1; NOTE: Assertions have been autogenerated by utils/update_test_checks.py 2; RUN: opt < %s -slp-vectorizer -instcombine -S -mtriple=x86_64-unknown-linux -mattr=+sse2 | FileCheck %s --check-prefixes=SSE 3; RUN: opt < %s -slp-vectorizer -instcombine -S -mtriple=x86_64-unknown-linux -mattr=+avx | FileCheck %s --check-prefixes=AVX 4; RUN: opt < %s -slp-vectorizer -instcombine -S -mtriple=x86_64-unknown-linux -mattr=+avx2 | FileCheck %s --check-prefixes=AVX 5; RUN: opt < %s -slp-vectorizer -instcombine -S -mtriple=x86_64-unknown-linux -mattr=+avx512f | FileCheck %s --check-prefixes=AVX 6; RUN: opt < %s -slp-vectorizer -instcombine -S -mtriple=x86_64-unknown-linux -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX 7 8 9@b = global [8 x i32] zeroinitializer, align 16 10@a = global [8 x i32] zeroinitializer, align 16 11 12define void @foo() { 13; SSE-LABEL: @foo( 14; SSE-NEXT: [[TMP1:%.*]] = load i32, i32* getelementptr inbounds ([8 x i32], [8 x i32]* @b, i64 0, i64 0), align 16 15; SSE-NEXT: store i32 [[TMP1]], i32* getelementptr inbounds ([8 x i32], [8 x i32]* @a, i64 0, i64 0), align 16 16; SSE-NEXT: [[TMP2:%.*]] = load i32, i32* getelementptr inbounds ([8 x i32], [8 x i32]* @b, i64 0, i64 2), align 8 17; SSE-NEXT: store i32 [[TMP2]], i32* getelementptr inbounds ([8 x i32], [8 x i32]* @a, i64 0, i64 1), align 4 18; SSE-NEXT: store i32 [[TMP1]], i32* getelementptr inbounds ([8 x i32], [8 x i32]* @a, i64 0, i64 2), align 8 19; SSE-NEXT: store i32 [[TMP2]], i32* getelementptr inbounds ([8 x i32], [8 x i32]* @a, i64 0, i64 3), align 4 20; SSE-NEXT: store i32 [[TMP1]], i32* getelementptr inbounds ([8 x i32], [8 x i32]* @a, i64 0, i64 4), align 16 21; SSE-NEXT: store i32 [[TMP2]], i32* getelementptr inbounds ([8 x i32], [8 x i32]* @a, i64 0, i64 5), align 4 22; SSE-NEXT: store i32 [[TMP1]], i32* getelementptr inbounds ([8 x i32], [8 x i32]* @a, i64 0, i64 6), align 8 23; SSE-NEXT: store i32 [[TMP2]], i32* getelementptr inbounds ([8 x i32], [8 x i32]* @a, i64 0, i64 7), align 4 24; SSE-NEXT: ret void 25; 26; AVX-LABEL: @foo( 27; AVX-NEXT: [[TMP1:%.*]] = call <2 x i32> @llvm.masked.gather.v2i32.v2p0i32(<2 x i32*> <i32* getelementptr inbounds ([8 x i32], [8 x i32]* @b, i64 0, i64 0), i32* getelementptr inbounds ([8 x i32], [8 x i32]* @b, i64 0, i64 2)>, i32 8, <2 x i1> <i1 true, i1 true>, <2 x i32> undef) 28; AVX-NEXT: [[SHUFFLE:%.*]] = shufflevector <2 x i32> [[TMP1]], <2 x i32> undef, <8 x i32> <i32 0, i32 1, i32 0, i32 1, i32 0, i32 1, i32 0, i32 1> 29; AVX-NEXT: store <8 x i32> [[SHUFFLE]], <8 x i32>* bitcast ([8 x i32]* @a to <8 x i32>*), align 16 30; AVX-NEXT: ret void 31; 32 %1 = load i32, i32* getelementptr inbounds ([8 x i32], [8 x i32]* @b, i64 0, i64 0), align 16 33 store i32 %1, i32* getelementptr inbounds ([8 x i32], [8 x i32]* @a, i64 0, i64 0), align 16 34 %2 = load i32, i32* getelementptr inbounds ([8 x i32], [8 x i32]* @b, i64 0, i64 2), align 8 35 store i32 %2, i32* getelementptr inbounds ([8 x i32], [8 x i32]* @a, i64 0, i64 1), align 4 36 store i32 %1, i32* getelementptr inbounds ([8 x i32], [8 x i32]* @a, i64 0, i64 2), align 8 37 store i32 %2, i32* getelementptr inbounds ([8 x i32], [8 x i32]* @a, i64 0, i64 3), align 4 38 store i32 %1, i32* getelementptr inbounds ([8 x i32], [8 x i32]* @a, i64 0, i64 4), align 16 39 store i32 %2, i32* getelementptr inbounds ([8 x i32], [8 x i32]* @a, i64 0, i64 5), align 4 40 store i32 %1, i32* getelementptr inbounds ([8 x i32], [8 x i32]* @a, i64 0, i64 6), align 8 41 store i32 %2, i32* getelementptr inbounds ([8 x i32], [8 x i32]* @a, i64 0, i64 7), align 4 42 ret void 43} 44