[X86] Add avx512bw command lines to the 256-bit vector idiv tests.

Some of the operations are being sign extended to 512 bits with avx512bw.

llvm-svn: 314272

diff --git a/llvm/test/CodeGen/X86/vector-idiv-sdiv-256.ll b/llvm/test/CodeGen/X86/vector-idiv-sdiv-256.ll
index ce0ec6c..5e38c57 100644 (file)
--- a/llvm/test/CodeGen/X86/vector-idiv-sdiv-256.ll
+++ b/llvm/test/CodeGen/X86/vector-idiv-sdiv-256.ll
@@ -1,6 +1,7 @@
 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
-; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx | FileCheck %s --check-prefix=AVX --check-prefix=AVX1
-; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx2 | FileCheck %s --check-prefix=AVX --check-prefix=AVX2
+; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx | FileCheck %s --check-prefix=AVX1
+; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx2 | FileCheck %s --check-prefix=AVX2 --check-prefix=AVX2NOBW
+; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx512bw | FileCheck %s --check-prefix=AVX2 --check-prefix=AVX512BW
 
 ;
 ; sdiv by 7
@@ -497,49 +498,80 @@ define <32 x i8> @test_rem7_32i8(<32 x i8> %a) nounwind {
 ; AVX1-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
 ; AVX1-NEXT:    retq
 ;
-; AVX2-LABEL: test_rem7_32i8:
-; AVX2:       # BB#0:
-; AVX2-NEXT:    vmovdqa {{.*#+}} ymm1 = [147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147]
-; AVX2-NEXT:    vextracti128 $1, %ymm1, %xmm2
-; AVX2-NEXT:    vpmovsxbw %xmm2, %ymm2
-; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm3
-; AVX2-NEXT:    vpmovsxbw %xmm3, %ymm3
-; AVX2-NEXT:    vpmullw %ymm2, %ymm3, %ymm2
-; AVX2-NEXT:    vpsrlw $8, %ymm2, %ymm2
-; AVX2-NEXT:    vpmovsxbw %xmm1, %ymm1
-; AVX2-NEXT:    vpmovsxbw %xmm0, %ymm3
-; AVX2-NEXT:    vpmullw %ymm1, %ymm3, %ymm1
-; AVX2-NEXT:    vpsrlw $8, %ymm1, %ymm1
-; AVX2-NEXT:    vperm2i128 {{.*#+}} ymm3 = ymm1[2,3],ymm2[2,3]
-; AVX2-NEXT:    vinserti128 $1, %xmm2, %ymm1, %ymm1
-; AVX2-NEXT:    vpackuswb %ymm3, %ymm1, %ymm1
-; AVX2-NEXT:    vpaddb %ymm0, %ymm1, %ymm1
-; AVX2-NEXT:    vpsrlw $2, %ymm1, %ymm2
-; AVX2-NEXT:    vpand {{.*}}(%rip), %ymm2, %ymm2
-; AVX2-NEXT:    vmovdqa {{.*#+}} ymm3 = [32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32]
-; AVX2-NEXT:    vpxor %ymm3, %ymm2, %ymm2
-; AVX2-NEXT:    vpsubb %ymm3, %ymm2, %ymm2
-; AVX2-NEXT:    vpsrlw $7, %ymm1, %ymm1
-; AVX2-NEXT:    vpand {{.*}}(%rip), %ymm1, %ymm1
-; AVX2-NEXT:    vpaddb %ymm1, %ymm2, %ymm1
-; AVX2-NEXT:    vextracti128 $1, %ymm1, %xmm2
-; AVX2-NEXT:    vpmovsxbw %xmm2, %ymm2
-; AVX2-NEXT:    vmovdqa {{.*#+}} ymm3 = [7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7]
-; AVX2-NEXT:    vpmullw %ymm3, %ymm2, %ymm2
-; AVX2-NEXT:    vextracti128 $1, %ymm2, %xmm4
-; AVX2-NEXT:    vmovdqa {{.*#+}} xmm5 = <0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u>
-; AVX2-NEXT:    vpshufb %xmm5, %xmm4, %xmm4
-; AVX2-NEXT:    vpshufb %xmm5, %xmm2, %xmm2
-; AVX2-NEXT:    vpunpcklqdq {{.*#+}} xmm2 = xmm2[0],xmm4[0]
-; AVX2-NEXT:    vpmovsxbw %xmm1, %ymm1
-; AVX2-NEXT:    vpmullw %ymm3, %ymm1, %ymm1
-; AVX2-NEXT:    vextracti128 $1, %ymm1, %xmm3
-; AVX2-NEXT:    vpshufb %xmm5, %xmm3, %xmm3
-; AVX2-NEXT:    vpshufb %xmm5, %xmm1, %xmm1
-; AVX2-NEXT:    vpunpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm3[0]
-; AVX2-NEXT:    vinserti128 $1, %xmm2, %ymm1, %ymm1
-; AVX2-NEXT:    vpsubb %ymm1, %ymm0, %ymm0
-; AVX2-NEXT:    retq
+; AVX2NOBW-LABEL: test_rem7_32i8:
+; AVX2NOBW:       # BB#0:
+; AVX2NOBW-NEXT:    vmovdqa {{.*#+}} ymm1 = [147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147]
+; AVX2NOBW-NEXT:    vextracti128 $1, %ymm1, %xmm2
+; AVX2NOBW-NEXT:    vpmovsxbw %xmm2, %ymm2
+; AVX2NOBW-NEXT:    vextracti128 $1, %ymm0, %xmm3
+; AVX2NOBW-NEXT:    vpmovsxbw %xmm3, %ymm3
+; AVX2NOBW-NEXT:    vpmullw %ymm2, %ymm3, %ymm2
+; AVX2NOBW-NEXT:    vpsrlw $8, %ymm2, %ymm2
+; AVX2NOBW-NEXT:    vpmovsxbw %xmm1, %ymm1
+; AVX2NOBW-NEXT:    vpmovsxbw %xmm0, %ymm3
+; AVX2NOBW-NEXT:    vpmullw %ymm1, %ymm3, %ymm1
+; AVX2NOBW-NEXT:    vpsrlw $8, %ymm1, %ymm1
+; AVX2NOBW-NEXT:    vperm2i128 {{.*#+}} ymm3 = ymm1[2,3],ymm2[2,3]
+; AVX2NOBW-NEXT:    vinserti128 $1, %xmm2, %ymm1, %ymm1
+; AVX2NOBW-NEXT:    vpackuswb %ymm3, %ymm1, %ymm1
+; AVX2NOBW-NEXT:    vpaddb %ymm0, %ymm1, %ymm1
+; AVX2NOBW-NEXT:    vpsrlw $2, %ymm1, %ymm2
+; AVX2NOBW-NEXT:    vpand {{.*}}(%rip), %ymm2, %ymm2
+; AVX2NOBW-NEXT:    vmovdqa {{.*#+}} ymm3 = [32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32]
+; AVX2NOBW-NEXT:    vpxor %ymm3, %ymm2, %ymm2
+; AVX2NOBW-NEXT:    vpsubb %ymm3, %ymm2, %ymm2
+; AVX2NOBW-NEXT:    vpsrlw $7, %ymm1, %ymm1
+; AVX2NOBW-NEXT:    vpand {{.*}}(%rip), %ymm1, %ymm1
+; AVX2NOBW-NEXT:    vpaddb %ymm1, %ymm2, %ymm1
+; AVX2NOBW-NEXT:    vextracti128 $1, %ymm1, %xmm2
+; AVX2NOBW-NEXT:    vpmovsxbw %xmm2, %ymm2
+; AVX2NOBW-NEXT:    vmovdqa {{.*#+}} ymm3 = [7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7]
+; AVX2NOBW-NEXT:    vpmullw %ymm3, %ymm2, %ymm2
+; AVX2NOBW-NEXT:    vextracti128 $1, %ymm2, %xmm4
+; AVX2NOBW-NEXT:    vmovdqa {{.*#+}} xmm5 = <0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u>
+; AVX2NOBW-NEXT:    vpshufb %xmm5, %xmm4, %xmm4
+; AVX2NOBW-NEXT:    vpshufb %xmm5, %xmm2, %xmm2
+; AVX2NOBW-NEXT:    vpunpcklqdq {{.*#+}} xmm2 = xmm2[0],xmm4[0]
+; AVX2NOBW-NEXT:    vpmovsxbw %xmm1, %ymm1
+; AVX2NOBW-NEXT:    vpmullw %ymm3, %ymm1, %ymm1
+; AVX2NOBW-NEXT:    vextracti128 $1, %ymm1, %xmm3
+; AVX2NOBW-NEXT:    vpshufb %xmm5, %xmm3, %xmm3
+; AVX2NOBW-NEXT:    vpshufb %xmm5, %xmm1, %xmm1
+; AVX2NOBW-NEXT:    vpunpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm3[0]
+; AVX2NOBW-NEXT:    vinserti128 $1, %xmm2, %ymm1, %ymm1
+; AVX2NOBW-NEXT:    vpsubb %ymm1, %ymm0, %ymm0
+; AVX2NOBW-NEXT:    retq
+;
+; AVX512BW-LABEL: test_rem7_32i8:
+; AVX512BW:       # BB#0:
+; AVX512BW-NEXT:    vmovdqa {{.*#+}} ymm1 = [147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147]
+; AVX512BW-NEXT:    vextracti128 $1, %ymm1, %xmm2
+; AVX512BW-NEXT:    vpmovsxbw %xmm2, %ymm2
+; AVX512BW-NEXT:    vextracti128 $1, %ymm0, %xmm3
+; AVX512BW-NEXT:    vpmovsxbw %xmm3, %ymm3
+; AVX512BW-NEXT:    vpmullw %ymm2, %ymm3, %ymm2
+; AVX512BW-NEXT:    vpsrlw $8, %ymm2, %ymm2
+; AVX512BW-NEXT:    vpmovsxbw %xmm1, %ymm1
+; AVX512BW-NEXT:    vpmovsxbw %xmm0, %ymm3
+; AVX512BW-NEXT:    vpmullw %ymm1, %ymm3, %ymm1
+; AVX512BW-NEXT:    vpsrlw $8, %ymm1, %ymm1
+; AVX512BW-NEXT:    vperm2i128 {{.*#+}} ymm3 = ymm1[2,3],ymm2[2,3]
+; AVX512BW-NEXT:    vinserti128 $1, %xmm2, %ymm1, %ymm1
+; AVX512BW-NEXT:    vpackuswb %ymm3, %ymm1, %ymm1
+; AVX512BW-NEXT:    vpaddb %ymm0, %ymm1, %ymm1
+; AVX512BW-NEXT:    vpsrlw $2, %ymm1, %ymm2
+; AVX512BW-NEXT:    vpand {{.*}}(%rip), %ymm2, %ymm2
+; AVX512BW-NEXT:    vmovdqa {{.*#+}} ymm3 = [32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32]
+; AVX512BW-NEXT:    vpxor %ymm3, %ymm2, %ymm2
+; AVX512BW-NEXT:    vpsubb %ymm3, %ymm2, %ymm2
+; AVX512BW-NEXT:    vpsrlw $7, %ymm1, %ymm1
+; AVX512BW-NEXT:    vpand {{.*}}(%rip), %ymm1, %ymm1
+; AVX512BW-NEXT:    vpaddb %ymm1, %ymm2, %ymm1
+; AVX512BW-NEXT:    vpmovsxbw %ymm1, %zmm1
+; AVX512BW-NEXT:    vpmullw {{.*}}(%rip), %zmm1, %zmm1
+; AVX512BW-NEXT:    vpmovwb %zmm1, %ymm1
+; AVX512BW-NEXT:    vpsubb %ymm1, %ymm0, %ymm0
+; AVX512BW-NEXT:    retq
   %res = srem <32 x i8> %a, <i8 7, i8 7, i8 7, i8 7,i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7,i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7,i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7,i8 7, i8 7, i8 7, i8 7>
   ret <32 x i8> %res
 }

diff --git a/llvm/test/CodeGen/X86/vector-idiv-udiv-256.ll b/llvm/test/CodeGen/X86/vector-idiv-udiv-256.ll
index b043311..0a2079f 100644 (file)
--- a/llvm/test/CodeGen/X86/vector-idiv-udiv-256.ll
+++ b/llvm/test/CodeGen/X86/vector-idiv-udiv-256.ll
@@ -1,6 +1,7 @@
 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
-; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx | FileCheck %s --check-prefix=AVX --check-prefix=AVX1
-; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx2 | FileCheck %s --check-prefix=AVX --check-prefix=AVX2
+; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx | FileCheck %s --check-prefix=AVX1
+; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx2 | FileCheck %s --check-prefix=AVX2 --check-prefix=AVX2NOBW
+; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx512bw | FileCheck %s --check-prefix=AVX2 --check-prefix=AVX512BW
 
 ;
 ; udiv by 7
@@ -506,46 +507,74 @@ define <32 x i8> @test_rem7_32i8(<32 x i8> %a) nounwind {
 ; AVX1-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
 ; AVX1-NEXT:    retq
 ;
-; AVX2-LABEL: test_rem7_32i8:
-; AVX2:       # BB#0:
-; AVX2-NEXT:    vmovdqa {{.*#+}} ymm1 = [37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37]
-; AVX2-NEXT:    vextracti128 $1, %ymm1, %xmm2
-; AVX2-NEXT:    vpmovzxbw {{.*#+}} ymm2 = xmm2[0],zero,xmm2[1],zero,xmm2[2],zero,xmm2[3],zero,xmm2[4],zero,xmm2[5],zero,xmm2[6],zero,xmm2[7],zero,xmm2[8],zero,xmm2[9],zero,xmm2[10],zero,xmm2[11],zero,xmm2[12],zero,xmm2[13],zero,xmm2[14],zero,xmm2[15],zero
-; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm3
-; AVX2-NEXT:    vpmovzxbw {{.*#+}} ymm3 = xmm3[0],zero,xmm3[1],zero,xmm3[2],zero,xmm3[3],zero,xmm3[4],zero,xmm3[5],zero,xmm3[6],zero,xmm3[7],zero,xmm3[8],zero,xmm3[9],zero,xmm3[10],zero,xmm3[11],zero,xmm3[12],zero,xmm3[13],zero,xmm3[14],zero,xmm3[15],zero
-; AVX2-NEXT:    vpmullw %ymm2, %ymm3, %ymm2
-; AVX2-NEXT:    vpsrlw $8, %ymm2, %ymm2
-; AVX2-NEXT:    vpmovzxbw {{.*#+}} ymm1 = xmm1[0],zero,xmm1[1],zero,xmm1[2],zero,xmm1[3],zero,xmm1[4],zero,xmm1[5],zero,xmm1[6],zero,xmm1[7],zero,xmm1[8],zero,xmm1[9],zero,xmm1[10],zero,xmm1[11],zero,xmm1[12],zero,xmm1[13],zero,xmm1[14],zero,xmm1[15],zero
-; AVX2-NEXT:    vpmovzxbw {{.*#+}} ymm3 = xmm0[0],zero,xmm0[1],zero,xmm0[2],zero,xmm0[3],zero,xmm0[4],zero,xmm0[5],zero,xmm0[6],zero,xmm0[7],zero,xmm0[8],zero,xmm0[9],zero,xmm0[10],zero,xmm0[11],zero,xmm0[12],zero,xmm0[13],zero,xmm0[14],zero,xmm0[15],zero
-; AVX2-NEXT:    vpmullw %ymm1, %ymm3, %ymm1
-; AVX2-NEXT:    vpsrlw $8, %ymm1, %ymm1
-; AVX2-NEXT:    vperm2i128 {{.*#+}} ymm3 = ymm1[2,3],ymm2[2,3]
-; AVX2-NEXT:    vinserti128 $1, %xmm2, %ymm1, %ymm1
-; AVX2-NEXT:    vpackuswb %ymm3, %ymm1, %ymm1
-; AVX2-NEXT:    vpsubb %ymm1, %ymm0, %ymm2
-; AVX2-NEXT:    vpsrlw $1, %ymm2, %ymm2
-; AVX2-NEXT:    vpand {{.*}}(%rip), %ymm2, %ymm2
-; AVX2-NEXT:    vpaddb %ymm1, %ymm2, %ymm1
-; AVX2-NEXT:    vpsrlw $2, %ymm1, %ymm1
-; AVX2-NEXT:    vpand {{.*}}(%rip), %ymm1, %ymm1
-; AVX2-NEXT:    vextracti128 $1, %ymm1, %xmm2
-; AVX2-NEXT:    vpmovsxbw %xmm2, %ymm2
-; AVX2-NEXT:    vmovdqa {{.*#+}} ymm3 = [7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7]
-; AVX2-NEXT:    vpmullw %ymm3, %ymm2, %ymm2
-; AVX2-NEXT:    vextracti128 $1, %ymm2, %xmm4
-; AVX2-NEXT:    vmovdqa {{.*#+}} xmm5 = <0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u>
-; AVX2-NEXT:    vpshufb %xmm5, %xmm4, %xmm4
-; AVX2-NEXT:    vpshufb %xmm5, %xmm2, %xmm2
-; AVX2-NEXT:    vpunpcklqdq {{.*#+}} xmm2 = xmm2[0],xmm4[0]
-; AVX2-NEXT:    vpmovsxbw %xmm1, %ymm1
-; AVX2-NEXT:    vpmullw %ymm3, %ymm1, %ymm1
-; AVX2-NEXT:    vextracti128 $1, %ymm1, %xmm3
-; AVX2-NEXT:    vpshufb %xmm5, %xmm3, %xmm3
-; AVX2-NEXT:    vpshufb %xmm5, %xmm1, %xmm1
-; AVX2-NEXT:    vpunpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm3[0]
-; AVX2-NEXT:    vinserti128 $1, %xmm2, %ymm1, %ymm1
-; AVX2-NEXT:    vpsubb %ymm1, %ymm0, %ymm0
-; AVX2-NEXT:    retq
+; AVX2NOBW-LABEL: test_rem7_32i8:
+; AVX2NOBW:       # BB#0:
+; AVX2NOBW-NEXT:    vmovdqa {{.*#+}} ymm1 = [37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37]
+; AVX2NOBW-NEXT:    vextracti128 $1, %ymm1, %xmm2
+; AVX2NOBW-NEXT:    vpmovzxbw {{.*#+}} ymm2 = xmm2[0],zero,xmm2[1],zero,xmm2[2],zero,xmm2[3],zero,xmm2[4],zero,xmm2[5],zero,xmm2[6],zero,xmm2[7],zero,xmm2[8],zero,xmm2[9],zero,xmm2[10],zero,xmm2[11],zero,xmm2[12],zero,xmm2[13],zero,xmm2[14],zero,xmm2[15],zero
+; AVX2NOBW-NEXT:    vextracti128 $1, %ymm0, %xmm3
+; AVX2NOBW-NEXT:    vpmovzxbw {{.*#+}} ymm3 = xmm3[0],zero,xmm3[1],zero,xmm3[2],zero,xmm3[3],zero,xmm3[4],zero,xmm3[5],zero,xmm3[6],zero,xmm3[7],zero,xmm3[8],zero,xmm3[9],zero,xmm3[10],zero,xmm3[11],zero,xmm3[12],zero,xmm3[13],zero,xmm3[14],zero,xmm3[15],zero
+; AVX2NOBW-NEXT:    vpmullw %ymm2, %ymm3, %ymm2
+; AVX2NOBW-NEXT:    vpsrlw $8, %ymm2, %ymm2
+; AVX2NOBW-NEXT:    vpmovzxbw {{.*#+}} ymm1 = xmm1[0],zero,xmm1[1],zero,xmm1[2],zero,xmm1[3],zero,xmm1[4],zero,xmm1[5],zero,xmm1[6],zero,xmm1[7],zero,xmm1[8],zero,xmm1[9],zero,xmm1[10],zero,xmm1[11],zero,xmm1[12],zero,xmm1[13],zero,xmm1[14],zero,xmm1[15],zero
+; AVX2NOBW-NEXT:    vpmovzxbw {{.*#+}} ymm3 = xmm0[0],zero,xmm0[1],zero,xmm0[2],zero,xmm0[3],zero,xmm0[4],zero,xmm0[5],zero,xmm0[6],zero,xmm0[7],zero,xmm0[8],zero,xmm0[9],zero,xmm0[10],zero,xmm0[11],zero,xmm0[12],zero,xmm0[13],zero,xmm0[14],zero,xmm0[15],zero
+; AVX2NOBW-NEXT:    vpmullw %ymm1, %ymm3, %ymm1
+; AVX2NOBW-NEXT:    vpsrlw $8, %ymm1, %ymm1
+; AVX2NOBW-NEXT:    vperm2i128 {{.*#+}} ymm3 = ymm1[2,3],ymm2[2,3]
+; AVX2NOBW-NEXT:    vinserti128 $1, %xmm2, %ymm1, %ymm1
+; AVX2NOBW-NEXT:    vpackuswb %ymm3, %ymm1, %ymm1
+; AVX2NOBW-NEXT:    vpsubb %ymm1, %ymm0, %ymm2
+; AVX2NOBW-NEXT:    vpsrlw $1, %ymm2, %ymm2
+; AVX2NOBW-NEXT:    vpand {{.*}}(%rip), %ymm2, %ymm2
+; AVX2NOBW-NEXT:    vpaddb %ymm1, %ymm2, %ymm1
+; AVX2NOBW-NEXT:    vpsrlw $2, %ymm1, %ymm1
+; AVX2NOBW-NEXT:    vpand {{.*}}(%rip), %ymm1, %ymm1
+; AVX2NOBW-NEXT:    vextracti128 $1, %ymm1, %xmm2
+; AVX2NOBW-NEXT:    vpmovsxbw %xmm2, %ymm2
+; AVX2NOBW-NEXT:    vmovdqa {{.*#+}} ymm3 = [7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7]
+; AVX2NOBW-NEXT:    vpmullw %ymm3, %ymm2, %ymm2
+; AVX2NOBW-NEXT:    vextracti128 $1, %ymm2, %xmm4
+; AVX2NOBW-NEXT:    vmovdqa {{.*#+}} xmm5 = <0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u>
+; AVX2NOBW-NEXT:    vpshufb %xmm5, %xmm4, %xmm4
+; AVX2NOBW-NEXT:    vpshufb %xmm5, %xmm2, %xmm2
+; AVX2NOBW-NEXT:    vpunpcklqdq {{.*#+}} xmm2 = xmm2[0],xmm4[0]
+; AVX2NOBW-NEXT:    vpmovsxbw %xmm1, %ymm1
+; AVX2NOBW-NEXT:    vpmullw %ymm3, %ymm1, %ymm1
+; AVX2NOBW-NEXT:    vextracti128 $1, %ymm1, %xmm3
+; AVX2NOBW-NEXT:    vpshufb %xmm5, %xmm3, %xmm3
+; AVX2NOBW-NEXT:    vpshufb %xmm5, %xmm1, %xmm1
+; AVX2NOBW-NEXT:    vpunpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm3[0]
+; AVX2NOBW-NEXT:    vinserti128 $1, %xmm2, %ymm1, %ymm1
+; AVX2NOBW-NEXT:    vpsubb %ymm1, %ymm0, %ymm0
+; AVX2NOBW-NEXT:    retq
+;
+; AVX512BW-LABEL: test_rem7_32i8:
+; AVX512BW:       # BB#0:
+; AVX512BW-NEXT:    vmovdqa {{.*#+}} ymm1 = [37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37]
+; AVX512BW-NEXT:    vextracti128 $1, %ymm1, %xmm2
+; AVX512BW-NEXT:    vpmovzxbw {{.*#+}} ymm2 = xmm2[0],zero,xmm2[1],zero,xmm2[2],zero,xmm2[3],zero,xmm2[4],zero,xmm2[5],zero,xmm2[6],zero,xmm2[7],zero,xmm2[8],zero,xmm2[9],zero,xmm2[10],zero,xmm2[11],zero,xmm2[12],zero,xmm2[13],zero,xmm2[14],zero,xmm2[15],zero
+; AVX512BW-NEXT:    vextracti128 $1, %ymm0, %xmm3
+; AVX512BW-NEXT:    vpmovzxbw {{.*#+}} ymm3 = xmm3[0],zero,xmm3[1],zero,xmm3[2],zero,xmm3[3],zero,xmm3[4],zero,xmm3[5],zero,xmm3[6],zero,xmm3[7],zero,xmm3[8],zero,xmm3[9],zero,xmm3[10],zero,xmm3[11],zero,xmm3[12],zero,xmm3[13],zero,xmm3[14],zero,xmm3[15],zero
+; AVX512BW-NEXT:    vpmullw %ymm2, %ymm3, %ymm2
+; AVX512BW-NEXT:    vpsrlw $8, %ymm2, %ymm2
+; AVX512BW-NEXT:    vpmovzxbw {{.*#+}} ymm1 = xmm1[0],zero,xmm1[1],zero,xmm1[2],zero,xmm1[3],zero,xmm1[4],zero,xmm1[5],zero,xmm1[6],zero,xmm1[7],zero,xmm1[8],zero,xmm1[9],zero,xmm1[10],zero,xmm1[11],zero,xmm1[12],zero,xmm1[13],zero,xmm1[14],zero,xmm1[15],zero
+; AVX512BW-NEXT:    vpmovzxbw {{.*#+}} ymm3 = xmm0[0],zero,xmm0[1],zero,xmm0[2],zero,xmm0[3],zero,xmm0[4],zero,xmm0[5],zero,xmm0[6],zero,xmm0[7],zero,xmm0[8],zero,xmm0[9],zero,xmm0[10],zero,xmm0[11],zero,xmm0[12],zero,xmm0[13],zero,xmm0[14],zero,xmm0[15],zero
+; AVX512BW-NEXT:    vpmullw %ymm1, %ymm3, %ymm1
+; AVX512BW-NEXT:    vpsrlw $8, %ymm1, %ymm1
+; AVX512BW-NEXT:    vperm2i128 {{.*#+}} ymm3 = ymm1[2,3],ymm2[2,3]
+; AVX512BW-NEXT:    vinserti128 $1, %xmm2, %ymm1, %ymm1
+; AVX512BW-NEXT:    vpackuswb %ymm3, %ymm1, %ymm1
+; AVX512BW-NEXT:    vpsubb %ymm1, %ymm0, %ymm2
+; AVX512BW-NEXT:    vpsrlw $1, %ymm2, %ymm2
+; AVX512BW-NEXT:    vpand {{.*}}(%rip), %ymm2, %ymm2
+; AVX512BW-NEXT:    vpaddb %ymm1, %ymm2, %ymm1
+; AVX512BW-NEXT:    vpsrlw $2, %ymm1, %ymm1
+; AVX512BW-NEXT:    vpand {{.*}}(%rip), %ymm1, %ymm1
+; AVX512BW-NEXT:    vpmovsxbw %ymm1, %zmm1
+; AVX512BW-NEXT:    vpmullw {{.*}}(%rip), %zmm1, %zmm1
+; AVX512BW-NEXT:    vpmovwb %zmm1, %ymm1
+; AVX512BW-NEXT:    vpsubb %ymm1, %ymm0, %ymm0
+; AVX512BW-NEXT:    retq
   %res = urem <32 x i8> %a, <i8 7, i8 7, i8 7, i8 7,i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7,i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7,i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7,i8 7, i8 7, i8 7, i8 7>
   ret <32 x i8> %res
 }