summaryrefslogtreecommitdiffstats
path: root/llvm/test/CodeGen/X86/vector-shuffle-combining-avx512bw.ll
diff options
context:
space:
mode:
Diffstat (limited to 'llvm/test/CodeGen/X86/vector-shuffle-combining-avx512bw.ll')
-rw-r--r--llvm/test/CodeGen/X86/vector-shuffle-combining-avx512bw.ll104
1 files changed, 50 insertions, 54 deletions
diff --git a/llvm/test/CodeGen/X86/vector-shuffle-combining-avx512bw.ll b/llvm/test/CodeGen/X86/vector-shuffle-combining-avx512bw.ll
index 64dd50639d3..b8d3824f1ca 100644
--- a/llvm/test/CodeGen/X86/vector-shuffle-combining-avx512bw.ll
+++ b/llvm/test/CodeGen/X86/vector-shuffle-combining-avx512bw.ll
@@ -124,21 +124,21 @@ define <8 x double> @combine_vpermt2var_8f64_identity(<8 x double> %x0, <8 x dou
define <8 x double> @combine_vpermt2var_8f64_identity_mask(<8 x double> %x0, <8 x double> %x1, i8 %m) {
; X32-LABEL: combine_vpermt2var_8f64_identity_mask:
; X32: # %bb.0:
-; X32-NEXT: vmovapd {{.*#+}} zmm2 = [7,0,6,0,5,0,4,0,3,0,2,0,1,0,0,0]
+; X32-NEXT: vmovapd {{.*#+}} zmm1 = [7,0,6,0,5,0,4,0,3,0,2,0,1,0,0,0]
; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
; X32-NEXT: kmovd %eax, %k1
-; X32-NEXT: vpermi2pd %zmm1, %zmm0, %zmm2 {%k1} {z}
+; X32-NEXT: vpermi2pd %zmm0, %zmm0, %zmm1 {%k1} {z}
; X32-NEXT: vmovapd {{.*#+}} zmm0 = [7,0,14,0,5,0,12,0,3,0,10,0,1,0,8,0]
-; X32-NEXT: vpermi2pd %zmm2, %zmm2, %zmm0 {%k1} {z}
+; X32-NEXT: vpermi2pd %zmm1, %zmm1, %zmm0 {%k1} {z}
; X32-NEXT: retl
;
; X64-LABEL: combine_vpermt2var_8f64_identity_mask:
; X64: # %bb.0:
-; X64-NEXT: vmovapd {{.*#+}} zmm2 = [7,6,5,4,3,2,1,0]
+; X64-NEXT: vmovapd {{.*#+}} zmm1 = [7,6,5,4,3,2,1,0]
; X64-NEXT: kmovd %edi, %k1
-; X64-NEXT: vpermi2pd %zmm1, %zmm0, %zmm2 {%k1} {z}
+; X64-NEXT: vpermi2pd %zmm0, %zmm0, %zmm1 {%k1} {z}
; X64-NEXT: vmovapd {{.*#+}} zmm0 = [7,14,5,12,3,10,1,8]
-; X64-NEXT: vpermi2pd %zmm2, %zmm2, %zmm0 {%k1} {z}
+; X64-NEXT: vpermi2pd %zmm1, %zmm1, %zmm0 {%k1} {z}
; X64-NEXT: retq
%res0 = call <8 x double> @llvm.x86.avx512.maskz.vpermt2var.pd.512(<8 x i64> <i64 7, i64 6, i64 5, i64 4, i64 3, i64 2, i64 1, i64 0>, <8 x double> %x0, <8 x double> %x1, i8 %m)
%res1 = call <8 x double> @llvm.x86.avx512.maskz.vpermt2var.pd.512(<8 x i64> <i64 7, i64 14, i64 5, i64 12, i64 3, i64 10, i64 1, i64 8>, <8 x double> %res0, <8 x double> %res0, i8 %m)
@@ -205,21 +205,21 @@ define <8 x i64> @combine_vpermt2var_8i64_identity(<8 x i64> %x0, <8 x i64> %x1)
define <8 x i64> @combine_vpermt2var_8i64_identity_mask(<8 x i64> %x0, <8 x i64> %x1, i8 %m) {
; X32-LABEL: combine_vpermt2var_8i64_identity_mask:
; X32: # %bb.0:
-; X32-NEXT: vmovdqa64 {{.*#+}} zmm2 = [7,0,6,0,5,0,4,0,3,0,2,0,1,0,0,0]
+; X32-NEXT: vmovdqa64 {{.*#+}} zmm1 = [7,0,6,0,5,0,4,0,3,0,2,0,1,0,0,0]
; X32-NEXT: movzbl {{[0-9]+}}(%esp), %eax
; X32-NEXT: kmovd %eax, %k1
-; X32-NEXT: vpermi2q %zmm1, %zmm0, %zmm2 {%k1} {z}
+; X32-NEXT: vpermi2q %zmm0, %zmm0, %zmm1 {%k1} {z}
; X32-NEXT: vmovdqa64 {{.*#+}} zmm0 = [7,0,14,0,5,0,12,0,3,0,10,0,1,0,8,0]
-; X32-NEXT: vpermi2q %zmm2, %zmm2, %zmm0 {%k1} {z}
+; X32-NEXT: vpermi2q %zmm1, %zmm1, %zmm0 {%k1} {z}
; X32-NEXT: retl
;
; X64-LABEL: combine_vpermt2var_8i64_identity_mask:
; X64: # %bb.0:
-; X64-NEXT: vmovdqa64 {{.*#+}} zmm2 = [7,6,5,4,3,2,1,0]
+; X64-NEXT: vmovdqa64 {{.*#+}} zmm1 = [7,6,5,4,3,2,1,0]
; X64-NEXT: kmovd %edi, %k1
-; X64-NEXT: vpermi2q %zmm1, %zmm0, %zmm2 {%k1} {z}
+; X64-NEXT: vpermi2q %zmm0, %zmm0, %zmm1 {%k1} {z}
; X64-NEXT: vmovdqa64 {{.*#+}} zmm0 = [7,14,5,12,3,10,1,8]
-; X64-NEXT: vpermi2q %zmm2, %zmm2, %zmm0 {%k1} {z}
+; X64-NEXT: vpermi2q %zmm1, %zmm1, %zmm0 {%k1} {z}
; X64-NEXT: retq
%res0 = call <8 x i64> @llvm.x86.avx512.maskz.vpermt2var.q.512(<8 x i64> <i64 7, i64 6, i64 5, i64 4, i64 3, i64 2, i64 1, i64 0>, <8 x i64> %x0, <8 x i64> %x1, i8 %m)
%res1 = call <8 x i64> @llvm.x86.avx512.maskz.vpermt2var.q.512(<8 x i64> <i64 7, i64 14, i64 5, i64 12, i64 3, i64 10, i64 1, i64 8>, <8 x i64> %res0, <8 x i64> %res0, i8 %m)
@@ -241,20 +241,20 @@ define <16 x float> @combine_vpermt2var_16f32_identity(<16 x float> %x0, <16 x f
define <16 x float> @combine_vpermt2var_16f32_identity_mask(<16 x float> %x0, <16 x float> %x1, i16 %m) {
; X32-LABEL: combine_vpermt2var_16f32_identity_mask:
; X32: # %bb.0:
-; X32-NEXT: vmovaps {{.*#+}} zmm2 = [15,14,13,12,11,10,9,8,7,6,5,4,3,2,1,0]
+; X32-NEXT: vmovaps {{.*#+}} zmm1 = [15,14,13,12,11,10,9,8,7,6,5,4,3,2,1,0]
; X32-NEXT: kmovw {{[0-9]+}}(%esp), %k1
-; X32-NEXT: vpermi2ps %zmm1, %zmm0, %zmm2 {%k1} {z}
+; X32-NEXT: vpermi2ps %zmm0, %zmm0, %zmm1 {%k1} {z}
; X32-NEXT: vmovaps {{.*#+}} zmm0 = [15,30,13,28,11,26,9,24,7,22,5,20,3,18,1,16]
-; X32-NEXT: vpermi2ps %zmm2, %zmm2, %zmm0 {%k1} {z}
+; X32-NEXT: vpermi2ps %zmm1, %zmm1, %zmm0 {%k1} {z}
; X32-NEXT: retl
;
; X64-LABEL: combine_vpermt2var_16f32_identity_mask:
; X64: # %bb.0:
-; X64-NEXT: vmovaps {{.*#+}} zmm2 = [15,14,13,12,11,10,9,8,7,6,5,4,3,2,1,0]
+; X64-NEXT: vmovaps {{.*#+}} zmm1 = [15,14,13,12,11,10,9,8,7,6,5,4,3,2,1,0]
; X64-NEXT: kmovd %edi, %k1
-; X64-NEXT: vpermi2ps %zmm1, %zmm0, %zmm2 {%k1} {z}
+; X64-NEXT: vpermi2ps %zmm0, %zmm0, %zmm1 {%k1} {z}
; X64-NEXT: vmovaps {{.*#+}} zmm0 = [15,30,13,28,11,26,9,24,7,22,5,20,3,18,1,16]
-; X64-NEXT: vpermi2ps %zmm2, %zmm2, %zmm0 {%k1} {z}
+; X64-NEXT: vpermi2ps %zmm1, %zmm1, %zmm0 {%k1} {z}
; X64-NEXT: retq
%res0 = call <16 x float> @llvm.x86.avx512.maskz.vpermt2var.ps.512(<16 x i32> <i32 15, i32 14, i32 13, i32 12, i32 11, i32 10, i32 9, i32 8, i32 7, i32 6, i32 5, i32 4, i32 3, i32 2, i32 1, i32 0>, <16 x float> %x0, <16 x float> %x1, i16 %m)
%res1 = call <16 x float> @llvm.x86.avx512.maskz.vpermt2var.ps.512(<16 x i32> <i32 15, i32 30, i32 13, i32 28, i32 11, i32 26, i32 9, i32 24, i32 7, i32 22, i32 5, i32 20, i32 3, i32 18, i32 1, i32 16>, <16 x float> %res0, <16 x float> %res0, i16 %m)
@@ -264,14 +264,14 @@ define <16 x float> @combine_vpermt2var_16f32_identity_mask(<16 x float> %x0, <1
define <16 x float> @combine_vpermt2var_16f32_vmovddup(<16 x float> %x0, <16 x float> %x1) {
; X32-LABEL: combine_vpermt2var_16f32_vmovddup:
; X32: # %bb.0:
-; X32-NEXT: vmovaps {{.*#+}} zmm2 = [0,1,0,1,4,5,4,5,8,9,8,9,12,13,12,13]
-; X32-NEXT: vpermt2ps %zmm1, %zmm2, %zmm0
+; X32-NEXT: vmovaps {{.*#+}} zmm1 = [0,1,0,1,4,5,4,5,8,9,8,9,12,13,12,13]
+; X32-NEXT: vpermt2ps %zmm0, %zmm1, %zmm0
; X32-NEXT: retl
;
; X64-LABEL: combine_vpermt2var_16f32_vmovddup:
; X64: # %bb.0:
-; X64-NEXT: vmovaps {{.*#+}} zmm2 = [0,1,0,1,4,5,4,5,8,9,8,9,12,13,12,13]
-; X64-NEXT: vpermt2ps %zmm1, %zmm2, %zmm0
+; X64-NEXT: vmovaps {{.*#+}} zmm1 = [0,1,0,1,4,5,4,5,8,9,8,9,12,13,12,13]
+; X64-NEXT: vpermt2ps %zmm0, %zmm1, %zmm0
; X64-NEXT: retq
%res0 = call <16 x float> @llvm.x86.avx512.maskz.vpermt2var.ps.512(<16 x i32> <i32 0, i32 1, i32 0, i32 1, i32 4, i32 5, i32 4, i32 5, i32 8, i32 9, i32 8, i32 9, i32 12, i32 13, i32 12, i32 13>, <16 x float> %x0, <16 x float> %x1, i16 -1)
ret <16 x float> %res0
@@ -280,18 +280,16 @@ define <16 x float> @combine_vpermt2var_16f32_vmovddup_load(<16 x float> *%p0, <
; X32-LABEL: combine_vpermt2var_16f32_vmovddup_load:
; X32: # %bb.0:
; X32-NEXT: movl {{[0-9]+}}(%esp), %eax
-; X32-NEXT: vmovaps (%eax), %zmm2
-; X32-NEXT: vmovaps {{.*#+}} zmm1 = [0,1,0,1,4,5,4,5,8,9,8,9,12,13,12,13]
-; X32-NEXT: vpermi2ps %zmm0, %zmm2, %zmm1
-; X32-NEXT: vmovaps %zmm1, %zmm0
+; X32-NEXT: vmovaps (%eax), %zmm1
+; X32-NEXT: vmovaps {{.*#+}} zmm0 = [0,1,0,1,4,5,4,5,8,9,8,9,12,13,12,13]
+; X32-NEXT: vpermi2ps %zmm0, %zmm1, %zmm0
; X32-NEXT: retl
;
; X64-LABEL: combine_vpermt2var_16f32_vmovddup_load:
; X64: # %bb.0:
-; X64-NEXT: vmovaps (%rdi), %zmm2
-; X64-NEXT: vmovaps {{.*#+}} zmm1 = [0,1,0,1,4,5,4,5,8,9,8,9,12,13,12,13]
-; X64-NEXT: vpermi2ps %zmm0, %zmm2, %zmm1
-; X64-NEXT: vmovaps %zmm1, %zmm0
+; X64-NEXT: vmovaps (%rdi), %zmm1
+; X64-NEXT: vmovaps {{.*#+}} zmm0 = [0,1,0,1,4,5,4,5,8,9,8,9,12,13,12,13]
+; X64-NEXT: vpermi2ps %zmm0, %zmm1, %zmm0
; X64-NEXT: retq
%x0 = load <16 x float>, <16 x float> *%p0
%res0 = call <16 x float> @llvm.x86.avx512.maskz.vpermt2var.ps.512(<16 x i32> <i32 0, i32 1, i32 0, i32 1, i32 4, i32 5, i32 4, i32 5, i32 8, i32 9, i32 8, i32 9, i32 12, i32 13, i32 12, i32 13>, <16 x float> %x0, <16 x float> %x1, i16 -1)
@@ -300,16 +298,16 @@ define <16 x float> @combine_vpermt2var_16f32_vmovddup_load(<16 x float> *%p0, <
define <16 x float> @combine_vpermt2var_16f32_vmovddup_mask(<16 x float> %x0, <16 x float> %x1, i16 %m) {
; X32-LABEL: combine_vpermt2var_16f32_vmovddup_mask:
; X32: # %bb.0:
-; X32-NEXT: vmovaps {{.*#+}} zmm2 = [0,1,0,1,4,5,4,5,8,9,8,9,12,13,12,13]
+; X32-NEXT: vmovaps {{.*#+}} zmm1 = [0,1,0,1,4,5,4,5,8,9,8,9,12,13,12,13]
; X32-NEXT: kmovw {{[0-9]+}}(%esp), %k1
-; X32-NEXT: vpermt2ps %zmm1, %zmm2, %zmm0 {%k1} {z}
+; X32-NEXT: vpermt2ps %zmm0, %zmm1, %zmm0 {%k1} {z}
; X32-NEXT: retl
;
; X64-LABEL: combine_vpermt2var_16f32_vmovddup_mask:
; X64: # %bb.0:
-; X64-NEXT: vmovaps {{.*#+}} zmm2 = [0,1,0,1,4,5,4,5,8,9,8,9,12,13,12,13]
+; X64-NEXT: vmovaps {{.*#+}} zmm1 = [0,1,0,1,4,5,4,5,8,9,8,9,12,13,12,13]
; X64-NEXT: kmovd %edi, %k1
-; X64-NEXT: vpermt2ps %zmm1, %zmm2, %zmm0 {%k1} {z}
+; X64-NEXT: vpermt2ps %zmm0, %zmm1, %zmm0 {%k1} {z}
; X64-NEXT: retq
%res0 = call <16 x float> @llvm.x86.avx512.maskz.vpermt2var.ps.512(<16 x i32> <i32 0, i32 1, i32 0, i32 1, i32 4, i32 5, i32 4, i32 5, i32 8, i32 9, i32 8, i32 9, i32 12, i32 13, i32 12, i32 13>, <16 x float> %x0, <16 x float> %x1, i16 %m)
ret <16 x float> %res0
@@ -318,20 +316,18 @@ define <16 x float> @combine_vpermt2var_16f32_vmovddup_mask_load(<16 x float> *%
; X32-LABEL: combine_vpermt2var_16f32_vmovddup_mask_load:
; X32: # %bb.0:
; X32-NEXT: movl {{[0-9]+}}(%esp), %eax
-; X32-NEXT: vmovaps (%eax), %zmm2
-; X32-NEXT: vmovaps {{.*#+}} zmm1 = [0,1,0,1,4,5,4,5,8,9,8,9,12,13,12,13]
+; X32-NEXT: vmovaps (%eax), %zmm1
+; X32-NEXT: vmovaps {{.*#+}} zmm0 = [0,1,0,1,4,5,4,5,8,9,8,9,12,13,12,13]
; X32-NEXT: kmovw {{[0-9]+}}(%esp), %k1
-; X32-NEXT: vpermi2ps %zmm0, %zmm2, %zmm1 {%k1} {z}
-; X32-NEXT: vmovaps %zmm1, %zmm0
+; X32-NEXT: vpermi2ps %zmm0, %zmm1, %zmm0 {%k1} {z}
; X32-NEXT: retl
;
; X64-LABEL: combine_vpermt2var_16f32_vmovddup_mask_load:
; X64: # %bb.0:
-; X64-NEXT: vmovaps (%rdi), %zmm2
-; X64-NEXT: vmovaps {{.*#+}} zmm1 = [0,1,0,1,4,5,4,5,8,9,8,9,12,13,12,13]
+; X64-NEXT: vmovaps (%rdi), %zmm1
+; X64-NEXT: vmovaps {{.*#+}} zmm0 = [0,1,0,1,4,5,4,5,8,9,8,9,12,13,12,13]
; X64-NEXT: kmovd %esi, %k1
-; X64-NEXT: vpermi2ps %zmm0, %zmm2, %zmm1 {%k1} {z}
-; X64-NEXT: vmovaps %zmm1, %zmm0
+; X64-NEXT: vpermi2ps %zmm0, %zmm1, %zmm0 {%k1} {z}
; X64-NEXT: retq
%x0 = load <16 x float>, <16 x float> *%p0
%res0 = call <16 x float> @llvm.x86.avx512.maskz.vpermt2var.ps.512(<16 x i32> <i32 0, i32 1, i32 0, i32 1, i32 4, i32 5, i32 4, i32 5, i32 8, i32 9, i32 8, i32 9, i32 12, i32 13, i32 12, i32 13>, <16 x float> %x0, <16 x float> %x1, i16 %m)
@@ -519,20 +515,20 @@ define <16 x i32> @combine_vpermt2var_16i32_identity(<16 x i32> %x0, <16 x i32>
define <16 x i32> @combine_vpermt2var_16i32_identity_mask(<16 x i32> %x0, <16 x i32> %x1, i16 %m) {
; X32-LABEL: combine_vpermt2var_16i32_identity_mask:
; X32: # %bb.0:
-; X32-NEXT: vmovdqa64 {{.*#+}} zmm2 = [15,14,13,12,11,10,9,8,7,6,5,4,3,2,1,0]
+; X32-NEXT: vmovdqa64 {{.*#+}} zmm1 = [15,14,13,12,11,10,9,8,7,6,5,4,3,2,1,0]
; X32-NEXT: kmovw {{[0-9]+}}(%esp), %k1
-; X32-NEXT: vpermi2d %zmm1, %zmm0, %zmm2 {%k1} {z}
+; X32-NEXT: vpermi2d %zmm0, %zmm0, %zmm1 {%k1} {z}
; X32-NEXT: vmovdqa64 {{.*#+}} zmm0 = [15,30,13,28,11,26,9,24,7,22,5,20,3,18,1,16]
-; X32-NEXT: vpermi2d %zmm2, %zmm2, %zmm0 {%k1} {z}
+; X32-NEXT: vpermi2d %zmm1, %zmm1, %zmm0 {%k1} {z}
; X32-NEXT: retl
;
; X64-LABEL: combine_vpermt2var_16i32_identity_mask:
; X64: # %bb.0:
-; X64-NEXT: vmovdqa64 {{.*#+}} zmm2 = [15,14,13,12,11,10,9,8,7,6,5,4,3,2,1,0]
+; X64-NEXT: vmovdqa64 {{.*#+}} zmm1 = [15,14,13,12,11,10,9,8,7,6,5,4,3,2,1,0]
; X64-NEXT: kmovd %edi, %k1
-; X64-NEXT: vpermi2d %zmm1, %zmm0, %zmm2 {%k1} {z}
+; X64-NEXT: vpermi2d %zmm0, %zmm0, %zmm1 {%k1} {z}
; X64-NEXT: vmovdqa64 {{.*#+}} zmm0 = [15,30,13,28,11,26,9,24,7,22,5,20,3,18,1,16]
-; X64-NEXT: vpermi2d %zmm2, %zmm2, %zmm0 {%k1} {z}
+; X64-NEXT: vpermi2d %zmm1, %zmm1, %zmm0 {%k1} {z}
; X64-NEXT: retq
%res0 = call <16 x i32> @llvm.x86.avx512.maskz.vpermt2var.d.512(<16 x i32> <i32 15, i32 14, i32 13, i32 12, i32 11, i32 10, i32 9, i32 8, i32 7, i32 6, i32 5, i32 4, i32 3, i32 2, i32 1, i32 0>, <16 x i32> %x0, <16 x i32> %x1, i16 %m)
%res1 = call <16 x i32> @llvm.x86.avx512.maskz.vpermt2var.d.512(<16 x i32> <i32 15, i32 30, i32 13, i32 28, i32 11, i32 26, i32 9, i32 24, i32 7, i32 22, i32 5, i32 20, i32 3, i32 18, i32 1, i32 16>, <16 x i32> %res0, <16 x i32> %res0, i16 %m)
@@ -554,20 +550,20 @@ define <32 x i16> @combine_vpermt2var_32i16_identity(<32 x i16> %x0, <32 x i16>
define <32 x i16> @combine_vpermt2var_32i16_identity_mask(<32 x i16> %x0, <32 x i16> %x1, i32 %m) {
; X32-LABEL: combine_vpermt2var_32i16_identity_mask:
; X32: # %bb.0:
-; X32-NEXT: vmovdqa64 {{.*#+}} zmm2 = [31,30,29,28,27,26,25,24,23,22,21,20,19,18,17,16,15,14,13,12,11,10,9,8,7,6,5,4,3,2,1,0]
+; X32-NEXT: vmovdqa64 {{.*#+}} zmm1 = [31,30,29,28,27,26,25,24,23,22,21,20,19,18,17,16,15,14,13,12,11,10,9,8,7,6,5,4,3,2,1,0]
; X32-NEXT: kmovd {{[0-9]+}}(%esp), %k1
-; X32-NEXT: vpermi2w %zmm1, %zmm0, %zmm2 {%k1} {z}
+; X32-NEXT: vpermi2w %zmm0, %zmm0, %zmm1 {%k1} {z}
; X32-NEXT: vmovdqa64 {{.*#+}} zmm0 = [63,30,61,28,59,26,57,24,55,22,53,20,51,18,49,16,47,46,13,44,11,42,9,40,7,38,5,36,3,34,1,32]
-; X32-NEXT: vpermi2w %zmm2, %zmm2, %zmm0 {%k1} {z}
+; X32-NEXT: vpermi2w %zmm1, %zmm1, %zmm0 {%k1} {z}
; X32-NEXT: retl
;
; X64-LABEL: combine_vpermt2var_32i16_identity_mask:
; X64: # %bb.0:
-; X64-NEXT: vmovdqa64 {{.*#+}} zmm2 = [31,30,29,28,27,26,25,24,23,22,21,20,19,18,17,16,15,14,13,12,11,10,9,8,7,6,5,4,3,2,1,0]
+; X64-NEXT: vmovdqa64 {{.*#+}} zmm1 = [31,30,29,28,27,26,25,24,23,22,21,20,19,18,17,16,15,14,13,12,11,10,9,8,7,6,5,4,3,2,1,0]
; X64-NEXT: kmovd %edi, %k1
-; X64-NEXT: vpermi2w %zmm1, %zmm0, %zmm2 {%k1} {z}
+; X64-NEXT: vpermi2w %zmm0, %zmm0, %zmm1 {%k1} {z}
; X64-NEXT: vmovdqa64 {{.*#+}} zmm0 = [63,30,61,28,59,26,57,24,55,22,53,20,51,18,49,16,47,46,13,44,11,42,9,40,7,38,5,36,3,34,1,32]
-; X64-NEXT: vpermi2w %zmm2, %zmm2, %zmm0 {%k1} {z}
+; X64-NEXT: vpermi2w %zmm1, %zmm1, %zmm0 {%k1} {z}
; X64-NEXT: retq
%res0 = call <32 x i16> @llvm.x86.avx512.maskz.vpermt2var.hi.512(<32 x i16> <i16 31, i16 30, i16 29, i16 28, i16 27, i16 26, i16 25, i16 24, i16 23, i16 22, i16 21, i16 20, i16 19, i16 18, i16 17, i16 16, i16 15, i16 14, i16 13, i16 12, i16 11, i16 10, i16 9, i16 8, i16 7, i16 6, i16 5, i16 4, i16 3, i16 2, i16 1, i16 0>, <32 x i16> %x0, <32 x i16> %x1, i32 %m)
%res1 = call <32 x i16> @llvm.x86.avx512.maskz.vpermt2var.hi.512(<32 x i16> <i16 63, i16 30, i16 61, i16 28, i16 59, i16 26, i16 57, i16 24, i16 55, i16 22, i16 53, i16 20, i16 51, i16 18, i16 49, i16 16, i16 47, i16 46, i16 13, i16 44, i16 11, i16 42, i16 9, i16 40, i16 7, i16 38, i16 5, i16 36, i16 3, i16 34, i16 1, i16 32>, <32 x i16> %res0, <32 x i16> %res0, i32 %m)
OpenPOWER on IntegriCloud