; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc < %s -mtriple=i686-unknown-unknown -mattr=+sse2 | FileCheck %s --check-prefixes=SSE,X86-SSE
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+sse2 | FileCheck %s --check-prefixes=SSE,X64-SSE
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+sse4.1 | FileCheck %s --check-prefixes=SSE,X64-SSE
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx | FileCheck %s --check-prefixes=AVX,AVX1
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX2
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx512f,+avx512bw | FileCheck %s --check-prefixes=AVX,AVX512
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx512f,+avx512bw,+avx512vl | FileCheck %s --check-prefixes=AVX,AVX512
;
; vXi64
;
define i64 @test_v2i64(<2 x i64> %a0) nounwind {
; X86-SSE-LABEL: test_v2i64:
; X86-SSE: # %bb.0:
; X86-SSE-NEXT: pshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
; X86-SSE-NEXT: pxor %xmm0, %xmm1
; X86-SSE-NEXT: movd %xmm1, %eax
; X86-SSE-NEXT: pshufd {{.*#+}} xmm0 = xmm1[1,1,1,1]
; X86-SSE-NEXT: movd %xmm0, %edx
; X86-SSE-NEXT: retl
;
; X64-SSE-LABEL: test_v2i64:
; X64-SSE: # %bb.0:
; X64-SSE-NEXT: pshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
; X64-SSE-NEXT: pxor %xmm0, %xmm1
; X64-SSE-NEXT: movq %xmm1, %rax
; X64-SSE-NEXT: retq
;
; AVX-LABEL: test_v2i64:
; AVX: # %bb.0:
; AVX-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
; AVX-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX-NEXT: vmovq %xmm0, %rax
; AVX-NEXT: retq
%1 = call i64 @llvm.vector.reduce.xor.v2i64(<2 x i64> %a0)
ret i64 %1
}
define i64 @test_v4i64(<4 x i64> %a0) nounwind {
; X86-SSE-LABEL: test_v4i64:
; X86-SSE: # %bb.0:
; X86-SSE-NEXT: pxor %xmm1, %xmm0
; X86-SSE-NEXT: pshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
; X86-SSE-NEXT: pxor %xmm0, %xmm1
; X86-SSE-NEXT: movd %xmm1, %eax
; X86-SSE-NEXT: pshufd {{.*#+}} xmm0 = xmm1[1,1,1,1]
; X86-SSE-NEXT: movd %xmm0, %edx
; X86-SSE-NEXT: retl
;
; X64-SSE-LABEL: test_v4i64:
; X64-SSE: # %bb.0:
; X64-SSE-NEXT: pxor %xmm1, %xmm0
; X64-SSE-NEXT: pshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
; X64-SSE-NEXT: pxor %xmm0, %xmm1
; X64-SSE-NEXT: movq %xmm1, %rax
; X64-SSE-NEXT: retq
;
; AVX1-LABEL: test_v4i64:
; AVX1: # %bb.0:
; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm1
; AVX1-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX1-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
; AVX1-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX1-NEXT: vmovq %xmm0, %rax
; AVX1-NEXT: vzeroupper
; AVX1-NEXT: retq
;
; AVX2-LABEL: test_v4i64:
; AVX2: # %bb.0:
; AVX2-NEXT: vextracti128 $1, %ymm0, %xmm1
; AVX2-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX2-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
; AVX2-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX2-NEXT: vmovq %xmm0, %rax
; AVX2-NEXT: vzeroupper
; AVX2-NEXT: retq
;
; AVX512-LABEL: test_v4i64:
; AVX512: # %bb.0:
; AVX512-NEXT: vextracti128 $1, %ymm0, %xmm1
; AVX512-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX512-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
; AVX512-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX512-NEXT: vmovq %xmm0, %rax
; AVX512-NEXT: vzeroupper
; AVX512-NEXT: retq
%1 = call i64 @llvm.vector.reduce.xor.v4i64(<4 x i64> %a0)
ret i64 %1
}
define i64 @test_v8i64(<8 x i64> %a0) nounwind {
; X86-SSE-LABEL: test_v8i64:
; X86-SSE: # %bb.0:
; X86-SSE-NEXT: pushl %ebp
; X86-SSE-NEXT: movl %esp, %ebp
; X86-SSE-NEXT: andl $-16, %esp
; X86-SSE-NEXT: subl $16, %esp
; X86-SSE-NEXT: pxor %xmm2, %xmm0
; X86-SSE-NEXT: pxor 8(%ebp), %xmm1
; X86-SSE-NEXT: pxor %xmm0, %xmm1
; X86-SSE-NEXT: pshufd {{.*#+}} xmm0 = xmm1[2,3,2,3]
; X86-SSE-NEXT: pxor %xmm1, %xmm0
; X86-SSE-NEXT: movd %xmm0, %eax
; X86-SSE-NEXT: pshufd {{.*#+}} xmm0 = xmm0[1,1,1,1]
; X86-SSE-NEXT: movd %xmm0, %edx
; X86-SSE-NEXT: movl %ebp, %esp
; X86-SSE-NEXT: popl %ebp
; X86-SSE-NEXT: retl
;
; X64-SSE-LABEL: test_v8i64:
; X64-SSE: # %bb.0:
; X64-SSE-NEXT: pxor %xmm3, %xmm1
; X64-SSE-NEXT: pxor %xmm2, %xmm0
; X64-SSE-NEXT: pxor %xmm1, %xmm0
; X64-SSE-NEXT: pshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
; X64-SSE-NEXT: pxor %xmm0, %xmm1
; X64-SSE-NEXT: movq %xmm1, %rax
; X64-SSE-NEXT: retq
;
; AVX1-LABEL: test_v8i64:
; AVX1: # %bb.0:
; AVX1-NEXT: vxorps %ymm1, %ymm0, %ymm0
; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm1
; AVX1-NEXT: vxorps %xmm1, %xmm0, %xmm0
; AVX1-NEXT: vshufps {{.*#+}} xmm1 = xmm0[2,3,2,3]
; AVX1-NEXT: vxorps %xmm1, %xmm0, %xmm0
; AVX1-NEXT: vmovq %xmm0, %rax
; AVX1-NEXT: vzeroupper
; AVX1-NEXT: retq
;
; AVX2-LABEL: test_v8i64:
; AVX2: # %bb.0:
; AVX2-NEXT: vpxor %ymm1, %ymm0, %ymm0
; AVX2-NEXT: vextracti128 $1, %ymm0, %xmm1
; AVX2-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX2-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
; AVX2-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX2-NEXT: vmovq %xmm0, %rax
; AVX2-NEXT: vzeroupper
; AVX2-NEXT: retq
;
; AVX512-LABEL: test_v8i64:
; AVX512: # %bb.0:
; AVX512-NEXT: vextracti64x4 $1, %zmm0, %ymm1
; AVX512-NEXT: vpxorq %zmm1, %zmm0, %zmm0
; AVX512-NEXT: vextracti128 $1, %ymm0, %xmm1
; AVX512-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX512-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
; AVX512-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX512-NEXT: vmovq %xmm0, %rax
; AVX512-NEXT: vzeroupper
; AVX512-NEXT: retq
%1 = call i64 @llvm.vector.reduce.xor.v8i64(<8 x i64> %a0)
ret i64 %1
}
define i64 @test_v16i64(<16 x i64> %a0) nounwind {
; X86-SSE-LABEL: test_v16i64:
; X86-SSE: # %bb.0:
; X86-SSE-NEXT: pushl %ebp
; X86-SSE-NEXT: movl %esp, %ebp
; X86-SSE-NEXT: andl $-16, %esp
; X86-SSE-NEXT: subl $16, %esp
; X86-SSE-NEXT: movdqa 8(%ebp), %xmm3
; X86-SSE-NEXT: pxor 56(%ebp), %xmm2
; X86-SSE-NEXT: pxor 24(%ebp), %xmm0
; X86-SSE-NEXT: pxor %xmm2, %xmm0
; X86-SSE-NEXT: pxor 72(%ebp), %xmm3
; X86-SSE-NEXT: pxor 40(%ebp), %xmm1
; X86-SSE-NEXT: pxor %xmm3, %xmm1
; X86-SSE-NEXT: pxor %xmm0, %xmm1
; X86-SSE-NEXT: pshufd {{.*#+}} xmm0 = xmm1[2,3,2,3]
; X86-SSE-NEXT: pxor %xmm1, %xmm0
; X86-SSE-NEXT: movd %xmm0, %eax
; X86-SSE-NEXT: pshufd {{.*#+}} xmm0 = xmm0[1,1,1,1]
; X86-SSE-NEXT: movd %xmm0, %edx
; X86-SSE-NEXT: movl %ebp, %esp
; X86-SSE-NEXT: popl %ebp
; X86-SSE-NEXT: retl
;
; X64-SSE-LABEL: test_v16i64:
; X64-SSE: # %bb.0:
; X64-SSE-NEXT: pxor %xmm6, %xmm2
; X64-SSE-NEXT: pxor %xmm4, %xmm0
; X64-SSE-NEXT: pxor %xmm2, %xmm0
; X64-SSE-NEXT: pxor %xmm7, %xmm3
; X64-SSE-NEXT: pxor %xmm5, %xmm1
; X64-SSE-NEXT: pxor %xmm3, %xmm1
; X64-SSE-NEXT: pxor %xmm0, %xmm1
; X64-SSE-NEXT: pshufd {{.*#+}} xmm0 = xmm1[2,3,2,3]
; X64-SSE-NEXT: pxor %xmm1, %xmm0
; X64-SSE-NEXT: movq %xmm0, %rax
; X64-SSE-NEXT: retq
;
; AVX1-LABEL: test_v16i64:
; AVX1: # %bb.0:
; AVX1-NEXT: vxorps %ymm3, %ymm1, %ymm1
; AVX1-NEXT: vxorps %ymm2, %ymm0, %ymm0
; AVX1-NEXT: vxorps %ymm1, %ymm0, %ymm0
; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm1
; AVX1-NEXT: vxorps %xmm1, %xmm0, %xmm0
; AVX1-NEXT: vshufps {{.*#+}} xmm1 = xmm0[2,3,2,3]
; AVX1-NEXT: vxorps %xmm1, %xmm0, %xmm0
; AVX1-NEXT: vmovq %xmm0, %rax
; AVX1-NEXT: vzeroupper
; AVX1-NEXT: retq
;
; AVX2-LABEL: test_v16i64:
; AVX2: # %bb.0:
; AVX2-NEXT: vpxor %ymm3, %ymm1, %ymm1
; AVX2-NEXT: vpxor %ymm2, %ymm0, %ymm0
; AVX2-NEXT: vpxor %ymm1, %ymm0, %ymm0
; AVX2-NEXT: vextracti128 $1, %ymm0, %xmm1
; AVX2-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX2-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
; AVX2-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX2-NEXT: vmovq %xmm0, %rax
; AVX2-NEXT: vzeroupper
; AVX2-NEXT: retq
;
; AVX512-LABEL: test_v16i64:
; AVX512: # %bb.0:
; AVX512-NEXT: vpxorq %zmm1, %zmm0, %zmm0
; AVX512-NEXT: vextracti64x4 $1, %zmm0, %ymm1
; AVX512-NEXT: vpxorq %zmm1, %zmm0, %zmm0
; AVX512-NEXT: vextracti128 $1, %ymm0, %xmm1
; AVX512-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX512-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
; AVX512-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX512-NEXT: vmovq %xmm0, %rax
; AVX512-NEXT: vzeroupper
; AVX512-NEXT: retq
%1 = call i64 @llvm.vector.reduce.xor.v16i64(<16 x i64> %a0)
ret i64 %1
}
;
; vXi32
;
define i32 @test_v2i32(<2 x i32> %a0) nounwind {
; SSE-LABEL: test_v2i32:
; SSE: # %bb.0:
; SSE-NEXT: pshufd {{.*#+}} xmm1 = xmm0[1,1,1,1]
; SSE-NEXT: pxor %xmm0, %xmm1
; SSE-NEXT: movd %xmm1, %eax
; SSE-NEXT: ret{{[l|q]}}
;
; AVX-LABEL: test_v2i32:
; AVX: # %bb.0:
; AVX-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[1,1,1,1]
; AVX-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX-NEXT: vmovd %xmm0, %eax
; AVX-NEXT: retq
%1 = call i32 @llvm.vector.reduce.xor.v2i32(<2 x i32> %a0)
ret i32 %1
}
define i32 @test_v4i32(<4 x i32> %a0) nounwind {
; SSE-LABEL: test_v4i32:
; SSE: # %bb.0:
; SSE-NEXT: pshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
; SSE-NEXT: pxor %xmm0, %xmm1
; SSE-NEXT: pshufd {{.*#+}} xmm0 = xmm1[1,1,1,1]
; SSE-NEXT: pxor %xmm1, %xmm0
; SSE-NEXT: movd %xmm0, %eax
; SSE-NEXT: ret{{[l|q]}}
;
; AVX-LABEL: test_v4i32:
; AVX: # %bb.0:
; AVX-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
; AVX-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[1,1,1,1]
; AVX-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX-NEXT: vmovd %xmm0, %eax
; AVX-NEXT: retq
%1 = call i32 @llvm.vector.reduce.xor.v4i32(<4 x i32> %a0)
ret i32 %1
}
define i32 @test_v8i32(<8 x i32> %a0) nounwind {
; SSE-LABEL: test_v8i32:
; SSE: # %bb.0:
; SSE-NEXT: pxor %xmm1, %xmm0
; SSE-NEXT: pshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
; SSE-NEXT: pxor %xmm0, %xmm1
; SSE-NEXT: pshufd {{.*#+}} xmm0 = xmm1[1,1,1,1]
; SSE-NEXT: pxor %xmm1, %xmm0
; SSE-NEXT: movd %xmm0, %eax
; SSE-NEXT: ret{{[l|q]}}
;
; AVX1-LABEL: test_v8i32:
; AVX1: # %bb.0:
; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm1
; AVX1-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX1-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
; AVX1-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX1-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[1,1,1,1]
; AVX1-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX1-NEXT: vmovd %xmm0, %eax
; AVX1-NEXT: vzeroupper
; AVX1-NEXT: retq
;
; AVX2-LABEL: test_v8i32:
; AVX2: # %bb.0:
; AVX2-NEXT: vextracti128 $1, %ymm0, %xmm1
; AVX2-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX2-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
; AVX2-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX2-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[1,1,1,1]
; AVX2-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX2-NEXT: vmovd %xmm0, %eax
; AVX2-NEXT: vzeroupper
; AVX2-NEXT: retq
;
; AVX512-LABEL: test_v8i32:
; AVX512: # %bb.0:
; AVX512-NEXT: vextracti128 $1, %ymm0, %xmm1
; AVX512-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX512-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
; AVX512-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX512-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[1,1,1,1]
; AVX512-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX512-NEXT: vmovd %xmm0, %eax
; AVX512-NEXT: vzeroupper
; AVX512-NEXT: retq
%1 = call i32 @llvm.vector.reduce.xor.v8i32(<8 x i32> %a0)
ret i32 %1
}
define i32 @test_v16i32(<16 x i32> %a0) nounwind {
; X86-SSE-LABEL: test_v16i32:
; X86-SSE: # %bb.0:
; X86-SSE-NEXT: pushl %ebp
; X86-SSE-NEXT: movl %esp, %ebp
; X86-SSE-NEXT: andl $-16, %esp
; X86-SSE-NEXT: subl $16, %esp
; X86-SSE-NEXT: pxor %xmm2, %xmm0
; X86-SSE-NEXT: pxor 8(%ebp), %xmm1
; X86-SSE-NEXT: pxor %xmm0, %xmm1
; X86-SSE-NEXT: pshufd {{.*#+}} xmm0 = xmm1[2,3,2,3]
; X86-SSE-NEXT: pxor %xmm1, %xmm0
; X86-SSE-NEXT: pshufd {{.*#+}} xmm1 = xmm0[1,1,1,1]
; X86-SSE-NEXT: pxor %xmm0, %xmm1
; X86-SSE-NEXT: movd %xmm1, %eax
; X86-SSE-NEXT: movl %ebp, %esp
; X86-SSE-NEXT: popl %ebp
; X86-SSE-NEXT: retl
;
; X64-SSE-LABEL: test_v16i32:
; X64-SSE: # %bb.0:
; X64-SSE-NEXT: pxor %xmm3, %xmm1
; X64-SSE-NEXT: pxor %xmm2, %xmm0
; X64-SSE-NEXT: pxor %xmm1, %xmm0
; X64-SSE-NEXT: pshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
; X64-SSE-NEXT: pxor %xmm0, %xmm1
; X64-SSE-NEXT: pshufd {{.*#+}} xmm0 = xmm1[1,1,1,1]
; X64-SSE-NEXT: pxor %xmm1, %xmm0
; X64-SSE-NEXT: movd %xmm0, %eax
; X64-SSE-NEXT: retq
;
; AVX1-LABEL: test_v16i32:
; AVX1: # %bb.0:
; AVX1-NEXT: vxorps %ymm1, %ymm0, %ymm0
; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm1
; AVX1-NEXT: vxorps %xmm1, %xmm0, %xmm0
; AVX1-NEXT: vshufps {{.*#+}} xmm1 = xmm0[2,3,2,3]
; AVX1-NEXT: vxorps %xmm1, %xmm0, %xmm0
; AVX1-NEXT: vshufps {{.*#+}} xmm1 = xmm0[1,1,1,1]
; AVX1-NEXT: vxorps %xmm1, %xmm0, %xmm0
; AVX1-NEXT: vmovd %xmm0, %eax
; AVX1-NEXT: vzeroupper
; AVX1-NEXT: retq
;
; AVX2-LABEL: test_v16i32:
; AVX2: # %bb.0:
; AVX2-NEXT: vpxor %ymm1, %ymm0, %ymm0
; AVX2-NEXT: vextracti128 $1, %ymm0, %xmm1
; AVX2-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX2-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
; AVX2-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX2-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[1,1,1,1]
; AVX2-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX2-NEXT: vmovd %xmm0, %eax
; AVX2-NEXT: vzeroupper
; AVX2-NEXT: retq
;
; AVX512-LABEL: test_v16i32:
; AVX512: # %bb.0:
; AVX512-NEXT: vextracti64x4 $1, %zmm0, %ymm1
; AVX512-NEXT: vpxord %zmm1, %zmm0, %zmm0
; AVX512-NEXT: vextracti128 $1, %ymm0, %xmm1
; AVX512-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX512-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
; AVX512-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX512-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[1,1,1,1]
; AVX512-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX512-NEXT: vmovd %xmm0, %eax
; AVX512-NEXT: vzeroupper
; AVX512-NEXT: retq
%1 = call i32 @llvm.vector.reduce.xor.v16i32(<16 x i32> %a0)
ret i32 %1
}
define i32 @test_v32i32(<32 x i32> %a0) nounwind {
; X86-SSE-LABEL: test_v32i32:
; X86-SSE: # %bb.0:
; X86-SSE-NEXT: pushl %ebp
; X86-SSE-NEXT: movl %esp, %ebp
; X86-SSE-NEXT: andl $-16, %esp
; X86-SSE-NEXT: subl $16, %esp
; X86-SSE-NEXT: movdqa 8(%ebp), %xmm3
; X86-SSE-NEXT: pxor 56(%ebp), %xmm2
; X86-SSE-NEXT: pxor 24(%ebp), %xmm0
; X86-SSE-NEXT: pxor %xmm2, %xmm0
; X86-SSE-NEXT: pxor 72(%ebp), %xmm3
; X86-SSE-NEXT: pxor 40(%ebp), %xmm1
; X86-SSE-NEXT: pxor %xmm3, %xmm1
; X86-SSE-NEXT: pxor %xmm0, %xmm1
; X86-SSE-NEXT: pshufd {{.*#+}} xmm0 = xmm1[2,3,2,3]
; X86-SSE-NEXT: pxor %xmm1, %xmm0
; X86-SSE-NEXT: pshufd {{.*#+}} xmm1 = xmm0[1,1,1,1]
; X86-SSE-NEXT: pxor %xmm0, %xmm1
; X86-SSE-NEXT: movd %xmm1, %eax
; X86-SSE-NEXT: movl %ebp, %esp
; X86-SSE-NEXT: popl %ebp
; X86-SSE-NEXT: retl
;
; X64-SSE-LABEL: test_v32i32:
; X64-SSE: # %bb.0:
; X64-SSE-NEXT: pxor %xmm6, %xmm2
; X64-SSE-NEXT: pxor %xmm4, %xmm0
; X64-SSE-NEXT: pxor %xmm2, %xmm0
; X64-SSE-NEXT: pxor %xmm7, %xmm3
; X64-SSE-NEXT: pxor %xmm5, %xmm1
; X64-SSE-NEXT: pxor %xmm3, %xmm1
; X64-SSE-NEXT: pxor %xmm0, %xmm1
; X64-SSE-NEXT: pshufd {{.*#+}} xmm0 = xmm1[2,3,2,3]
; X64-SSE-NEXT: pxor %xmm1, %xmm0
; X64-SSE-NEXT: pshufd {{.*#+}} xmm1 = xmm0[1,1,1,1]
; X64-SSE-NEXT: pxor %xmm0, %xmm1
; X64-SSE-NEXT: movd %xmm1, %eax
; X64-SSE-NEXT: retq
;
; AVX1-LABEL: test_v32i32:
; AVX1: # %bb.0:
; AVX1-NEXT: vxorps %ymm3, %ymm1, %ymm1
; AVX1-NEXT: vxorps %ymm2, %ymm0, %ymm0
; AVX1-NEXT: vxorps %ymm1, %ymm0, %ymm0
; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm1
; AVX1-NEXT: vxorps %xmm1, %xmm0, %xmm0
; AVX1-NEXT: vshufps {{.*#+}} xmm1 = xmm0[2,3,2,3]
; AVX1-NEXT: vxorps %xmm1, %xmm0, %xmm0
; AVX1-NEXT: vshufps {{.*#+}} xmm1 = xmm0[1,1,1,1]
; AVX1-NEXT: vxorps %xmm1, %xmm0, %xmm0
; AVX1-NEXT: vmovd %xmm0, %eax
; AVX1-NEXT: vzeroupper
; AVX1-NEXT: retq
;
; AVX2-LABEL: test_v32i32:
; AVX2: # %bb.0:
; AVX2-NEXT: vpxor %ymm3, %ymm1, %ymm1
; AVX2-NEXT: vpxor %ymm2, %ymm0, %ymm0
; AVX2-NEXT: vpxor %ymm1, %ymm0, %ymm0
; AVX2-NEXT: vextracti128 $1, %ymm0, %xmm1
; AVX2-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX2-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
; AVX2-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX2-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[1,1,1,1]
; AVX2-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX2-NEXT: vmovd %xmm0, %eax
; AVX2-NEXT: vzeroupper
; AVX2-NEXT: retq
;
; AVX512-LABEL: test_v32i32:
; AVX512: # %bb.0:
; AVX512-NEXT: vpxord %zmm1, %zmm0, %zmm0
; AVX512-NEXT: vextracti64x4 $1, %zmm0, %ymm1
; AVX512-NEXT: vpxord %zmm1, %zmm0, %zmm0
; AVX512-NEXT: vextracti128 $1, %ymm0, %xmm1
; AVX512-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX512-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
; AVX512-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX512-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[1,1,1,1]
; AVX512-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX512-NEXT: vmovd %xmm0, %eax
; AVX512-NEXT: vzeroupper
; AVX512-NEXT: retq
%1 = call i32 @llvm.vector.reduce.xor.v32i32(<32 x i32> %a0)
ret i32 %1
}
;
; vXi16
;
define i16 @test_v2i16(<2 x i16> %a0) nounwind {
; SSE-LABEL: test_v2i16:
; SSE: # %bb.0:
; SSE-NEXT: movdqa %xmm0, %xmm1
; SSE-NEXT: psrld $16, %xmm1
; SSE-NEXT: pxor %xmm0, %xmm1
; SSE-NEXT: movd %xmm1, %eax
; SSE-NEXT: # kill: def $ax killed $ax killed $eax
; SSE-NEXT: ret{{[l|q]}}
;
; AVX-LABEL: test_v2i16:
; AVX: # %bb.0:
; AVX-NEXT: vpsrld $16, %xmm0, %xmm1
; AVX-NEXT: vpxor %xmm0, %xmm1, %xmm0
; AVX-NEXT: vmovd %xmm0, %eax
; AVX-NEXT: # kill: def $ax killed $ax killed $eax
; AVX-NEXT: retq
%1 = call i16 @llvm.vector.reduce.xor.v2i16(<2 x i16> %a0)
ret i16 %1
}
define i16 @test_v4i16(<4 x i16> %a0) nounwind {
; SSE-LABEL: test_v4i16:
; SSE: # %bb.0:
; SSE-NEXT: pshufd {{.*#+}} xmm1 = xmm0[1,1,1,1]
; SSE-NEXT: pxor %xmm0, %xmm1
; SSE-NEXT: movdqa %xmm1, %xmm0
; SSE-NEXT: psrld $16, %xmm0
; SSE-NEXT: pxor %xmm1, %xmm0
; SSE-NEXT: movd %xmm0, %eax
; SSE-NEXT: # kill: def $ax killed $ax killed $eax
; SSE-NEXT: ret{{[l|q]}}
;
; AVX-LABEL: test_v4i16:
; AVX: # %bb.0:
; AVX-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[1,1,1,1]
; AVX-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX-NEXT: vpsrld $16, %xmm0, %xmm1
; AVX-NEXT: vpxor %xmm0, %xmm1, %xmm0
; AVX-NEXT: vmovd %xmm0, %eax
; AVX-NEXT: # kill: def $ax killed $ax killed $eax
; AVX-NEXT: retq
%1 = call i16 @llvm.vector.reduce.xor.v4i16(<4 x i16> %a0)
ret i16 %1
}
define i16 @test_v8i16(<8 x i16> %a0) nounwind {
; SSE-LABEL: test_v8i16:
; SSE: # %bb.0:
; SSE-NEXT: pshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
; SSE-NEXT: pxor %xmm0, %xmm1
; SSE-NEXT: pshufd {{.*#+}} xmm0 = xmm1[1,1,1,1]
; SSE-NEXT: pxor %xmm1, %xmm0
; SSE-NEXT: movdqa %xmm0, %xmm1
; SSE-NEXT: psrld $16, %xmm1
; SSE-NEXT: pxor %xmm0, %xmm1
; SSE-NEXT: movd %xmm1, %eax
; SSE-NEXT: # kill: def $ax killed $ax killed $eax
; SSE-NEXT: ret{{[l|q]}}
;
; AVX-LABEL: test_v8i16:
; AVX: # %bb.0:
; AVX-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
; AVX-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[1,1,1,1]
; AVX-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX-NEXT: vpsrld $16, %xmm0, %xmm1
; AVX-NEXT: vpxor %xmm0, %xmm1, %xmm0
; AVX-NEXT: vmovd %xmm0, %eax
; AVX-NEXT: # kill: def $ax killed $ax killed $eax
; AVX-NEXT: retq
%1 = call i16 @llvm.vector.reduce.xor.v8i16(<8 x i16> %a0)
ret i16 %1
}
define i16 @test_v16i16(<16 x i16> %a0) nounwind {
; SSE-LABEL: test_v16i16:
; SSE: # %bb.0:
; SSE-NEXT: pxor %xmm1, %xmm0
; SSE-NEXT: pshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
; SSE-NEXT: pxor %xmm0, %xmm1
; SSE-NEXT: pshufd {{.*#+}} xmm0 = xmm1[1,1,1,1]
; SSE-NEXT: pxor %xmm1, %xmm0
; SSE-NEXT: movdqa %xmm0, %xmm1
; SSE-NEXT: psrld $16, %xmm1
; SSE-NEXT: pxor %xmm0, %xmm1
; SSE-NEXT: movd %xmm1, %eax
; SSE-NEXT: # kill: def $ax killed $ax killed $eax
; SSE-NEXT: ret{{[l|q]}}
;
; AVX1-LABEL: test_v16i16:
; AVX1: # %bb.0:
; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm1
; AVX1-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX1-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
; AVX1-NEXT: vpxor %xmm0, %xmm1, %xmm0
; AVX1-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[1,1,1,1]
; AVX1-NEXT: vpxor %xmm0, %xmm1, %xmm0
; AVX1-NEXT: vpsrld $16, %xmm0, %xmm1
; AVX1-NEXT: vpxor %xmm0, %xmm1, %xmm0
; AVX1-NEXT: vmovd %xmm0, %eax
; AVX1-NEXT: # kill: def $ax killed $ax killed $eax
; AVX1-NEXT: vzeroupper
; AVX1-NEXT: retq
;
; AVX2-LABEL: test_v16i16:
; AVX2: # %bb.0:
; AVX2-NEXT: vextracti128 $1, %ymm0, %xmm1
; AVX2-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX2-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
; AVX2-NEXT: vpxor %xmm0, %xmm1, %xmm0
; AVX2-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[1,1,1,1]
; AVX2-NEXT: vpxor %xmm0, %xmm1, %xmm0
; AVX2-NEXT: vpsrld $16, %xmm0, %xmm1
; AVX2-NEXT: vpxor %xmm0, %xmm1, %xmm0
; AVX2-NEXT: vmovd %xmm0, %eax
; AVX2-NEXT: # kill: def $ax killed $ax killed $eax
; AVX2-NEXT: vzeroupper
; AVX2-NEXT: retq
;
; AVX512-LABEL: test_v16i16:
; AVX512: # %bb.0:
; AVX512-NEXT: vextracti128 $1, %ymm0, %xmm1
; AVX512-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX512-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
; AVX512-NEXT: vpxor %xmm0, %xmm1, %xmm0
; AVX512-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[1,1,1,1]
; AVX512-NEXT: vpxor %xmm0, %xmm1, %xmm0
; AVX512-NEXT: vpsrld $16, %xmm0, %xmm1
; AVX512-NEXT: vpxor %xmm0, %xmm1, %xmm0
; AVX512-NEXT: vmovd %xmm0, %eax
; AVX512-NEXT: # kill: def $ax killed $ax killed $eax
; AVX512-NEXT: vzeroupper
; AVX512-NEXT: retq
%1 = call i16 @llvm.vector.reduce.xor.v16i16(<16 x i16> %a0)
ret i16 %1
}
define i16 @test_v32i16(<32 x i16> %a0) nounwind {
; X86-SSE-LABEL: test_v32i16:
; X86-SSE: # %bb.0:
; X86-SSE-NEXT: pushl %ebp
; X86-SSE-NEXT: movl %esp, %ebp
; X86-SSE-NEXT: andl $-16, %esp
; X86-SSE-NEXT: subl $16, %esp
; X86-SSE-NEXT: pxor %xmm2, %xmm0
; X86-SSE-NEXT: pxor 8(%ebp), %xmm1
; X86-SSE-NEXT: pxor %xmm0, %xmm1
; X86-SSE-NEXT: pshufd {{.*#+}} xmm0 = xmm1[2,3,2,3]
; X86-SSE-NEXT: pxor %xmm1, %xmm0
; X86-SSE-NEXT: pshufd {{.*#+}} xmm1 = xmm0[1,1,1,1]
; X86-SSE-NEXT: pxor %xmm0, %xmm1
; X86-SSE-NEXT: movdqa %xmm1, %xmm0
; X86-SSE-NEXT: psrld $16, %xmm0
; X86-SSE-NEXT: pxor %xmm1, %xmm0
; X86-SSE-NEXT: movd %xmm0, %eax
; X86-SSE-NEXT: # kill: def $ax killed $ax killed $eax
; X86-SSE-NEXT: movl %ebp, %esp
; X86-SSE-NEXT: popl %ebp
; X86-SSE-NEXT: retl
;
; X64-SSE-LABEL: test_v32i16:
; X64-SSE: # %bb.0:
; X64-SSE-NEXT: pxor %xmm3, %xmm1
; X64-SSE-NEXT: pxor %xmm2, %xmm0
; X64-SSE-NEXT: pxor %xmm1, %xmm0
; X64-SSE-NEXT: pshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
; X64-SSE-NEXT: pxor %xmm0, %xmm1
; X64-SSE-NEXT: pshufd {{.*#+}} xmm0 = xmm1[1,1,1,1]
; X64-SSE-NEXT: pxor %xmm1, %xmm0
; X64-SSE-NEXT: movdqa %xmm0, %xmm1
; X64-SSE-NEXT: psrld $16, %xmm1
; X64-SSE-NEXT: pxor %xmm0, %xmm1
; X64-SSE-NEXT: movd %xmm1, %eax
; X64-SSE-NEXT: # kill: def $ax killed $ax killed $eax
; X64-SSE-NEXT: retq
;
; AVX1-LABEL: test_v32i16:
; AVX1: # %bb.0:
; AVX1-NEXT: vxorps %ymm1, %ymm0, %ymm0
; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm1
; AVX1-NEXT: vxorps %xmm1, %xmm0, %xmm0
; AVX1-NEXT: vshufps {{.*#+}} xmm1 = xmm0[2,3,2,3]
; AVX1-NEXT: vxorps %xmm0, %xmm1, %xmm0
; AVX1-NEXT: vshufps {{.*#+}} xmm1 = xmm0[1,1,1,1]
; AVX1-NEXT: vxorps %xmm0, %xmm1, %xmm0
; AVX1-NEXT: vpsrld $16, %xmm0, %xmm1
; AVX1-NEXT: vpxor %xmm0, %xmm1, %xmm0
; AVX1-NEXT: vmovd %xmm0, %eax
; AVX1-NEXT: # kill: def $ax killed $ax killed $eax
; AVX1-NEXT: vzeroupper
; AVX1-NEXT: retq
;
; AVX2-LABEL: test_v32i16:
; AVX2: # %bb.0:
; AVX2-NEXT: vpxor %ymm1, %ymm0, %ymm0
; AVX2-NEXT: vextracti128 $1, %ymm0, %xmm1
; AVX2-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX2-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
; AVX2-NEXT: vpxor %xmm0, %xmm1, %xmm0
; AVX2-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[1,1,1,1]
; AVX2-NEXT: vpxor %xmm0, %xmm1, %xmm0
; AVX2-NEXT: vpsrld $16, %xmm0, %xmm1
; AVX2-NEXT: vpxor %xmm0, %xmm1, %xmm0
; AVX2-NEXT: vmovd %xmm0, %eax
; AVX2-NEXT: # kill: def $ax killed $ax killed $eax
; AVX2-NEXT: vzeroupper
; AVX2-NEXT: retq
;
; AVX512-LABEL: test_v32i16:
; AVX512: # %bb.0:
; AVX512-NEXT: vextracti64x4 $1, %zmm0, %ymm1
; AVX512-NEXT: vpxorq %zmm1, %zmm0, %zmm0
; AVX512-NEXT: vextracti128 $1, %ymm0, %xmm1
; AVX512-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX512-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
; AVX512-NEXT: vpxor %xmm0, %xmm1, %xmm0
; AVX512-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[1,1,1,1]
; AVX512-NEXT: vpxor %xmm0, %xmm1, %xmm0
; AVX512-NEXT: vpsrld $16, %xmm0, %xmm1
; AVX512-NEXT: vpxor %xmm0, %xmm1, %xmm0
; AVX512-NEXT: vmovd %xmm0, %eax
; AVX512-NEXT: # kill: def $ax killed $ax killed $eax
; AVX512-NEXT: vzeroupper
; AVX512-NEXT: retq
%1 = call i16 @llvm.vector.reduce.xor.v32i16(<32 x i16> %a0)
ret i16 %1
}
define i16 @test_v64i16(<64 x i16> %a0) nounwind {
; X86-SSE-LABEL: test_v64i16:
; X86-SSE: # %bb.0:
; X86-SSE-NEXT: pushl %ebp
; X86-SSE-NEXT: movl %esp, %ebp
; X86-SSE-NEXT: andl $-16, %esp
; X86-SSE-NEXT: subl $16, %esp
; X86-SSE-NEXT: movdqa 8(%ebp), %xmm3
; X86-SSE-NEXT: pxor 56(%ebp), %xmm2
; X86-SSE-NEXT: pxor 24(%ebp), %xmm0
; X86-SSE-NEXT: pxor %xmm2, %xmm0
; X86-SSE-NEXT: pxor 72(%ebp), %xmm3
; X86-SSE-NEXT: pxor 40(%ebp), %xmm1
; X86-SSE-NEXT: pxor %xmm3, %xmm1
; X86-SSE-NEXT: pxor %xmm0, %xmm1
; X86-SSE-NEXT: pshufd {{.*#+}} xmm0 = xmm1[2,3,2,3]
; X86-SSE-NEXT: pxor %xmm1, %xmm0
; X86-SSE-NEXT: pshufd {{.*#+}} xmm1 = xmm0[1,1,1,1]
; X86-SSE-NEXT: pxor %xmm0, %xmm1
; X86-SSE-NEXT: movdqa %xmm1, %xmm0
; X86-SSE-NEXT: psrld $16, %xmm0
; X86-SSE-NEXT: pxor %xmm1, %xmm0
; X86-SSE-NEXT: movd %xmm0, %eax
; X86-SSE-NEXT: # kill: def $ax killed $ax killed $eax
; X86-SSE-NEXT: movl %ebp, %esp
; X86-SSE-NEXT: popl %ebp
; X86-SSE-NEXT: retl
;
; X64-SSE-LABEL: test_v64i16:
; X64-SSE: # %bb.0:
; X64-SSE-NEXT: pxor %xmm6, %xmm2
; X64-SSE-NEXT: pxor %xmm4, %xmm0
; X64-SSE-NEXT: pxor %xmm2, %xmm0
; X64-SSE-NEXT: pxor %xmm7, %xmm3
; X64-SSE-NEXT: pxor %xmm5, %xmm1
; X64-SSE-NEXT: pxor %xmm3, %xmm1
; X64-SSE-NEXT: pxor %xmm0, %xmm1
; X64-SSE-NEXT: pshufd {{.*#+}} xmm0 = xmm1[2,3,2,3]
; X64-SSE-NEXT: pxor %xmm1, %xmm0
; X64-SSE-NEXT: pshufd {{.*#+}} xmm1 = xmm0[1,1,1,1]
; X64-SSE-NEXT: pxor %xmm0, %xmm1
; X64-SSE-NEXT: movdqa %xmm1, %xmm0
; X64-SSE-NEXT: psrld $16, %xmm0
; X64-SSE-NEXT: pxor %xmm1, %xmm0
; X64-SSE-NEXT: movd %xmm0, %eax
; X64-SSE-NEXT: # kill: def $ax killed $ax killed $eax
; X64-SSE-NEXT: retq
;
; AVX1-LABEL: test_v64i16:
; AVX1: # %bb.0:
; AVX1-NEXT: vxorps %ymm3, %ymm1, %ymm1
; AVX1-NEXT: vxorps %ymm2, %ymm0, %ymm0
; AVX1-NEXT: vxorps %ymm1, %ymm0, %ymm0
; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm1
; AVX1-NEXT: vxorps %xmm1, %xmm0, %xmm0
; AVX1-NEXT: vshufps {{.*#+}} xmm1 = xmm0[2,3,2,3]
; AVX1-NEXT: vxorps %xmm0, %xmm1, %xmm0
; AVX1-NEXT: vshufps {{.*#+}} xmm1 = xmm0[1,1,1,1]
; AVX1-NEXT: vxorps %xmm0, %xmm1, %xmm0
; AVX1-NEXT: vpsrld $16, %xmm0, %xmm1
; AVX1-NEXT: vpxor %xmm0, %xmm1, %xmm0
; AVX1-NEXT: vmovd %xmm0, %eax
; AVX1-NEXT: # kill: def $ax killed $ax killed $eax
; AVX1-NEXT: vzeroupper
; AVX1-NEXT: retq
;
; AVX2-LABEL: test_v64i16:
; AVX2: # %bb.0:
; AVX2-NEXT: vpxor %ymm3, %ymm1, %ymm1
; AVX2-NEXT: vpxor %ymm2, %ymm0, %ymm0
; AVX2-NEXT: vpxor %ymm1, %ymm0, %ymm0
; AVX2-NEXT: vextracti128 $1, %ymm0, %xmm1
; AVX2-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX2-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
; AVX2-NEXT: vpxor %xmm0, %xmm1, %xmm0
; AVX2-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[1,1,1,1]
; AVX2-NEXT: vpxor %xmm0, %xmm1, %xmm0
; AVX2-NEXT: vpsrld $16, %xmm0, %xmm1
; AVX2-NEXT: vpxor %xmm0, %xmm1, %xmm0
; AVX2-NEXT: vmovd %xmm0, %eax
; AVX2-NEXT: # kill: def $ax killed $ax killed $eax
; AVX2-NEXT: vzeroupper
; AVX2-NEXT: retq
;
; AVX512-LABEL: test_v64i16:
; AVX512: # %bb.0:
; AVX512-NEXT: vpxorq %zmm1, %zmm0, %zmm0
; AVX512-NEXT: vextracti64x4 $1, %zmm0, %ymm1
; AVX512-NEXT: vpxorq %zmm1, %zmm0, %zmm0
; AVX512-NEXT: vextracti128 $1, %ymm0, %xmm1
; AVX512-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX512-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
; AVX512-NEXT: vpxor %xmm0, %xmm1, %xmm0
; AVX512-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[1,1,1,1]
; AVX512-NEXT: vpxor %xmm0, %xmm1, %xmm0
; AVX512-NEXT: vpsrld $16, %xmm0, %xmm1
; AVX512-NEXT: vpxor %xmm0, %xmm1, %xmm0
; AVX512-NEXT: vmovd %xmm0, %eax
; AVX512-NEXT: # kill: def $ax killed $ax killed $eax
; AVX512-NEXT: vzeroupper
; AVX512-NEXT: retq
%1 = call i16 @llvm.vector.reduce.xor.v64i16(<64 x i16> %a0)
ret i16 %1
}
;
; vXi8
;
define i8 @test_v2i8(<2 x i8> %a0) nounwind {
; SSE-LABEL: test_v2i8:
; SSE: # %bb.0:
; SSE-NEXT: movdqa %xmm0, %xmm1
; SSE-NEXT: psrlw $8, %xmm1
; SSE-NEXT: pxor %xmm0, %xmm1
; SSE-NEXT: movd %xmm1, %eax
; SSE-NEXT: # kill: def $al killed $al killed $eax
; SSE-NEXT: ret{{[l|q]}}
;
; AVX-LABEL: test_v2i8:
; AVX: # %bb.0:
; AVX-NEXT: vpsrlw $8, %xmm0, %xmm1
; AVX-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX-NEXT: vmovd %xmm0, %eax
; AVX-NEXT: # kill: def $al killed $al killed $eax
; AVX-NEXT: retq
%1 = call i8 @llvm.vector.reduce.xor.v2i8(<2 x i8> %a0)
ret i8 %1
}
define i8 @test_v4i8(<4 x i8> %a0) nounwind {
; SSE-LABEL: test_v4i8:
; SSE: # %bb.0:
; SSE-NEXT: movdqa %xmm0, %xmm1
; SSE-NEXT: psrld $16, %xmm1
; SSE-NEXT: pxor %xmm0, %xmm1
; SSE-NEXT: movdqa %xmm1, %xmm0
; SSE-NEXT: psrlw $8, %xmm0
; SSE-NEXT: pxor %xmm1, %xmm0
; SSE-NEXT: movd %xmm0, %eax
; SSE-NEXT: # kill: def $al killed $al killed $eax
; SSE-NEXT: ret{{[l|q]}}
;
; AVX-LABEL: test_v4i8:
; AVX: # %bb.0:
; AVX-NEXT: vpsrld $16, %xmm0, %xmm1
; AVX-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX-NEXT: vpsrlw $8, %xmm0, %xmm1
; AVX-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX-NEXT: vmovd %xmm0, %eax
; AVX-NEXT: # kill: def $al killed $al killed $eax
; AVX-NEXT: retq
%1 = call i8 @llvm.vector.reduce.xor.v4i8(<4 x i8> %a0)
ret i8 %1
}
define i8 @test_v8i8(<8 x i8> %a0) nounwind {
; SSE-LABEL: test_v8i8:
; SSE: # %bb.0:
; SSE-NEXT: pshufd {{.*#+}} xmm1 = xmm0[1,1,1,1]
; SSE-NEXT: pxor %xmm0, %xmm1
; SSE-NEXT: movdqa %xmm1, %xmm0
; SSE-NEXT: psrld $16, %xmm0
; SSE-NEXT: pxor %xmm1, %xmm0
; SSE-NEXT: movdqa %xmm0, %xmm1
; SSE-NEXT: psrlw $8, %xmm1
; SSE-NEXT: pxor %xmm0, %xmm1
; SSE-NEXT: movd %xmm1, %eax
; SSE-NEXT: # kill: def $al killed $al killed $eax
; SSE-NEXT: ret{{[l|q]}}
;
; AVX-LABEL: test_v8i8:
; AVX: # %bb.0:
; AVX-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[1,1,1,1]
; AVX-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX-NEXT: vpsrld $16, %xmm0, %xmm1
; AVX-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX-NEXT: vpsrlw $8, %xmm0, %xmm1
; AVX-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX-NEXT: vmovd %xmm0, %eax
; AVX-NEXT: # kill: def $al killed $al killed $eax
; AVX-NEXT: retq
%1 = call i8 @llvm.vector.reduce.xor.v8i8(<8 x i8> %a0)
ret i8 %1
}
define i8 @test_v16i8(<16 x i8> %a0) nounwind {
; SSE-LABEL: test_v16i8:
; SSE: # %bb.0:
; SSE-NEXT: pshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
; SSE-NEXT: pxor %xmm0, %xmm1
; SSE-NEXT: pshufd {{.*#+}} xmm0 = xmm1[1,1,1,1]
; SSE-NEXT: pxor %xmm1, %xmm0
; SSE-NEXT: movdqa %xmm0, %xmm1
; SSE-NEXT: psrld $16, %xmm1
; SSE-NEXT: pxor %xmm0, %xmm1
; SSE-NEXT: movdqa %xmm1, %xmm0
; SSE-NEXT: psrlw $8, %xmm0
; SSE-NEXT: pxor %xmm1, %xmm0
; SSE-NEXT: movd %xmm0, %eax
; SSE-NEXT: # kill: def $al killed $al killed $eax
; SSE-NEXT: ret{{[l|q]}}
;
; AVX-LABEL: test_v16i8:
; AVX: # %bb.0:
; AVX-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
; AVX-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[1,1,1,1]
; AVX-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX-NEXT: vpsrld $16, %xmm0, %xmm1
; AVX-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX-NEXT: vpsrlw $8, %xmm0, %xmm1
; AVX-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX-NEXT: vmovd %xmm0, %eax
; AVX-NEXT: # kill: def $al killed $al killed $eax
; AVX-NEXT: retq
%1 = call i8 @llvm.vector.reduce.xor.v16i8(<16 x i8> %a0)
ret i8 %1
}
define i8 @test_v32i8(<32 x i8> %a0) nounwind {
; SSE-LABEL: test_v32i8:
; SSE: # %bb.0:
; SSE-NEXT: pxor %xmm1, %xmm0
; SSE-NEXT: pshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
; SSE-NEXT: pxor %xmm0, %xmm1
; SSE-NEXT: pshufd {{.*#+}} xmm0 = xmm1[1,1,1,1]
; SSE-NEXT: pxor %xmm1, %xmm0
; SSE-NEXT: movdqa %xmm0, %xmm1
; SSE-NEXT: psrld $16, %xmm1
; SSE-NEXT: pxor %xmm0, %xmm1
; SSE-NEXT: movdqa %xmm1, %xmm0
; SSE-NEXT: psrlw $8, %xmm0
; SSE-NEXT: pxor %xmm1, %xmm0
; SSE-NEXT: movd %xmm0, %eax
; SSE-NEXT: # kill: def $al killed $al killed $eax
; SSE-NEXT: ret{{[l|q]}}
;
; AVX1-LABEL: test_v32i8:
; AVX1: # %bb.0:
; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm1
; AVX1-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX1-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
; AVX1-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX1-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[1,1,1,1]
; AVX1-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX1-NEXT: vpsrld $16, %xmm0, %xmm1
; AVX1-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX1-NEXT: vpsrlw $8, %xmm0, %xmm1
; AVX1-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX1-NEXT: vmovd %xmm0, %eax
; AVX1-NEXT: # kill: def $al killed $al killed $eax
; AVX1-NEXT: vzeroupper
; AVX1-NEXT: retq
;
; AVX2-LABEL: test_v32i8:
; AVX2: # %bb.0:
; AVX2-NEXT: vextracti128 $1, %ymm0, %xmm1
; AVX2-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX2-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
; AVX2-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX2-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[1,1,1,1]
; AVX2-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX2-NEXT: vpsrld $16, %xmm0, %xmm1
; AVX2-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX2-NEXT: vpsrlw $8, %xmm0, %xmm1
; AVX2-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX2-NEXT: vmovd %xmm0, %eax
; AVX2-NEXT: # kill: def $al killed $al killed $eax
; AVX2-NEXT: vzeroupper
; AVX2-NEXT: retq
;
; AVX512-LABEL: test_v32i8:
; AVX512: # %bb.0:
; AVX512-NEXT: vextracti128 $1, %ymm0, %xmm1
; AVX512-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX512-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
; AVX512-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX512-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[1,1,1,1]
; AVX512-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX512-NEXT: vpsrld $16, %xmm0, %xmm1
; AVX512-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX512-NEXT: vpsrlw $8, %xmm0, %xmm1
; AVX512-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX512-NEXT: vmovd %xmm0, %eax
; AVX512-NEXT: # kill: def $al killed $al killed $eax
; AVX512-NEXT: vzeroupper
; AVX512-NEXT: retq
%1 = call i8 @llvm.vector.reduce.xor.v32i8(<32 x i8> %a0)
ret i8 %1
}
define i8 @test_v64i8(<64 x i8> %a0) nounwind {
; X86-SSE-LABEL: test_v64i8:
; X86-SSE: # %bb.0:
; X86-SSE-NEXT: pushl %ebp
; X86-SSE-NEXT: movl %esp, %ebp
; X86-SSE-NEXT: andl $-16, %esp
; X86-SSE-NEXT: subl $16, %esp
; X86-SSE-NEXT: pxor %xmm2, %xmm0
; X86-SSE-NEXT: pxor 8(%ebp), %xmm1
; X86-SSE-NEXT: pxor %xmm0, %xmm1
; X86-SSE-NEXT: pshufd {{.*#+}} xmm0 = xmm1[2,3,2,3]
; X86-SSE-NEXT: pxor %xmm1, %xmm0
; X86-SSE-NEXT: pshufd {{.*#+}} xmm1 = xmm0[1,1,1,1]
; X86-SSE-NEXT: pxor %xmm0, %xmm1
; X86-SSE-NEXT: movdqa %xmm1, %xmm0
; X86-SSE-NEXT: psrld $16, %xmm0
; X86-SSE-NEXT: pxor %xmm1, %xmm0
; X86-SSE-NEXT: movdqa %xmm0, %xmm1
; X86-SSE-NEXT: psrlw $8, %xmm1
; X86-SSE-NEXT: pxor %xmm0, %xmm1
; X86-SSE-NEXT: movd %xmm1, %eax
; X86-SSE-NEXT: # kill: def $al killed $al killed $eax
; X86-SSE-NEXT: movl %ebp, %esp
; X86-SSE-NEXT: popl %ebp
; X86-SSE-NEXT: retl
;
; X64-SSE-LABEL: test_v64i8:
; X64-SSE: # %bb.0:
; X64-SSE-NEXT: pxor %xmm3, %xmm1
; X64-SSE-NEXT: pxor %xmm2, %xmm0
; X64-SSE-NEXT: pxor %xmm1, %xmm0
; X64-SSE-NEXT: pshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
; X64-SSE-NEXT: pxor %xmm0, %xmm1
; X64-SSE-NEXT: pshufd {{.*#+}} xmm0 = xmm1[1,1,1,1]
; X64-SSE-NEXT: pxor %xmm1, %xmm0
; X64-SSE-NEXT: movdqa %xmm0, %xmm1
; X64-SSE-NEXT: psrld $16, %xmm1
; X64-SSE-NEXT: pxor %xmm0, %xmm1
; X64-SSE-NEXT: movdqa %xmm1, %xmm0
; X64-SSE-NEXT: psrlw $8, %xmm0
; X64-SSE-NEXT: pxor %xmm1, %xmm0
; X64-SSE-NEXT: movd %xmm0, %eax
; X64-SSE-NEXT: # kill: def $al killed $al killed $eax
; X64-SSE-NEXT: retq
;
; AVX1-LABEL: test_v64i8:
; AVX1: # %bb.0:
; AVX1-NEXT: vxorps %ymm1, %ymm0, %ymm0
; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm1
; AVX1-NEXT: vxorps %xmm1, %xmm0, %xmm0
; AVX1-NEXT: vshufps {{.*#+}} xmm1 = xmm0[2,3,2,3]
; AVX1-NEXT: vxorps %xmm1, %xmm0, %xmm0
; AVX1-NEXT: vshufps {{.*#+}} xmm1 = xmm0[1,1,1,1]
; AVX1-NEXT: vxorps %xmm1, %xmm0, %xmm0
; AVX1-NEXT: vpsrld $16, %xmm0, %xmm1
; AVX1-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX1-NEXT: vpsrlw $8, %xmm0, %xmm1
; AVX1-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX1-NEXT: vmovd %xmm0, %eax
; AVX1-NEXT: # kill: def $al killed $al killed $eax
; AVX1-NEXT: vzeroupper
; AVX1-NEXT: retq
;
; AVX2-LABEL: test_v64i8:
; AVX2: # %bb.0:
; AVX2-NEXT: vpxor %ymm1, %ymm0, %ymm0
; AVX2-NEXT: vextracti128 $1, %ymm0, %xmm1
; AVX2-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX2-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
; AVX2-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX2-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[1,1,1,1]
; AVX2-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX2-NEXT: vpsrld $16, %xmm0, %xmm1
; AVX2-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX2-NEXT: vpsrlw $8, %xmm0, %xmm1
; AVX2-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX2-NEXT: vmovd %xmm0, %eax
; AVX2-NEXT: # kill: def $al killed $al killed $eax
; AVX2-NEXT: vzeroupper
; AVX2-NEXT: retq
;
; AVX512-LABEL: test_v64i8:
; AVX512: # %bb.0:
; AVX512-NEXT: vextracti64x4 $1, %zmm0, %ymm1
; AVX512-NEXT: vpxorq %zmm1, %zmm0, %zmm0
; AVX512-NEXT: vextracti128 $1, %ymm0, %xmm1
; AVX512-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX512-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
; AVX512-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX512-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[1,1,1,1]
; AVX512-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX512-NEXT: vpsrld $16, %xmm0, %xmm1
; AVX512-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX512-NEXT: vpsrlw $8, %xmm0, %xmm1
; AVX512-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX512-NEXT: vmovd %xmm0, %eax
; AVX512-NEXT: # kill: def $al killed $al killed $eax
; AVX512-NEXT: vzeroupper
; AVX512-NEXT: retq
%1 = call i8 @llvm.vector.reduce.xor.v64i8(<64 x i8> %a0)
ret i8 %1
}
define i8 @test_v128i8(<128 x i8> %a0) nounwind {
; X86-SSE-LABEL: test_v128i8:
; X86-SSE: # %bb.0:
; X86-SSE-NEXT: pushl %ebp
; X86-SSE-NEXT: movl %esp, %ebp
; X86-SSE-NEXT: andl $-16, %esp
; X86-SSE-NEXT: subl $16, %esp
; X86-SSE-NEXT: movdqa 8(%ebp), %xmm3
; X86-SSE-NEXT: pxor 56(%ebp), %xmm2
; X86-SSE-NEXT: pxor 24(%ebp), %xmm0
; X86-SSE-NEXT: pxor %xmm2, %xmm0
; X86-SSE-NEXT: pxor 72(%ebp), %xmm3
; X86-SSE-NEXT: pxor 40(%ebp), %xmm1
; X86-SSE-NEXT: pxor %xmm3, %xmm1
; X86-SSE-NEXT: pxor %xmm0, %xmm1
; X86-SSE-NEXT: pshufd {{.*#+}} xmm0 = xmm1[2,3,2,3]
; X86-SSE-NEXT: pxor %xmm1, %xmm0
; X86-SSE-NEXT: pshufd {{.*#+}} xmm1 = xmm0[1,1,1,1]
; X86-SSE-NEXT: pxor %xmm0, %xmm1
; X86-SSE-NEXT: movdqa %xmm1, %xmm0
; X86-SSE-NEXT: psrld $16, %xmm0
; X86-SSE-NEXT: pxor %xmm1, %xmm0
; X86-SSE-NEXT: movdqa %xmm0, %xmm1
; X86-SSE-NEXT: psrlw $8, %xmm1
; X86-SSE-NEXT: pxor %xmm0, %xmm1
; X86-SSE-NEXT: movd %xmm1, %eax
; X86-SSE-NEXT: # kill: def $al killed $al killed $eax
; X86-SSE-NEXT: movl %ebp, %esp
; X86-SSE-NEXT: popl %ebp
; X86-SSE-NEXT: retl
;
; X64-SSE-LABEL: test_v128i8:
; X64-SSE: # %bb.0:
; X64-SSE-NEXT: pxor %xmm6, %xmm2
; X64-SSE-NEXT: pxor %xmm4, %xmm0
; X64-SSE-NEXT: pxor %xmm2, %xmm0
; X64-SSE-NEXT: pxor %xmm7, %xmm3
; X64-SSE-NEXT: pxor %xmm5, %xmm1
; X64-SSE-NEXT: pxor %xmm3, %xmm1
; X64-SSE-NEXT: pxor %xmm0, %xmm1
; X64-SSE-NEXT: pshufd {{.*#+}} xmm0 = xmm1[2,3,2,3]
; X64-SSE-NEXT: pxor %xmm1, %xmm0
; X64-SSE-NEXT: pshufd {{.*#+}} xmm1 = xmm0[1,1,1,1]
; X64-SSE-NEXT: pxor %xmm0, %xmm1
; X64-SSE-NEXT: movdqa %xmm1, %xmm0
; X64-SSE-NEXT: psrld $16, %xmm0
; X64-SSE-NEXT: pxor %xmm1, %xmm0
; X64-SSE-NEXT: movdqa %xmm0, %xmm1
; X64-SSE-NEXT: psrlw $8, %xmm1
; X64-SSE-NEXT: pxor %xmm0, %xmm1
; X64-SSE-NEXT: movd %xmm1, %eax
; X64-SSE-NEXT: # kill: def $al killed $al killed $eax
; X64-SSE-NEXT: retq
;
; AVX1-LABEL: test_v128i8:
; AVX1: # %bb.0:
; AVX1-NEXT: vxorps %ymm3, %ymm1, %ymm1
; AVX1-NEXT: vxorps %ymm2, %ymm0, %ymm0
; AVX1-NEXT: vxorps %ymm1, %ymm0, %ymm0
; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm1
; AVX1-NEXT: vxorps %xmm1, %xmm0, %xmm0
; AVX1-NEXT: vshufps {{.*#+}} xmm1 = xmm0[2,3,2,3]
; AVX1-NEXT: vxorps %xmm1, %xmm0, %xmm0
; AVX1-NEXT: vshufps {{.*#+}} xmm1 = xmm0[1,1,1,1]
; AVX1-NEXT: vxorps %xmm1, %xmm0, %xmm0
; AVX1-NEXT: vpsrld $16, %xmm0, %xmm1
; AVX1-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX1-NEXT: vpsrlw $8, %xmm0, %xmm1
; AVX1-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX1-NEXT: vmovd %xmm0, %eax
; AVX1-NEXT: # kill: def $al killed $al killed $eax
; AVX1-NEXT: vzeroupper
; AVX1-NEXT: retq
;
; AVX2-LABEL: test_v128i8:
; AVX2: # %bb.0:
; AVX2-NEXT: vpxor %ymm3, %ymm1, %ymm1
; AVX2-NEXT: vpxor %ymm2, %ymm0, %ymm0
; AVX2-NEXT: vpxor %ymm1, %ymm0, %ymm0
; AVX2-NEXT: vextracti128 $1, %ymm0, %xmm1
; AVX2-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX2-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
; AVX2-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX2-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[1,1,1,1]
; AVX2-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX2-NEXT: vpsrld $16, %xmm0, %xmm1
; AVX2-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX2-NEXT: vpsrlw $8, %xmm0, %xmm1
; AVX2-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX2-NEXT: vmovd %xmm0, %eax
; AVX2-NEXT: # kill: def $al killed $al killed $eax
; AVX2-NEXT: vzeroupper
; AVX2-NEXT: retq
;
; AVX512-LABEL: test_v128i8:
; AVX512: # %bb.0:
; AVX512-NEXT: vpxorq %zmm1, %zmm0, %zmm0
; AVX512-NEXT: vextracti64x4 $1, %zmm0, %ymm1
; AVX512-NEXT: vpxorq %zmm1, %zmm0, %zmm0
; AVX512-NEXT: vextracti128 $1, %ymm0, %xmm1
; AVX512-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX512-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
; AVX512-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX512-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[1,1,1,1]
; AVX512-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX512-NEXT: vpsrld $16, %xmm0, %xmm1
; AVX512-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX512-NEXT: vpsrlw $8, %xmm0, %xmm1
; AVX512-NEXT: vpxor %xmm1, %xmm0, %xmm0
; AVX512-NEXT: vmovd %xmm0, %eax
; AVX512-NEXT: # kill: def $al killed $al killed $eax
; AVX512-NEXT: vzeroupper
; AVX512-NEXT: retq
%1 = call i8 @llvm.vector.reduce.xor.v128i8(<128 x i8> %a0)
ret i8 %1
}
declare i64 @llvm.vector.reduce.xor.v2i64(<2 x i64>)
declare i64 @llvm.vector.reduce.xor.v4i64(<4 x i64>)
declare i64 @llvm.vector.reduce.xor.v8i64(<8 x i64>)
declare i64 @llvm.vector.reduce.xor.v16i64(<16 x i64>)
declare i32 @llvm.vector.reduce.xor.v2i32(<2 x i32>)
declare i32 @llvm.vector.reduce.xor.v4i32(<4 x i32>)
declare i32 @llvm.vector.reduce.xor.v8i32(<8 x i32>)
declare i32 @llvm.vector.reduce.xor.v16i32(<16 x i32>)
declare i32 @llvm.vector.reduce.xor.v32i32(<32 x i32>)
declare i16 @llvm.vector.reduce.xor.v2i16(<2 x i16>)
declare i16 @llvm.vector.reduce.xor.v4i16(<4 x i16>)
declare i16 @llvm.vector.reduce.xor.v8i16(<8 x i16>)
declare i16 @llvm.vector.reduce.xor.v16i16(<16 x i16>)
declare i16 @llvm.vector.reduce.xor.v32i16(<32 x i16>)
declare i16 @llvm.vector.reduce.xor.v64i16(<64 x i16>)
declare i8 @llvm.vector.reduce.xor.v2i8(<2 x i8>)
declare i8 @llvm.vector.reduce.xor.v4i8(<4 x i8>)
declare i8 @llvm.vector.reduce.xor.v8i8(<8 x i8>)
declare i8 @llvm.vector.reduce.xor.v16i8(<16 x i8>)
declare i8 @llvm.vector.reduce.xor.v32i8(<32 x i8>)
declare i8 @llvm.vector.reduce.xor.v64i8(<64 x i8>)
declare i8 @llvm.vector.reduce.xor.v128i8(<128 x i8>)
Codebase Browser
llvm