[X86] Add test coverage for #143238

Author: RKSimon

llvm/test/CodeGen/X86/vector-idiv-sdiv-128.ll

@@ -1118,3 +1118,101 @@ define <16 x i8> @test_rem_variable_16i8(<16 x i8> %a, <16 x i8> %b) nounwind {
   %res = srem <16 x i8> %a, %b
   ret <16 x i8> %res
 }
+
+define <16 x i8> @PR143238(<16 x i8> %a0) {
+; SSE-LABEL: PR143238:
+; SSE:       # %bb.0:
+; SSE-NEXT:    pxor %xmm1, %xmm1
+; SSE-NEXT:    pxor %xmm2, %xmm2
+; SSE-NEXT:    punpckhbw {{.*#+}} xmm2 = xmm2[8],xmm0[8],xmm2[9],xmm0[9],xmm2[10],xmm0[10],xmm2[11],xmm0[11],xmm2[12],xmm0[12],xmm2[13],xmm0[13],xmm2[14],xmm0[14],xmm2[15],xmm0[15]
+; SSE-NEXT:    pmulhw {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm2 # [26368,47872,11008,20224,37632,35072,33024,30976]
+; SSE-NEXT:    psrlw $8, %xmm2
+; SSE-NEXT:    punpcklbw {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1],xmm1[2],xmm0[2],xmm1[3],xmm0[3],xmm1[4],xmm0[4],xmm1[5],xmm0[5],xmm1[6],xmm0[6],xmm1[7],xmm0[7]
+; SSE-NEXT:    pmulhw {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm1 # [33024,22016,33024,26368,11008,37632,33024,14592]
+; SSE-NEXT:    psrlw $8, %xmm1
+; SSE-NEXT:    packuswb %xmm2, %xmm1
+; SSE-NEXT:    pand {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0
+; SSE-NEXT:    paddb %xmm1, %xmm0
+; SSE-NEXT:    movdqa %xmm0, %xmm1
+; SSE-NEXT:    punpckhbw {{.*#+}} xmm1 = xmm1[8],xmm0[8],xmm1[9],xmm0[9],xmm1[10],xmm0[10],xmm1[11],xmm0[11],xmm1[12],xmm0[12],xmm1[13],xmm0[13],xmm1[14],xmm0[14],xmm1[15],xmm0[15]
+; SSE-NEXT:    psraw $8, %xmm1
+; SSE-NEXT:    pmullw {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm1 # [64,32,128,64,32,32,32,32]
+; SSE-NEXT:    psrlw $8, %xmm1
+; SSE-NEXT:    movdqa %xmm0, %xmm2
+; SSE-NEXT:    punpcklbw {{.*#+}} xmm2 = xmm2[0],xmm0[0],xmm2[1],xmm0[1],xmm2[2],xmm0[2],xmm2[3],xmm0[3],xmm2[4],xmm0[4],xmm2[5],xmm0[5],xmm2[6],xmm0[6],xmm2[7],xmm0[7]
+; SSE-NEXT:    psraw $8, %xmm2
+; SSE-NEXT:    pmullw {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm2 # [256,256,128,128,256,64,64,128]
+; SSE-NEXT:    psrlw $8, %xmm2
+; SSE-NEXT:    packuswb %xmm1, %xmm2
+; SSE-NEXT:    psrlw $7, %xmm0
+; SSE-NEXT:    paddb %xmm2, %xmm0
+; SSE-NEXT:    pand {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0
+; SSE-NEXT:    retq
+;
+; AVX1-LABEL: PR143238:
+; AVX1:       # %bb.0:
+; AVX1-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX1-NEXT:    vpunpckhbw {{.*#+}} xmm2 = xmm1[8],xmm0[8],xmm1[9],xmm0[9],xmm1[10],xmm0[10],xmm1[11],xmm0[11],xmm1[12],xmm0[12],xmm1[13],xmm0[13],xmm1[14],xmm0[14],xmm1[15],xmm0[15]
+; AVX1-NEXT:    vpmulhw {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm2, %xmm2 # [26368,47872,11008,20224,37632,35072,33024,30976]
+; AVX1-NEXT:    vpsrlw $8, %xmm2, %xmm2
+; AVX1-NEXT:    vpunpcklbw {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1],xmm1[2],xmm0[2],xmm1[3],xmm0[3],xmm1[4],xmm0[4],xmm1[5],xmm0[5],xmm1[6],xmm0[6],xmm1[7],xmm0[7]
+; AVX1-NEXT:    vpmulhw {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm1, %xmm1 # [33024,22016,33024,26368,11008,37632,33024,14592]
+; AVX1-NEXT:    vpsrlw $8, %xmm1, %xmm1
+; AVX1-NEXT:    vpackuswb %xmm2, %xmm1, %xmm1
+; AVX1-NEXT:    vpand {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0, %xmm0
+; AVX1-NEXT:    vpaddb %xmm0, %xmm1, %xmm0
+; AVX1-NEXT:    vpunpckhbw {{.*#+}} xmm1 = xmm0[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15]
+; AVX1-NEXT:    vpsraw $8, %xmm1, %xmm1
+; AVX1-NEXT:    vpmullw {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm1, %xmm1 # [64,32,128,64,32,32,32,32]
+; AVX1-NEXT:    vpsrlw $8, %xmm1, %xmm1
+; AVX1-NEXT:    vpunpcklbw {{.*#+}} xmm2 = xmm0[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; AVX1-NEXT:    vpsraw $8, %xmm2, %xmm2
+; AVX1-NEXT:    vpmullw {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm2, %xmm2 # [256,256,128,128,256,64,64,128]
+; AVX1-NEXT:    vpsrlw $8, %xmm2, %xmm2
+; AVX1-NEXT:    vpackuswb %xmm1, %xmm2, %xmm1
+; AVX1-NEXT:    vpsrlw $7, %xmm0, %xmm0
+; AVX1-NEXT:    vpaddb %xmm0, %xmm1, %xmm0
+; AVX1-NEXT:    vpand {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0, %xmm0
+; AVX1-NEXT:    retq
+;
+; AVX2NOBW-LABEL: PR143238:
+; AVX2NOBW:       # %bb.0:
+; AVX2NOBW-NEXT:    vpmovsxbw %xmm0, %ymm1
+; AVX2NOBW-NEXT:    vpmullw {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %ymm1, %ymm1 # [65409,86,65409,103,43,65427,65409,57,103,65467,43,79,65427,65417,65409,121]
+; AVX2NOBW-NEXT:    vpsrlw $8, %ymm1, %ymm1
+; AVX2NOBW-NEXT:    vextracti128 $1, %ymm1, %xmm2
+; AVX2NOBW-NEXT:    vpackuswb %xmm2, %xmm1, %xmm1
+; AVX2NOBW-NEXT:    vpand {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0, %xmm0
+; AVX2NOBW-NEXT:    vpaddb %xmm0, %xmm1, %xmm0
+; AVX2NOBW-NEXT:    vpmovsxbw %xmm0, %ymm1
+; AVX2NOBW-NEXT:    vpmullw {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %ymm1, %ymm1 # [256,256,128,128,256,64,64,128,64,32,128,64,32,32,32,32]
+; AVX2NOBW-NEXT:    vpsrlw $8, %ymm1, %ymm1
+; AVX2NOBW-NEXT:    vextracti128 $1, %ymm1, %xmm2
+; AVX2NOBW-NEXT:    vpackuswb %xmm2, %xmm1, %xmm1
+; AVX2NOBW-NEXT:    vpsrlw $7, %xmm0, %xmm0
+; AVX2NOBW-NEXT:    vpaddb %xmm0, %xmm1, %xmm0
+; AVX2NOBW-NEXT:    vpand {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0, %xmm0
+; AVX2NOBW-NEXT:    vzeroupper
+; AVX2NOBW-NEXT:    retq
+;
+; AVX512BW-LABEL: PR143238:
+; AVX512BW:       # %bb.0:
+; AVX512BW-NEXT:    vpmovsxbw {{.*#+}} ymm1 = [0,0,1,1,0,2,2,1,2,3,1,2,3,3,3,3]
+; AVX512BW-NEXT:    vpmovsxbw %xmm0, %ymm2
+; AVX512BW-NEXT:    vpmullw {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %ymm2, %ymm2 # [65409,86,65409,103,43,65427,65409,57,103,65467,43,79,65427,65417,65409,121]
+; AVX512BW-NEXT:    vpsrlw $8, %ymm2, %ymm2
+; AVX512BW-NEXT:    vpmovwb %zmm2, %ymm2
+; AVX512BW-NEXT:    vpand {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0, %xmm0
+; AVX512BW-NEXT:    vpaddb %xmm0, %xmm2, %xmm0
+; AVX512BW-NEXT:    vpmovsxbw %xmm0, %ymm2
+; AVX512BW-NEXT:    vpsrlvw %zmm1, %zmm2, %zmm1
+; AVX512BW-NEXT:    vpmovwb %zmm1, %ymm1
+; AVX512BW-NEXT:    vpsrlw $7, %xmm0, %xmm0
+; AVX512BW-NEXT:    vpaddb %xmm0, %xmm1, %xmm0
+; AVX512BW-NEXT:    vpand {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0, %xmm0
+; AVX512BW-NEXT:    vzeroupper
+; AVX512BW-NEXT:    retq
+  %sdiv = sdiv <16 x i8> %a0, <i8 2, i8 3, i8 4, i8 5, i8 6, i8 7, i8 8, i8 9, i8 10, i8 11, i8 12, i8 13, i8 14, i8 15, i8 16, i8 17>
+  %mask = and <16 x i8> %sdiv, splat (i8 1)
+  ret <16 x i8> %mask
+}