[llvm] r270229 - [X86][AVX] Sync with clang/test/CodeGen/avx-builtins.c
Simon Pilgrim via llvm-commits
llvm-commits at lists.llvm.org
Fri May 20 09:05:55 PDT 2016
Author: rksimon
Date: Fri May 20 11:05:55 2016
New Revision: 270229
URL: http://llvm.org/viewvc/llvm-project?rev=270229&view=rev
Log:
[X86][AVX] Sync with clang/test/CodeGen/avx-builtins.c
Modified:
llvm/trunk/test/CodeGen/X86/avx-intrinsics-fast-isel.ll
Modified: llvm/trunk/test/CodeGen/X86/avx-intrinsics-fast-isel.ll
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/X86/avx-intrinsics-fast-isel.ll?rev=270229&r1=270228&r2=270229&view=diff
==============================================================================
--- llvm/trunk/test/CodeGen/X86/avx-intrinsics-fast-isel.ll (original)
+++ llvm/trunk/test/CodeGen/X86/avx-intrinsics-fast-isel.ll Fri May 20 11:05:55 2016
@@ -4,7 +4,7 @@
; NOTE: This should use IR equivalent to what is generated by clang/test/CodeGen/avx-builtins.c
-define <4 x double> @test_mm256_add_pd(<4 x double> %a0, <4 x double> %a1) {
+define <4 x double> @test_mm256_add_pd(<4 x double> %a0, <4 x double> %a1) nounwind {
; X32-LABEL: test_mm256_add_pd:
; X32: # BB#0:
; X32-NEXT: vaddpd %ymm1, %ymm0, %ymm0
@@ -18,7 +18,7 @@ define <4 x double> @test_mm256_add_pd(<
ret <4 x double> %res
}
-define <8 x float> @test_mm256_add_ps(<8 x float> %a0, <8 x float> %a1) {
+define <8 x float> @test_mm256_add_ps(<8 x float> %a0, <8 x float> %a1) nounwind {
; X32-LABEL: test_mm256_add_ps:
; X32: # BB#0:
; X32-NEXT: vaddps %ymm1, %ymm0, %ymm0
@@ -32,7 +32,7 @@ define <8 x float> @test_mm256_add_ps(<8
ret <8 x float> %res
}
-define <4 x double> @test_mm256_addsub_pd(<4 x double> %a0, <4 x double> %a1) {
+define <4 x double> @test_mm256_addsub_pd(<4 x double> %a0, <4 x double> %a1) nounwind {
; X32-LABEL: test_mm256_addsub_pd:
; X32: # BB#0:
; X32-NEXT: vaddsubpd %ymm1, %ymm0, %ymm0
@@ -47,7 +47,7 @@ define <4 x double> @test_mm256_addsub_p
}
declare <4 x double> @llvm.x86.avx.addsub.pd.256(<4 x double>, <4 x double>) nounwind readnone
-define <8 x float> @test_mm256_addsub_ps(<8 x float> %a0, <8 x float> %a1) {
+define <8 x float> @test_mm256_addsub_ps(<8 x float> %a0, <8 x float> %a1) nounwind {
; X32-LABEL: test_mm256_addsub_ps:
; X32: # BB#0:
; X32-NEXT: vaddsubps %ymm1, %ymm0, %ymm0
@@ -62,7 +62,7 @@ define <8 x float> @test_mm256_addsub_ps
}
declare <8 x float> @llvm.x86.avx.addsub.ps.256(<8 x float>, <8 x float>) nounwind readnone
-define <4 x double> @test_mm256_and_pd(<4 x double> %a0, <4 x double> %a1) {
+define <4 x double> @test_mm256_and_pd(<4 x double> %a0, <4 x double> %a1) nounwind {
; X32-LABEL: test_mm256_and_pd:
; X32: # BB#0:
; X32-NEXT: vandps %ymm1, %ymm0, %ymm0
@@ -79,7 +79,7 @@ define <4 x double> @test_mm256_and_pd(<
ret <4 x double> %bc
}
-define <8 x float> @test_mm256_and_ps(<8 x float> %a0, <8 x float> %a1) {
+define <8 x float> @test_mm256_and_ps(<8 x float> %a0, <8 x float> %a1) nounwind {
; X32-LABEL: test_mm256_and_ps:
; X32: # BB#0:
; X32-NEXT: vandps %ymm1, %ymm0, %ymm0
@@ -96,7 +96,7 @@ define <8 x float> @test_mm256_and_ps(<8
ret <8 x float> %bc
}
-define <4 x double> @test_mm256_andnot_pd(<4 x double> %a0, <4 x double> %a1) {
+define <4 x double> @test_mm256_andnot_pd(<4 x double> %a0, <4 x double> %a1) nounwind {
; X32-LABEL: test_mm256_andnot_pd:
; X32: # BB#0:
; X32-NEXT: vpcmpeqd %xmm2, %xmm2, %xmm2
@@ -120,7 +120,7 @@ define <4 x double> @test_mm256_andnot_p
ret <4 x double> %bc
}
-define <8 x float> @test_mm256_andnot_ps(<8 x float> %a0, <8 x float> %a1) {
+define <8 x float> @test_mm256_andnot_ps(<8 x float> %a0, <8 x float> %a1) nounwind {
; X32-LABEL: test_mm256_andnot_ps:
; X32: # BB#0:
; X32-NEXT: vandnps %ymm1, %ymm0, %ymm0
@@ -138,7 +138,7 @@ define <8 x float> @test_mm256_andnot_ps
ret <8 x float> %bc
}
-define <4 x double> @test_mm256_blend_pd(<4 x double> %a0, <4 x double> %a1) {
+define <4 x double> @test_mm256_blend_pd(<4 x double> %a0, <4 x double> %a1) nounwind {
; X32-LABEL: test_mm256_blend_pd:
; X32: # BB#0:
; X32-NEXT: vblendpd {{.*#+}} ymm0 = ymm0[0],ymm1[1,2],ymm0[3]
@@ -152,7 +152,7 @@ define <4 x double> @test_mm256_blend_pd
ret <4 x double> %res
}
-define <8 x float> @test_mm256_blend_ps(<8 x float> %a0, <8 x float> %a1) {
+define <8 x float> @test_mm256_blend_ps(<8 x float> %a0, <8 x float> %a1) nounwind {
; X32-LABEL: test_mm256_blend_ps:
; X32: # BB#0:
; X32-NEXT: vblendps {{.*#+}} ymm0 = ymm0[0],ymm1[1,2,3],ymm0[4,5,6],ymm1[7]
@@ -166,7 +166,7 @@ define <8 x float> @test_mm256_blend_ps(
ret <8 x float> %res
}
-define <4 x double> @test_mm256_blendv_pd(<4 x double> %a0, <4 x double> %a1, <4 x double> %a2) {
+define <4 x double> @test_mm256_blendv_pd(<4 x double> %a0, <4 x double> %a1, <4 x double> %a2) nounwind {
; X32-LABEL: test_mm256_blendv_pd:
; X32: # BB#0:
; X32-NEXT: vblendvpd %ymm2, %ymm1, %ymm0, %ymm0
@@ -176,12 +176,12 @@ define <4 x double> @test_mm256_blendv_p
; X64: # BB#0:
; X64-NEXT: vblendvpd %ymm2, %ymm1, %ymm0, %ymm0
; X64-NEXT: retq
- %res = call <4 x double> @llvm.x86.avx.blendv.pd.256(<4 x double> %a0, <4 x double> %a1, <4 x double> %a2) ; <<4 x double>> [#uses=1]
+ %res = call <4 x double> @llvm.x86.avx.blendv.pd.256(<4 x double> %a0, <4 x double> %a1, <4 x double> %a2)
ret <4 x double> %res
}
declare <4 x double> @llvm.x86.avx.blendv.pd.256(<4 x double>, <4 x double>, <4 x double>) nounwind readnone
-define <8 x float> @test_mm256_blendv_ps(<8 x float> %a0, <8 x float> %a1, <8 x float> %a2) {
+define <8 x float> @test_mm256_blendv_ps(<8 x float> %a0, <8 x float> %a1, <8 x float> %a2) nounwind {
; X32-LABEL: test_mm256_blendv_ps:
; X32: # BB#0:
; X32-NEXT: vblendvps %ymm2, %ymm1, %ymm0, %ymm0
@@ -191,12 +191,12 @@ define <8 x float> @test_mm256_blendv_ps
; X64: # BB#0:
; X64-NEXT: vblendvps %ymm2, %ymm1, %ymm0, %ymm0
; X64-NEXT: retq
- %res = call <8 x float> @llvm.x86.avx.blendv.ps.256(<8 x float> %a0, <8 x float> %a1, <8 x float> %a2) ; <<8 x float>> [#uses=1]
+ %res = call <8 x float> @llvm.x86.avx.blendv.ps.256(<8 x float> %a0, <8 x float> %a1, <8 x float> %a2)
ret <8 x float> %res
}
declare <8 x float> @llvm.x86.avx.blendv.ps.256(<8 x float>, <8 x float>, <8 x float>) nounwind readnone
-define <4 x double> @test_mm256_broadcast_pd(i8* %a0) {
+define <4 x double> @test_mm256_broadcast_pd(<2 x double>* %a0) nounwind {
; X32-LABEL: test_mm256_broadcast_pd:
; X32: # BB#0:
; X32-NEXT: movl {{[0-9]+}}(%esp), %eax
@@ -207,12 +207,13 @@ define <4 x double> @test_mm256_broadcas
; X64: # BB#0:
; X64-NEXT: vbroadcastf128 (%rdi), %ymm0
; X64-NEXT: retq
- %res = call <4 x double> @llvm.x86.avx.vbroadcastf128.pd.256(i8* %a0) ; <<4 x double>> [#uses=1]
+ %arg0 = bitcast <2 x double>* %a0 to i8*
+ %res = call <4 x double> @llvm.x86.avx.vbroadcastf128.pd.256(i8* %arg0)
ret <4 x double> %res
}
declare <4 x double> @llvm.x86.avx.vbroadcastf128.pd.256(i8*) nounwind readonly
-define <8 x float> @test_mm256_broadcast_ps(i8* %a0) {
+define <8 x float> @test_mm256_broadcast_ps(<4 x float>* %a0) nounwind {
; X32-LABEL: test_mm256_broadcast_ps:
; X32: # BB#0:
; X32-NEXT: movl {{[0-9]+}}(%esp), %eax
@@ -223,257 +224,3050 @@ define <8 x float> @test_mm256_broadcast
; X64: # BB#0:
; X64-NEXT: vbroadcastf128 (%rdi), %ymm0
; X64-NEXT: retq
- %res = call <8 x float> @llvm.x86.avx.vbroadcastf128.ps.256(i8* %a0) ; <<8 x float>> [#uses=1]
+ %arg0 = bitcast <4 x float>* %a0 to i8*
+ %res = call <8 x float> @llvm.x86.avx.vbroadcastf128.ps.256(i8* %arg0)
ret <8 x float> %res
}
declare <8 x float> @llvm.x86.avx.vbroadcastf128.ps.256(i8*) nounwind readonly
-; TODO test_mm256_broadcast_sd
-; TODO test_mm_broadcast_ss
-; TODO test_mm256_broadcast_sd
-; TODO test_mm256_castpd_ps
-; TODO test_mm256_castpd_si256
-; TODO test_mm256_castpd128_pd256
-; TODO test_mm256_castpd256_pd128
-; TODO test_mm256_castps_pd
-; TODO test_mm256_castps_si256
-; TODO test_mm256_castps128_ps256
-; TODO test_mm256_castps256_ps128
-; TODO test_mm256_castsi128_si256
-; TODO test_mm256_castsi256_pd
-; TODO test_mm256_castsi256_ps
-; TODO test_mm256_castsi256_si128
-; TODO test_mm256_ceil_pd
-; TODO test_mm256_ceil_ps
-; TODO test_mm_cmp_pd
-; TODO test_mm256_cmp_pd
-; TODO test _mm_cmp_ps
-; TODO test _mm256_cmp_ps
-; TODO test_mm_cmp_sd
-; TODO test_mm_cmp_ss
-; TODO test_mm256_cvtepi32_pd
-; TODO test_mm256_cvtepi32_ps
-; TODO test_mm256_cvtpd_epi32
-; TODO test_mm256_cvtpd_ps
-; TODO test_mm256_cvtps_epi32
-; TODO test_mm256_cvtps_pd
-; TODO test_mm256_cvttpd_epi32
-; TODO test_mm256_cvttps_epi32
+define <4 x double> @test_mm256_broadcast_sd(double* %a0) nounwind {
+; X32-LABEL: test_mm256_broadcast_sd:
+; X32: # BB#0:
+; X32-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vbroadcastsd (%eax), %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_broadcast_sd:
+; X64: # BB#0:
+; X64-NEXT: vbroadcastsd (%rdi), %ymm0
+; X64-NEXT: retq
+ %ld = load double, double* %a0
+ %ins0 = insertelement <4 x double> undef, double %ld, i32 0
+ %ins1 = insertelement <4 x double> %ins0, double %ld, i32 1
+ %ins2 = insertelement <4 x double> %ins1, double %ld, i32 2
+ %ins3 = insertelement <4 x double> %ins2, double %ld, i32 3
+ ret <4 x double> %ins3
+}
+
+define <4 x float> @test_mm_broadcast_ss(float* %a0) nounwind {
+; X32-LABEL: test_mm_broadcast_ss:
+; X32: # BB#0:
+; X32-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vbroadcastss (%eax), %xmm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm_broadcast_ss:
+; X64: # BB#0:
+; X64-NEXT: vbroadcastss (%rdi), %xmm0
+; X64-NEXT: retq
+ %ld = load float, float* %a0
+ %ins0 = insertelement <4 x float> undef, float %ld, i32 0
+ %ins1 = insertelement <4 x float> %ins0, float %ld, i32 1
+ %ins2 = insertelement <4 x float> %ins1, float %ld, i32 2
+ %ins3 = insertelement <4 x float> %ins2, float %ld, i32 3
+ ret <4 x float> %ins3
+}
+
+define <8 x float> @test_mm256_broadcast_ss(float* %a0) nounwind {
+; X32-LABEL: test_mm256_broadcast_ss:
+; X32: # BB#0:
+; X32-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vbroadcastss (%eax), %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_broadcast_ss:
+; X64: # BB#0:
+; X64-NEXT: vbroadcastss (%rdi), %ymm0
+; X64-NEXT: retq
+ %ld = load float, float* %a0
+ %ins0 = insertelement <8 x float> undef, float %ld, i32 0
+ %ins1 = insertelement <8 x float> %ins0, float %ld, i32 1
+ %ins2 = insertelement <8 x float> %ins1, float %ld, i32 2
+ %ins3 = insertelement <8 x float> %ins2, float %ld, i32 3
+ %ins4 = insertelement <8 x float> %ins3, float %ld, i32 4
+ %ins5 = insertelement <8 x float> %ins4, float %ld, i32 5
+ %ins6 = insertelement <8 x float> %ins5, float %ld, i32 6
+ %ins7 = insertelement <8 x float> %ins6, float %ld, i32 7
+ ret <8 x float> %ins7
+}
+
+define <8 x float> @test_mm256_castpd_ps(<4 x double> %a0) nounwind {
+; X32-LABEL: test_mm256_castpd_ps:
+; X32: # BB#0:
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_castpd_ps:
+; X64: # BB#0:
+; X64-NEXT: retq
+ %res = bitcast <4 x double> %a0 to <8 x float>
+ ret <8 x float> %res
+}
+
+define <4 x i64> @test_mm256_castpd_si256(<4 x double> %a0) nounwind {
+; X32-LABEL: test_mm256_castpd_si256:
+; X32: # BB#0:
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_castpd_si256:
+; X64: # BB#0:
+; X64-NEXT: retq
+ %res = bitcast <4 x double> %a0 to <4 x i64>
+ ret <4 x i64> %res
+}
+
+define <4 x double> @test_mm256_castpd128_pd256(<2 x double> %a0) nounwind {
+; X32-LABEL: test_mm256_castpd128_pd256:
+; X32: # BB#0:
+; X32-NEXT: vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_castpd128_pd256:
+; X64: # BB#0:
+; X64-NEXT: vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; X64-NEXT: retq
+ %res = shufflevector <2 x double> %a0, <2 x double> %a0, <4 x i32> <i32 0, i32 1, i32 undef, i32 undef>
+ ret <4 x double> %res
+}
+
+define <2 x double> @test_mm256_castpd256_pd128(<4 x double> %a0) nounwind {
+; X32-LABEL: test_mm256_castpd256_pd128:
+; X32: # BB#0:
+; X32-NEXT: vzeroupper
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_castpd256_pd128:
+; X64: # BB#0:
+; X64-NEXT: vzeroupper
+; X64-NEXT: retq
+ %res = shufflevector <4 x double> %a0, <4 x double> %a0, <2 x i32> <i32 0, i32 1>
+ ret <2 x double> %res
+}
+
+define <4 x double> @test_mm256_castps_pd(<8 x float> %a0) nounwind {
+; X32-LABEL: test_mm256_castps_pd:
+; X32: # BB#0:
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_castps_pd:
+; X64: # BB#0:
+; X64-NEXT: retq
+ %res = bitcast <8 x float> %a0 to <4 x double>
+ ret <4 x double> %res
+}
+
+define <4 x i64> @test_mm256_castps_si256(<8 x float> %a0) nounwind {
+; X32-LABEL: test_mm256_castps_si256:
+; X32: # BB#0:
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_castps_si256:
+; X64: # BB#0:
+; X64-NEXT: retq
+ %res = bitcast <8 x float> %a0 to <4 x i64>
+ ret <4 x i64> %res
+}
+
+define <8 x float> @test_mm256_castps128_ps256(<4 x float> %a0) nounwind {
+; X32-LABEL: test_mm256_castps128_ps256:
+; X32: # BB#0:
+; X32-NEXT: vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_castps128_ps256:
+; X64: # BB#0:
+; X64-NEXT: vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; X64-NEXT: retq
+ %res = shufflevector <4 x float> %a0, <4 x float> %a0, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef>
+ ret <8 x float> %res
+}
+
+define <4 x float> @test_mm256_castps256_ps128(<8 x float> %a0) nounwind {
+; X32-LABEL: test_mm256_castps256_ps128:
+; X32: # BB#0:
+; X32-NEXT: vzeroupper
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_castps256_ps128:
+; X64: # BB#0:
+; X64-NEXT: vzeroupper
+; X64-NEXT: retq
+ %res = shufflevector <8 x float> %a0, <8 x float> %a0, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
+ ret <4 x float> %res
+}
+
+define <4 x i64> @test_mm256_castsi128_si256(<2 x i64> %a0) nounwind {
+; X32-LABEL: test_mm256_castsi128_si256:
+; X32: # BB#0:
+; X32-NEXT: vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_castsi128_si256:
+; X64: # BB#0:
+; X64-NEXT: vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; X64-NEXT: retq
+ %res = shufflevector <2 x i64> %a0, <2 x i64> %a0, <4 x i32> <i32 0, i32 1, i32 undef, i32 undef>
+ ret <4 x i64> %res
+}
+
+define <4 x double> @test_mm256_castsi256_pd(<4 x i64> %a0) nounwind {
+; X32-LABEL: test_mm256_castsi256_pd:
+; X32: # BB#0:
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_castsi256_pd:
+; X64: # BB#0:
+; X64-NEXT: retq
+ %res = bitcast <4 x i64> %a0 to <4 x double>
+ ret <4 x double> %res
+}
+
+define <8 x float> @test_mm256_castsi256_ps(<4 x i64> %a0) nounwind {
+; X32-LABEL: test_mm256_castsi256_ps:
+; X32: # BB#0:
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_castsi256_ps:
+; X64: # BB#0:
+; X64-NEXT: retq
+ %res = bitcast <4 x i64> %a0 to <8 x float>
+ ret <8 x float> %res
+}
+
+define <2 x i64> @test_mm256_castsi256_si128(<4 x i64> %a0) nounwind {
+; X32-LABEL: test_mm256_castsi256_si128:
+; X32: # BB#0:
+; X32-NEXT: vzeroupper
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_castsi256_si128:
+; X64: # BB#0:
+; X64-NEXT: vzeroupper
+; X64-NEXT: retq
+ %res = shufflevector <4 x i64> %a0, <4 x i64> %a0, <2 x i32> <i32 0, i32 1>
+ ret <2 x i64> %res
+}
+
+define <4 x double> @test_mm256_ceil_pd(<4 x double> %a0) nounwind {
+; X32-LABEL: test_mm256_ceil_pd:
+; X32: # BB#0:
+; X32-NEXT: vroundpd $2, %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_ceil_pd:
+; X64: # BB#0:
+; X64-NEXT: vroundpd $2, %ymm0, %ymm0
+; X64-NEXT: retq
+ %res = call <4 x double> @llvm.x86.avx.round.pd.256(<4 x double> %a0, i32 2)
+ ret <4 x double> %res
+}
+declare <4 x double> @llvm.x86.avx.round.pd.256(<4 x double>, i32) nounwind readnone
+
+define <8 x float> @test_mm256_ceil_ps(<8 x float> %a0) nounwind {
+; X32-LABEL: test_mm256_ceil_ps:
+; X32: # BB#0:
+; X32-NEXT: vroundps $2, %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_ceil_ps:
+; X64: # BB#0:
+; X64-NEXT: vroundps $2, %ymm0, %ymm0
+; X64-NEXT: retq
+ %res = call <8 x float> @llvm.x86.avx.round.ps.256(<8 x float> %a0, i32 2)
+ ret <8 x float> %res
+}
+declare <8 x float> @llvm.x86.avx.round.ps.256(<8 x float>, i32) nounwind readnone
+
+define <2 x double> @test_mm_cmp_pd(<2 x double> %a0, <2 x double> %a1) nounwind {
+; X32-LABEL: test_mm_cmp_pd:
+; X32: # BB#0:
+; X32-NEXT: vcmpgepd %xmm1, %xmm0, %xmm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm_cmp_pd:
+; X64: # BB#0:
+; X64-NEXT: vcmpgepd %xmm1, %xmm0, %xmm0
+; X64-NEXT: retq
+ %res = call <2 x double> @llvm.x86.sse2.cmp.pd(<2 x double> %a0, <2 x double> %a1, i8 13)
+ ret <2 x double> %res
+}
+declare <2 x double> @llvm.x86.sse2.cmp.pd(<2 x double>, <2 x double>, i8) nounwind readnone
+
+define <4 x double> @test_mm256_cmp_pd(<4 x double> %a0, <4 x double> %a1) nounwind {
+; X32-LABEL: test_mm256_cmp_pd:
+; X32: # BB#0:
+; X32-NEXT: vcmpgepd %ymm1, %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_cmp_pd:
+; X64: # BB#0:
+; X64-NEXT: vcmpgepd %ymm1, %ymm0, %ymm0
+; X64-NEXT: retq
+ %res = call <4 x double> @llvm.x86.avx.cmp.pd.256(<4 x double> %a0, <4 x double> %a1, i8 13)
+ ret <4 x double> %res
+}
+declare <4 x double> @llvm.x86.avx.cmp.pd.256(<4 x double>, <4 x double>, i8) nounwind readnone
+
+define <4 x float> @test_mm_cmp_ps(<4 x float> %a0, <4 x float> %a1) nounwind {
+; X32-LABEL: test_mm_cmp_ps:
+; X32: # BB#0:
+; X32-NEXT: vcmpgeps %xmm1, %xmm0, %xmm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm_cmp_ps:
+; X64: # BB#0:
+; X64-NEXT: vcmpgeps %xmm1, %xmm0, %xmm0
+; X64-NEXT: retq
+ %res = call <4 x float> @llvm.x86.sse.cmp.ps(<4 x float> %a0, <4 x float> %a1, i8 13)
+ ret <4 x float> %res
+}
+declare <4 x float> @llvm.x86.sse.cmp.ps(<4 x float>, <4 x float>, i8) nounwind readnone
+
+define <8 x float> @test_mm256_cmp_ps(<8 x float> %a0, <8 x float> %a1) nounwind {
+; X32-LABEL: test_mm256_cmp_ps:
+; X32: # BB#0:
+; X32-NEXT: vcmpgeps %ymm1, %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_cmp_ps:
+; X64: # BB#0:
+; X64-NEXT: vcmpgeps %ymm1, %ymm0, %ymm0
+; X64-NEXT: retq
+ %res = call <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float> %a0, <8 x float> %a1, i8 13)
+ ret <8 x float> %res
+}
+declare <8 x float> @llvm.x86.avx.cmp.ps.256(<8 x float>, <8 x float>, i8) nounwind readnone
+
+define <2 x double> @test_mm_cmp_sd(<2 x double> %a0, <2 x double> %a1) nounwind {
+; X32-LABEL: test_mm_cmp_sd:
+; X32: # BB#0:
+; X32-NEXT: vcmpgesd %xmm1, %xmm0, %xmm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm_cmp_sd:
+; X64: # BB#0:
+; X64-NEXT: vcmpgesd %xmm1, %xmm0, %xmm0
+; X64-NEXT: retq
+ %res = call <2 x double> @llvm.x86.sse2.cmp.sd(<2 x double> %a0, <2 x double> %a1, i8 13)
+ ret <2 x double> %res
+}
+declare <2 x double> @llvm.x86.sse2.cmp.sd(<2 x double>, <2 x double>, i8) nounwind readnone
+
+define <4 x float> @test_mm_cmp_ss(<4 x float> %a0, <4 x float> %a1) nounwind {
+; X32-LABEL: test_mm_cmp_ss:
+; X32: # BB#0:
+; X32-NEXT: vcmpgess %xmm1, %xmm0, %xmm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm_cmp_ss:
+; X64: # BB#0:
+; X64-NEXT: vcmpgess %xmm1, %xmm0, %xmm0
+; X64-NEXT: retq
+ %res = call <4 x float> @llvm.x86.sse.cmp.ss(<4 x float> %a0, <4 x float> %a1, i8 13)
+ ret <4 x float> %res
+}
+declare <4 x float> @llvm.x86.sse.cmp.ss(<4 x float>, <4 x float>, i8) nounwind readnone
+
+define <4 x double> @test_mm256_cvtepi32_pd(<2 x i64> %a0) nounwind {
+; X32-LABEL: test_mm256_cvtepi32_pd:
+; X32: # BB#0:
+; X32-NEXT: vcvtdq2pd %xmm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_cvtepi32_pd:
+; X64: # BB#0:
+; X64-NEXT: vcvtdq2pd %xmm0, %ymm0
+; X64-NEXT: retq
+ %arg0 = bitcast <2 x i64> %a0 to <4 x i32>
+ %res = call <4 x double> @llvm.x86.avx.cvtdq2.pd.256(<4 x i32> %arg0)
+ ret <4 x double> %res
+}
+declare <4 x double> @llvm.x86.avx.cvtdq2.pd.256(<4 x i32>) nounwind readnone
+
+define <8 x float> @test_mm256_cvtepi32_ps(<4 x i64> %a0) nounwind {
+; X32-LABEL: test_mm256_cvtepi32_ps:
+; X32: # BB#0:
+; X32-NEXT: vcvtdq2ps %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_cvtepi32_ps:
+; X64: # BB#0:
+; X64-NEXT: vcvtdq2ps %ymm0, %ymm0
+; X64-NEXT: retq
+ %arg0 = bitcast <4 x i64> %a0 to <8 x i32>
+ %res = call <8 x float> @llvm.x86.avx.cvtdq2.ps.256(<8 x i32> %arg0)
+ ret <8 x float> %res
+}
+declare <8 x float> @llvm.x86.avx.cvtdq2.ps.256(<8 x i32>) nounwind readnone
+
+define <2 x i64> @test_mm256_cvtpd_epi32(<4 x double> %a0) nounwind {
+; X32-LABEL: test_mm256_cvtpd_epi32:
+; X32: # BB#0:
+; X32-NEXT: vcvtpd2dqy %ymm0, %xmm0
+; X32-NEXT: vzeroupper
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_cvtpd_epi32:
+; X64: # BB#0:
+; X64-NEXT: vcvtpd2dqy %ymm0, %xmm0
+; X64-NEXT: vzeroupper
+; X64-NEXT: retq
+ %cvt = call <4 x i32> @llvm.x86.avx.cvt.pd2dq.256(<4 x double> %a0)
+ %res = bitcast <4 x i32> %cvt to <2 x i64>
+ ret <2 x i64> %res
+}
+declare <4 x i32> @llvm.x86.avx.cvt.pd2dq.256(<4 x double>) nounwind readnone
+
+define <4 x float> @test_mm256_cvtpd_ps(<4 x double> %a0) nounwind {
+; X32-LABEL: test_mm256_cvtpd_ps:
+; X32: # BB#0:
+; X32-NEXT: vcvtpd2psy %ymm0, %xmm0
+; X32-NEXT: vzeroupper
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_cvtpd_ps:
+; X64: # BB#0:
+; X64-NEXT: vcvtpd2psy %ymm0, %xmm0
+; X64-NEXT: vzeroupper
+; X64-NEXT: retq
+ %res = call <4 x float> @llvm.x86.avx.cvt.pd2.ps.256(<4 x double> %a0)
+ ret <4 x float> %res
+}
+declare <4 x float> @llvm.x86.avx.cvt.pd2.ps.256(<4 x double>) nounwind readnone
+
+define <4 x i64> @test_mm256_cvtps_epi32(<8 x float> %a0) nounwind {
+; X32-LABEL: test_mm256_cvtps_epi32:
+; X32: # BB#0:
+; X32-NEXT: vcvtps2dq %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_cvtps_epi32:
+; X64: # BB#0:
+; X64-NEXT: vcvtps2dq %ymm0, %ymm0
+; X64-NEXT: retq
+ %cvt = call <8 x i32> @llvm.x86.avx.cvt.ps2dq.256(<8 x float> %a0)
+ %res = bitcast <8 x i32> %cvt to <4 x i64>
+ ret <4 x i64> %res
+}
+declare <8 x i32> @llvm.x86.avx.cvt.ps2dq.256(<8 x float>) nounwind readnone
+
+define <4 x double> @test_mm256_cvtps_pd(<4 x float> %a0) nounwind {
+; X32-LABEL: test_mm256_cvtps_pd:
+; X32: # BB#0:
+; X32-NEXT: vcvtps2pd %xmm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_cvtps_pd:
+; X64: # BB#0:
+; X64-NEXT: vcvtps2pd %xmm0, %ymm0
+; X64-NEXT: retq
+ %res = call <4 x double> @llvm.x86.avx.cvt.ps2.pd.256(<4 x float> %a0)
+ ret <4 x double> %res
+}
+declare <4 x double> @llvm.x86.avx.cvt.ps2.pd.256(<4 x float>) nounwind readnone
+
+define <2 x i64> @test_mm256_cvttpd_epi32(<4 x double> %a0) nounwind {
+; X32-LABEL: test_mm256_cvttpd_epi32:
+; X32: # BB#0:
+; X32-NEXT: vcvttpd2dqy %ymm0, %xmm0
+; X32-NEXT: vzeroupper
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_cvttpd_epi32:
+; X64: # BB#0:
+; X64-NEXT: vcvttpd2dqy %ymm0, %xmm0
+; X64-NEXT: vzeroupper
+; X64-NEXT: retq
+ %cvt = call <4 x i32> @llvm.x86.avx.cvtt.pd2dq.256(<4 x double> %a0)
+ %res = bitcast <4 x i32> %cvt to <2 x i64>
+ ret <2 x i64> %res
+}
+declare <4 x i32> @llvm.x86.avx.cvtt.pd2dq.256(<4 x double>) nounwind readnone
+
+define <4 x i64> @test_mm256_cvttps_epi32(<8 x float> %a0) nounwind {
+; X32-LABEL: test_mm256_cvttps_epi32:
+; X32: # BB#0:
+; X32-NEXT: vcvttps2dq %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_cvttps_epi32:
+; X64: # BB#0:
+; X64-NEXT: vcvttps2dq %ymm0, %ymm0
+; X64-NEXT: retq
+ %cvt = call <8 x i32> @llvm.x86.avx.cvtt.ps2dq.256(<8 x float> %a0)
+ %res = bitcast <8 x i32> %cvt to <4 x i64>
+ ret <4 x i64> %res
+}
+declare <8 x i32> @llvm.x86.avx.cvtt.ps2dq.256(<8 x float>) nounwind readnone
+
+define <4 x double> @test_mm256_div_pd(<4 x double> %a0, <4 x double> %a1) nounwind {
+; X32-LABEL: test_mm256_div_pd:
+; X32: # BB#0:
+; X32-NEXT: vdivpd %ymm1, %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_div_pd:
+; X64: # BB#0:
+; X64-NEXT: vdivpd %ymm1, %ymm0, %ymm0
+; X64-NEXT: retq
+ %res = fdiv <4 x double> %a0, %a1
+ ret <4 x double> %res
+}
+
+define <8 x float> @test_mm256_div_ps(<8 x float> %a0, <8 x float> %a1) nounwind {
+; X32-LABEL: test_mm256_div_ps:
+; X32: # BB#0:
+; X32-NEXT: vdivps %ymm1, %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_div_ps:
+; X64: # BB#0:
+; X64-NEXT: vdivps %ymm1, %ymm0, %ymm0
+; X64-NEXT: retq
+ %res = fdiv <8 x float> %a0, %a1
+ ret <8 x float> %res
+}
+
+define <8 x float> @test_mm256_dp_ps(<8 x float> %a0, <8 x float> %a1) nounwind {
+; X32-LABEL: test_mm256_dp_ps:
+; X32: # BB#0:
+; X32-NEXT: vdpps $7, %ymm1, %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_dp_ps:
+; X64: # BB#0:
+; X64-NEXT: vdpps $7, %ymm1, %ymm0, %ymm0
+; X64-NEXT: retq
+ %res = call <8 x float> @llvm.x86.avx.dp.ps.256(<8 x float> %a0, <8 x float> %a1, i8 7)
+ ret <8 x float> %res
+}
+declare <8 x float> @llvm.x86.avx.dp.ps.256(<8 x float>, <8 x float>, i8) nounwind readnone
+
+define i32 @test_mm256_extract_epi8(<4 x i64> %a0) nounwind {
+; X32-LABEL: test_mm256_extract_epi8:
+; X32: # BB#0:
+; X32-NEXT: vextractf128 $1, %ymm0, %xmm0
+; X32-NEXT: vpextrb $15, %xmm0, %eax
+; X32-NEXT: movzbl %al, %eax
+; X32-NEXT: vzeroupper
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_extract_epi8:
+; X64: # BB#0:
+; X64-NEXT: vextractf128 $1, %ymm0, %xmm0
+; X64-NEXT: vpextrb $15, %xmm0, %eax
+; X64-NEXT: movzbl %al, %eax
+; X64-NEXT: vzeroupper
+; X64-NEXT: retq
+ %arg0 = bitcast <4 x i64> %a0 to <32 x i8>
+ %ext = extractelement <32 x i8> %arg0, i32 31
+ %res = zext i8 %ext to i32
+ ret i32 %res
+}
+
+define i32 @test_mm256_extract_epi16(<4 x i64> %a0) nounwind {
+; X32-LABEL: test_mm256_extract_epi16:
+; X32: # BB#0:
+; X32-NEXT: vextractf128 $1, %ymm0, %xmm0
+; X32-NEXT: vpextrw $3, %xmm0, %eax
+; X32-NEXT: movzwl %ax, %eax
+; X32-NEXT: vzeroupper
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_extract_epi16:
+; X64: # BB#0:
+; X64-NEXT: vextractf128 $1, %ymm0, %xmm0
+; X64-NEXT: vpextrw $3, %xmm0, %eax
+; X64-NEXT: movzwl %ax, %eax
+; X64-NEXT: vzeroupper
+; X64-NEXT: retq
+ %arg0 = bitcast <4 x i64> %a0 to <16 x i16>
+ %ext = extractelement <16 x i16> %arg0, i32 11
+ %res = zext i16 %ext to i32
+ ret i32 %res
+}
+
+define i32 @test_mm256_extract_epi32(<4 x i64> %a0) nounwind {
+; X32-LABEL: test_mm256_extract_epi32:
+; X32: # BB#0:
+; X32-NEXT: vextractf128 $1, %ymm0, %xmm0
+; X32-NEXT: vpextrd $1, %xmm0, %eax
+; X32-NEXT: vzeroupper
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_extract_epi32:
+; X64: # BB#0:
+; X64-NEXT: vextractf128 $1, %ymm0, %xmm0
+; X64-NEXT: vpextrd $1, %xmm0, %eax
+; X64-NEXT: vzeroupper
+; X64-NEXT: retq
+ %arg0 = bitcast <4 x i64> %a0 to <8 x i32>
+ %res = extractelement <8 x i32> %arg0, i32 5
+ ret i32 %res
+}
+
+define i64 @test_mm256_extract_epi64(<4 x i64> %a0) nounwind {
+; X32-LABEL: test_mm256_extract_epi64:
+; X32: # BB#0:
+; X32-NEXT: vextractf128 $1, %ymm0, %xmm0
+; X32-NEXT: vpextrd $2, %xmm0, %eax
+; X32-NEXT: vpextrd $3, %xmm0, %edx
+; X32-NEXT: vzeroupper
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_extract_epi64:
+; X64: # BB#0:
+; X64-NEXT: vextractf128 $1, %ymm0, %xmm0
+; X64-NEXT: vpextrq $1, %xmm0, %rax
+; X64-NEXT: vzeroupper
+; X64-NEXT: retq
+ %res = extractelement <4 x i64> %a0, i32 3
+ ret i64 %res
+}
+
+define <2 x double> @test_mm256_extractf128_pd(<4 x double> %a0) nounwind {
+; X32-LABEL: test_mm256_extractf128_pd:
+; X32: # BB#0:
+; X32-NEXT: vextractf128 $1, %ymm0, %xmm0
+; X32-NEXT: vzeroupper
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_extractf128_pd:
+; X64: # BB#0:
+; X64-NEXT: vextractf128 $1, %ymm0, %xmm0
+; X64-NEXT: vzeroupper
+; X64-NEXT: retq
+ %res = shufflevector <4 x double> %a0, <4 x double> %a0, <2 x i32> <i32 2, i32 3>
+ ret <2 x double> %res
+}
+
+define <4 x float> @test_mm256_extractf128_ps(<8 x float> %a0) nounwind {
+; X32-LABEL: test_mm256_extractf128_ps:
+; X32: # BB#0:
+; X32-NEXT: vextractf128 $1, %ymm0, %xmm0
+; X32-NEXT: vzeroupper
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_extractf128_ps:
+; X64: # BB#0:
+; X64-NEXT: vextractf128 $1, %ymm0, %xmm0
+; X64-NEXT: vzeroupper
+; X64-NEXT: retq
+ %res = shufflevector <8 x float> %a0, <8 x float> %a0, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
+ ret <4 x float> %res
+}
+
+define <2 x i64> @test_mm256_extractf128_si256(<4 x i64> %a0) nounwind {
+; X32-LABEL: test_mm256_extractf128_si256:
+; X32: # BB#0:
+; X32-NEXT: vextractf128 $1, %ymm0, %xmm0
+; X32-NEXT: vzeroupper
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_extractf128_si256:
+; X64: # BB#0:
+; X64-NEXT: vextractf128 $1, %ymm0, %xmm0
+; X64-NEXT: vzeroupper
+; X64-NEXT: retq
+ %res = shufflevector <4 x i64> %a0, <4 x i64> %a0, <2 x i32> <i32 2, i32 3>
+ ret <2 x i64> %res
+}
+
+define <4 x double> @test_mm256_floor_pd(<4 x double> %a0) nounwind {
+; X32-LABEL: test_mm256_floor_pd:
+; X32: # BB#0:
+; X32-NEXT: vroundpd $1, %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_floor_pd:
+; X64: # BB#0:
+; X64-NEXT: vroundpd $1, %ymm0, %ymm0
+; X64-NEXT: retq
+ %res = call <4 x double> @llvm.x86.avx.round.pd.256(<4 x double> %a0, i32 1)
+ ret <4 x double> %res
+}
+
+define <8 x float> @test_mm256_floor_ps(<8 x float> %a0) nounwind {
+; X32-LABEL: test_mm256_floor_ps:
+; X32: # BB#0:
+; X32-NEXT: vroundps $1, %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_floor_ps:
+; X64: # BB#0:
+; X64-NEXT: vroundps $1, %ymm0, %ymm0
+; X64-NEXT: retq
+ %res = call <8 x float> @llvm.x86.avx.round.ps.256(<8 x float> %a0, i32 1)
+ ret <8 x float> %res
+}
+
+define <4 x double> @test_mm256_hadd_pd(<4 x double> %a0, <4 x double> %a1) nounwind {
+; X32-LABEL: test_mm256_hadd_pd:
+; X32: # BB#0:
+; X32-NEXT: vhaddpd %ymm1, %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_hadd_pd:
+; X64: # BB#0:
+; X64-NEXT: vhaddpd %ymm1, %ymm0, %ymm0
+; X64-NEXT: retq
+ %res = call <4 x double> @llvm.x86.avx.hadd.pd.256(<4 x double> %a0, <4 x double> %a1)
+ ret <4 x double> %res
+}
+declare <4 x double> @llvm.x86.avx.hadd.pd.256(<4 x double>, <4 x double>) nounwind readnone
+
+define <8 x float> @test_mm256_hadd_ps(<8 x float> %a0, <8 x float> %a1) nounwind {
+; X32-LABEL: test_mm256_hadd_ps:
+; X32: # BB#0:
+; X32-NEXT: vhaddps %ymm1, %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_hadd_ps:
+; X64: # BB#0:
+; X64-NEXT: vhaddps %ymm1, %ymm0, %ymm0
+; X64-NEXT: retq
+ %res = call <8 x float> @llvm.x86.avx.hadd.ps.256(<8 x float> %a0, <8 x float> %a1)
+ ret <8 x float> %res
+}
+declare <8 x float> @llvm.x86.avx.hadd.ps.256(<8 x float>, <8 x float>) nounwind readnone
+
+define <4 x double> @test_mm256_hsub_pd(<4 x double> %a0, <4 x double> %a1) nounwind {
+; X32-LABEL: test_mm256_hsub_pd:
+; X32: # BB#0:
+; X32-NEXT: vhsubpd %ymm1, %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_hsub_pd:
+; X64: # BB#0:
+; X64-NEXT: vhsubpd %ymm1, %ymm0, %ymm0
+; X64-NEXT: retq
+ %res = call <4 x double> @llvm.x86.avx.hsub.pd.256(<4 x double> %a0, <4 x double> %a1)
+ ret <4 x double> %res
+}
+declare <4 x double> @llvm.x86.avx.hsub.pd.256(<4 x double>, <4 x double>) nounwind readnone
+
+define <8 x float> @test_mm256_hsub_ps(<8 x float> %a0, <8 x float> %a1) nounwind {
+; X32-LABEL: test_mm256_hsub_ps:
+; X32: # BB#0:
+; X32-NEXT: vhsubps %ymm1, %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_hsub_ps:
+; X64: # BB#0:
+; X64-NEXT: vhsubps %ymm1, %ymm0, %ymm0
+; X64-NEXT: retq
+ %res = call <8 x float> @llvm.x86.avx.hsub.ps.256(<8 x float> %a0, <8 x float> %a1)
+ ret <8 x float> %res
+}
+declare <8 x float> @llvm.x86.avx.hsub.ps.256(<8 x float>, <8 x float>) nounwind readnone
+
+define <4 x i64> @test_mm256_insert_epi8(<4 x i64> %a0, i8 %a1) nounwind {
+; X32-LABEL: test_mm256_insert_epi8:
+; X32: # BB#0:
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $4, %eax, %xmm0, %xmm1
+; X32-NEXT: vblendps {{.*#+}} ymm0 = ymm1[0,1,2,3],ymm0[4,5,6,7]
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_insert_epi8:
+; X64: # BB#0:
+; X64-NEXT: movzbl %dil, %eax
+; X64-NEXT: vpinsrb $4, %eax, %xmm0, %xmm1
+; X64-NEXT: vblendps {{.*#+}} ymm0 = ymm1[0,1,2,3],ymm0[4,5,6,7]
+; X64-NEXT: retq
+ %arg0 = bitcast <4 x i64> %a0 to <32 x i8>
+ %res = insertelement <32 x i8> %arg0, i8 %a1, i32 4
+ %bc = bitcast <32 x i8> %res to <4 x i64>
+ ret <4 x i64> %bc
+}
+
+define <4 x i64> @test_mm256_insert_epi16(<4 x i64> %a0, i16 %a1) nounwind {
+; X32-LABEL: test_mm256_insert_epi16:
+; X32: # BB#0:
+; X32-NEXT: movw {{[0-9]+}}(%esp), %ax
+; X32-NEXT: vextractf128 $1, %ymm0, %xmm1
+; X32-NEXT: vpinsrw $6, %eax, %xmm1, %xmm1
+; X32-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_insert_epi16:
+; X64: # BB#0:
+; X64-NEXT: vextractf128 $1, %ymm0, %xmm1
+; X64-NEXT: vpinsrw $6, %edi, %xmm1, %xmm1
+; X64-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; X64-NEXT: retq
+ %arg0 = bitcast <4 x i64> %a0 to <16 x i16>
+ %res = insertelement <16 x i16> %arg0, i16 %a1, i32 14
+ %bc = bitcast <16 x i16> %res to <4 x i64>
+ ret <4 x i64> %bc
+}
+
+define <4 x i64> @test_mm256_insert_epi32(<4 x i64> %a0, i32 %a1) nounwind {
+; X32-LABEL: test_mm256_insert_epi32:
+; X32: # BB#0:
+; X32-NEXT: vpinsrd $3, {{[0-9]+}}(%esp), %xmm0, %xmm1
+; X32-NEXT: vblendps {{.*#+}} ymm0 = ymm1[0,1,2,3],ymm0[4,5,6,7]
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_insert_epi32:
+; X64: # BB#0:
+; X64-NEXT: vpinsrd $3, %edi, %xmm0, %xmm1
+; X64-NEXT: vblendps {{.*#+}} ymm0 = ymm1[0,1,2,3],ymm0[4,5,6,7]
+; X64-NEXT: retq
+ %arg0 = bitcast <4 x i64> %a0 to <8 x i32>
+ %res = insertelement <8 x i32> %arg0, i32 %a1, i32 3
+ %bc = bitcast <8 x i32> %res to <4 x i64>
+ ret <4 x i64> %bc
+}
+
+define <4 x i64> @test_mm256_insert_epi64(<4 x i64> %a0, i64 %a1) nounwind {
+; X32-LABEL: test_mm256_insert_epi64:
+; X32: # BB#0:
+; X32-NEXT: vextractf128 $1, %ymm0, %xmm1
+; X32-NEXT: vpinsrd $2, {{[0-9]+}}(%esp), %xmm1, %xmm1
+; X32-NEXT: vpinsrd $3, {{[0-9]+}}(%esp), %xmm1, %xmm2
+; X32-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; X32-NEXT: vinsertf128 $1, %xmm2, %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_insert_epi64:
+; X64: # BB#0:
+; X64-NEXT: vextractf128 $1, %ymm0, %xmm1
+; X64-NEXT: vpinsrq $1, %rdi, %xmm1, %xmm1
+; X64-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; X64-NEXT: retq
+ %res = insertelement <4 x i64> %a0, i64 %a1, i32 3
+ ret <4 x i64> %res
+}
+
+define <4 x double> @test_mm256_insertf128_pd(<4 x double> %a0, <2 x double> %a1) nounwind {
+; X32-LABEL: test_mm256_insertf128_pd:
+; X32: # BB#0:
+; X32-NEXT: vblendpd {{.*#+}} ymm0 = ymm1[0,1],ymm0[2,3]
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_insertf128_pd:
+; X64: # BB#0:
+; X64-NEXT: vblendpd {{.*#+}} ymm0 = ymm1[0,1],ymm0[2,3]
+; X64-NEXT: retq
+ %ext = shufflevector <2 x double> %a1, <2 x double> %a1, <4 x i32> <i32 0, i32 1, i32 undef, i32 undef>
+ %res = shufflevector <4 x double> %a0, <4 x double> %ext, <4 x i32> <i32 4, i32 5, i32 2, i32 3>
+ ret <4 x double> %res
+}
+
+define <8 x float> @test_mm256_insertf128_ps(<8 x float> %a0, <4 x float> %a1) nounwind {
+; X32-LABEL: test_mm256_insertf128_ps:
+; X32: # BB#0:
+; X32-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_insertf128_ps:
+; X64: # BB#0:
+; X64-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; X64-NEXT: retq
+ %ext = shufflevector <4 x float> %a1, <4 x float> %a1, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef>
+ %res = shufflevector <8 x float> %a0, <8 x float> %ext, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 8, i32 9, i32 10, i32 11>
+ ret <8 x float> %res
+}
+
+define <4 x i64> @test_mm256_insertf128_si256(<4 x i64> %a0, <2 x i64> %a1) nounwind {
+; X32-LABEL: test_mm256_insertf128_si256:
+; X32: # BB#0:
+; X32-NEXT: vblendpd {{.*#+}} ymm0 = ymm1[0,1],ymm0[2,3]
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_insertf128_si256:
+; X64: # BB#0:
+; X64-NEXT: vblendpd {{.*#+}} ymm0 = ymm1[0,1],ymm0[2,3]
+; X64-NEXT: retq
+ %ext = shufflevector <2 x i64> %a1, <2 x i64> %a1, <4 x i32> <i32 0, i32 1, i32 undef, i32 undef>
+ %res = shufflevector <4 x i64> %a0, <4 x i64> %ext, <4 x i32> <i32 4, i32 5, i32 2, i32 3>
+ ret <4 x i64> %res
+}
+
+define <4 x i64> @test_mm256_lddqu_si256(<4 x i64>* %a0) nounwind {
+; X32-LABEL: test_mm256_lddqu_si256:
+; X32: # BB#0:
+; X32-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vlddqu (%eax), %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_lddqu_si256:
+; X64: # BB#0:
+; X64-NEXT: vlddqu (%rdi), %ymm0
+; X64-NEXT: retq
+ %arg0 = bitcast <4 x i64>* %a0 to i8*
+ %res = call <32 x i8> @llvm.x86.avx.ldu.dq.256(i8* %arg0)
+ %bc = bitcast <32 x i8> %res to <4 x i64>
+ ret <4 x i64> %bc
+}
+declare <32 x i8> @llvm.x86.avx.ldu.dq.256(i8*) nounwind readnone
+
+define <4 x double> @test_mm256_load_pd(double* %a0) nounwind {
+; X32-LABEL: test_mm256_load_pd:
+; X32: # BB#0:
+; X32-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vmovaps (%eax), %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_load_pd:
+; X64: # BB#0:
+; X64-NEXT: vmovaps (%rdi), %ymm0
+; X64-NEXT: retq
+ %arg0 = bitcast double* %a0 to <4 x double>*
+ %res = load <4 x double>, <4 x double>* %arg0, align 32
+ ret <4 x double> %res
+}
+
+define <8 x float> @test_mm256_load_ps(float* %a0) nounwind {
+; X32-LABEL: test_mm256_load_ps:
+; X32: # BB#0:
+; X32-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vmovaps (%eax), %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_load_ps:
+; X64: # BB#0:
+; X64-NEXT: vmovaps (%rdi), %ymm0
+; X64-NEXT: retq
+ %arg0 = bitcast float* %a0 to <8 x float>*
+ %res = load <8 x float>, <8 x float>* %arg0, align 32
+ ret <8 x float> %res
+}
+
+define <4 x i64> @test_mm256_load_si256(<4 x i64>* %a0) nounwind {
+; X32-LABEL: test_mm256_load_si256:
+; X32: # BB#0:
+; X32-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vmovaps (%eax), %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_load_si256:
+; X64: # BB#0:
+; X64-NEXT: vmovaps (%rdi), %ymm0
+; X64-NEXT: retq
+ %res = load <4 x i64>, <4 x i64>* %a0, align 32
+ ret <4 x i64> %res
+}
+
+define <4 x double> @test_mm256_loadu_pd(double* %a0) nounwind {
+; X32-LABEL: test_mm256_loadu_pd:
+; X32: # BB#0:
+; X32-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vmovups (%eax), %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_loadu_pd:
+; X64: # BB#0:
+; X64-NEXT: vmovups (%rdi), %ymm0
+; X64-NEXT: retq
+ %arg0 = bitcast double* %a0 to <4 x double>*
+ %res = load <4 x double>, <4 x double>* %arg0, align 1
+ ret <4 x double> %res
+}
+
+define <8 x float> @test_mm256_loadu_ps(float* %a0) nounwind {
+; X32-LABEL: test_mm256_loadu_ps:
+; X32: # BB#0:
+; X32-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vmovups (%eax), %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_loadu_ps:
+; X64: # BB#0:
+; X64-NEXT: vmovups (%rdi), %ymm0
+; X64-NEXT: retq
+ %arg0 = bitcast float* %a0 to <8 x float>*
+ %res = load <8 x float>, <8 x float>* %arg0, align 1
+ ret <8 x float> %res
+}
+
+define <4 x i64> @test_mm256_loadu_si256(<4 x i64>* %a0) nounwind {
+; X32-LABEL: test_mm256_loadu_si256:
+; X32: # BB#0:
+; X32-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vmovups (%eax), %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_loadu_si256:
+; X64: # BB#0:
+; X64-NEXT: vmovups (%rdi), %ymm0
+; X64-NEXT: retq
+ %res = load <4 x i64>, <4 x i64>* %a0, align 1
+ ret <4 x i64> %res
+}
+
+define <8 x float> @test_mm256_loadu2_m128(float* %a0, float* %a1) nounwind {
+; X32-LABEL: test_mm256_loadu2_m128:
+; X32: # BB#0:
+; X32-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: movl {{[0-9]+}}(%esp), %ecx
+; X32-NEXT: vmovups (%eax), %xmm0
+; X32-NEXT: vinsertf128 $1, (%ecx), %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_loadu2_m128:
+; X64: # BB#0:
+; X64-NEXT: vmovups (%rsi), %xmm0
+; X64-NEXT: vinsertf128 $1, (%rdi), %ymm0, %ymm0
+; X64-NEXT: retq
+ %arg0 = bitcast float* %a0 to <4 x float>*
+ %hi4 = load <4 x float>, <4 x float>* %arg0, align 1
+ %hi8 = shufflevector <4 x float> %hi4, <4 x float> %hi4, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef>
+ %arg1 = bitcast float* %a1 to <4 x float>*
+ %lo4 = load <4 x float>, <4 x float>* %arg1, align 1
+ %lo8 = shufflevector <4 x float> %lo4, <4 x float> %lo4, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef>
+ %res = shufflevector <8 x float> %lo8, <8 x float> %hi8, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 8, i32 9, i32 10, i32 11>
+ ret <8 x float> %res
+}
+
+define <4 x double> @test_mm256_loadu2_m128d(double* %a0, double* %a1) nounwind {
+; X32-LABEL: test_mm256_loadu2_m128d:
+; X32: # BB#0:
+; X32-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: movl {{[0-9]+}}(%esp), %ecx
+; X32-NEXT: vmovups (%eax), %xmm0
+; X32-NEXT: vinsertf128 $1, (%ecx), %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_loadu2_m128d:
+; X64: # BB#0:
+; X64-NEXT: vmovups (%rsi), %xmm0
+; X64-NEXT: vinsertf128 $1, (%rdi), %ymm0, %ymm0
+; X64-NEXT: retq
+ %arg0 = bitcast double* %a0 to <2 x double>*
+ %hi2 = load <2 x double>, <2 x double>* %arg0, align 1
+ %hi4 = shufflevector <2 x double> %hi2, <2 x double> %hi2, <4 x i32> <i32 0, i32 1, i32 undef, i32 undef>
+ %arg1 = bitcast double* %a1 to <2 x double>*
+ %lo2 = load <2 x double>, <2 x double>* %arg1, align 1
+ %lo4 = shufflevector <2 x double> %lo2, <2 x double> %lo2, <4 x i32> <i32 0, i32 1, i32 undef, i32 undef>
+ %res = shufflevector <4 x double> %lo4, <4 x double> %hi4, <4 x i32> <i32 0, i32 1, i32 4, i32 5>
+ ret <4 x double> %res
+}
+
+define <4 x i64> @test_mm256_loadu2_m128i(i64* %a0, i64* %a1) nounwind {
+; X32-LABEL: test_mm256_loadu2_m128i:
+; X32: # BB#0:
+; X32-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: movl {{[0-9]+}}(%esp), %ecx
+; X32-NEXT: vmovups (%eax), %xmm0
+; X32-NEXT: vinsertf128 $1, (%ecx), %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_loadu2_m128i:
+; X64: # BB#0:
+; X64-NEXT: vmovups (%rsi), %xmm0
+; X64-NEXT: vinsertf128 $1, (%rdi), %ymm0, %ymm0
+; X64-NEXT: retq
+ %arg0 = bitcast i64* %a0 to <2 x i64>*
+ %hi2 = load <2 x i64>, <2 x i64>* %arg0, align 1
+ %hi4 = shufflevector <2 x i64> %hi2, <2 x i64> %hi2, <4 x i32> <i32 0, i32 1, i32 undef, i32 undef>
+ %arg1 = bitcast i64* %a1 to <2 x i64>*
+ %lo2 = load <2 x i64>, <2 x i64>* %arg1, align 1
+ %lo4 = shufflevector <2 x i64> %lo2, <2 x i64> %lo2, <4 x i32> <i32 0, i32 1, i32 undef, i32 undef>
+ %res = shufflevector <4 x i64> %lo4, <4 x i64> %hi4, <4 x i32> <i32 0, i32 1, i32 4, i32 5>
+ ret <4 x i64> %res
+}
+
+define <2 x double> @test_mm_maskload_pd(double* %a0, <2 x i64> %a1) nounwind {
+; X32-LABEL: test_mm_maskload_pd:
+; X32: # BB#0:
+; X32-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vmaskmovpd (%eax), %xmm0, %xmm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm_maskload_pd:
+; X64: # BB#0:
+; X64-NEXT: vmaskmovpd (%rdi), %xmm0, %xmm0
+; X64-NEXT: retq
+ %arg0 = bitcast double* %a0 to i8*
+ %res = call <2 x double> @llvm.x86.avx.maskload.pd(i8* %arg0, <2 x i64> %a1)
+ ret <2 x double> %res
+}
+declare <2 x double> @llvm.x86.avx.maskload.pd(i8*, <2 x i64>) nounwind readnone
+
+define <4 x double> @test_mm256_maskload_pd(double* %a0, <4 x i64> %a1) nounwind {
+; X32-LABEL: test_mm256_maskload_pd:
+; X32: # BB#0:
+; X32-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vmaskmovpd (%eax), %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_maskload_pd:
+; X64: # BB#0:
+; X64-NEXT: vmaskmovpd (%rdi), %ymm0, %ymm0
+; X64-NEXT: retq
+ %arg0 = bitcast double* %a0 to i8*
+ %res = call <4 x double> @llvm.x86.avx.maskload.pd.256(i8* %arg0, <4 x i64> %a1)
+ ret <4 x double> %res
+}
+declare <4 x double> @llvm.x86.avx.maskload.pd.256(i8*, <4 x i64>) nounwind readnone
+
+define <4 x float> @test_mm_maskload_ps(float* %a0, <2 x i64> %a1) nounwind {
+; X32-LABEL: test_mm_maskload_ps:
+; X32: # BB#0:
+; X32-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vmaskmovps (%eax), %xmm0, %xmm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm_maskload_ps:
+; X64: # BB#0:
+; X64-NEXT: vmaskmovps (%rdi), %xmm0, %xmm0
+; X64-NEXT: retq
+ %arg0 = bitcast float* %a0 to i8*
+ %arg1 = bitcast <2 x i64> %a1 to <4 x i32>
+ %res = call <4 x float> @llvm.x86.avx.maskload.ps(i8* %arg0, <4 x i32> %arg1)
+ ret <4 x float> %res
+}
+declare <4 x float> @llvm.x86.avx.maskload.ps(i8*, <4 x i32>) nounwind readnone
+
+define <8 x float> @test_mm256_maskload_ps(float* %a0, <4 x i64> %a1) nounwind {
+; X32-LABEL: test_mm256_maskload_ps:
+; X32: # BB#0:
+; X32-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vmaskmovps (%eax), %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_maskload_ps:
+; X64: # BB#0:
+; X64-NEXT: vmaskmovps (%rdi), %ymm0, %ymm0
+; X64-NEXT: retq
+ %arg0 = bitcast float* %a0 to i8*
+ %arg1 = bitcast <4 x i64> %a1 to <8 x i32>
+ %res = call <8 x float> @llvm.x86.avx.maskload.ps.256(i8* %arg0, <8 x i32> %arg1)
+ ret <8 x float> %res
+}
+declare <8 x float> @llvm.x86.avx.maskload.ps.256(i8*, <8 x i32>) nounwind readnone
+
+define void @test_mm_maskstore_pd(double* %a0, <2 x i64> %a1, <2 x double> %a2) nounwind {
+; X32-LABEL: test_mm_maskstore_pd:
+; X32: # BB#0:
+; X32-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vmaskmovpd %xmm1, %xmm0, (%eax)
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm_maskstore_pd:
+; X64: # BB#0:
+; X64-NEXT: vmaskmovpd %xmm1, %xmm0, (%rdi)
+; X64-NEXT: retq
+ %arg0 = bitcast double* %a0 to i8*
+ call void @llvm.x86.avx.maskstore.pd(i8* %arg0, <2 x i64> %a1, <2 x double> %a2)
+ ret void
+}
+declare void @llvm.x86.avx.maskstore.pd(i8*, <2 x i64>, <2 x double>) nounwind readnone
+
+define void @test_mm256_maskstore_pd(double* %a0, <4 x i64> %a1, <4 x double> %a2) nounwind {
+; X32-LABEL: test_mm256_maskstore_pd:
+; X32: # BB#0:
+; X32-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vmaskmovpd %ymm1, %ymm0, (%eax)
+; X32-NEXT: vzeroupper
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_maskstore_pd:
+; X64: # BB#0:
+; X64-NEXT: vmaskmovpd %ymm1, %ymm0, (%rdi)
+; X64-NEXT: vzeroupper
+; X64-NEXT: retq
+ %arg0 = bitcast double* %a0 to i8*
+ call void @llvm.x86.avx.maskstore.pd.256(i8* %arg0, <4 x i64> %a1, <4 x double> %a2)
+ ret void
+}
+declare void @llvm.x86.avx.maskstore.pd.256(i8*, <4 x i64>, <4 x double>) nounwind readnone
+
+define void @test_mm_maskstore_ps(float* %a0, <2 x i64> %a1, <4 x float> %a2) nounwind {
+; X32-LABEL: test_mm_maskstore_ps:
+; X32: # BB#0:
+; X32-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vmaskmovps %xmm1, %xmm0, (%eax)
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm_maskstore_ps:
+; X64: # BB#0:
+; X64-NEXT: vmaskmovps %xmm1, %xmm0, (%rdi)
+; X64-NEXT: retq
+ %arg0 = bitcast float* %a0 to i8*
+ %arg1 = bitcast <2 x i64> %a1 to <4 x i32>
+ call void @llvm.x86.avx.maskstore.ps(i8* %arg0, <4 x i32> %arg1, <4 x float> %a2)
+ ret void
+}
+declare void @llvm.x86.avx.maskstore.ps(i8*, <4 x i32>, <4 x float>) nounwind readnone
+
+define void @test_mm256_maskstore_ps(float* %a0, <4 x i64> %a1, <8 x float> %a2) nounwind {
+; X32-LABEL: test_mm256_maskstore_ps:
+; X32: # BB#0:
+; X32-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vmaskmovps %ymm1, %ymm0, (%eax)
+; X32-NEXT: vzeroupper
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_maskstore_ps:
+; X64: # BB#0:
+; X64-NEXT: vmaskmovps %ymm1, %ymm0, (%rdi)
+; X64-NEXT: vzeroupper
+; X64-NEXT: retq
+ %arg0 = bitcast float* %a0 to i8*
+ %arg1 = bitcast <4 x i64> %a1 to <8 x i32>
+ call void @llvm.x86.avx.maskstore.ps.256(i8* %arg0, <8 x i32> %arg1, <8 x float> %a2)
+ ret void
+}
+declare void @llvm.x86.avx.maskstore.ps.256(i8*, <8 x i32>, <8 x float>) nounwind readnone
+
+define <4 x double> @test_mm256_max_pd(<4 x double> %a0, <4 x double> %a1) nounwind {
+; X32-LABEL: test_mm256_max_pd:
+; X32: # BB#0:
+; X32-NEXT: vmaxpd %ymm1, %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_max_pd:
+; X64: # BB#0:
+; X64-NEXT: vmaxpd %ymm1, %ymm0, %ymm0
+; X64-NEXT: retq
+ %res = call <4 x double> @llvm.x86.avx.max.pd.256(<4 x double> %a0, <4 x double> %a1)
+ ret <4 x double> %res
+}
+declare <4 x double> @llvm.x86.avx.max.pd.256(<4 x double>, <4 x double>) nounwind readnone
+
+define <8 x float> @test_mm256_max_ps(<8 x float> %a0, <8 x float> %a1) nounwind {
+; X32-LABEL: test_mm256_max_ps:
+; X32: # BB#0:
+; X32-NEXT: vmaxps %ymm1, %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_max_ps:
+; X64: # BB#0:
+; X64-NEXT: vmaxps %ymm1, %ymm0, %ymm0
+; X64-NEXT: retq
+ %res = call <8 x float> @llvm.x86.avx.max.ps.256(<8 x float> %a0, <8 x float> %a1)
+ ret <8 x float> %res
+}
+declare <8 x float> @llvm.x86.avx.max.ps.256(<8 x float>, <8 x float>) nounwind readnone
+
+define <4 x double> @test_mm256_min_pd(<4 x double> %a0, <4 x double> %a1) nounwind {
+; X32-LABEL: test_mm256_min_pd:
+; X32: # BB#0:
+; X32-NEXT: vminpd %ymm1, %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_min_pd:
+; X64: # BB#0:
+; X64-NEXT: vminpd %ymm1, %ymm0, %ymm0
+; X64-NEXT: retq
+ %res = call <4 x double> @llvm.x86.avx.min.pd.256(<4 x double> %a0, <4 x double> %a1)
+ ret <4 x double> %res
+}
+declare <4 x double> @llvm.x86.avx.min.pd.256(<4 x double>, <4 x double>) nounwind readnone
+
+define <8 x float> @test_mm256_min_ps(<8 x float> %a0, <8 x float> %a1) nounwind {
+; X32-LABEL: test_mm256_min_ps:
+; X32: # BB#0:
+; X32-NEXT: vminps %ymm1, %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_min_ps:
+; X64: # BB#0:
+; X64-NEXT: vminps %ymm1, %ymm0, %ymm0
+; X64-NEXT: retq
+ %res = call <8 x float> @llvm.x86.avx.min.ps.256(<8 x float> %a0, <8 x float> %a1)
+ ret <8 x float> %res
+}
+declare <8 x float> @llvm.x86.avx.min.ps.256(<8 x float>, <8 x float>) nounwind readnone
+
+define <4 x double> @test_mm256_movedup_pd(<4 x double> %a0) nounwind {
+; X32-LABEL: test_mm256_movedup_pd:
+; X32: # BB#0:
+; X32-NEXT: vmovddup {{.*#+}} ymm0 = ymm0[0,0,2,2]
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_movedup_pd:
+; X64: # BB#0:
+; X64-NEXT: vmovddup {{.*#+}} ymm0 = ymm0[0,0,2,2]
+; X64-NEXT: retq
+ %res = shufflevector <4 x double> %a0, <4 x double> %a0, <4 x i32> <i32 0, i32 0, i32 2, i32 2>
+ ret <4 x double> %res
+}
+
+define <8 x float> @test_mm256_movehdup_ps(<8 x float> %a0) nounwind {
+; X32-LABEL: test_mm256_movehdup_ps:
+; X32: # BB#0:
+; X32-NEXT: vmovshdup {{.*#+}} ymm0 = ymm0[1,1,3,3,5,5,7,7]
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_movehdup_ps:
+; X64: # BB#0:
+; X64-NEXT: vmovshdup {{.*#+}} ymm0 = ymm0[1,1,3,3,5,5,7,7]
+; X64-NEXT: retq
+ %res = shufflevector <8 x float> %a0, <8 x float> %a0, <8 x i32> <i32 1, i32 1, i32 3, i32 3, i32 5, i32 5, i32 7, i32 7>
+ ret <8 x float> %res
+}
+
+define <8 x float> @test_mm256_moveldup_ps(<8 x float> %a0) nounwind {
+; X32-LABEL: test_mm256_moveldup_ps:
+; X32: # BB#0:
+; X32-NEXT: vmovsldup {{.*#+}} ymm0 = ymm0[0,0,2,2,4,4,6,6]
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_moveldup_ps:
+; X64: # BB#0:
+; X64-NEXT: vmovsldup {{.*#+}} ymm0 = ymm0[0,0,2,2,4,4,6,6]
+; X64-NEXT: retq
+ %res = shufflevector <8 x float> %a0, <8 x float> %a0, <8 x i32> <i32 0, i32 0, i32 2, i32 2, i32 4, i32 4, i32 6, i32 6>
+ ret <8 x float> %res
+}
+
+define i32 @test_mm256_movemask_pd(<4 x double> %a0) nounwind {
+; X32-LABEL: test_mm256_movemask_pd:
+; X32: # BB#0:
+; X32-NEXT: vmovmskpd %ymm0, %eax
+; X32-NEXT: vzeroupper
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_movemask_pd:
+; X64: # BB#0:
+; X64-NEXT: vmovmskpd %ymm0, %eax
+; X64-NEXT: vzeroupper
+; X64-NEXT: retq
+ %res = call i32 @llvm.x86.avx.movmsk.pd.256(<4 x double> %a0)
+ ret i32 %res
+}
+declare i32 @llvm.x86.avx.movmsk.pd.256(<4 x double>) nounwind readnone
+
+define i32 @test_mm256_movemask_ps(<8 x float> %a0) nounwind {
+; X32-LABEL: test_mm256_movemask_ps:
+; X32: # BB#0:
+; X32-NEXT: vmovmskps %ymm0, %eax
+; X32-NEXT: vzeroupper
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_movemask_ps:
+; X64: # BB#0:
+; X64-NEXT: vmovmskps %ymm0, %eax
+; X64-NEXT: vzeroupper
+; X64-NEXT: retq
+ %res = call i32 @llvm.x86.avx.movmsk.ps.256(<8 x float> %a0)
+ ret i32 %res
+}
+declare i32 @llvm.x86.avx.movmsk.ps.256(<8 x float>) nounwind readnone
+
+define <4 x double> @test_mm256_mul_pd(<4 x double> %a0, <4 x double> %a1) nounwind {
+; X32-LABEL: test_mm256_mul_pd:
+; X32: # BB#0:
+; X32-NEXT: vmulpd %ymm1, %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_mul_pd:
+; X64: # BB#0:
+; X64-NEXT: vmulpd %ymm1, %ymm0, %ymm0
+; X64-NEXT: retq
+ %res = fmul <4 x double> %a0, %a1
+ ret <4 x double> %res
+}
+
+define <8 x float> @test_mm256_mul_ps(<8 x float> %a0, <8 x float> %a1) nounwind {
+; X32-LABEL: test_mm256_mul_ps:
+; X32: # BB#0:
+; X32-NEXT: vmulps %ymm1, %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_mul_ps:
+; X64: # BB#0:
+; X64-NEXT: vmulps %ymm1, %ymm0, %ymm0
+; X64-NEXT: retq
+ %res = fmul <8 x float> %a0, %a1
+ ret <8 x float> %res
+}
+
+define <4 x double> @test_mm256_or_pd(<4 x double> %a0, <4 x double> %a1) nounwind {
+; X32-LABEL: test_mm256_or_pd:
+; X32: # BB#0:
+; X32-NEXT: vorps %ymm1, %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_or_pd:
+; X64: # BB#0:
+; X64-NEXT: vorps %ymm1, %ymm0, %ymm0
+; X64-NEXT: retq
+ %1 = bitcast <4 x double> %a0 to <4 x i64>
+ %2 = bitcast <4 x double> %a1 to <4 x i64>
+ %res = or <4 x i64> %1, %2
+ %bc = bitcast <4 x i64> %res to <4 x double>
+ ret <4 x double> %bc
+}
+
+define <8 x float> @test_mm256_or_ps(<8 x float> %a0, <8 x float> %a1) nounwind {
+; X32-LABEL: test_mm256_or_ps:
+; X32: # BB#0:
+; X32-NEXT: vorps %ymm1, %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_or_ps:
+; X64: # BB#0:
+; X64-NEXT: vorps %ymm1, %ymm0, %ymm0
+; X64-NEXT: retq
+ %1 = bitcast <8 x float> %a0 to <8 x i32>
+ %2 = bitcast <8 x float> %a1 to <8 x i32>
+ %res = or <8 x i32> %1, %2
+ %bc = bitcast <8 x i32> %res to <8 x float>
+ ret <8 x float> %bc
+}
+
+define <2 x double> @test_mm_permute_pd(<2 x double> %a0) nounwind {
+; X32-LABEL: test_mm_permute_pd:
+; X32: # BB#0:
+; X32-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm_permute_pd:
+; X64: # BB#0:
+; X64-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
+; X64-NEXT: retq
+ %res = shufflevector <2 x double> %a0, <2 x double> %a0, <2 x i32> <i32 1, i32 0>
+ ret <2 x double> %res
+}
+
+define <4 x double> @test_mm256_permute_pd(<4 x double> %a0) nounwind {
+; X32-LABEL: test_mm256_permute_pd:
+; X32: # BB#0:
+; X32-NEXT: vpermilpd {{.*#+}} ymm0 = ymm0[1,0,3,2]
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_permute_pd:
+; X64: # BB#0:
+; X64-NEXT: vpermilpd {{.*#+}} ymm0 = ymm0[1,0,3,2]
+; X64-NEXT: retq
+ %res = shufflevector <4 x double> %a0, <4 x double> %a0, <4 x i32> <i32 1, i32 0, i32 3, i32 2>
+ ret <4 x double> %res
+}
+
+define <4 x float> @test_mm_permute_ps(<4 x float> %a0) nounwind {
+; X32-LABEL: test_mm_permute_ps:
+; X32: # BB#0:
+; X32-NEXT: vpermilps {{.*#+}} xmm0 = xmm0[3,2,1,0]
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm_permute_ps:
+; X64: # BB#0:
+; X64-NEXT: vpermilps {{.*#+}} xmm0 = xmm0[3,2,1,0]
+; X64-NEXT: retq
+ %res = shufflevector <4 x float> %a0, <4 x float> %a0, <4 x i32> <i32 3, i32 2, i32 1, i32 0>
+ ret <4 x float> %res
+}
+
+define <4 x float> @test2_mm_permute_ps(<4 x float> %a0) nounwind {
+; X32-LABEL: test2_mm_permute_ps:
+; X32: # BB#0:
+; X32-NEXT: vpermilps {{.*#+}} xmm0 = xmm0[2,1,2,3]
+; X32-NEXT: retl
+;
+; X64-LABEL: test2_mm_permute_ps:
+; X64: # BB#0:
+; X64-NEXT: vpermilps {{.*#+}} xmm0 = xmm0[2,1,2,3]
+; X64-NEXT: retq
+ %res = shufflevector <4 x float> %a0, <4 x float> %a0, <4 x i32> <i32 2, i32 1, i32 2, i32 3>
+ ret <4 x float> %res
+}
+
+define <8 x float> @test_mm256_permute_ps(<8 x float> %a0) nounwind {
+; X32-LABEL: test_mm256_permute_ps:
+; X32: # BB#0:
+; X32-NEXT: vpermilps {{.*#+}} ymm0 = ymm0[3,2,1,0,7,6,5,4]
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_permute_ps:
+; X64: # BB#0:
+; X64-NEXT: vpermilps {{.*#+}} ymm0 = ymm0[3,2,1,0,7,6,5,4]
+; X64-NEXT: retq
+ %res = shufflevector <8 x float> %a0, <8 x float> %a0, <8 x i32> <i32 3, i32 2, i32 1, i32 0, i32 7, i32 6, i32 5, i32 4>
+ ret <8 x float> %res
+}
+
+define <4 x double> @test_mm256_permute2f128_pd(<4 x double> %a0, <4 x double> %a1) nounwind {
+; X32-LABEL: test_mm256_permute2f128_pd:
+; X32: # BB#0:
+; X32-NEXT: vperm2f128 {{.*#+}} ymm0 = zero,zero,ymm1[0,1]
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_permute2f128_pd:
+; X64: # BB#0:
+; X64-NEXT: vperm2f128 {{.*#+}} ymm0 = zero,zero,ymm1[0,1]
+; X64-NEXT: retq
+ %res = call <4 x double> @llvm.x86.avx.vperm2f128.pd.256(<4 x double> %a0, <4 x double> %a1, i8 44)
+ ret <4 x double> %res
+}
+declare <4 x double> @llvm.x86.avx.vperm2f128.pd.256(<4 x double>, <4 x double>, i8) nounwind readnone
+
+; PR26667
+define <8 x float> @test_mm256_permute2f128_ps(<8 x float> %a0, <8 x float> %a1) nounwind {
+; X32-LABEL: test_mm256_permute2f128_ps:
+; X32: # BB#0:
+; X32-NEXT: vmovaps %ymm1, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_permute2f128_ps:
+; X64: # BB#0:
+; X64-NEXT: vmovaps %ymm1, %ymm0
+; X64-NEXT: retq
+ %res = call <8 x float> @llvm.x86.avx.vperm2f128.ps.256(<8 x float> %a0, <8 x float> %a1, i8 50)
+ ret <8 x float> %res
+}
+declare <8 x float> @llvm.x86.avx.vperm2f128.ps.256(<8 x float>, <8 x float>, i8) nounwind readnone
+
+define <4 x i64> @test_mm256_permute2f128_si256(<4 x i64> %a0, <4 x i64> %a1) nounwind {
+; X32-LABEL: test_mm256_permute2f128_si256:
+; X32: # BB#0:
+; X32-NEXT: vperm2f128 {{.*#+}} ymm0 = ymm1[2,3,0,1]
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_permute2f128_si256:
+; X64: # BB#0:
+; X64-NEXT: vperm2f128 {{.*#+}} ymm0 = ymm1[2,3,0,1]
+; X64-NEXT: retq
+ %1 = bitcast <4 x i64> %a0 to <8 x i32>
+ %2 = bitcast <4 x i64> %a1 to <8 x i32>
+ %res = call <8 x i32> @llvm.x86.avx.vperm2f128.si.256(<8 x i32> %1, <8 x i32> %2, i8 35)
+ %bc = bitcast <8 x i32> %res to <4 x i64>
+ ret <4 x i64> %bc
+}
+declare <8 x i32> @llvm.x86.avx.vperm2f128.si.256(<8 x i32>, <8 x i32>, i8) nounwind readnone
+
+define <2 x double> @test_mm_permutevar_pd(<2 x double> %a0, <2 x i64> %a1) nounwind {
+; X32-LABEL: test_mm_permutevar_pd:
+; X32: # BB#0:
+; X32-NEXT: vpermilpd %xmm1, %xmm0, %xmm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm_permutevar_pd:
+; X64: # BB#0:
+; X64-NEXT: vpermilpd %xmm1, %xmm0, %xmm0
+; X64-NEXT: retq
+ %res = call <2 x double> @llvm.x86.avx.vpermilvar.pd(<2 x double> %a0, <2 x i64> %a1)
+ ret <2 x double> %res
+}
+declare <2 x double> @llvm.x86.avx.vpermilvar.pd(<2 x double>, <2 x i64>) nounwind readnone
+
+define <4 x double> @test_mm256_permutevar_pd(<4 x double> %a0, <4 x i64> %a1) nounwind {
+; X32-LABEL: test_mm256_permutevar_pd:
+; X32: # BB#0:
+; X32-NEXT: vpermilpd %ymm1, %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_permutevar_pd:
+; X64: # BB#0:
+; X64-NEXT: vpermilpd %ymm1, %ymm0, %ymm0
+; X64-NEXT: retq
+ %res = call <4 x double> @llvm.x86.avx.vpermilvar.pd.256(<4 x double> %a0, <4 x i64> %a1)
+ ret <4 x double> %res
+}
+declare <4 x double> @llvm.x86.avx.vpermilvar.pd.256(<4 x double>, <4 x i64>) nounwind readnone
+
+define <4 x float> @test_mm_permutevar_ps(<4 x float> %a0, <2 x i64> %a1) nounwind {
+; X32-LABEL: test_mm_permutevar_ps:
+; X32: # BB#0:
+; X32-NEXT: vpermilps %xmm1, %xmm0, %xmm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm_permutevar_ps:
+; X64: # BB#0:
+; X64-NEXT: vpermilps %xmm1, %xmm0, %xmm0
+; X64-NEXT: retq
+ %arg1 = bitcast <2 x i64> %a1 to <4 x i32>
+ %res = call <4 x float> @llvm.x86.avx.vpermilvar.ps(<4 x float> %a0, <4 x i32> %arg1)
+ ret <4 x float> %res
+}
+declare <4 x float> @llvm.x86.avx.vpermilvar.ps(<4 x float>, <4 x i32>) nounwind readnone
+
+define <8 x float> @test_mm256_permutevar_ps(<8 x float> %a0, <4 x i64> %a1) nounwind {
+; X32-LABEL: test_mm256_permutevar_ps:
+; X32: # BB#0:
+; X32-NEXT: vpermilps %ymm1, %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_permutevar_ps:
+; X64: # BB#0:
+; X64-NEXT: vpermilps %ymm1, %ymm0, %ymm0
+; X64-NEXT: retq
+ %arg1 = bitcast <4 x i64> %a1 to <8 x i32>
+ %res = call <8 x float> @llvm.x86.avx.vpermilvar.ps.256(<8 x float> %a0, <8 x i32> %arg1)
+ ret <8 x float> %res
+}
+declare <8 x float> @llvm.x86.avx.vpermilvar.ps.256(<8 x float>, <8 x i32>) nounwind readnone
+
+define <8 x float> @test_mm256_rcp_ps(<8 x float> %a0) nounwind {
+; X32-LABEL: test_mm256_rcp_ps:
+; X32: # BB#0:
+; X32-NEXT: vrcpps %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_rcp_ps:
+; X64: # BB#0:
+; X64-NEXT: vrcpps %ymm0, %ymm0
+; X64-NEXT: retq
+ %res = call <8 x float> @llvm.x86.avx.rcp.ps.256(<8 x float> %a0)
+ ret <8 x float> %res
+}
+declare <8 x float> @llvm.x86.avx.rcp.ps.256(<8 x float>) nounwind readnone
+
+define <4 x double> @test_mm256_round_pd(<4 x double> %a0) nounwind {
+; X32-LABEL: test_mm256_round_pd:
+; X32: # BB#0:
+; X32-NEXT: vroundpd $4, %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_round_pd:
+; X64: # BB#0:
+; X64-NEXT: vroundpd $4, %ymm0, %ymm0
+; X64-NEXT: retq
+ %res = call <4 x double> @llvm.x86.avx.round.pd.256(<4 x double> %a0, i32 4)
+ ret <4 x double> %res
+}
+
+define <8 x float> @test_mm256_round_ps(<8 x float> %a0) nounwind {
+; X32-LABEL: test_mm256_round_ps:
+; X32: # BB#0:
+; X32-NEXT: vroundps $4, %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_round_ps:
+; X64: # BB#0:
+; X64-NEXT: vroundps $4, %ymm0, %ymm0
+; X64-NEXT: retq
+ %res = call <8 x float> @llvm.x86.avx.round.ps.256(<8 x float> %a0, i32 4)
+ ret <8 x float> %res
+}
+
+define <8 x float> @test_mm256_rsqrt_ps(<8 x float> %a0) nounwind {
+; X32-LABEL: test_mm256_rsqrt_ps:
+; X32: # BB#0:
+; X32-NEXT: vrsqrtps %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_rsqrt_ps:
+; X64: # BB#0:
+; X64-NEXT: vrsqrtps %ymm0, %ymm0
+; X64-NEXT: retq
+ %res = call <8 x float> @llvm.x86.avx.rsqrt.ps.256(<8 x float> %a0)
+ ret <8 x float> %res
+}
+declare <8 x float> @llvm.x86.avx.rsqrt.ps.256(<8 x float>) nounwind readnone
+
+define <4 x i64> @test_mm256_set_epi8(i8 %a0, i8 %a1, i8 %a2, i8 %a3, i8 %a4, i8 %a5, i8 %a6, i8 %a7, i8 %a8, i8 %a9, i8 %a10, i8 %a11, i8 %a12, i8 %a13, i8 %a14, i8 %a15, i8 %a16, i8 %a17, i8 %a18, i8 %a19, i8 %a20, i8 %a21, i8 %a22, i8 %a23, i8 %a24, i8 %a25, i8 %a26, i8 %a27, i8 %a28, i8 %a29, i8 %a30, i8 %a31) nounwind {
+; X32-LABEL: test_mm256_set_epi8:
+; X32: # BB#0:
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %ecx
+; X32-NEXT: vmovd %ecx, %xmm0
+; X32-NEXT: vpinsrb $1, %eax, %xmm0, %xmm0
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $2, %eax, %xmm0, %xmm0
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $3, %eax, %xmm0, %xmm0
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $4, %eax, %xmm0, %xmm0
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $5, %eax, %xmm0, %xmm0
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $6, %eax, %xmm0, %xmm0
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $7, %eax, %xmm0, %xmm0
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $8, %eax, %xmm0, %xmm0
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $9, %eax, %xmm0, %xmm0
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $10, %eax, %xmm0, %xmm0
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $11, %eax, %xmm0, %xmm0
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $12, %eax, %xmm0, %xmm0
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $13, %eax, %xmm0, %xmm0
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $14, %eax, %xmm0, %xmm0
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $15, %eax, %xmm0, %xmm0
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %ecx
+; X32-NEXT: vmovd %ecx, %xmm1
+; X32-NEXT: vpinsrb $1, %eax, %xmm1, %xmm1
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $2, %eax, %xmm1, %xmm1
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $3, %eax, %xmm1, %xmm1
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $4, %eax, %xmm1, %xmm1
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $5, %eax, %xmm1, %xmm1
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $6, %eax, %xmm1, %xmm1
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $7, %eax, %xmm1, %xmm1
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $8, %eax, %xmm1, %xmm1
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $9, %eax, %xmm1, %xmm1
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $10, %eax, %xmm1, %xmm1
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $11, %eax, %xmm1, %xmm1
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $12, %eax, %xmm1, %xmm1
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $13, %eax, %xmm1, %xmm1
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $14, %eax, %xmm1, %xmm1
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $15, %eax, %xmm1, %xmm1
+; X32-NEXT: vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_set_epi8:
+; X64: # BB#0:
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %r10d
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vmovd %eax, %xmm0
+; X64-NEXT: vpinsrb $1, %r10d, %xmm0, %xmm0
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vpinsrb $2, %eax, %xmm0, %xmm0
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vpinsrb $3, %eax, %xmm0, %xmm0
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vpinsrb $4, %eax, %xmm0, %xmm0
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vpinsrb $5, %eax, %xmm0, %xmm0
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vpinsrb $6, %eax, %xmm0, %xmm0
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vpinsrb $7, %eax, %xmm0, %xmm0
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vpinsrb $8, %eax, %xmm0, %xmm0
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vpinsrb $9, %eax, %xmm0, %xmm0
+; X64-NEXT: movzbl %r9b, %eax
+; X64-NEXT: vpinsrb $10, %eax, %xmm0, %xmm0
+; X64-NEXT: movzbl %r8b, %eax
+; X64-NEXT: vpinsrb $11, %eax, %xmm0, %xmm0
+; X64-NEXT: movzbl %cl, %eax
+; X64-NEXT: vpinsrb $12, %eax, %xmm0, %xmm0
+; X64-NEXT: movzbl %dl, %eax
+; X64-NEXT: vpinsrb $13, %eax, %xmm0, %xmm0
+; X64-NEXT: movzbl %sil, %eax
+; X64-NEXT: vpinsrb $14, %eax, %xmm0, %xmm0
+; X64-NEXT: movzbl %dil, %eax
+; X64-NEXT: vpinsrb $15, %eax, %xmm0, %xmm0
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %ecx
+; X64-NEXT: vmovd %ecx, %xmm1
+; X64-NEXT: vpinsrb $1, %eax, %xmm1, %xmm1
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vpinsrb $2, %eax, %xmm1, %xmm1
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vpinsrb $3, %eax, %xmm1, %xmm1
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vpinsrb $4, %eax, %xmm1, %xmm1
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vpinsrb $5, %eax, %xmm1, %xmm1
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vpinsrb $6, %eax, %xmm1, %xmm1
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vpinsrb $7, %eax, %xmm1, %xmm1
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vpinsrb $8, %eax, %xmm1, %xmm1
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vpinsrb $9, %eax, %xmm1, %xmm1
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vpinsrb $10, %eax, %xmm1, %xmm1
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vpinsrb $11, %eax, %xmm1, %xmm1
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vpinsrb $12, %eax, %xmm1, %xmm1
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vpinsrb $13, %eax, %xmm1, %xmm1
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vpinsrb $14, %eax, %xmm1, %xmm1
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vpinsrb $15, %eax, %xmm1, %xmm1
+; X64-NEXT: vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; X64-NEXT: retq
+ %res0 = insertelement <32 x i8> undef, i8 %a31, i32 0
+ %res1 = insertelement <32 x i8> %res0, i8 %a30, i32 1
+ %res2 = insertelement <32 x i8> %res1, i8 %a29, i32 2
+ %res3 = insertelement <32 x i8> %res2, i8 %a28, i32 3
+ %res4 = insertelement <32 x i8> %res3, i8 %a27, i32 4
+ %res5 = insertelement <32 x i8> %res4, i8 %a26, i32 5
+ %res6 = insertelement <32 x i8> %res5, i8 %a25, i32 6
+ %res7 = insertelement <32 x i8> %res6, i8 %a24, i32 7
+ %res8 = insertelement <32 x i8> %res7, i8 %a23, i32 8
+ %res9 = insertelement <32 x i8> %res8, i8 %a22, i32 9
+ %res10 = insertelement <32 x i8> %res9, i8 %a21, i32 10
+ %res11 = insertelement <32 x i8> %res10, i8 %a20, i32 11
+ %res12 = insertelement <32 x i8> %res11, i8 %a19, i32 12
+ %res13 = insertelement <32 x i8> %res12, i8 %a18, i32 13
+ %res14 = insertelement <32 x i8> %res13, i8 %a17, i32 14
+ %res15 = insertelement <32 x i8> %res14, i8 %a16, i32 15
+ %res16 = insertelement <32 x i8> %res15, i8 %a15, i32 16
+ %res17 = insertelement <32 x i8> %res16, i8 %a14, i32 17
+ %res18 = insertelement <32 x i8> %res17, i8 %a13, i32 18
+ %res19 = insertelement <32 x i8> %res18, i8 %a12, i32 19
+ %res20 = insertelement <32 x i8> %res19, i8 %a11, i32 20
+ %res21 = insertelement <32 x i8> %res20, i8 %a10, i32 21
+ %res22 = insertelement <32 x i8> %res21, i8 %a9 , i32 22
+ %res23 = insertelement <32 x i8> %res22, i8 %a8 , i32 23
+ %res24 = insertelement <32 x i8> %res23, i8 %a7 , i32 24
+ %res25 = insertelement <32 x i8> %res24, i8 %a6 , i32 25
+ %res26 = insertelement <32 x i8> %res25, i8 %a5 , i32 26
+ %res27 = insertelement <32 x i8> %res26, i8 %a4 , i32 27
+ %res28 = insertelement <32 x i8> %res27, i8 %a3 , i32 28
+ %res29 = insertelement <32 x i8> %res28, i8 %a2 , i32 29
+ %res30 = insertelement <32 x i8> %res29, i8 %a1 , i32 30
+ %res31 = insertelement <32 x i8> %res30, i8 %a0 , i32 31
+ %res = bitcast <32 x i8> %res31 to <4 x i64>
+ ret <4 x i64> %res
+}
+
+define <4 x i64> @test_mm256_set_epi16(i16 %a0, i16 %a1, i16 %a2, i16 %a3, i16 %a4, i16 %a5, i16 %a6, i16 %a7, i16 %a8, i16 %a9, i16 %a10, i16 %a11, i16 %a12, i16 %a13, i16 %a14, i16 %a15) nounwind {
+; X32-LABEL: test_mm256_set_epi16:
+; X32: # BB#0:
+; X32-NEXT: movw {{[0-9]+}}(%esp), %ax
+; X32-NEXT: vmovd %eax, %xmm0
+; X32-NEXT: movw {{[0-9]+}}(%esp), %ax
+; X32-NEXT: vpinsrw $1, %eax, %xmm0, %xmm0
+; X32-NEXT: movw {{[0-9]+}}(%esp), %ax
+; X32-NEXT: vpinsrw $2, %eax, %xmm0, %xmm0
+; X32-NEXT: movw {{[0-9]+}}(%esp), %ax
+; X32-NEXT: vpinsrw $3, %eax, %xmm0, %xmm0
+; X32-NEXT: movw {{[0-9]+}}(%esp), %ax
+; X32-NEXT: vpinsrw $4, %eax, %xmm0, %xmm0
+; X32-NEXT: movw {{[0-9]+}}(%esp), %ax
+; X32-NEXT: vpinsrw $5, %eax, %xmm0, %xmm0
+; X32-NEXT: movw {{[0-9]+}}(%esp), %ax
+; X32-NEXT: vpinsrw $6, %eax, %xmm0, %xmm0
+; X32-NEXT: movw {{[0-9]+}}(%esp), %ax
+; X32-NEXT: vpinsrw $7, %eax, %xmm0, %xmm0
+; X32-NEXT: movw {{[0-9]+}}(%esp), %ax
+; X32-NEXT: vmovd %eax, %xmm1
+; X32-NEXT: movw {{[0-9]+}}(%esp), %ax
+; X32-NEXT: vpinsrw $1, %eax, %xmm1, %xmm1
+; X32-NEXT: movw {{[0-9]+}}(%esp), %ax
+; X32-NEXT: vpinsrw $2, %eax, %xmm1, %xmm1
+; X32-NEXT: movw {{[0-9]+}}(%esp), %ax
+; X32-NEXT: vpinsrw $3, %eax, %xmm1, %xmm1
+; X32-NEXT: movw {{[0-9]+}}(%esp), %ax
+; X32-NEXT: vpinsrw $4, %eax, %xmm1, %xmm1
+; X32-NEXT: movw {{[0-9]+}}(%esp), %ax
+; X32-NEXT: vpinsrw $5, %eax, %xmm1, %xmm1
+; X32-NEXT: movw {{[0-9]+}}(%esp), %ax
+; X32-NEXT: vpinsrw $6, %eax, %xmm1, %xmm1
+; X32-NEXT: movw {{[0-9]+}}(%esp), %ax
+; X32-NEXT: vpinsrw $7, %eax, %xmm1, %xmm1
+; X32-NEXT: vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_set_epi16:
+; X64: # BB#0:
+; X64-NEXT: movw {{[0-9]+}}(%rsp), %ax
+; X64-NEXT: vmovd %eax, %xmm0
+; X64-NEXT: movw {{[0-9]+}}(%rsp), %ax
+; X64-NEXT: vpinsrw $1, %eax, %xmm0, %xmm0
+; X64-NEXT: vpinsrw $2, %r9d, %xmm0, %xmm0
+; X64-NEXT: vpinsrw $3, %r8d, %xmm0, %xmm0
+; X64-NEXT: vpinsrw $4, %ecx, %xmm0, %xmm0
+; X64-NEXT: vpinsrw $5, %edx, %xmm0, %xmm0
+; X64-NEXT: vpinsrw $6, %esi, %xmm0, %xmm0
+; X64-NEXT: vpinsrw $7, %edi, %xmm0, %xmm0
+; X64-NEXT: movw {{[0-9]+}}(%rsp), %ax
+; X64-NEXT: vmovd %eax, %xmm1
+; X64-NEXT: movw {{[0-9]+}}(%rsp), %ax
+; X64-NEXT: vpinsrw $1, %eax, %xmm1, %xmm1
+; X64-NEXT: movw {{[0-9]+}}(%rsp), %ax
+; X64-NEXT: vpinsrw $2, %eax, %xmm1, %xmm1
+; X64-NEXT: movw {{[0-9]+}}(%rsp), %ax
+; X64-NEXT: vpinsrw $3, %eax, %xmm1, %xmm1
+; X64-NEXT: movw {{[0-9]+}}(%rsp), %ax
+; X64-NEXT: vpinsrw $4, %eax, %xmm1, %xmm1
+; X64-NEXT: movw {{[0-9]+}}(%rsp), %ax
+; X64-NEXT: vpinsrw $5, %eax, %xmm1, %xmm1
+; X64-NEXT: movw {{[0-9]+}}(%rsp), %ax
+; X64-NEXT: vpinsrw $6, %eax, %xmm1, %xmm1
+; X64-NEXT: movw {{[0-9]+}}(%rsp), %ax
+; X64-NEXT: vpinsrw $7, %eax, %xmm1, %xmm1
+; X64-NEXT: vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; X64-NEXT: retq
+ %res0 = insertelement <16 x i16> undef, i16 %a15, i32 0
+ %res1 = insertelement <16 x i16> %res0, i16 %a14, i32 1
+ %res2 = insertelement <16 x i16> %res1, i16 %a13, i32 2
+ %res3 = insertelement <16 x i16> %res2, i16 %a12, i32 3
+ %res4 = insertelement <16 x i16> %res3, i16 %a11, i32 4
+ %res5 = insertelement <16 x i16> %res4, i16 %a10, i32 5
+ %res6 = insertelement <16 x i16> %res5, i16 %a9 , i32 6
+ %res7 = insertelement <16 x i16> %res6, i16 %a8 , i32 7
+ %res8 = insertelement <16 x i16> %res7, i16 %a7 , i32 8
+ %res9 = insertelement <16 x i16> %res8, i16 %a6 , i32 9
+ %res10 = insertelement <16 x i16> %res9, i16 %a5 , i32 10
+ %res11 = insertelement <16 x i16> %res10, i16 %a4 , i32 11
+ %res12 = insertelement <16 x i16> %res11, i16 %a3 , i32 12
+ %res13 = insertelement <16 x i16> %res12, i16 %a2 , i32 13
+ %res14 = insertelement <16 x i16> %res13, i16 %a1 , i32 14
+ %res15 = insertelement <16 x i16> %res14, i16 %a0 , i32 15
+ %res = bitcast <16 x i16> %res15 to <4 x i64>
+ ret <4 x i64> %res
+}
+
+define <4 x i64> @test_mm256_set_epi32(i32 %a0, i32 %a1, i32 %a2, i32 %a3, i32 %a4, i32 %a5, i32 %a6, i32 %a7) nounwind {
+; X32-LABEL: test_mm256_set_epi32:
+; X32: # BB#0:
+; X32-NEXT: vmovd {{.*#+}} xmm0 = mem[0],zero,zero,zero
+; X32-NEXT: vpinsrd $1, {{[0-9]+}}(%esp), %xmm0, %xmm0
+; X32-NEXT: vpinsrd $2, {{[0-9]+}}(%esp), %xmm0, %xmm0
+; X32-NEXT: vpinsrd $3, {{[0-9]+}}(%esp), %xmm0, %xmm0
+; X32-NEXT: vmovd {{.*#+}} xmm1 = mem[0],zero,zero,zero
+; X32-NEXT: vpinsrd $1, {{[0-9]+}}(%esp), %xmm1, %xmm1
+; X32-NEXT: vpinsrd $2, {{[0-9]+}}(%esp), %xmm1, %xmm1
+; X32-NEXT: vpinsrd $3, {{[0-9]+}}(%esp), %xmm1, %xmm1
+; X32-NEXT: vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_set_epi32:
+; X64: # BB#0:
+; X64-NEXT: vmovd %ecx, %xmm0
+; X64-NEXT: vpinsrd $1, %edx, %xmm0, %xmm0
+; X64-NEXT: vpinsrd $2, %esi, %xmm0, %xmm0
+; X64-NEXT: vpinsrd $3, %edi, %xmm0, %xmm0
+; X64-NEXT: vmovd {{.*#+}} xmm1 = mem[0],zero,zero,zero
+; X64-NEXT: vpinsrd $1, {{[0-9]+}}(%rsp), %xmm1, %xmm1
+; X64-NEXT: vpinsrd $2, %r9d, %xmm1, %xmm1
+; X64-NEXT: vpinsrd $3, %r8d, %xmm1, %xmm1
+; X64-NEXT: vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; X64-NEXT: retq
+ %res0 = insertelement <8 x i32> undef, i32 %a7, i32 0
+ %res1 = insertelement <8 x i32> %res0, i32 %a6, i32 1
+ %res2 = insertelement <8 x i32> %res1, i32 %a5, i32 2
+ %res3 = insertelement <8 x i32> %res2, i32 %a4, i32 3
+ %res4 = insertelement <8 x i32> %res3, i32 %a3, i32 4
+ %res5 = insertelement <8 x i32> %res4, i32 %a2, i32 5
+ %res6 = insertelement <8 x i32> %res5, i32 %a1, i32 6
+ %res7 = insertelement <8 x i32> %res6, i32 %a0, i32 7
+ %res = bitcast <8 x i32> %res7 to <4 x i64>
+ ret <4 x i64> %res
+}
+
+define <4 x i64> @test_mm256_set_epi64x(i64 %a0, i64 %a1, i64 %a2, i64 %a3) nounwind {
+; X32-LABEL: test_mm256_set_epi64x:
+; X32: # BB#0:
+; X32-NEXT: vmovd {{.*#+}} xmm0 = mem[0],zero,zero,zero
+; X32-NEXT: vpinsrd $1, {{[0-9]+}}(%esp), %xmm0, %xmm0
+; X32-NEXT: vpinsrd $2, {{[0-9]+}}(%esp), %xmm0, %xmm0
+; X32-NEXT: vpinsrd $3, {{[0-9]+}}(%esp), %xmm0, %xmm0
+; X32-NEXT: vmovd {{.*#+}} xmm1 = mem[0],zero,zero,zero
+; X32-NEXT: vpinsrd $1, {{[0-9]+}}(%esp), %xmm1, %xmm1
+; X32-NEXT: vpinsrd $2, {{[0-9]+}}(%esp), %xmm1, %xmm1
+; X32-NEXT: vpinsrd $3, {{[0-9]+}}(%esp), %xmm1, %xmm1
+; X32-NEXT: vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_set_epi64x:
+; X64: # BB#0:
+; X64-NEXT: vmovq %rdi, %xmm0
+; X64-NEXT: vmovq %rsi, %xmm1
+; X64-NEXT: vpunpcklqdq {{.*#+}} xmm0 = xmm1[0],xmm0[0]
+; X64-NEXT: vmovq %rdx, %xmm1
+; X64-NEXT: vmovq %rcx, %xmm2
+; X64-NEXT: vpunpcklqdq {{.*#+}} xmm1 = xmm2[0],xmm1[0]
+; X64-NEXT: vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; X64-NEXT: retq
+ %res0 = insertelement <4 x i64> undef, i64 %a3, i32 0
+ %res1 = insertelement <4 x i64> %res0, i64 %a2, i32 1
+ %res2 = insertelement <4 x i64> %res1, i64 %a1, i32 2
+ %res3 = insertelement <4 x i64> %res2, i64 %a0, i32 3
+ ret <4 x i64> %res3
+}
+
+define <8 x float> @test_mm256_set_m128(<4 x float> %a0, <4 x float> %a1) nounwind {
+; X32-LABEL: test_mm256_set_m128:
+; X32: # BB#0:
+; X32-NEXT: vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_set_m128:
+; X64: # BB#0:
+; X64-NEXT: vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; X64-NEXT: retq
+ %res = shufflevector <4 x float> %a1, <4 x float> %a0, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
+ ret <8 x float> %res
+}
+
+define <4 x double> @test_mm256_set_m128d(<2 x double> %a0, <2 x double> %a1) nounwind {
+; X32-LABEL: test_mm256_set_m128d:
+; X32: # BB#0:
+; X32-NEXT: vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_set_m128d:
+; X64: # BB#0:
+; X64-NEXT: vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; X64-NEXT: retq
+ %arg0 = bitcast <2 x double> %a0 to <4 x float>
+ %arg1 = bitcast <2 x double> %a1 to <4 x float>
+ %res = shufflevector <4 x float> %arg1, <4 x float> %arg0, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
+ %bc = bitcast <8 x float> %res to <4 x double>
+ ret <4 x double> %bc
+}
+
+define <4 x i64> @test_mm256_set_m128i(<2 x i64> %a0, <2 x i64> %a1) nounwind {
+; X32-LABEL: test_mm256_set_m128i:
+; X32: # BB#0:
+; X32-NEXT: vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_set_m128i:
+; X64: # BB#0:
+; X64-NEXT: vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; X64-NEXT: retq
+ %arg0 = bitcast <2 x i64> %a0 to <4 x float>
+ %arg1 = bitcast <2 x i64> %a1 to <4 x float>
+ %res = shufflevector <4 x float> %arg1, <4 x float> %arg0, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
+ %bc = bitcast <8 x float> %res to <4 x i64>
+ ret <4 x i64> %bc
+}
+
+define <4 x double> @test_mm256_set_pd(double %a0, double %a1, double %a2, double %a3) nounwind {
+; X32-LABEL: test_mm256_set_pd:
+; X32: # BB#0:
+; X32-NEXT: vmovsd {{.*#+}} xmm0 = mem[0],zero
+; X32-NEXT: vmovsd {{.*#+}} xmm1 = mem[0],zero
+; X32-NEXT: vmovsd {{.*#+}} xmm2 = mem[0],zero
+; X32-NEXT: vmovsd {{.*#+}} xmm3 = mem[0],zero
+; X32-NEXT: vunpcklpd {{.*#+}} xmm2 = xmm2[0],xmm3[0]
+; X32-NEXT: vunpcklpd {{.*#+}} xmm0 = xmm0[0],xmm1[0]
+; X32-NEXT: vinsertf128 $1, %xmm2, %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_set_pd:
+; X64: # BB#0:
+; X64-NEXT: vunpcklpd {{.*#+}} xmm0 = xmm1[0],xmm0[0]
+; X64-NEXT: vunpcklpd {{.*#+}} xmm1 = xmm3[0],xmm2[0]
+; X64-NEXT: vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; X64-NEXT: retq
+ %res0 = insertelement <4 x double> undef, double %a3, i32 0
+ %res1 = insertelement <4 x double> %res0, double %a2, i32 1
+ %res2 = insertelement <4 x double> %res1, double %a1, i32 2
+ %res3 = insertelement <4 x double> %res2, double %a0, i32 3
+ ret <4 x double> %res3
+}
+
+define <8 x float> @test_mm256_set_ps(float %a0, float %a1, float %a2, float %a3, float %a4, float %a5, float %a6, float %a7) nounwind {
+; X32-LABEL: test_mm256_set_ps:
+; X32: # BB#0:
+; X32-NEXT: vmovss {{.*#+}} xmm0 = mem[0],zero,zero,zero
+; X32-NEXT: vmovss {{.*#+}} xmm1 = mem[0],zero,zero,zero
+; X32-NEXT: vmovss {{.*#+}} xmm2 = mem[0],zero,zero,zero
+; X32-NEXT: vmovss {{.*#+}} xmm3 = mem[0],zero,zero,zero
+; X32-NEXT: vmovss {{.*#+}} xmm4 = mem[0],zero,zero,zero
+; X32-NEXT: vmovss {{.*#+}} xmm5 = mem[0],zero,zero,zero
+; X32-NEXT: vmovss {{.*#+}} xmm6 = mem[0],zero,zero,zero
+; X32-NEXT: vmovss {{.*#+}} xmm7 = mem[0],zero,zero,zero
+; X32-NEXT: vinsertps {{.*#+}} xmm4 = xmm4[0],xmm5[0],xmm4[2,3]
+; X32-NEXT: vinsertps {{.*#+}} xmm4 = xmm4[0,1],xmm6[0],xmm4[3]
+; X32-NEXT: vinsertps {{.*#+}} xmm4 = xmm4[0,1,2],xmm7[0]
+; X32-NEXT: vinsertps {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[2,3]
+; X32-NEXT: vinsertps {{.*#+}} xmm0 = xmm0[0,1],xmm2[0],xmm0[3]
+; X32-NEXT: vinsertps {{.*#+}} xmm0 = xmm0[0,1,2],xmm3[0]
+; X32-NEXT: vinsertf128 $1, %xmm4, %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_set_ps:
+; X64: # BB#0:
+; X64-NEXT: vinsertps {{.*#+}} xmm2 = xmm3[0],xmm2[0],xmm3[2,3]
+; X64-NEXT: vinsertps {{.*#+}} xmm1 = xmm2[0,1],xmm1[0],xmm2[3]
+; X64-NEXT: vinsertps {{.*#+}} xmm0 = xmm1[0,1,2],xmm0[0]
+; X64-NEXT: vinsertps {{.*#+}} xmm1 = xmm7[0],xmm6[0],xmm7[2,3]
+; X64-NEXT: vinsertps {{.*#+}} xmm1 = xmm1[0,1],xmm5[0],xmm1[3]
+; X64-NEXT: vinsertps {{.*#+}} xmm1 = xmm1[0,1,2],xmm4[0]
+; X64-NEXT: vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; X64-NEXT: retq
+ %res0 = insertelement <8 x float> undef, float %a7, i32 0
+ %res1 = insertelement <8 x float> %res0, float %a6, i32 1
+ %res2 = insertelement <8 x float> %res1, float %a5, i32 2
+ %res3 = insertelement <8 x float> %res2, float %a4, i32 3
+ %res4 = insertelement <8 x float> %res3, float %a3, i32 4
+ %res5 = insertelement <8 x float> %res4, float %a2, i32 5
+ %res6 = insertelement <8 x float> %res5, float %a1, i32 6
+ %res7 = insertelement <8 x float> %res6, float %a0, i32 7
+ ret <8 x float> %res7
+}
+
+define <4 x i64> @test_mm256_set1_epi8(i8 %a0) nounwind {
+; X32-LABEL: test_mm256_set1_epi8:
+; X32: # BB#0:
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vmovd %eax, %xmm0
+; X32-NEXT: vpxor %xmm1, %xmm1, %xmm1
+; X32-NEXT: vpshufb %xmm1, %xmm0, %xmm0
+; X32-NEXT: vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_set1_epi8:
+; X64: # BB#0:
+; X64-NEXT: movzbl %dil, %eax
+; X64-NEXT: vmovd %eax, %xmm0
+; X64-NEXT: vpxor %xmm1, %xmm1, %xmm1
+; X64-NEXT: vpshufb %xmm1, %xmm0, %xmm0
+; X64-NEXT: vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; X64-NEXT: retq
+ %res0 = insertelement <32 x i8> undef, i8 %a0, i32 0
+ %res1 = insertelement <32 x i8> %res0, i8 %a0, i32 1
+ %res2 = insertelement <32 x i8> %res1, i8 %a0, i32 2
+ %res3 = insertelement <32 x i8> %res2, i8 %a0, i32 3
+ %res4 = insertelement <32 x i8> %res3, i8 %a0, i32 4
+ %res5 = insertelement <32 x i8> %res4, i8 %a0, i32 5
+ %res6 = insertelement <32 x i8> %res5, i8 %a0, i32 6
+ %res7 = insertelement <32 x i8> %res6, i8 %a0, i32 7
+ %res8 = insertelement <32 x i8> %res7, i8 %a0, i32 8
+ %res9 = insertelement <32 x i8> %res8, i8 %a0, i32 9
+ %res10 = insertelement <32 x i8> %res9, i8 %a0, i32 10
+ %res11 = insertelement <32 x i8> %res10, i8 %a0, i32 11
+ %res12 = insertelement <32 x i8> %res11, i8 %a0, i32 12
+ %res13 = insertelement <32 x i8> %res12, i8 %a0, i32 13
+ %res14 = insertelement <32 x i8> %res13, i8 %a0, i32 14
+ %res15 = insertelement <32 x i8> %res14, i8 %a0, i32 15
+ %res16 = insertelement <32 x i8> %res15, i8 %a0, i32 16
+ %res17 = insertelement <32 x i8> %res16, i8 %a0, i32 17
+ %res18 = insertelement <32 x i8> %res17, i8 %a0, i32 18
+ %res19 = insertelement <32 x i8> %res18, i8 %a0, i32 19
+ %res20 = insertelement <32 x i8> %res19, i8 %a0, i32 20
+ %res21 = insertelement <32 x i8> %res20, i8 %a0, i32 21
+ %res22 = insertelement <32 x i8> %res21, i8 %a0, i32 22
+ %res23 = insertelement <32 x i8> %res22, i8 %a0, i32 23
+ %res24 = insertelement <32 x i8> %res23, i8 %a0, i32 24
+ %res25 = insertelement <32 x i8> %res24, i8 %a0, i32 25
+ %res26 = insertelement <32 x i8> %res25, i8 %a0, i32 26
+ %res27 = insertelement <32 x i8> %res26, i8 %a0, i32 27
+ %res28 = insertelement <32 x i8> %res27, i8 %a0, i32 28
+ %res29 = insertelement <32 x i8> %res28, i8 %a0, i32 29
+ %res30 = insertelement <32 x i8> %res29, i8 %a0, i32 30
+ %res31 = insertelement <32 x i8> %res30, i8 %a0, i32 31
+ %res = bitcast <32 x i8> %res31 to <4 x i64>
+ ret <4 x i64> %res
+}
+
+define <4 x i64> @test_mm256_set1_epi16(i16 %a0) nounwind {
+; X32-LABEL: test_mm256_set1_epi16:
+; X32: # BB#0:
+; X32-NEXT: movw {{[0-9]+}}(%esp), %ax
+; X32-NEXT: vmovd %eax, %xmm0
+; X32-NEXT: vpshuflw {{.*#+}} xmm0 = xmm0[0,0,0,0,4,5,6,7]
+; X32-NEXT: vpshufd {{.*#+}} xmm0 = xmm0[0,0,1,1]
+; X32-NEXT: vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_set1_epi16:
+; X64: # BB#0:
+; X64-NEXT: vmovd %edi, %xmm0
+; X64-NEXT: vpshuflw {{.*#+}} xmm0 = xmm0[0,0,0,0,4,5,6,7]
+; X64-NEXT: vpshufd {{.*#+}} xmm0 = xmm0[0,0,1,1]
+; X64-NEXT: vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; X64-NEXT: retq
+ %res0 = insertelement <16 x i16> undef, i16 %a0, i32 0
+ %res1 = insertelement <16 x i16> %res0, i16 %a0, i32 1
+ %res2 = insertelement <16 x i16> %res1, i16 %a0, i32 2
+ %res3 = insertelement <16 x i16> %res2, i16 %a0, i32 3
+ %res4 = insertelement <16 x i16> %res3, i16 %a0, i32 4
+ %res5 = insertelement <16 x i16> %res4, i16 %a0, i32 5
+ %res6 = insertelement <16 x i16> %res5, i16 %a0, i32 6
+ %res7 = insertelement <16 x i16> %res6, i16 %a0, i32 7
+ %res8 = insertelement <16 x i16> %res7, i16 %a0, i32 8
+ %res9 = insertelement <16 x i16> %res8, i16 %a0, i32 9
+ %res10 = insertelement <16 x i16> %res9, i16 %a0, i32 10
+ %res11 = insertelement <16 x i16> %res10, i16 %a0, i32 11
+ %res12 = insertelement <16 x i16> %res11, i16 %a0, i32 12
+ %res13 = insertelement <16 x i16> %res12, i16 %a0, i32 13
+ %res14 = insertelement <16 x i16> %res13, i16 %a0, i32 14
+ %res15 = insertelement <16 x i16> %res14, i16 %a0, i32 15
+ %res = bitcast <16 x i16> %res15 to <4 x i64>
+ ret <4 x i64> %res
+}
+
+define <4 x i64> @test_mm256_set1_epi32(i32 %a0) nounwind {
+; X32-LABEL: test_mm256_set1_epi32:
+; X32: # BB#0:
+; X32-NEXT: vmovd {{.*#+}} xmm0 = mem[0],zero,zero,zero
+; X32-NEXT: vpermilps {{.*#+}} xmm0 = xmm0[0,0,0,0]
+; X32-NEXT: vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_set1_epi32:
+; X64: # BB#0:
+; X64-NEXT: vmovd %edi, %xmm0
+; X64-NEXT: vpermilps {{.*#+}} xmm0 = xmm0[0,0,0,0]
+; X64-NEXT: vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; X64-NEXT: retq
+ %res0 = insertelement <8 x i32> undef, i32 %a0, i32 0
+ %res1 = insertelement <8 x i32> %res0, i32 %a0, i32 1
+ %res2 = insertelement <8 x i32> %res1, i32 %a0, i32 2
+ %res3 = insertelement <8 x i32> %res2, i32 %a0, i32 3
+ %res4 = insertelement <8 x i32> %res3, i32 %a0, i32 4
+ %res5 = insertelement <8 x i32> %res4, i32 %a0, i32 5
+ %res6 = insertelement <8 x i32> %res5, i32 %a0, i32 6
+ %res7 = insertelement <8 x i32> %res6, i32 %a0, i32 7
+ %res = bitcast <8 x i32> %res7 to <4 x i64>
+ ret <4 x i64> %res
+}
+
+define <4 x i64> @test_mm256_set1_epi64x(i64 %a0) nounwind {
+; X32-LABEL: test_mm256_set1_epi64x:
+; X32: # BB#0:
+; X32-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: movl {{[0-9]+}}(%esp), %ecx
+; X32-NEXT: vmovd %ecx, %xmm0
+; X32-NEXT: vpinsrd $1, %eax, %xmm0, %xmm0
+; X32-NEXT: vpinsrd $2, %ecx, %xmm0, %xmm0
+; X32-NEXT: vpinsrd $3, %eax, %xmm0, %xmm0
+; X32-NEXT: vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_set1_epi64x:
+; X64: # BB#0:
+; X64-NEXT: vmovq %rdi, %xmm0
+; X64-NEXT: vmovddup {{.*#+}} xmm0 = xmm0[0,0]
+; X64-NEXT: vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; X64-NEXT: retq
+ %res0 = insertelement <4 x i64> undef, i64 %a0, i32 0
+ %res1 = insertelement <4 x i64> %res0, i64 %a0, i32 1
+ %res2 = insertelement <4 x i64> %res1, i64 %a0, i32 2
+ %res3 = insertelement <4 x i64> %res2, i64 %a0, i32 3
+ ret <4 x i64> %res3
+}
+
+define <4 x double> @test_mm256_set1_pd(double %a0) nounwind {
+; X32-LABEL: test_mm256_set1_pd:
+; X32: # BB#0:
+; X32-NEXT: vmovsd {{.*#+}} xmm0 = mem[0],zero
+; X32-NEXT: vmovddup {{.*#+}} xmm0 = xmm0[0,0]
+; X32-NEXT: vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_set1_pd:
+; X64: # BB#0:
+; X64-NEXT: vmovddup {{.*#+}} xmm0 = xmm0[0,0]
+; X64-NEXT: vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; X64-NEXT: retq
+ %res0 = insertelement <4 x double> undef, double %a0, i32 0
+ %res1 = insertelement <4 x double> %res0, double %a0, i32 1
+ %res2 = insertelement <4 x double> %res1, double %a0, i32 2
+ %res3 = insertelement <4 x double> %res2, double %a0, i32 3
+ ret <4 x double> %res3
+}
+
+define <8 x float> @test_mm256_set1_ps(float %a0) nounwind {
+; X32-LABEL: test_mm256_set1_ps:
+; X32: # BB#0:
+; X32-NEXT: vmovss {{.*#+}} xmm0 = mem[0],zero,zero,zero
+; X32-NEXT: vpermilps {{.*#+}} xmm0 = xmm0[0,0,0,0]
+; X32-NEXT: vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_set1_ps:
+; X64: # BB#0:
+; X64-NEXT: vpermilps {{.*#+}} xmm0 = xmm0[0,0,0,0]
+; X64-NEXT: vinsertf128 $1, %xmm0, %ymm0, %ymm0
+; X64-NEXT: retq
+ %res0 = insertelement <8 x float> undef, float %a0, i32 0
+ %res1 = insertelement <8 x float> %res0, float %a0, i32 1
+ %res2 = insertelement <8 x float> %res1, float %a0, i32 2
+ %res3 = insertelement <8 x float> %res2, float %a0, i32 3
+ %res4 = insertelement <8 x float> %res3, float %a0, i32 4
+ %res5 = insertelement <8 x float> %res4, float %a0, i32 5
+ %res6 = insertelement <8 x float> %res5, float %a0, i32 6
+ %res7 = insertelement <8 x float> %res6, float %a0, i32 7
+ ret <8 x float> %res7
+}
+
+define <4 x i64> @test_mm256_setr_epi8(i8 %a0, i8 %a1, i8 %a2, i8 %a3, i8 %a4, i8 %a5, i8 %a6, i8 %a7, i8 %a8, i8 %a9, i8 %a10, i8 %a11, i8 %a12, i8 %a13, i8 %a14, i8 %a15, i8 %a16, i8 %a17, i8 %a18, i8 %a19, i8 %a20, i8 %a21, i8 %a22, i8 %a23, i8 %a24, i8 %a25, i8 %a26, i8 %a27, i8 %a28, i8 %a29, i8 %a30, i8 %a31) nounwind {
+; X32-LABEL: test_mm256_setr_epi8:
+; X32: # BB#0:
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %ecx
+; X32-NEXT: vmovd %ecx, %xmm0
+; X32-NEXT: vpinsrb $1, %eax, %xmm0, %xmm0
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $2, %eax, %xmm0, %xmm0
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $3, %eax, %xmm0, %xmm0
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $4, %eax, %xmm0, %xmm0
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $5, %eax, %xmm0, %xmm0
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $6, %eax, %xmm0, %xmm0
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $7, %eax, %xmm0, %xmm0
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $8, %eax, %xmm0, %xmm0
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $9, %eax, %xmm0, %xmm0
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $10, %eax, %xmm0, %xmm0
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $11, %eax, %xmm0, %xmm0
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $12, %eax, %xmm0, %xmm0
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $13, %eax, %xmm0, %xmm0
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $14, %eax, %xmm0, %xmm0
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $15, %eax, %xmm0, %xmm0
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %ecx
+; X32-NEXT: vmovd %ecx, %xmm1
+; X32-NEXT: vpinsrb $1, %eax, %xmm1, %xmm1
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $2, %eax, %xmm1, %xmm1
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $3, %eax, %xmm1, %xmm1
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $4, %eax, %xmm1, %xmm1
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $5, %eax, %xmm1, %xmm1
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $6, %eax, %xmm1, %xmm1
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $7, %eax, %xmm1, %xmm1
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $8, %eax, %xmm1, %xmm1
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $9, %eax, %xmm1, %xmm1
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $10, %eax, %xmm1, %xmm1
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $11, %eax, %xmm1, %xmm1
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $12, %eax, %xmm1, %xmm1
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $13, %eax, %xmm1, %xmm1
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $14, %eax, %xmm1, %xmm1
+; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vpinsrb $15, %eax, %xmm1, %xmm1
+; X32-NEXT: vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_setr_epi8:
+; X64: # BB#0:
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %r10d
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vmovd %eax, %xmm0
+; X64-NEXT: vpinsrb $1, %r10d, %xmm0, %xmm0
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vpinsrb $2, %eax, %xmm0, %xmm0
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vpinsrb $3, %eax, %xmm0, %xmm0
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vpinsrb $4, %eax, %xmm0, %xmm0
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vpinsrb $5, %eax, %xmm0, %xmm0
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vpinsrb $6, %eax, %xmm0, %xmm0
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vpinsrb $7, %eax, %xmm0, %xmm0
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vpinsrb $8, %eax, %xmm0, %xmm0
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vpinsrb $9, %eax, %xmm0, %xmm0
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vpinsrb $10, %eax, %xmm0, %xmm0
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vpinsrb $11, %eax, %xmm0, %xmm0
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vpinsrb $12, %eax, %xmm0, %xmm0
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vpinsrb $13, %eax, %xmm0, %xmm0
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vpinsrb $14, %eax, %xmm0, %xmm0
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vpinsrb $15, %eax, %xmm0, %xmm0
+; X64-NEXT: movzbl %sil, %eax
+; X64-NEXT: movzbl %dil, %esi
+; X64-NEXT: vmovd %esi, %xmm1
+; X64-NEXT: vpinsrb $1, %eax, %xmm1, %xmm1
+; X64-NEXT: movzbl %dl, %eax
+; X64-NEXT: vpinsrb $2, %eax, %xmm1, %xmm1
+; X64-NEXT: movzbl %cl, %eax
+; X64-NEXT: vpinsrb $3, %eax, %xmm1, %xmm1
+; X64-NEXT: movzbl %r8b, %eax
+; X64-NEXT: vpinsrb $4, %eax, %xmm1, %xmm1
+; X64-NEXT: movzbl %r9b, %eax
+; X64-NEXT: vpinsrb $5, %eax, %xmm1, %xmm1
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vpinsrb $6, %eax, %xmm1, %xmm1
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vpinsrb $7, %eax, %xmm1, %xmm1
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vpinsrb $8, %eax, %xmm1, %xmm1
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vpinsrb $9, %eax, %xmm1, %xmm1
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vpinsrb $10, %eax, %xmm1, %xmm1
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vpinsrb $11, %eax, %xmm1, %xmm1
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vpinsrb $12, %eax, %xmm1, %xmm1
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vpinsrb $13, %eax, %xmm1, %xmm1
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vpinsrb $14, %eax, %xmm1, %xmm1
+; X64-NEXT: movzbl {{[0-9]+}}(%rsp), %eax
+; X64-NEXT: vpinsrb $15, %eax, %xmm1, %xmm1
+; X64-NEXT: vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; X64-NEXT: retq
+ %res0 = insertelement <32 x i8> undef, i8 %a0 , i32 0
+ %res1 = insertelement <32 x i8> %res0, i8 %a1 , i32 1
+ %res2 = insertelement <32 x i8> %res1, i8 %a2 , i32 2
+ %res3 = insertelement <32 x i8> %res2, i8 %a3 , i32 3
+ %res4 = insertelement <32 x i8> %res3, i8 %a4 , i32 4
+ %res5 = insertelement <32 x i8> %res4, i8 %a5 , i32 5
+ %res6 = insertelement <32 x i8> %res5, i8 %a6 , i32 6
+ %res7 = insertelement <32 x i8> %res6, i8 %a7 , i32 7
+ %res8 = insertelement <32 x i8> %res7, i8 %a8 , i32 8
+ %res9 = insertelement <32 x i8> %res8, i8 %a9 , i32 9
+ %res10 = insertelement <32 x i8> %res9, i8 %a10, i32 10
+ %res11 = insertelement <32 x i8> %res10, i8 %a11, i32 11
+ %res12 = insertelement <32 x i8> %res11, i8 %a12, i32 12
+ %res13 = insertelement <32 x i8> %res12, i8 %a13, i32 13
+ %res14 = insertelement <32 x i8> %res13, i8 %a14, i32 14
+ %res15 = insertelement <32 x i8> %res14, i8 %a15, i32 15
+ %res16 = insertelement <32 x i8> %res15, i8 %a16, i32 16
+ %res17 = insertelement <32 x i8> %res16, i8 %a17, i32 17
+ %res18 = insertelement <32 x i8> %res17, i8 %a18, i32 18
+ %res19 = insertelement <32 x i8> %res18, i8 %a19, i32 19
+ %res20 = insertelement <32 x i8> %res19, i8 %a20, i32 20
+ %res21 = insertelement <32 x i8> %res20, i8 %a21, i32 21
+ %res22 = insertelement <32 x i8> %res21, i8 %a22, i32 22
+ %res23 = insertelement <32 x i8> %res22, i8 %a23, i32 23
+ %res24 = insertelement <32 x i8> %res23, i8 %a24, i32 24
+ %res25 = insertelement <32 x i8> %res24, i8 %a25, i32 25
+ %res26 = insertelement <32 x i8> %res25, i8 %a26, i32 26
+ %res27 = insertelement <32 x i8> %res26, i8 %a27, i32 27
+ %res28 = insertelement <32 x i8> %res27, i8 %a28, i32 28
+ %res29 = insertelement <32 x i8> %res28, i8 %a29, i32 29
+ %res30 = insertelement <32 x i8> %res29, i8 %a30, i32 30
+ %res31 = insertelement <32 x i8> %res30, i8 %a31, i32 31
+ %res = bitcast <32 x i8> %res31 to <4 x i64>
+ ret <4 x i64> %res
+}
+
+define <4 x i64> @test_mm256_setr_epi16(i16 %a0, i16 %a1, i16 %a2, i16 %a3, i16 %a4, i16 %a5, i16 %a6, i16 %a7, i16 %a8, i16 %a9, i16 %a10, i16 %a11, i16 %a12, i16 %a13, i16 %a14, i16 %a15) nounwind {
+; X32-LABEL: test_mm256_setr_epi16:
+; X32: # BB#0:
+; X32-NEXT: movw {{[0-9]+}}(%esp), %ax
+; X32-NEXT: vmovd %eax, %xmm0
+; X32-NEXT: movw {{[0-9]+}}(%esp), %ax
+; X32-NEXT: vpinsrw $1, %eax, %xmm0, %xmm0
+; X32-NEXT: movw {{[0-9]+}}(%esp), %ax
+; X32-NEXT: vpinsrw $2, %eax, %xmm0, %xmm0
+; X32-NEXT: movw {{[0-9]+}}(%esp), %ax
+; X32-NEXT: vpinsrw $3, %eax, %xmm0, %xmm0
+; X32-NEXT: movw {{[0-9]+}}(%esp), %ax
+; X32-NEXT: vpinsrw $4, %eax, %xmm0, %xmm0
+; X32-NEXT: movw {{[0-9]+}}(%esp), %ax
+; X32-NEXT: vpinsrw $5, %eax, %xmm0, %xmm0
+; X32-NEXT: movw {{[0-9]+}}(%esp), %ax
+; X32-NEXT: vpinsrw $6, %eax, %xmm0, %xmm0
+; X32-NEXT: movw {{[0-9]+}}(%esp), %ax
+; X32-NEXT: vpinsrw $7, %eax, %xmm0, %xmm0
+; X32-NEXT: movw {{[0-9]+}}(%esp), %ax
+; X32-NEXT: vmovd %eax, %xmm1
+; X32-NEXT: movw {{[0-9]+}}(%esp), %ax
+; X32-NEXT: vpinsrw $1, %eax, %xmm1, %xmm1
+; X32-NEXT: movw {{[0-9]+}}(%esp), %ax
+; X32-NEXT: vpinsrw $2, %eax, %xmm1, %xmm1
+; X32-NEXT: movw {{[0-9]+}}(%esp), %ax
+; X32-NEXT: vpinsrw $3, %eax, %xmm1, %xmm1
+; X32-NEXT: movw {{[0-9]+}}(%esp), %ax
+; X32-NEXT: vpinsrw $4, %eax, %xmm1, %xmm1
+; X32-NEXT: movw {{[0-9]+}}(%esp), %ax
+; X32-NEXT: vpinsrw $5, %eax, %xmm1, %xmm1
+; X32-NEXT: movw {{[0-9]+}}(%esp), %ax
+; X32-NEXT: vpinsrw $6, %eax, %xmm1, %xmm1
+; X32-NEXT: movw {{[0-9]+}}(%esp), %ax
+; X32-NEXT: vpinsrw $7, %eax, %xmm1, %xmm1
+; X32-NEXT: vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_setr_epi16:
+; X64: # BB#0:
+; X64-NEXT: movw {{[0-9]+}}(%rsp), %ax
+; X64-NEXT: vmovd %eax, %xmm0
+; X64-NEXT: movw {{[0-9]+}}(%rsp), %ax
+; X64-NEXT: vpinsrw $1, %eax, %xmm0, %xmm0
+; X64-NEXT: movw {{[0-9]+}}(%rsp), %ax
+; X64-NEXT: vpinsrw $2, %eax, %xmm0, %xmm0
+; X64-NEXT: movw {{[0-9]+}}(%rsp), %ax
+; X64-NEXT: vpinsrw $3, %eax, %xmm0, %xmm0
+; X64-NEXT: movw {{[0-9]+}}(%rsp), %ax
+; X64-NEXT: vpinsrw $4, %eax, %xmm0, %xmm0
+; X64-NEXT: movw {{[0-9]+}}(%rsp), %ax
+; X64-NEXT: vpinsrw $5, %eax, %xmm0, %xmm0
+; X64-NEXT: movw {{[0-9]+}}(%rsp), %ax
+; X64-NEXT: vpinsrw $6, %eax, %xmm0, %xmm0
+; X64-NEXT: movw {{[0-9]+}}(%rsp), %ax
+; X64-NEXT: vpinsrw $7, %eax, %xmm0, %xmm0
+; X64-NEXT: vmovd %edi, %xmm1
+; X64-NEXT: vpinsrw $1, %esi, %xmm1, %xmm1
+; X64-NEXT: vpinsrw $2, %edx, %xmm1, %xmm1
+; X64-NEXT: vpinsrw $3, %ecx, %xmm1, %xmm1
+; X64-NEXT: vpinsrw $4, %r8d, %xmm1, %xmm1
+; X64-NEXT: vpinsrw $5, %r9d, %xmm1, %xmm1
+; X64-NEXT: movw {{[0-9]+}}(%rsp), %ax
+; X64-NEXT: vpinsrw $6, %eax, %xmm1, %xmm1
+; X64-NEXT: movw {{[0-9]+}}(%rsp), %ax
+; X64-NEXT: vpinsrw $7, %eax, %xmm1, %xmm1
+; X64-NEXT: vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; X64-NEXT: retq
+ %res0 = insertelement <16 x i16> undef, i16 %a0 , i32 0
+ %res1 = insertelement <16 x i16> %res0, i16 %a1 , i32 1
+ %res2 = insertelement <16 x i16> %res1, i16 %a2 , i32 2
+ %res3 = insertelement <16 x i16> %res2, i16 %a3 , i32 3
+ %res4 = insertelement <16 x i16> %res3, i16 %a4 , i32 4
+ %res5 = insertelement <16 x i16> %res4, i16 %a5 , i32 5
+ %res6 = insertelement <16 x i16> %res5, i16 %a6 , i32 6
+ %res7 = insertelement <16 x i16> %res6, i16 %a7 , i32 7
+ %res8 = insertelement <16 x i16> %res7, i16 %a8 , i32 8
+ %res9 = insertelement <16 x i16> %res8, i16 %a9 , i32 9
+ %res10 = insertelement <16 x i16> %res9, i16 %a10, i32 10
+ %res11 = insertelement <16 x i16> %res10, i16 %a11, i32 11
+ %res12 = insertelement <16 x i16> %res11, i16 %a12, i32 12
+ %res13 = insertelement <16 x i16> %res12, i16 %a13, i32 13
+ %res14 = insertelement <16 x i16> %res13, i16 %a14, i32 14
+ %res15 = insertelement <16 x i16> %res14, i16 %a15, i32 15
+ %res = bitcast <16 x i16> %res15 to <4 x i64>
+ ret <4 x i64> %res
+}
+
+define <4 x i64> @test_mm256_setr_epi32(i32 %a0, i32 %a1, i32 %a2, i32 %a3, i32 %a4, i32 %a5, i32 %a6, i32 %a7) nounwind {
+; X32-LABEL: test_mm256_setr_epi32:
+; X32: # BB#0:
+; X32-NEXT: vmovd {{.*#+}} xmm0 = mem[0],zero,zero,zero
+; X32-NEXT: vpinsrd $1, {{[0-9]+}}(%esp), %xmm0, %xmm0
+; X32-NEXT: vpinsrd $2, {{[0-9]+}}(%esp), %xmm0, %xmm0
+; X32-NEXT: vpinsrd $3, {{[0-9]+}}(%esp), %xmm0, %xmm0
+; X32-NEXT: vmovd {{.*#+}} xmm1 = mem[0],zero,zero,zero
+; X32-NEXT: vpinsrd $1, {{[0-9]+}}(%esp), %xmm1, %xmm1
+; X32-NEXT: vpinsrd $2, {{[0-9]+}}(%esp), %xmm1, %xmm1
+; X32-NEXT: vpinsrd $3, {{[0-9]+}}(%esp), %xmm1, %xmm1
+; X32-NEXT: vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_setr_epi32:
+; X64: # BB#0:
+; X64-NEXT: vmovd %r8d, %xmm0
+; X64-NEXT: vpinsrd $1, %r9d, %xmm0, %xmm0
+; X64-NEXT: vpinsrd $2, {{[0-9]+}}(%rsp), %xmm0, %xmm0
+; X64-NEXT: vpinsrd $3, {{[0-9]+}}(%rsp), %xmm0, %xmm0
+; X64-NEXT: vmovd %edi, %xmm1
+; X64-NEXT: vpinsrd $1, %esi, %xmm1, %xmm1
+; X64-NEXT: vpinsrd $2, %edx, %xmm1, %xmm1
+; X64-NEXT: vpinsrd $3, %ecx, %xmm1, %xmm1
+; X64-NEXT: vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; X64-NEXT: retq
+ %res0 = insertelement <8 x i32> undef, i32 %a0, i32 0
+ %res1 = insertelement <8 x i32> %res0, i32 %a1, i32 1
+ %res2 = insertelement <8 x i32> %res1, i32 %a2, i32 2
+ %res3 = insertelement <8 x i32> %res2, i32 %a3, i32 3
+ %res4 = insertelement <8 x i32> %res3, i32 %a4, i32 4
+ %res5 = insertelement <8 x i32> %res4, i32 %a5, i32 5
+ %res6 = insertelement <8 x i32> %res5, i32 %a6, i32 6
+ %res7 = insertelement <8 x i32> %res6, i32 %a7, i32 7
+ %res = bitcast <8 x i32> %res7 to <4 x i64>
+ ret <4 x i64> %res
+}
+
+define <4 x i64> @test_mm256_setr_epi64x(i64 %a0, i64 %a1, i64 %a2, i64 %a3) nounwind {
+; X32-LABEL: test_mm256_setr_epi64x:
+; X32: # BB#0:
+; X32-NEXT: vmovd {{.*#+}} xmm0 = mem[0],zero,zero,zero
+; X32-NEXT: vpinsrd $1, {{[0-9]+}}(%esp), %xmm0, %xmm0
+; X32-NEXT: vpinsrd $2, {{[0-9]+}}(%esp), %xmm0, %xmm0
+; X32-NEXT: vpinsrd $3, {{[0-9]+}}(%esp), %xmm0, %xmm0
+; X32-NEXT: vmovd {{.*#+}} xmm1 = mem[0],zero,zero,zero
+; X32-NEXT: vpinsrd $1, {{[0-9]+}}(%esp), %xmm1, %xmm1
+; X32-NEXT: vpinsrd $2, {{[0-9]+}}(%esp), %xmm1, %xmm1
+; X32-NEXT: vpinsrd $3, {{[0-9]+}}(%esp), %xmm1, %xmm1
+; X32-NEXT: vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_setr_epi64x:
+; X64: # BB#0:
+; X64-NEXT: vmovq %rcx, %xmm0
+; X64-NEXT: vmovq %rdx, %xmm1
+; X64-NEXT: vpunpcklqdq {{.*#+}} xmm0 = xmm1[0],xmm0[0]
+; X64-NEXT: vmovq %rsi, %xmm1
+; X64-NEXT: vmovq %rdi, %xmm2
+; X64-NEXT: vpunpcklqdq {{.*#+}} xmm1 = xmm2[0],xmm1[0]
+; X64-NEXT: vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; X64-NEXT: retq
+ %res0 = insertelement <4 x i64> undef, i64 %a0, i32 0
+ %res1 = insertelement <4 x i64> %res0, i64 %a1, i32 1
+ %res2 = insertelement <4 x i64> %res1, i64 %a2, i32 2
+ %res3 = insertelement <4 x i64> %res2, i64 %a3, i32 3
+ ret <4 x i64> %res3
+}
+
+define <8 x float> @test_mm256_setr_m128(<4 x float> %a0, <4 x float> %a1) nounwind {
+; X32-LABEL: test_mm256_setr_m128:
+; X32: # BB#0:
+; X32-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_setr_m128:
+; X64: # BB#0:
+; X64-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; X64-NEXT: retq
+ %res = shufflevector <4 x float> %a0, <4 x float> %a1, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
+ ret <8 x float> %res
+}
+
+define <4 x double> @test_mm256_setr_m128d(<2 x double> %a0, <2 x double> %a1) nounwind {
+; X32-LABEL: test_mm256_setr_m128d:
+; X32: # BB#0:
+; X32-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_setr_m128d:
+; X64: # BB#0:
+; X64-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; X64-NEXT: retq
+ %arg0 = bitcast <2 x double> %a0 to <4 x float>
+ %arg1 = bitcast <2 x double> %a1 to <4 x float>
+ %res = shufflevector <4 x float> %arg0, <4 x float> %arg1, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
+ %bc = bitcast <8 x float> %res to <4 x double>
+ ret <4 x double> %bc
+}
+
+define <4 x i64> @test_mm256_setr_m128i(<2 x i64> %a0, <2 x i64> %a1) nounwind {
+; X32-LABEL: test_mm256_setr_m128i:
+; X32: # BB#0:
+; X32-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_setr_m128i:
+; X64: # BB#0:
+; X64-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; X64-NEXT: retq
+ %arg0 = bitcast <2 x i64> %a0 to <4 x float>
+ %arg1 = bitcast <2 x i64> %a1 to <4 x float>
+ %res = shufflevector <4 x float> %arg0, <4 x float> %arg1, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
+ %bc = bitcast <8 x float> %res to <4 x i64>
+ ret <4 x i64> %bc
+}
+
+define <4 x double> @test_mm256_setr_pd(double %a0, double %a1, double %a2, double %a3) nounwind {
+; X32-LABEL: test_mm256_setr_pd:
+; X32: # BB#0:
+; X32-NEXT: vmovsd {{.*#+}} xmm0 = mem[0],zero
+; X32-NEXT: vmovsd {{.*#+}} xmm1 = mem[0],zero
+; X32-NEXT: vmovsd {{.*#+}} xmm2 = mem[0],zero
+; X32-NEXT: vmovsd {{.*#+}} xmm3 = mem[0],zero
+; X32-NEXT: vunpcklpd {{.*#+}} xmm0 = xmm1[0],xmm0[0]
+; X32-NEXT: vunpcklpd {{.*#+}} xmm1 = xmm3[0],xmm2[0]
+; X32-NEXT: vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_setr_pd:
+; X64: # BB#0:
+; X64-NEXT: vunpcklpd {{.*#+}} xmm2 = xmm2[0],xmm3[0]
+; X64-NEXT: vunpcklpd {{.*#+}} xmm0 = xmm0[0],xmm1[0]
+; X64-NEXT: vinsertf128 $1, %xmm2, %ymm0, %ymm0
+; X64-NEXT: retq
+ %res0 = insertelement <4 x double> undef, double %a0, i32 0
+ %res1 = insertelement <4 x double> %res0, double %a1, i32 1
+ %res2 = insertelement <4 x double> %res1, double %a2, i32 2
+ %res3 = insertelement <4 x double> %res2, double %a3, i32 3
+ ret <4 x double> %res3
+}
+
+define <8 x float> @test_mm256_setr_ps(float %a0, float %a1, float %a2, float %a3, float %a4, float %a5, float %a6, float %a7) nounwind {
+; X32-LABEL: test_mm256_setr_ps:
+; X32: # BB#0:
+; X32-NEXT: vmovss {{.*#+}} xmm0 = mem[0],zero,zero,zero
+; X32-NEXT: vmovss {{.*#+}} xmm1 = mem[0],zero,zero,zero
+; X32-NEXT: vmovss {{.*#+}} xmm2 = mem[0],zero,zero,zero
+; X32-NEXT: vmovss {{.*#+}} xmm3 = mem[0],zero,zero,zero
+; X32-NEXT: vmovss {{.*#+}} xmm4 = mem[0],zero,zero,zero
+; X32-NEXT: vmovss {{.*#+}} xmm5 = mem[0],zero,zero,zero
+; X32-NEXT: vmovss {{.*#+}} xmm6 = mem[0],zero,zero,zero
+; X32-NEXT: vmovss {{.*#+}} xmm7 = mem[0],zero,zero,zero
+; X32-NEXT: vinsertps {{.*#+}} xmm2 = xmm3[0],xmm2[0],xmm3[2,3]
+; X32-NEXT: vinsertps {{.*#+}} xmm1 = xmm2[0,1],xmm1[0],xmm2[3]
+; X32-NEXT: vinsertps {{.*#+}} xmm0 = xmm1[0,1,2],xmm0[0]
+; X32-NEXT: vinsertps {{.*#+}} xmm1 = xmm7[0],xmm6[0],xmm7[2,3]
+; X32-NEXT: vinsertps {{.*#+}} xmm1 = xmm1[0,1],xmm5[0],xmm1[3]
+; X32-NEXT: vinsertps {{.*#+}} xmm1 = xmm1[0,1,2],xmm4[0]
+; X32-NEXT: vinsertf128 $1, %xmm0, %ymm1, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_setr_ps:
+; X64: # BB#0:
+; X64-NEXT: vinsertps {{.*#+}} xmm4 = xmm4[0],xmm5[0],xmm4[2,3]
+; X64-NEXT: vinsertps {{.*#+}} xmm4 = xmm4[0,1],xmm6[0],xmm4[3]
+; X64-NEXT: vinsertps {{.*#+}} xmm4 = xmm4[0,1,2],xmm7[0]
+; X64-NEXT: vinsertps {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[2,3]
+; X64-NEXT: vinsertps {{.*#+}} xmm0 = xmm0[0,1],xmm2[0],xmm0[3]
+; X64-NEXT: vinsertps {{.*#+}} xmm0 = xmm0[0,1,2],xmm3[0]
+; X64-NEXT: vinsertf128 $1, %xmm4, %ymm0, %ymm0
+; X64-NEXT: retq
+ %res0 = insertelement <8 x float> undef, float %a0, i32 0
+ %res1 = insertelement <8 x float> %res0, float %a1, i32 1
+ %res2 = insertelement <8 x float> %res1, float %a2, i32 2
+ %res3 = insertelement <8 x float> %res2, float %a3, i32 3
+ %res4 = insertelement <8 x float> %res3, float %a4, i32 4
+ %res5 = insertelement <8 x float> %res4, float %a5, i32 5
+ %res6 = insertelement <8 x float> %res5, float %a6, i32 6
+ %res7 = insertelement <8 x float> %res6, float %a7, i32 7
+ ret <8 x float> %res7
+}
-define <4 x double> @test_mm256_div_pd(<4 x double> %a0, <4 x double> %a1) {
-; X32-LABEL: test_mm256_div_pd:
+define <4 x double> @test_mm256_setzero_pd() nounwind {
+; X32-LABEL: test_mm256_setzero_pd:
; X32: # BB#0:
-; X32-NEXT: vdivpd %ymm1, %ymm0, %ymm0
+; X32-NEXT: vxorps %ymm0, %ymm0, %ymm0
; X32-NEXT: retl
;
-; X64-LABEL: test_mm256_div_pd:
+; X64-LABEL: test_mm256_setzero_pd:
; X64: # BB#0:
-; X64-NEXT: vdivpd %ymm1, %ymm0, %ymm0
+; X64-NEXT: vxorps %ymm0, %ymm0, %ymm0
; X64-NEXT: retq
- %res = fdiv <4 x double> %a0, %a1
- ret <4 x double> %res
+ ret <4 x double> zeroinitializer
}
-define <8 x float> @test_mm256_div_ps(<8 x float> %a0, <8 x float> %a1) {
-; X32-LABEL: test_mm256_div_ps:
+define <8 x float> @test_mm256_setzero_ps() nounwind {
+; X32-LABEL: test_mm256_setzero_ps:
; X32: # BB#0:
-; X32-NEXT: vdivps %ymm1, %ymm0, %ymm0
+; X32-NEXT: vxorps %ymm0, %ymm0, %ymm0
; X32-NEXT: retl
;
-; X64-LABEL: test_mm256_div_ps:
+; X64-LABEL: test_mm256_setzero_ps:
; X64: # BB#0:
-; X64-NEXT: vdivps %ymm1, %ymm0, %ymm0
+; X64-NEXT: vxorps %ymm0, %ymm0, %ymm0
; X64-NEXT: retq
- %res = fdiv <8 x float> %a0, %a1
- ret <8 x float> %res
+ ret <8 x float> zeroinitializer
}
-; TODO test_mm256_dp_ps
-; TODO test_mm256_extract_epi16
-; TODO test_mm256_extract_epi32
-; TODO test_mm256_extract_epi64
-; TODO test_mm256_extract_epi8
-; TODO test_mm256_extractf128_pd
-; TODO test_mm256_extractf128_ps
-; TODO test_mm256_extractf128_si256
-; TODO test_mm256_floor_pd
-; TODO test_mm256_floor_ps
-; TODO test_mm256_hadd_pd
-; TODO test_mm256_hadd_ps
-; TODO test_mm256_hsub_pd
-; TODO test_mm256_hsub_ps
-; TODO test_mm256_insert_epi16
-; TODO test_mm256_insert_epi32
-; TODO test_mm256_insert_epi64
-; TODO test_mm256_insert_epi8
-; TODO test_mm256_insertf128_pd
-; TODO test_mm256_insertf128_ps
-; TODO test_mm256_insertf128_si256
-; TODO test_mm256_lddqu_si256
-; TODO test_mm256_load_pd
-; TODO test_mm256_load_ps
-; TODO test_mm256_load_si256
-; TODO test_mm256_loadu_pd
-; TODO test_mm256_loadu_ps
-; TODO test_mm256_loadu_si256
-; TODO test_mm256_loadu2_m128
-; TODO test_mm256_loadu2_m128d
-; TODO test_mm256_loadu2_m128i
-; TODO test_mm_maskload_pd
-; TODO test_mm256_maskload_pd
-; TODO test_mm_maskload_ps
-; TODO test_mm256_maskload_ps
-; TODO test_mm_maskstore_pd
-; TODO test_mm256_maskstore_pd
-; TODO test_mm_maskstore_ps
-; TODO test_mm256_maskstore_ps
-; TODO test_mm256_max_pd
-; TODO test_mm256_max_ps
-; TODO test_mm256_min_pd
-; TODO test_mm256_min_ps
-; TODO test_mm256_movedup_pd
-; TODO test_mm256_movehdup_ps
-; TODO test_mm256_moveldup_ps
-; TODO test_mm256_movemask_pd
-; TODO test_mm256_movemask_ps
+define <4 x i64> @test_mm256_setzero_si256() nounwind {
+; X32-LABEL: test_mm256_setzero_si256:
+; X32: # BB#0:
+; X32-NEXT: vxorps %ymm0, %ymm0, %ymm0
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_setzero_si256:
+; X64: # BB#0:
+; X64-NEXT: vxorps %ymm0, %ymm0, %ymm0
+; X64-NEXT: retq
+ ret <4 x i64> zeroinitializer
+}
-define <4 x double> @test_mm256_mul_pd(<4 x double> %a0, <4 x double> %a1) {
-; X32-LABEL: test_mm256_mul_pd:
+define <4 x double> @test_mm256_shuffle_pd(<4 x double> %a0, <4 x double> %a1) nounwind {
+; X32-LABEL: test_mm256_shuffle_pd:
; X32: # BB#0:
-; X32-NEXT: vmulpd %ymm1, %ymm0, %ymm0
+; X32-NEXT: vunpcklpd {{.*#+}} ymm0 = ymm0[0],ymm1[0],ymm0[2],ymm1[2]
; X32-NEXT: retl
;
-; X64-LABEL: test_mm256_mul_pd:
+; X64-LABEL: test_mm256_shuffle_pd:
; X64: # BB#0:
-; X64-NEXT: vmulpd %ymm1, %ymm0, %ymm0
+; X64-NEXT: vunpcklpd {{.*#+}} ymm0 = ymm0[0],ymm1[0],ymm0[2],ymm1[2]
; X64-NEXT: retq
- %res = fmul <4 x double> %a0, %a1
+ %res = shufflevector <4 x double> %a0, <4 x double> %a1, <4 x i32> <i32 0, i32 4, i32 2, i32 6>
ret <4 x double> %res
}
-define <8 x float> @test_mm256_mul_ps(<8 x float> %a0, <8 x float> %a1) {
-; X32-LABEL: test_mm256_mul_ps:
+define <8 x float> @test_mm256_shuffle_ps(<8 x float> %a0, <8 x float> %a1) nounwind {
+; X32-LABEL: test_mm256_shuffle_ps:
; X32: # BB#0:
-; X32-NEXT: vmulps %ymm1, %ymm0, %ymm0
+; X32-NEXT: vshufps {{.*#+}} ymm0 = ymm0[0,0],ymm1[0,0],ymm0[4,4],ymm1[4,4]
; X32-NEXT: retl
;
-; X64-LABEL: test_mm256_mul_ps:
+; X64-LABEL: test_mm256_shuffle_ps:
; X64: # BB#0:
-; X64-NEXT: vmulps %ymm1, %ymm0, %ymm0
+; X64-NEXT: vshufps {{.*#+}} ymm0 = ymm0[0,0],ymm1[0,0],ymm0[4,4],ymm1[4,4]
; X64-NEXT: retq
- %res = fmul <8 x float> %a0, %a1
+ %res = shufflevector <8 x float> %a0, <8 x float> %a1, <8 x i32> <i32 0, i32 0, i32 8, i32 8, i32 4, i32 4, i32 12, i32 12>
ret <8 x float> %res
}
-; TODO test_mm256_or_pd
-; TODO test_mm256_or_ps
-; TODO test_mm_permute_pd
-; TODO test_mm256_permute_pd
-; TODO test_mm_permute_ps
-; TODO test_mm256_permute_ps
-
-define <4 x double> @test_mm256_permute2f128_pd(<4 x double> %a0, <4 x double> %a1) {
-; X32-LABEL: test_mm256_permute2f128_pd:
+define <4 x double> @test_mm256_sqrt_pd(<4 x double> %a0) nounwind {
+; X32-LABEL: test_mm256_sqrt_pd:
; X32: # BB#0:
-; X32-NEXT: vperm2f128 {{.*#+}} ymm0 = zero,zero,ymm1[0,1]
+; X32-NEXT: vsqrtpd %ymm0, %ymm0
; X32-NEXT: retl
;
-; X64-LABEL: test_mm256_permute2f128_pd:
+; X64-LABEL: test_mm256_sqrt_pd:
; X64: # BB#0:
-; X64-NEXT: vperm2f128 {{.*#+}} ymm0 = zero,zero,ymm1[0,1]
+; X64-NEXT: vsqrtpd %ymm0, %ymm0
; X64-NEXT: retq
- %res = call <4 x double> @llvm.x86.avx.vperm2f128.pd.256(<4 x double> %a0, <4 x double> %a1, i8 44)
+ %res = call <4 x double> @llvm.x86.avx.sqrt.pd.256(<4 x double> %a0)
ret <4 x double> %res
}
-declare <4 x double> @llvm.x86.avx.vperm2f128.pd.256(<4 x double>, <4 x double>, i8) nounwind readnone
+declare <4 x double> @llvm.x86.avx.sqrt.pd.256(<4 x double>) nounwind readnone
-; PR26667
-define <8 x float> @test_mm256_permute2f128_ps(<8 x float> %a0, <8 x float> %a1) {
-; X32-LABEL: test_mm256_permute2f128_ps:
+define <8 x float> @test_mm256_sqrt_ps(<8 x float> %a0) nounwind {
+; X32-LABEL: test_mm256_sqrt_ps:
; X32: # BB#0:
-; X32-NEXT: vmovaps %ymm1, %ymm0
+; X32-NEXT: vsqrtps %ymm0, %ymm0
; X32-NEXT: retl
;
-; X64-LABEL: test_mm256_permute2f128_ps:
+; X64-LABEL: test_mm256_sqrt_ps:
; X64: # BB#0:
-; X64-NEXT: vmovaps %ymm1, %ymm0
+; X64-NEXT: vsqrtps %ymm0, %ymm0
; X64-NEXT: retq
- %res = call <8 x float> @llvm.x86.avx.vperm2f128.ps.256(<8 x float> %a0, <8 x float> %a1, i8 50)
+ %res = call <8 x float> @llvm.x86.avx.sqrt.ps.256(<8 x float> %a0)
ret <8 x float> %res
}
-declare <8 x float> @llvm.x86.avx.vperm2f128.ps.256(<8 x float>, <8 x float>, i8) nounwind readnone
+declare <8 x float> @llvm.x86.avx.sqrt.ps.256(<8 x float>) nounwind readnone
-define <4 x i64> @test_mm256_permute2f128_si256(<4 x i64> %a0, <4 x i64> %a1) {
-; X32-LABEL: test_mm256_permute2f128_si256:
+define void @test_mm256_store_pd(double* %a0, <4 x double> %a1) nounwind {
+; X32-LABEL: test_mm256_store_pd:
; X32: # BB#0:
-; X32-NEXT: vperm2f128 {{.*#+}} ymm0 = ymm1[2,3,0,1]
+; X32-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vmovaps %ymm0, (%eax)
+; X32-NEXT: vzeroupper
; X32-NEXT: retl
;
-; X64-LABEL: test_mm256_permute2f128_si256:
+; X64-LABEL: test_mm256_store_pd:
; X64: # BB#0:
-; X64-NEXT: vperm2f128 {{.*#+}} ymm0 = ymm1[2,3,0,1]
+; X64-NEXT: vmovaps %ymm0, (%rdi)
+; X64-NEXT: vzeroupper
; X64-NEXT: retq
- %1 = bitcast <4 x i64> %a0 to <8 x i32>
- %2 = bitcast <4 x i64> %a1 to <8 x i32>
- %res = call <8 x i32> @llvm.x86.avx.vperm2f128.si.256(<8 x i32> %1, <8 x i32> %2, i8 35)
- %bc = bitcast <8 x i32> %res to <4 x i64>
- ret <4 x i64> %bc
+ %arg0 = bitcast double* %a0 to <4 x double>*
+ store <4 x double> %a1, <4 x double>* %arg0, align 32
+ ret void
+}
+
+define void @test_mm256_store_ps(float* %a0, <8 x float> %a1) nounwind {
+; X32-LABEL: test_mm256_store_ps:
+; X32: # BB#0:
+; X32-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vmovaps %ymm0, (%eax)
+; X32-NEXT: vzeroupper
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_store_ps:
+; X64: # BB#0:
+; X64-NEXT: vmovaps %ymm0, (%rdi)
+; X64-NEXT: vzeroupper
+; X64-NEXT: retq
+ %arg0 = bitcast float* %a0 to <8 x float>*
+ store <8 x float> %a1, <8 x float>* %arg0, align 32
+ ret void
+}
+
+define void @test_mm256_store_si256(<4 x i64>* %a0, <4 x i64> %a1) nounwind {
+; X32-LABEL: test_mm256_store_si256:
+; X32: # BB#0:
+; X32-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vmovaps %ymm0, (%eax)
+; X32-NEXT: vzeroupper
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_store_si256:
+; X64: # BB#0:
+; X64-NEXT: vmovaps %ymm0, (%rdi)
+; X64-NEXT: vzeroupper
+; X64-NEXT: retq
+ store <4 x i64> %a1, <4 x i64>* %a0, align 32
+ ret void
+}
+
+define void @test_mm256_storeu_pd(double* %a0, <4 x double> %a1) nounwind {
+; X32-LABEL: test_mm256_storeu_pd:
+; X32: # BB#0:
+; X32-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vmovups %ymm0, (%eax)
+; X32-NEXT: vzeroupper
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_storeu_pd:
+; X64: # BB#0:
+; X64-NEXT: vmovups %ymm0, (%rdi)
+; X64-NEXT: vzeroupper
+; X64-NEXT: retq
+ %arg0 = bitcast double* %a0 to <4 x double>*
+ store <4 x double> %a1, <4 x double>* %arg0, align 1
+ ret void
+}
+
+define void @test_mm256_storeu_ps(float* %a0, <8 x float> %a1) nounwind {
+; X32-LABEL: test_mm256_storeu_ps:
+; X32: # BB#0:
+; X32-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vmovups %ymm0, (%eax)
+; X32-NEXT: vzeroupper
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_storeu_ps:
+; X64: # BB#0:
+; X64-NEXT: vmovups %ymm0, (%rdi)
+; X64-NEXT: vzeroupper
+; X64-NEXT: retq
+ %arg0 = bitcast float* %a0 to <8 x float>*
+ store <8 x float> %a1, <8 x float>* %arg0, align 1
+ ret void
+}
+
+define void @test_mm256_storeu_si256(<4 x i64>* %a0, <4 x i64> %a1) nounwind {
+; X32-LABEL: test_mm256_storeu_si256:
+; X32: # BB#0:
+; X32-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vmovups %ymm0, (%eax)
+; X32-NEXT: vzeroupper
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_storeu_si256:
+; X64: # BB#0:
+; X64-NEXT: vmovups %ymm0, (%rdi)
+; X64-NEXT: vzeroupper
+; X64-NEXT: retq
+ store <4 x i64> %a1, <4 x i64>* %a0, align 1
+ ret void
+}
+
+define void @test_mm256_storeu2_m128(float* %a0, float* %a1, <8 x float> %a2) nounwind {
+; X32-LABEL: test_mm256_storeu2_m128:
+; X32: # BB#0:
+; X32-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: movl {{[0-9]+}}(%esp), %ecx
+; X32-NEXT: vmovups %xmm0, (%ecx)
+; X32-NEXT: vextractf128 $1, %ymm0, %xmm0
+; X32-NEXT: vmovups %xmm0, (%eax)
+; X32-NEXT: vzeroupper
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_storeu2_m128:
+; X64: # BB#0:
+; X64-NEXT: vmovups %xmm0, (%rdi)
+; X64-NEXT: vextractf128 $1, %ymm0, %xmm0
+; X64-NEXT: vmovups %xmm0, (%rsi)
+; X64-NEXT: vzeroupper
+; X64-NEXT: retq
+ %arg0 = bitcast float* %a0 to i8*
+ %lo = shufflevector <8 x float> %a2, <8 x float> %a2, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
+ call void @llvm.x86.sse.storeu.ps(i8* %arg0, <4 x float> %lo)
+ %arg1 = bitcast float* %a1 to i8*
+ %hi = shufflevector <8 x float> %a2, <8 x float> %a2, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
+ call void @llvm.x86.sse.storeu.ps(i8* %arg1, <4 x float> %hi)
+ ret void
+}
+declare void @llvm.x86.sse.storeu.ps(i8*, <4 x float>) nounwind readnone
+
+define void @test_mm256_storeu2_m128d(double* %a0, double* %a1, <4 x double> %a2) nounwind {
+; X32-LABEL: test_mm256_storeu2_m128d:
+; X32: # BB#0:
+; X32-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: movl {{[0-9]+}}(%esp), %ecx
+; X32-NEXT: vmovups %xmm0, (%ecx)
+; X32-NEXT: vextractf128 $1, %ymm0, %xmm0
+; X32-NEXT: vmovups %xmm0, (%eax)
+; X32-NEXT: vzeroupper
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_storeu2_m128d:
+; X64: # BB#0:
+; X64-NEXT: vmovups %xmm0, (%rdi)
+; X64-NEXT: vextractf128 $1, %ymm0, %xmm0
+; X64-NEXT: vmovups %xmm0, (%rsi)
+; X64-NEXT: vzeroupper
+; X64-NEXT: retq
+ %arg0 = bitcast double* %a0 to i8*
+ %lo = shufflevector <4 x double> %a2, <4 x double> %a2, <2 x i32> <i32 0, i32 1>
+ call void @llvm.x86.sse2.storeu.pd(i8* %arg0, <2 x double> %lo)
+ %arg1 = bitcast double* %a1 to i8*
+ %hi = shufflevector <4 x double> %a2, <4 x double> %a2, <2 x i32> <i32 2, i32 3>
+ call void @llvm.x86.sse2.storeu.pd(i8* %arg1, <2 x double> %hi)
+ ret void
+}
+declare void @llvm.x86.sse2.storeu.pd(i8*, <2 x double>) nounwind readnone
+
+define void @test_mm256_storeu2_m128i(<2 x i64>* %a0, <2 x i64>* %a1, <4 x i64> %a2) nounwind {
+; X32-LABEL: test_mm256_storeu2_m128i:
+; X32: # BB#0:
+; X32-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: movl {{[0-9]+}}(%esp), %ecx
+; X32-NEXT: vmovups %xmm0, (%ecx)
+; X32-NEXT: vextractf128 $1, %ymm0, %xmm0
+; X32-NEXT: vmovups %xmm0, (%eax)
+; X32-NEXT: vzeroupper
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_storeu2_m128i:
+; X64: # BB#0:
+; X64-NEXT: vmovups %xmm0, (%rdi)
+; X64-NEXT: vextractf128 $1, %ymm0, %xmm0
+; X64-NEXT: vmovups %xmm0, (%rsi)
+; X64-NEXT: vzeroupper
+; X64-NEXT: retq
+ %arg0 = bitcast <2 x i64>* %a0 to i8*
+ %lo2 = shufflevector <4 x i64> %a2, <4 x i64> %a2, <2 x i32> <i32 0, i32 1>
+ %lo = bitcast <2 x i64> %lo2 to <16 x i8>
+ call void @llvm.x86.sse2.storeu.dq(i8* %arg0, <16 x i8> %lo)
+ %arg1 = bitcast <2 x i64>* %a1 to i8*
+ %hi2 = shufflevector <4 x i64> %a2, <4 x i64> %a2, <2 x i32> <i32 2, i32 3>
+ %hi = bitcast <2 x i64> %hi2 to <16 x i8>
+ call void @llvm.x86.sse2.storeu.dq(i8* %arg1, <16 x i8> %hi)
+ ret void
+}
+declare void @llvm.x86.sse2.storeu.dq(i8*, <16 x i8>) nounwind readnone
+
+define void @test_mm256_stream_pd(double *%a0, <4 x double> %a1) nounwind {
+; X32-LABEL: test_mm256_stream_pd:
+; X32: # BB#0:
+; X32-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vmovntps %ymm0, (%eax)
+; X32-NEXT: vzeroupper
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_stream_pd:
+; X64: # BB#0:
+; X64-NEXT: vmovntps %ymm0, (%rdi)
+; X64-NEXT: vzeroupper
+; X64-NEXT: retq
+ %arg0 = bitcast double* %a0 to <4 x double>*
+ store <4 x double> %a1, <4 x double>* %arg0, align 32, !nontemporal !0
+ ret void
+}
+
+define void @test_mm256_stream_ps(float *%a0, <8 x float> %a1) nounwind {
+; X32-LABEL: test_mm256_stream_ps:
+; X32: # BB#0:
+; X32-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vmovntps %ymm0, (%eax)
+; X32-NEXT: vzeroupper
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_stream_ps:
+; X64: # BB#0:
+; X64-NEXT: vmovntps %ymm0, (%rdi)
+; X64-NEXT: vzeroupper
+; X64-NEXT: retq
+ %arg0 = bitcast float* %a0 to <8 x float>*
+ store <8 x float> %a1, <8 x float>* %arg0, align 32, !nontemporal !0
+ ret void
}
-declare <8 x i32> @llvm.x86.avx.vperm2f128.si.256(<8 x i32>, <8 x i32>, i8) nounwind readnone
-; TODO test_mm_permutevar_pd
-; TODO test_mm256_permutevar_pd
-; TODO test_mm_permutevar_ps
-; TODO test_mm256_permutevar_ps
-; TODO test_mm256_rcp_ps
-; TODO test_mm256_round_pd
-; TODO test_mm256_round_ps
-; TODO test_mm256_rsqrt_ps
-; TODO test_mm256_set_epi16
-; TODO test_mm256_set_epi32
-; TODO test_mm256_set_epi64x
-; TODO test_mm256_set_epi8
-; TODO test_mm256_set_m128
-; TODO test_mm256_set_m128d
-; TODO test_mm256_set_m128i
-; TODO test_mm256_set_pd
-; TODO test_mm256_set_ps
-; TODO test_mm256_set1_epi16
-; TODO test_mm256_set1_epi32
-; TODO test_mm256_set1_epi64x
-; TODO test_mm256_set1_epi8
-; TODO test_mm256_set1_pd
-; TODO test_mm256_set1_ps
-; TODO test_mm256_setr_epi16
-; TODO test_mm256_setr_epi32
-; TODO test_mm256_setr_epi64x
-; TODO test_mm256_setr_epi8
-; TODO test_mm256_setr_m128
-; TODO test_mm256_setr_m128d
-; TODO test_mm256_setr_m128i
-; TODO test_mm256_setr_pd
-; TODO test_mm256_setr_ps
-; TODO test_mm256_setzero_pd
-; TODO test_mm256_setzero_ps
-; TODO test_mm256_setzero_si256
-; TODO test_mm256_shuffle_pd
-; TODO test_mm256_shuffle_ps
-; TODO test_mm256_sqrt_pd
-; TODO test_mm256_sqrt_ps
-; TODO test_mm256_store_pd
-; TODO test_mm256_store_ps
-; TODO test_mm256_store_si256
-; TODO test_mm256_storeu_pd
-; TODO test_mm256_storeu_ps
-; TODO test_mm256_storeu_si256
-; TODO test_mm256_storeu2_m128
-; TODO test_mm256_storeu2_m128d
-; TODO test_mm256_storeu2_m128i
-; TODO test_mm256_stream_pd
-; TODO test_mm256_stream_ps
-; TODO test_mm256_stream_si256
+define void @test_mm256_stream_si256(<4 x i64> *%a0, <4 x i64> %a1) nounwind {
+; X32-LABEL: test_mm256_stream_si256:
+; X32: # BB#0:
+; X32-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT: vmovntps %ymm0, (%eax)
+; X32-NEXT: vzeroupper
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_stream_si256:
+; X64: # BB#0:
+; X64-NEXT: vmovntps %ymm0, (%rdi)
+; X64-NEXT: vzeroupper
+; X64-NEXT: retq
+ store <4 x i64> %a1, <4 x i64>* %a0, align 32, !nontemporal !0
+ ret void
+}
-define <4 x double> @test_mm256_sub_pd(<4 x double> %a0, <4 x double> %a1) {
+define <4 x double> @test_mm256_sub_pd(<4 x double> %a0, <4 x double> %a1) nounwind {
; X32-LABEL: test_mm256_sub_pd:
; X32: # BB#0:
; X32-NEXT: vsubpd %ymm1, %ymm0, %ymm0
@@ -487,7 +3281,7 @@ define <4 x double> @test_mm256_sub_pd(<
ret <4 x double> %res
}
-define <8 x float> @test_mm256_sub_ps(<8 x float> %a0, <8 x float> %a1) {
+define <8 x float> @test_mm256_sub_ps(<8 x float> %a0, <8 x float> %a1) nounwind {
; X32-LABEL: test_mm256_sub_ps:
; X32: # BB#0:
; X32-NEXT: vsubps %ymm1, %ymm0, %ymm0
@@ -501,78 +3295,343 @@ define <8 x float> @test_mm256_sub_ps(<8
ret <8 x float> %res
}
-; TODO test_mm_testc_pd
-; TODO test_mm256_testc_pd
-; TODO test_mm_testc_ps
-; TODO test_mm256_testc_ps
-; TODO test_mm256_testc_si256
-; TODO test_mm_testnzc_pd
-; TODO test_mm256_testnzc_pd
-; TODO test_mm_testnzc_ps
-; TODO test_mm256_testnzc_ps
-; TODO test_mm256_testnzc_si256
-; TODO test_mm_testz_pd
-; TODO test_mm256_testz_pd
-; TODO test_mm_testz_ps
-; TODO test_mm256_testz_ps
-; TODO test_mm256_testz_si256
+define i32 @test_mm_testc_pd(<2 x double> %a0, <2 x double> %a1) nounwind {
+; X32-LABEL: test_mm_testc_pd:
+; X32: # BB#0:
+; X32-NEXT: vtestpd %xmm1, %xmm0
+; X32-NEXT: sbbl %eax, %eax
+; X32-NEXT: andl $1, %eax
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm_testc_pd:
+; X64: # BB#0:
+; X64-NEXT: vtestpd %xmm1, %xmm0
+; X64-NEXT: sbbl %eax, %eax
+; X64-NEXT: andl $1, %eax
+; X64-NEXT: retq
+ %res = call i32 @llvm.x86.avx.vtestc.pd(<2 x double> %a0, <2 x double> %a1)
+ ret i32 %res
+}
+declare i32 @llvm.x86.avx.vtestc.pd(<2 x double>, <2 x double>) nounwind readnone
-define <2 x double> @test_mm_undefined_pd() {
-; X32-LABEL: test_mm_undefined_pd:
+define i32 @test_mm256_testc_pd(<4 x double> %a0, <4 x double> %a1) nounwind {
+; X32-LABEL: test_mm256_testc_pd:
; X32: # BB#0:
+; X32-NEXT: vtestpd %ymm1, %ymm0
+; X32-NEXT: sbbl %eax, %eax
+; X32-NEXT: andl $1, %eax
+; X32-NEXT: vzeroupper
; X32-NEXT: retl
;
-; X64-LABEL: test_mm_undefined_pd:
+; X64-LABEL: test_mm256_testc_pd:
; X64: # BB#0:
+; X64-NEXT: vtestpd %ymm1, %ymm0
+; X64-NEXT: sbbl %eax, %eax
+; X64-NEXT: andl $1, %eax
+; X64-NEXT: vzeroupper
; X64-NEXT: retq
- ret <2 x double> undef
+ %res = call i32 @llvm.x86.avx.vtestc.pd.256(<4 x double> %a0, <4 x double> %a1)
+ ret i32 %res
}
+declare i32 @llvm.x86.avx.vtestc.pd.256(<4 x double>, <4 x double>) nounwind readnone
-define <4 x double> @test_mm256_undefined_pd() {
-; X32-LABEL: test_mm256_undefined_pd:
+define i32 @test_mm_testc_ps(<4 x float> %a0, <4 x float> %a1) nounwind {
+; X32-LABEL: test_mm_testc_ps:
; X32: # BB#0:
+; X32-NEXT: vtestps %xmm1, %xmm0
+; X32-NEXT: sbbl %eax, %eax
+; X32-NEXT: andl $1, %eax
; X32-NEXT: retl
;
-; X64-LABEL: test_mm256_undefined_pd:
+; X64-LABEL: test_mm_testc_ps:
; X64: # BB#0:
+; X64-NEXT: vtestps %xmm1, %xmm0
+; X64-NEXT: sbbl %eax, %eax
+; X64-NEXT: andl $1, %eax
; X64-NEXT: retq
- ret <4 x double> undef
+ %res = call i32 @llvm.x86.avx.vtestc.ps(<4 x float> %a0, <4 x float> %a1)
+ ret i32 %res
}
+declare i32 @llvm.x86.avx.vtestc.ps(<4 x float>, <4 x float>) nounwind readnone
-define <4 x float> @test_mm_undefined_ps() {
-; X32-LABEL: test_mm_undefined_ps:
+define i32 @test_mm256_testc_ps(<8 x float> %a0, <8 x float> %a1) nounwind {
+; X32-LABEL: test_mm256_testc_ps:
; X32: # BB#0:
+; X32-NEXT: vtestps %ymm1, %ymm0
+; X32-NEXT: sbbl %eax, %eax
+; X32-NEXT: andl $1, %eax
+; X32-NEXT: vzeroupper
; X32-NEXT: retl
;
-; X64-LABEL: test_mm_undefined_ps:
+; X64-LABEL: test_mm256_testc_ps:
; X64: # BB#0:
+; X64-NEXT: vtestps %ymm1, %ymm0
+; X64-NEXT: sbbl %eax, %eax
+; X64-NEXT: andl $1, %eax
+; X64-NEXT: vzeroupper
; X64-NEXT: retq
- ret <4 x float> undef
+ %res = call i32 @llvm.x86.avx.vtestc.ps.256(<8 x float> %a0, <8 x float> %a1)
+ ret i32 %res
}
+declare i32 @llvm.x86.avx.vtestc.ps.256(<8 x float>, <8 x float>) nounwind readnone
-define <8 x float> @test_mm256_undefined_ps() {
-; X32-LABEL: test_mm256_undefined_ps:
+define i32 @test_mm256_testc_si256(<4 x i64> %a0, <4 x i64> %a1) nounwind {
+; X32-LABEL: test_mm256_testc_si256:
; X32: # BB#0:
+; X32-NEXT: vptest %ymm1, %ymm0
+; X32-NEXT: sbbl %eax, %eax
+; X32-NEXT: andl $1, %eax
+; X32-NEXT: vzeroupper
; X32-NEXT: retl
;
-; X64-LABEL: test_mm256_undefined_ps:
+; X64-LABEL: test_mm256_testc_si256:
; X64: # BB#0:
+; X64-NEXT: vptest %ymm1, %ymm0
+; X64-NEXT: sbbl %eax, %eax
+; X64-NEXT: andl $1, %eax
+; X64-NEXT: vzeroupper
; X64-NEXT: retq
- ret <8 x float> undef
+ %res = call i32 @llvm.x86.avx.ptestc.256(<4 x i64> %a0, <4 x i64> %a1)
+ ret i32 %res
+}
+declare i32 @llvm.x86.avx.ptestc.256(<4 x i64>, <4 x i64>) nounwind readnone
+
+define i32 @test_mm_testnzc_pd(<2 x double> %a0, <2 x double> %a1) nounwind {
+; X32-LABEL: test_mm_testnzc_pd:
+; X32: # BB#0:
+; X32-NEXT: vtestpd %xmm1, %xmm0
+; X32-NEXT: seta %al
+; X32-NEXT: movzbl %al, %eax
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm_testnzc_pd:
+; X64: # BB#0:
+; X64-NEXT: vtestpd %xmm1, %xmm0
+; X64-NEXT: seta %al
+; X64-NEXT: movzbl %al, %eax
+; X64-NEXT: retq
+ %res = call i32 @llvm.x86.avx.vtestnzc.pd(<2 x double> %a0, <2 x double> %a1)
+ ret i32 %res
+}
+declare i32 @llvm.x86.avx.vtestnzc.pd(<2 x double>, <2 x double>) nounwind readnone
+
+define i32 @test_mm256_testnzc_pd(<4 x double> %a0, <4 x double> %a1) nounwind {
+; X32-LABEL: test_mm256_testnzc_pd:
+; X32: # BB#0:
+; X32-NEXT: vtestpd %ymm1, %ymm0
+; X32-NEXT: seta %al
+; X32-NEXT: movzbl %al, %eax
+; X32-NEXT: vzeroupper
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_testnzc_pd:
+; X64: # BB#0:
+; X64-NEXT: vtestpd %ymm1, %ymm0
+; X64-NEXT: seta %al
+; X64-NEXT: movzbl %al, %eax
+; X64-NEXT: vzeroupper
+; X64-NEXT: retq
+ %res = call i32 @llvm.x86.avx.vtestnzc.pd.256(<4 x double> %a0, <4 x double> %a1)
+ ret i32 %res
+}
+declare i32 @llvm.x86.avx.vtestnzc.pd.256(<4 x double>, <4 x double>) nounwind readnone
+
+define i32 @test_mm_testnzc_ps(<4 x float> %a0, <4 x float> %a1) nounwind {
+; X32-LABEL: test_mm_testnzc_ps:
+; X32: # BB#0:
+; X32-NEXT: vtestps %xmm1, %xmm0
+; X32-NEXT: seta %al
+; X32-NEXT: movzbl %al, %eax
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm_testnzc_ps:
+; X64: # BB#0:
+; X64-NEXT: vtestps %xmm1, %xmm0
+; X64-NEXT: seta %al
+; X64-NEXT: movzbl %al, %eax
+; X64-NEXT: retq
+ %res = call i32 @llvm.x86.avx.vtestnzc.ps(<4 x float> %a0, <4 x float> %a1)
+ ret i32 %res
+}
+declare i32 @llvm.x86.avx.vtestnzc.ps(<4 x float>, <4 x float>) nounwind readnone
+
+define i32 @test_mm256_testnzc_ps(<8 x float> %a0, <8 x float> %a1) nounwind {
+; X32-LABEL: test_mm256_testnzc_ps:
+; X32: # BB#0:
+; X32-NEXT: vtestps %ymm1, %ymm0
+; X32-NEXT: seta %al
+; X32-NEXT: movzbl %al, %eax
+; X32-NEXT: vzeroupper
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_testnzc_ps:
+; X64: # BB#0:
+; X64-NEXT: vtestps %ymm1, %ymm0
+; X64-NEXT: seta %al
+; X64-NEXT: movzbl %al, %eax
+; X64-NEXT: vzeroupper
+; X64-NEXT: retq
+ %res = call i32 @llvm.x86.avx.vtestnzc.ps.256(<8 x float> %a0, <8 x float> %a1)
+ ret i32 %res
+}
+declare i32 @llvm.x86.avx.vtestnzc.ps.256(<8 x float>, <8 x float>) nounwind readnone
+
+define i32 @test_mm256_testnzc_si256(<4 x i64> %a0, <4 x i64> %a1) nounwind {
+; X32-LABEL: test_mm256_testnzc_si256:
+; X32: # BB#0:
+; X32-NEXT: vptest %ymm1, %ymm0
+; X32-NEXT: seta %al
+; X32-NEXT: movzbl %al, %eax
+; X32-NEXT: vzeroupper
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_testnzc_si256:
+; X64: # BB#0:
+; X64-NEXT: vptest %ymm1, %ymm0
+; X64-NEXT: seta %al
+; X64-NEXT: movzbl %al, %eax
+; X64-NEXT: vzeroupper
+; X64-NEXT: retq
+ %res = call i32 @llvm.x86.avx.ptestnzc.256(<4 x i64> %a0, <4 x i64> %a1)
+ ret i32 %res
+}
+declare i32 @llvm.x86.avx.ptestnzc.256(<4 x i64>, <4 x i64>) nounwind readnone
+
+define i32 @test_mm_testz_pd(<2 x double> %a0, <2 x double> %a1) nounwind {
+; X32-LABEL: test_mm_testz_pd:
+; X32: # BB#0:
+; X32-NEXT: vtestpd %xmm1, %xmm0
+; X32-NEXT: sete %al
+; X32-NEXT: movzbl %al, %eax
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm_testz_pd:
+; X64: # BB#0:
+; X64-NEXT: vtestpd %xmm1, %xmm0
+; X64-NEXT: sete %al
+; X64-NEXT: movzbl %al, %eax
+; X64-NEXT: retq
+ %res = call i32 @llvm.x86.avx.vtestz.pd(<2 x double> %a0, <2 x double> %a1)
+ ret i32 %res
+}
+declare i32 @llvm.x86.avx.vtestz.pd(<2 x double>, <2 x double>) nounwind readnone
+
+define i32 @test_mm256_testz_pd(<4 x double> %a0, <4 x double> %a1) nounwind {
+; X32-LABEL: test_mm256_testz_pd:
+; X32: # BB#0:
+; X32-NEXT: vtestpd %ymm1, %ymm0
+; X32-NEXT: sete %al
+; X32-NEXT: movzbl %al, %eax
+; X32-NEXT: vzeroupper
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_testz_pd:
+; X64: # BB#0:
+; X64-NEXT: vtestpd %ymm1, %ymm0
+; X64-NEXT: sete %al
+; X64-NEXT: movzbl %al, %eax
+; X64-NEXT: vzeroupper
+; X64-NEXT: retq
+ %res = call i32 @llvm.x86.avx.vtestz.pd.256(<4 x double> %a0, <4 x double> %a1)
+ ret i32 %res
+}
+declare i32 @llvm.x86.avx.vtestz.pd.256(<4 x double>, <4 x double>) nounwind readnone
+
+define i32 @test_mm_testz_ps(<4 x float> %a0, <4 x float> %a1) nounwind {
+; X32-LABEL: test_mm_testz_ps:
+; X32: # BB#0:
+; X32-NEXT: vtestps %xmm1, %xmm0
+; X32-NEXT: sete %al
+; X32-NEXT: movzbl %al, %eax
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm_testz_ps:
+; X64: # BB#0:
+; X64-NEXT: vtestps %xmm1, %xmm0
+; X64-NEXT: sete %al
+; X64-NEXT: movzbl %al, %eax
+; X64-NEXT: retq
+ %res = call i32 @llvm.x86.avx.vtestz.ps(<4 x float> %a0, <4 x float> %a1)
+ ret i32 %res
+}
+declare i32 @llvm.x86.avx.vtestz.ps(<4 x float>, <4 x float>) nounwind readnone
+
+define i32 @test_mm256_testz_ps(<8 x float> %a0, <8 x float> %a1) nounwind {
+; X32-LABEL: test_mm256_testz_ps:
+; X32: # BB#0:
+; X32-NEXT: vtestps %ymm1, %ymm0
+; X32-NEXT: sete %al
+; X32-NEXT: movzbl %al, %eax
+; X32-NEXT: vzeroupper
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_testz_ps:
+; X64: # BB#0:
+; X64-NEXT: vtestps %ymm1, %ymm0
+; X64-NEXT: sete %al
+; X64-NEXT: movzbl %al, %eax
+; X64-NEXT: vzeroupper
+; X64-NEXT: retq
+ %res = call i32 @llvm.x86.avx.vtestz.ps.256(<8 x float> %a0, <8 x float> %a1)
+ ret i32 %res
+}
+declare i32 @llvm.x86.avx.vtestz.ps.256(<8 x float>, <8 x float>) nounwind readnone
+
+define i32 @test_mm256_testz_si256(<4 x i64> %a0, <4 x i64> %a1) nounwind {
+; X32-LABEL: test_mm256_testz_si256:
+; X32: # BB#0:
+; X32-NEXT: vptest %ymm1, %ymm0
+; X32-NEXT: sete %al
+; X32-NEXT: movzbl %al, %eax
+; X32-NEXT: vzeroupper
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_testz_si256:
+; X64: # BB#0:
+; X64-NEXT: vptest %ymm1, %ymm0
+; X64-NEXT: sete %al
+; X64-NEXT: movzbl %al, %eax
+; X64-NEXT: vzeroupper
+; X64-NEXT: retq
+ %res = call i32 @llvm.x86.avx.ptestz.256(<4 x i64> %a0, <4 x i64> %a1)
+ ret i32 %res
+}
+declare i32 @llvm.x86.avx.ptestz.256(<4 x i64>, <4 x i64>) nounwind readnone
+
+define <2 x double> @test_mm_undefined_pd() nounwind {
+; X32-LABEL: test_mm_undefined_pd:
+; X32: # BB#0:
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm_undefined_pd:
+; X64: # BB#0:
+; X64-NEXT: retq
+ ret <2 x double> undef
+}
+
+define <4 x double> @test_mm256_undefined_pd() nounwind {
+; X32-LABEL: test_mm256_undefined_pd:
+; X32: # BB#0:
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_undefined_pd:
+; X64: # BB#0:
+; X64-NEXT: retq
+ ret <4 x double> undef
}
-define <2 x i64> @test_mm_undefined_si256() {
-; X32-LABEL: test_mm_undefined_si256:
+define <8 x float> @test_mm256_undefined_ps() nounwind {
+; X32-LABEL: test_mm256_undefined_ps:
; X32: # BB#0:
; X32-NEXT: retl
;
-; X64-LABEL: test_mm_undefined_si256:
+; X64-LABEL: test_mm256_undefined_ps:
; X64: # BB#0:
; X64-NEXT: retq
- ret <2 x i64> undef
+ ret <8 x float> undef
}
-define <4 x i64> @test_mm256_undefined_si256() {
+define <4 x i64> @test_mm256_undefined_si256() nounwind {
; X32-LABEL: test_mm256_undefined_si256:
; X32: # BB#0:
; X32-NEXT: retl
@@ -583,7 +3642,7 @@ define <4 x i64> @test_mm256_undefined_s
ret <4 x i64> undef
}
-define <4 x double> @test_mm256_unpackhi_pd(<4 x double> %a0, <4 x double> %a1) {
+define <4 x double> @test_mm256_unpackhi_pd(<4 x double> %a0, <4 x double> %a1) nounwind {
; X32-LABEL: test_mm256_unpackhi_pd:
; X32: # BB#0:
; X32-NEXT: vunpckhpd {{.*#+}} ymm0 = ymm0[1],ymm1[1],ymm0[3],ymm1[3]
@@ -597,7 +3656,7 @@ define <4 x double> @test_mm256_unpackhi
ret <4 x double> %res
}
-define <8 x float> @test_mm256_unpackhi_ps(<8 x float> %a0, <8 x float> %a1) {
+define <8 x float> @test_mm256_unpackhi_ps(<8 x float> %a0, <8 x float> %a1) nounwind {
; X32-LABEL: test_mm256_unpackhi_ps:
; X32: # BB#0:
; X32-NEXT: vunpckhps {{.*#+}} ymm0 = ymm0[2],ymm1[2],ymm0[3],ymm1[3],ymm0[6],ymm1[6],ymm0[7],ymm1[7]
@@ -611,7 +3670,7 @@ define <8 x float> @test_mm256_unpackhi_
ret <8 x float> %res
}
-define <4 x double> @test_mm256_unpacklo_pd(<4 x double> %a0, <4 x double> %a1) {
+define <4 x double> @test_mm256_unpacklo_pd(<4 x double> %a0, <4 x double> %a1) nounwind {
; X32-LABEL: test_mm256_unpacklo_pd:
; X32: # BB#0:
; X32-NEXT: vunpcklpd {{.*#+}} ymm0 = ymm0[0],ymm1[0],ymm0[2],ymm1[2]
@@ -625,7 +3684,7 @@ define <4 x double> @test_mm256_unpacklo
ret <4 x double> %res
}
-define <8 x float> @test_mm256_unpacklo_ps(<8 x float> %a0, <8 x float> %a1) {
+define <8 x float> @test_mm256_unpacklo_ps(<8 x float> %a0, <8 x float> %a1) nounwind {
; X32-LABEL: test_mm256_unpacklo_ps:
; X32: # BB#0:
; X32-NEXT: vunpcklps {{.*#+}} ymm0 = ymm0[0],ymm1[0],ymm0[1],ymm1[1],ymm0[4],ymm1[4],ymm0[5],ymm1[5]
@@ -639,7 +3698,7 @@ define <8 x float> @test_mm256_unpacklo_
ret <8 x float> %res
}
-define <4 x double> @test_mm256_xor_pd(<4 x double> %a0, <4 x double> %a1) {
+define <4 x double> @test_mm256_xor_pd(<4 x double> %a0, <4 x double> %a1) nounwind {
; X32-LABEL: test_mm256_xor_pd:
; X32: # BB#0:
; X32-NEXT: vxorps %ymm1, %ymm0, %ymm0
@@ -656,7 +3715,7 @@ define <4 x double> @test_mm256_xor_pd(<
ret <4 x double> %bc
}
-define <8 x float> @test_mm256_xor_ps(<8 x float> %a0, <8 x float> %a1) {
+define <8 x float> @test_mm256_xor_ps(<8 x float> %a0, <8 x float> %a1) nounwind {
; X32-LABEL: test_mm256_xor_ps:
; X32: # BB#0:
; X32-NEXT: vxorps %ymm1, %ymm0, %ymm0
@@ -673,5 +3732,38 @@ define <8 x float> @test_mm256_xor_ps(<8
ret <8 x float> %bc
}
-; TODO test_mm256_zeroall
-; TODO test_mm256_zeroupper
+define void @test_mm256_zeroall() nounwind {
+; X32-LABEL: test_mm256_zeroall:
+; X32: # BB#0:
+; X32-NEXT: vzeroall
+; X32-NEXT: vzeroupper
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_zeroall:
+; X64: # BB#0:
+; X64-NEXT: vzeroall
+; X64-NEXT: vzeroupper
+; X64-NEXT: retq
+ call void @llvm.x86.avx.vzeroall()
+ ret void
+}
+declare void @llvm.x86.avx.vzeroall() nounwind readnone
+
+define void @test_mm256_zeroupper() nounwind {
+; X32-LABEL: test_mm256_zeroupper:
+; X32: # BB#0:
+; X32-NEXT: vzeroupper
+; X32-NEXT: vzeroupper
+; X32-NEXT: retl
+;
+; X64-LABEL: test_mm256_zeroupper:
+; X64: # BB#0:
+; X64-NEXT: vzeroupper
+; X64-NEXT: vzeroupper
+; X64-NEXT: retq
+ call void @llvm.x86.avx.vzeroupper()
+ ret void
+}
+declare void @llvm.x86.avx.vzeroupper() nounwind readnone
+
+!0 = !{i32 1}
More information about the llvm-commits
mailing list