swizzle(<4 x double> %a, <4 x double> %b) {
; SSE-LABEL: @test_v4f64_partial_swizzle(
; SSE-NEXT: [[B2:%.*]] = extractelement <4 x double> [[B:%.*]], i64 2
; SSE-NEXT: [[B3:%.*]] = extractelement <4 x double> [[B]], i64 3
; SSE-NEXT: [[TMP1:%.*]] = shufflevector <4 x double> [[A:%.*]], <4 x double> [[B]], <2 x i32> <i32 0, i32 4>
; SSE-NEXT: [[TMP2:%.*]] = shufflevector <4 x double> [[A]], <4 x double> [[B]], <2 x i32> <i32 1, i32 5>
; SSE-NEXT: [[TMP3:%.*]] = fadd <2 x double> [[TMP1]], [[TMP2]]
; SSE-NEXT: [[R3:%.*]] = fadd double [[B2]], [[B3]]
; SSE-NEXT: [[TMP4:%.*]] = shufflevector <2 x double> [[TMP3]], <2 x double> poison, <4 x i32> <i32 0, i32 poison, i32 1, i32 poison>
; SSE-NEXT: [[R03:%.*]] = insertelement <4 x double> [[TMP4]], double [[R3]], i64 3
; SSE-NEXT: ret <4 x double> [[R03]]
;
; SLM-LABEL: @test_v4f64_partial_swizzle(
; SLM-NEXT: [[B2:%.*]] = extractelement <4 x double> [[B:%.*]], i64 2
; SLM-NEXT: [[B3:%.*]] = extractelement <4 x double> [[B]], i64 3
; SLM-NEXT: [[TMP1:%.*]] = shufflevector <4 x double> [[A:%.*]], <4 x double> [[B]], <2 x i32> <i32 0, i32 4>
; SLM-NEXT: [[TMP2:%.*]] = shufflevector <4 x double> [[A]], <4 x double> [[B]], <2 x i32> <i32 1, i32 5>
; SLM-NEXT: [[TMP3:%.*]] = fadd <2 x double> [[TMP1]], [[TMP2]]
; SLM-NEXT: [[R3:%.*]] = fadd double [[B2]], [[B3]]
; SLM-NEXT: [[TMP4:%.*]] = shufflevector <2 x double> [[TMP3]], <2 x double> poison, <4 x i32> <i32 0, i32 poison, i32 1, i32 poison>
; SLM-NEXT: [[R03:%.*]] = insertelement <4 x double> [[TMP4]], double [[R3]], i64 3
; SLM-NEXT: ret <4 x double> [[R03]]
;
; AVX1-LABEL: @test_v4f64_partial_swizzle(
; AVX1-NEXT: [[A0:%.*]] = extractelement <4 x double> [[A:%.*]], i64 0
; AVX1-NEXT: [[A1:%.*]] = extractelement <4 x double> [[A]], i64 1
; AVX1-NEXT: [[R0:%.*]] = fadd double [[A0]], [[A1]]
; AVX1-NEXT: [[TMP1:%.*]] = shufflevector <4 x double> [[B:%.*]], <4 x double> poison, <2 x i32> <i32 1, i32 2>
; AVX1-NEXT: [[TMP2:%.*]] = shufflevector <4 x double> [[B]], <4 x double> poison, <2 x i32> <i32 0, i32 3>
; AVX1-NEXT: [[TMP3:%.*]] = fadd <2 x double> [[TMP1]], [[TMP2]]
; AVX1-NEXT: [[R00:%.*]] = insertelement <4 x double> poison, double [[R0]], i64 0
; AVX1-NEXT: [[TMP4:%.*]] = shufflevector <2 x double> [[TMP3]], <2 x double> poison, <4 x i32> <i32 0, i32 1, i32 poison, i32 poison>
; AVX1-NEXT: [[R031:%.*]] = shufflevector <4 x double> [[R00]], <4 x double> [[TMP4]], <4 x i32> <i32 0, i32 poison, i32 4, i32 5>
; AVX1-NEXT: ret <4 x double> [[R031]]
;
; AVX2-LABEL: @test_v4f64_partial_swizzle(
; AVX2-NEXT: [[A0:%.*]] = extractelement <4 x double> [[A:%.*]], i64 0
; AVX2-NEXT: [[A1:%.*]] = extractelement <4 x double> [[A]], i64 1
; AVX2-NEXT: [[R0:%.*]] = fadd double [[A0]], [[A1]]
; AVX2-NEXT: [[TMP1:%.*]] = shufflevector <4 x double> [[B:%.*]], <4 x double> poison, <2 x i32> <i32 1, i32 2>
; AVX2-NEXT: [[TMP2:%.*]] = shufflevector <4 x double> [[B]], <4 x double> poison, <2 x i32> <i32 0, i32 3>
; AVX2-NEXT: [[TMP3:%.*]] = fadd <2 x double> [[TMP1]], [[TMP2]]
; AVX2-NEXT: [[R00:%.*]] = insertelement <4 x double> poison, double [[R0]], i64 0
; AVX2-NEXT: [[TMP4:%.*]] = shufflevector <2 x double> [[TMP3]], <2 x double> poison, <4 x i32> <i32 0, i32 1, i32 poison, i32 poison>
; AVX2-NEXT: [[R031:%.*]] = shufflevector <4 x double> [[R00]], <4 x double> [[TMP4]], <4 x i32> <i32 0, i32 poison, i32 4, i32 5>
; AVX2-NEXT: ret <4 x double> [[R031]]
;
; AVX512-LABEL: @test_v4f64_partial_swizzle(
; AVX512-NEXT: [[B2:%.*]] = extractelement <4 x double> [[B:%.*]], i64 2
; AVX512-NEXT: [[B3:%.*]] = extractelement <4 x double> [[B]], i64 3
; AVX512-NEXT: [[TMP1:%.*]] = shufflevector <4 x double> [[A:%.*]], <4 x double> [[B]], <2 x i32> <i32 0, i32 4>
; AVX512-NEXT: [[TMP2:%.*]] = shufflevector <4 x double> [[A]], <4 x double> [[B]], <2 x i32> <i32 1, i32 5>
; AVX512-NEXT: [[TMP3:%.*]] = fadd <2 x double> [[TMP1]], [[TMP2]]
; AVX512-NEXT: [[R3:%.*]] = fadd double [[B2]], [[B3]]
; AVX512-NEXT: [[TMP4:%.*]] = shufflevector <2 x double> [[TMP3]], <2 x double> poison, <4 x i32> <i32 0, i32 poison, i32 1, i32 poison>
; AVX512-NEXT: [[R03:%.*]] = insertelement <4 x double> [[TMP4]], double [[R3]], i64 3
; AVX512-NEXT: ret <4 x double> [[R03]]
; CHECK-LABEL: @test_v4f64_partial_swizzle(
; CHECK-NEXT: [[B2:%.*]] = extractelement <4 x double> [[B:%.*]], i64 2
; CHECK-NEXT: [[B3:%.*]] = extractelement <4 x double> [[B]], i64 3
; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <4 x double> [[A:%.*]], <4 x double> [[B]], <2 x i32> <i32 0, i32 4>
; CHECK-NEXT: [[TMP2:%.*]] = shufflevector <4 x double> [[A]], <4 x double> [[B]], <2 x i32> <i32 1, i32 5>
; CHECK-NEXT: [[TMP3:%.*]] = fadd <2 x double> [[TMP1]], [[TMP2]]
; CHECK-NEXT: [[R3:%.*]] = fadd double [[B2]], [[B3]]
; CHECK-NEXT: [[TMP4:%.*]] = shufflevector <2 x double> [[TMP3]], <2 x double> poison, <4 x i32> <i32 0, i32 poison, i32 1, i32 poison>
; CHECK-NEXT: [[R03:%.*]] = insertelement <4 x double> [[TMP4]], double [[R3]], i64 3
; CHECK-NEXT: ret <4 x double> [[R03]]
;
%a0 = extractelement <4 x double> %a, i64 0
%a1 = extractelement <4 x double> %a, i64 1
Expand Down
Loading
Loading