[X86][AVX] getTargetShuffleMask - don't decode VBROADCAST(EXTRACT_SUBVECTOR(X,0)) patterns.

getTargetShuffleMask is used by the various "SimplifyDemanded" folds so we can't assume that the bypassed extract_subvector can be safely simplified - getFauxShuffleMask performs a more general decode that allows us to more safely catch many of these cases so the impact is minimal.

Author: RKSimon

llvm/lib/Target/X86/X86ISelLowering.cpp

@@ -6916,25 +6916,16 @@ static bool getTargetShuffleMask(SDNode *N, MVT VT, bool AllowSentinelZero,
     DecodeZeroMoveLowMask(NumElems, Mask);
     IsUnary = true;
     break;
-  case X86ISD::VBROADCAST: {
-    SDValue N0 = N->getOperand(0);
-    // See if we're broadcasting from index 0 of an EXTRACT_SUBVECTOR. If so,
-    // add the pre-extracted value to the Ops vector.
-    if (N0.getOpcode() == ISD::EXTRACT_SUBVECTOR &&
-        N0.getOperand(0).getValueType() == VT &&
-        N0.getConstantOperandVal(1) == 0)
-      Ops.push_back(N0.getOperand(0));
-
-    // We only decode broadcasts of same-sized vectors, unless the broadcast
-    // came from an extract from the original width. If we found one, we
-    // pushed it the Ops vector above.
-    if (N0.getValueType() == VT || !Ops.empty()) {
+  case X86ISD::VBROADCAST:
+    // We only decode broadcasts of same-sized vectors, peeking through to
+    // extracted subvectors is likely to cause hasOneUse issues with
+    // SimplifyDemandedBits etc.
+    if (N->getOperand(0).getValueType() == VT) {
       DecodeVectorBroadcast(NumElems, Mask);
       IsUnary = true;
       break;
     }
     return false;
-  }
   case X86ISD::VPERMILPV: {
     assert(N->getOperand(0).getValueType() == VT && "Unexpected value type");
     IsUnary = true;

llvm/test/CodeGen/X86/vector-fshl-256.ll

@@ -1092,12 +1092,12 @@ define <8 x i32> @splatvar_funnnel_v8i32(<8 x i32> %x, <8 x i32> %y, <8 x i32> %
 ; AVX2-NEXT:    vpand %xmm3, %xmm2, %xmm2
 ; AVX2-NEXT:    vpmovzxdq {{.*#+}} xmm3 = xmm2[0],zero,xmm2[1],zero
 ; AVX2-NEXT:    vpslld %xmm3, %ymm0, %ymm3
+; AVX2-NEXT:    vpbroadcastd %xmm2, %ymm2
 ; AVX2-NEXT:    vpbroadcastd {{.*#+}} xmm4 = [32,32,32,32]
 ; AVX2-NEXT:    vpsubd %xmm2, %xmm4, %xmm4
 ; AVX2-NEXT:    vpmovzxdq {{.*#+}} xmm4 = xmm4[0],zero,xmm4[1],zero
 ; AVX2-NEXT:    vpsrld %xmm4, %ymm1, %ymm1
 ; AVX2-NEXT:    vpor %ymm1, %ymm3, %ymm1
-; AVX2-NEXT:    vpbroadcastd %xmm2, %ymm2
 ; AVX2-NEXT:    vpxor %xmm3, %xmm3, %xmm3
 ; AVX2-NEXT:    vpcmpeqd %ymm3, %ymm2, %ymm2
 ; AVX2-NEXT:    vblendvps %ymm2, %ymm0, %ymm1, %ymm0
@@ -1110,12 +1110,12 @@ define <8 x i32> @splatvar_funnnel_v8i32(<8 x i32> %x, <8 x i32> %y, <8 x i32> %
 ; AVX512F-NEXT:    vpand %xmm3, %xmm2, %xmm2
 ; AVX512F-NEXT:    vpmovzxdq {{.*#+}} xmm3 = xmm2[0],zero,xmm2[1],zero
 ; AVX512F-NEXT:    vpslld %xmm3, %ymm0, %ymm3
+; AVX512F-NEXT:    vpbroadcastd %xmm2, %ymm2
 ; AVX512F-NEXT:    vpbroadcastd {{.*#+}} xmm4 = [32,32,32,32]
 ; AVX512F-NEXT:    vpsubd %xmm2, %xmm4, %xmm4
 ; AVX512F-NEXT:    vpmovzxdq {{.*#+}} xmm4 = xmm4[0],zero,xmm4[1],zero
 ; AVX512F-NEXT:    vpsrld %xmm4, %ymm1, %ymm1
 ; AVX512F-NEXT:    vpor %ymm1, %ymm3, %ymm1
-; AVX512F-NEXT:    vpbroadcastd %xmm2, %ymm2
 ; AVX512F-NEXT:    vptestnmd %zmm2, %zmm2, %k1
 ; AVX512F-NEXT:    vmovdqa32 %zmm0, %zmm1 {%k1}
 ; AVX512F-NEXT:    vmovdqa %ymm1, %ymm0
@@ -1126,12 +1126,12 @@ define <8 x i32> @splatvar_funnnel_v8i32(<8 x i32> %x, <8 x i32> %y, <8 x i32> %
 ; AVX512VL-NEXT:    vpandd {{.*}}(%rip){1to4}, %xmm2, %xmm2
 ; AVX512VL-NEXT:    vpmovzxdq {{.*#+}} xmm3 = xmm2[0],zero,xmm2[1],zero
 ; AVX512VL-NEXT:    vpslld %xmm3, %ymm0, %ymm3
+; AVX512VL-NEXT:    vpbroadcastd %xmm2, %ymm2
 ; AVX512VL-NEXT:    vpbroadcastd {{.*#+}} xmm4 = [32,32,32,32]
 ; AVX512VL-NEXT:    vpsubd %xmm2, %xmm4, %xmm4
 ; AVX512VL-NEXT:    vpmovzxdq {{.*#+}} xmm4 = xmm4[0],zero,xmm4[1],zero
 ; AVX512VL-NEXT:    vpsrld %xmm4, %ymm1, %ymm1
 ; AVX512VL-NEXT:    vpor %ymm1, %ymm3, %ymm1
-; AVX512VL-NEXT:    vpbroadcastd %xmm2, %ymm2
 ; AVX512VL-NEXT:    vptestnmd %ymm2, %ymm2, %k1
 ; AVX512VL-NEXT:    vmovdqa32 %ymm0, %ymm1 {%k1}
 ; AVX512VL-NEXT:    vmovdqa %ymm1, %ymm0
@@ -1144,12 +1144,12 @@ define <8 x i32> @splatvar_funnnel_v8i32(<8 x i32> %x, <8 x i32> %y, <8 x i32> %
 ; AVX512BW-NEXT:    vpand %xmm3, %xmm2, %xmm2
 ; AVX512BW-NEXT:    vpmovzxdq {{.*#+}} xmm3 = xmm2[0],zero,xmm2[1],zero
 ; AVX512BW-NEXT:    vpslld %xmm3, %ymm0, %ymm3
+; AVX512BW-NEXT:    vpbroadcastd %xmm2, %ymm2
 ; AVX512BW-NEXT:    vpbroadcastd {{.*#+}} xmm4 = [32,32,32,32]
 ; AVX512BW-NEXT:    vpsubd %xmm2, %xmm4, %xmm4
 ; AVX512BW-NEXT:    vpmovzxdq {{.*#+}} xmm4 = xmm4[0],zero,xmm4[1],zero
 ; AVX512BW-NEXT:    vpsrld %xmm4, %ymm1, %ymm1
 ; AVX512BW-NEXT:    vpor %ymm1, %ymm3, %ymm1
-; AVX512BW-NEXT:    vpbroadcastd %xmm2, %ymm2
 ; AVX512BW-NEXT:    vptestnmd %zmm2, %zmm2, %k1
 ; AVX512BW-NEXT:    vmovdqa32 %zmm0, %zmm1 {%k1}
 ; AVX512BW-NEXT:    vmovdqa %ymm1, %ymm0
@@ -1162,12 +1162,12 @@ define <8 x i32> @splatvar_funnnel_v8i32(<8 x i32> %x, <8 x i32> %y, <8 x i32> %
 ; AVX512VBMI2-NEXT:    vpand %xmm3, %xmm2, %xmm2
 ; AVX512VBMI2-NEXT:    vpmovzxdq {{.*#+}} xmm3 = xmm2[0],zero,xmm2[1],zero
 ; AVX512VBMI2-NEXT:    vpslld %xmm3, %ymm0, %ymm3
+; AVX512VBMI2-NEXT:    vpbroadcastd %xmm2, %ymm2
 ; AVX512VBMI2-NEXT:    vpbroadcastd {{.*#+}} xmm4 = [32,32,32,32]
 ; AVX512VBMI2-NEXT:    vpsubd %xmm2, %xmm4, %xmm4
 ; AVX512VBMI2-NEXT:    vpmovzxdq {{.*#+}} xmm4 = xmm4[0],zero,xmm4[1],zero
 ; AVX512VBMI2-NEXT:    vpsrld %xmm4, %ymm1, %ymm1
 ; AVX512VBMI2-NEXT:    vpor %ymm1, %ymm3, %ymm1
-; AVX512VBMI2-NEXT:    vpbroadcastd %xmm2, %ymm2
 ; AVX512VBMI2-NEXT:    vptestnmd %zmm2, %zmm2, %k1
 ; AVX512VBMI2-NEXT:    vmovdqa32 %zmm0, %zmm1 {%k1}
 ; AVX512VBMI2-NEXT:    vmovdqa %ymm1, %ymm0
@@ -1178,12 +1178,12 @@ define <8 x i32> @splatvar_funnnel_v8i32(<8 x i32> %x, <8 x i32> %y, <8 x i32> %
 ; AVX512VLBW-NEXT:    vpandd {{.*}}(%rip){1to4}, %xmm2, %xmm2
 ; AVX512VLBW-NEXT:    vpmovzxdq {{.*#+}} xmm3 = xmm2[0],zero,xmm2[1],zero
 ; AVX512VLBW-NEXT:    vpslld %xmm3, %ymm0, %ymm3
+; AVX512VLBW-NEXT:    vpbroadcastd %xmm2, %ymm2
 ; AVX512VLBW-NEXT:    vpbroadcastd {{.*#+}} xmm4 = [32,32,32,32]
 ; AVX512VLBW-NEXT:    vpsubd %xmm2, %xmm4, %xmm4
 ; AVX512VLBW-NEXT:    vpmovzxdq {{.*#+}} xmm4 = xmm4[0],zero,xmm4[1],zero
 ; AVX512VLBW-NEXT:    vpsrld %xmm4, %ymm1, %ymm1
 ; AVX512VLBW-NEXT:    vpor %ymm1, %ymm3, %ymm1
-; AVX512VLBW-NEXT:    vpbroadcastd %xmm2, %ymm2
 ; AVX512VLBW-NEXT:    vptestnmd %ymm2, %ymm2, %k1
 ; AVX512VLBW-NEXT:    vmovdqa32 %ymm0, %ymm1 {%k1}
 ; AVX512VLBW-NEXT:    vmovdqa %ymm1, %ymm0
@@ -1224,12 +1224,12 @@ define <8 x i32> @splatvar_funnnel_v8i32(<8 x i32> %x, <8 x i32> %y, <8 x i32> %
 ; XOPAVX2-NEXT:    vpand %xmm3, %xmm2, %xmm2
 ; XOPAVX2-NEXT:    vpmovzxdq {{.*#+}} xmm3 = xmm2[0],zero,xmm2[1],zero
 ; XOPAVX2-NEXT:    vpslld %xmm3, %ymm0, %ymm3
+; XOPAVX2-NEXT:    vpbroadcastd %xmm2, %ymm2
 ; XOPAVX2-NEXT:    vpbroadcastd {{.*#+}} xmm4 = [32,32,32,32]
 ; XOPAVX2-NEXT:    vpsubd %xmm2, %xmm4, %xmm4
 ; XOPAVX2-NEXT:    vpmovzxdq {{.*#+}} xmm4 = xmm4[0],zero,xmm4[1],zero
 ; XOPAVX2-NEXT:    vpsrld %xmm4, %ymm1, %ymm1
 ; XOPAVX2-NEXT:    vpor %ymm1, %ymm3, %ymm1
-; XOPAVX2-NEXT:    vpbroadcastd %xmm2, %ymm2
 ; XOPAVX2-NEXT:    vpxor %xmm3, %xmm3, %xmm3
 ; XOPAVX2-NEXT:    vpcmpeqd %ymm3, %ymm2, %ymm2
 ; XOPAVX2-NEXT:    vblendvps %ymm2, %ymm0, %ymm1, %ymm0
@@ -1271,12 +1271,12 @@ define <16 x i16> @splatvar_funnnel_v16i16(<16 x i16> %x, <16 x i16> %y, <16 x i
 ; AVX2-NEXT:    vpand {{.*}}(%rip), %xmm2, %xmm2
 ; AVX2-NEXT:    vpmovzxwq {{.*#+}} xmm3 = xmm2[0],zero,zero,zero,xmm2[1],zero,zero,zero
 ; AVX2-NEXT:    vpsllw %xmm3, %ymm0, %ymm3
+; AVX2-NEXT:    vpbroadcastw %xmm2, %ymm2
 ; AVX2-NEXT:    vmovdqa {{.*#+}} xmm4 = [16,16,16,16,16,16,16,16]
 ; AVX2-NEXT:    vpsubw %xmm2, %xmm4, %xmm4
 ; AVX2-NEXT:    vpmovzxwq {{.*#+}} xmm4 = xmm4[0],zero,zero,zero,xmm4[1],zero,zero,zero
 ; AVX2-NEXT:    vpsrlw %xmm4, %ymm1, %ymm1
 ; AVX2-NEXT:    vpor %ymm1, %ymm3, %ymm1
-; AVX2-NEXT:    vpbroadcastw %xmm2, %ymm2
 ; AVX2-NEXT:    vpxor %xmm3, %xmm3, %xmm3
 ; AVX2-NEXT:    vpcmpeqw %ymm3, %ymm2, %ymm2
 ; AVX2-NEXT:    vpblendvb %ymm2, %ymm0, %ymm1, %ymm0
@@ -1287,12 +1287,12 @@ define <16 x i16> @splatvar_funnnel_v16i16(<16 x i16> %x, <16 x i16> %y, <16 x i
 ; AVX512F-NEXT:    vpand {{.*}}(%rip), %xmm2, %xmm2
 ; AVX512F-NEXT:    vpmovzxwq {{.*#+}} xmm3 = xmm2[0],zero,zero,zero,xmm2[1],zero,zero,zero
 ; AVX512F-NEXT:    vpsllw %xmm3, %ymm0, %ymm3
+; AVX512F-NEXT:    vpbroadcastw %xmm2, %ymm2
 ; AVX512F-NEXT:    vmovdqa {{.*#+}} xmm4 = [16,16,16,16,16,16,16,16]
 ; AVX512F-NEXT:    vpsubw %xmm2, %xmm4, %xmm4
 ; AVX512F-NEXT:    vpmovzxwq {{.*#+}} xmm4 = xmm4[0],zero,zero,zero,xmm4[1],zero,zero,zero
 ; AVX512F-NEXT:    vpsrlw %xmm4, %ymm1, %ymm1
 ; AVX512F-NEXT:    vpor %ymm1, %ymm3, %ymm1
-; AVX512F-NEXT:    vpbroadcastw %xmm2, %ymm2
 ; AVX512F-NEXT:    vpxor %xmm3, %xmm3, %xmm3
 ; AVX512F-NEXT:    vpcmpeqw %ymm3, %ymm2, %ymm2
 ; AVX512F-NEXT:    vpblendvb %ymm2, %ymm0, %ymm1, %ymm0
@@ -1303,12 +1303,12 @@ define <16 x i16> @splatvar_funnnel_v16i16(<16 x i16> %x, <16 x i16> %y, <16 x i
 ; AVX512VL-NEXT:    vpand {{.*}}(%rip), %xmm2, %xmm2
 ; AVX512VL-NEXT:    vpmovzxwq {{.*#+}} xmm3 = xmm2[0],zero,zero,zero,xmm2[1],zero,zero,zero
 ; AVX512VL-NEXT:    vpsllw %xmm3, %ymm0, %ymm3
+; AVX512VL-NEXT:    vpbroadcastw %xmm2, %ymm2
 ; AVX512VL-NEXT:    vmovdqa {{.*#+}} xmm4 = [16,16,16,16,16,16,16,16]
 ; AVX512VL-NEXT:    vpsubw %xmm2, %xmm4, %xmm4
 ; AVX512VL-NEXT:    vpmovzxwq {{.*#+}} xmm4 = xmm4[0],zero,zero,zero,xmm4[1],zero,zero,zero
 ; AVX512VL-NEXT:    vpsrlw %xmm4, %ymm1, %ymm1
 ; AVX512VL-NEXT:    vpor %ymm1, %ymm3, %ymm1
-; AVX512VL-NEXT:    vpbroadcastw %xmm2, %ymm2
 ; AVX512VL-NEXT:    vpxor %xmm3, %xmm3, %xmm3
 ; AVX512VL-NEXT:    vpcmpeqw %ymm3, %ymm2, %ymm2
 ; AVX512VL-NEXT:    vpblendvb %ymm2, %ymm0, %ymm1, %ymm0
@@ -1320,12 +1320,12 @@ define <16 x i16> @splatvar_funnnel_v16i16(<16 x i16> %x, <16 x i16> %y, <16 x i
 ; AVX512BW-NEXT:    vpand {{.*}}(%rip), %xmm2, %xmm2
 ; AVX512BW-NEXT:    vpmovzxwq {{.*#+}} xmm3 = xmm2[0],zero,zero,zero,xmm2[1],zero,zero,zero
 ; AVX512BW-NEXT:    vpsllw %xmm3, %ymm0, %ymm3
+; AVX512BW-NEXT:    vpbroadcastw %xmm2, %ymm2
 ; AVX512BW-NEXT:    vmovdqa {{.*#+}} xmm4 = [16,16,16,16,16,16,16,16]
 ; AVX512BW-NEXT:    vpsubw %xmm2, %xmm4, %xmm4
 ; AVX512BW-NEXT:    vpmovzxwq {{.*#+}} xmm4 = xmm4[0],zero,zero,zero,xmm4[1],zero,zero,zero
 ; AVX512BW-NEXT:    vpsrlw %xmm4, %ymm1, %ymm1
 ; AVX512BW-NEXT:    vpor %ymm1, %ymm3, %ymm1
-; AVX512BW-NEXT:    vpbroadcastw %xmm2, %ymm2
 ; AVX512BW-NEXT:    vptestnmw %zmm2, %zmm2, %k1
 ; AVX512BW-NEXT:    vmovdqu16 %zmm0, %zmm1 {%k1}
 ; AVX512BW-NEXT:    vmovdqa %ymm1, %ymm0
@@ -1337,12 +1337,12 @@ define <16 x i16> @splatvar_funnnel_v16i16(<16 x i16> %x, <16 x i16> %y, <16 x i
 ; AVX512VBMI2-NEXT:    vpand {{.*}}(%rip), %xmm2, %xmm2
 ; AVX512VBMI2-NEXT:    vpmovzxwq {{.*#+}} xmm3 = xmm2[0],zero,zero,zero,xmm2[1],zero,zero,zero
 ; AVX512VBMI2-NEXT:    vpsllw %xmm3, %ymm0, %ymm3
+; AVX512VBMI2-NEXT:    vpbroadcastw %xmm2, %ymm2
 ; AVX512VBMI2-NEXT:    vmovdqa {{.*#+}} xmm4 = [16,16,16,16,16,16,16,16]
 ; AVX512VBMI2-NEXT:    vpsubw %xmm2, %xmm4, %xmm4
 ; AVX512VBMI2-NEXT:    vpmovzxwq {{.*#+}} xmm4 = xmm4[0],zero,zero,zero,xmm4[1],zero,zero,zero
 ; AVX512VBMI2-NEXT:    vpsrlw %xmm4, %ymm1, %ymm1
 ; AVX512VBMI2-NEXT:    vpor %ymm1, %ymm3, %ymm1
-; AVX512VBMI2-NEXT:    vpbroadcastw %xmm2, %ymm2
 ; AVX512VBMI2-NEXT:    vptestnmw %zmm2, %zmm2, %k1
 ; AVX512VBMI2-NEXT:    vmovdqu16 %zmm0, %zmm1 {%k1}
 ; AVX512VBMI2-NEXT:    vmovdqa %ymm1, %ymm0
@@ -1353,12 +1353,12 @@ define <16 x i16> @splatvar_funnnel_v16i16(<16 x i16> %x, <16 x i16> %y, <16 x i
 ; AVX512VLBW-NEXT:    vpand {{.*}}(%rip), %xmm2, %xmm2
 ; AVX512VLBW-NEXT:    vpmovzxwq {{.*#+}} xmm3 = xmm2[0],zero,zero,zero,xmm2[1],zero,zero,zero
 ; AVX512VLBW-NEXT:    vpsllw %xmm3, %ymm0, %ymm3
+; AVX512VLBW-NEXT:    vpbroadcastw %xmm2, %ymm2
 ; AVX512VLBW-NEXT:    vmovdqa {{.*#+}} xmm4 = [16,16,16,16,16,16,16,16]
 ; AVX512VLBW-NEXT:    vpsubw %xmm2, %xmm4, %xmm4
 ; AVX512VLBW-NEXT:    vpmovzxwq {{.*#+}} xmm4 = xmm4[0],zero,zero,zero,xmm4[1],zero,zero,zero
 ; AVX512VLBW-NEXT:    vpsrlw %xmm4, %ymm1, %ymm1
 ; AVX512VLBW-NEXT:    vpor %ymm1, %ymm3, %ymm1
-; AVX512VLBW-NEXT:    vpbroadcastw %xmm2, %ymm2
 ; AVX512VLBW-NEXT:    vptestnmw %ymm2, %ymm2, %k1
 ; AVX512VLBW-NEXT:    vmovdqu16 %ymm0, %ymm1 {%k1}
 ; AVX512VLBW-NEXT:    vmovdqa %ymm1, %ymm0
@@ -1399,12 +1399,12 @@ define <16 x i16> @splatvar_funnnel_v16i16(<16 x i16> %x, <16 x i16> %y, <16 x i
 ; XOPAVX2-NEXT:    vpand {{.*}}(%rip), %xmm2, %xmm2
 ; XOPAVX2-NEXT:    vpmovzxwq {{.*#+}} xmm3 = xmm2[0],zero,zero,zero,xmm2[1],zero,zero,zero
 ; XOPAVX2-NEXT:    vpsllw %xmm3, %ymm0, %ymm3
+; XOPAVX2-NEXT:    vpbroadcastw %xmm2, %ymm2
 ; XOPAVX2-NEXT:    vmovdqa {{.*#+}} xmm4 = [16,16,16,16,16,16,16,16]
 ; XOPAVX2-NEXT:    vpsubw %xmm2, %xmm4, %xmm4
 ; XOPAVX2-NEXT:    vpmovzxwq {{.*#+}} xmm4 = xmm4[0],zero,zero,zero,xmm4[1],zero,zero,zero
 ; XOPAVX2-NEXT:    vpsrlw %xmm4, %ymm1, %ymm1
 ; XOPAVX2-NEXT:    vpor %ymm1, %ymm3, %ymm1
-; XOPAVX2-NEXT:    vpbroadcastw %xmm2, %ymm2
 ; XOPAVX2-NEXT:    vpxor %xmm3, %xmm3, %xmm3
 ; XOPAVX2-NEXT:    vpcmpeqw %ymm3, %ymm2, %ymm2
 ; XOPAVX2-NEXT:    vpblendvb %ymm2, %ymm0, %ymm1, %ymm0
@@ -1458,6 +1458,7 @@ define <32 x i8> @splatvar_funnnel_v32i8(<32 x i8> %x, <32 x i8> %y, <32 x i8> %
 ; AVX2-NEXT:    vpsllw %xmm3, %xmm5, %xmm3
 ; AVX2-NEXT:    vpbroadcastb %xmm3, %ymm3
 ; AVX2-NEXT:    vpand %ymm3, %ymm4, %ymm3
+; AVX2-NEXT:    vpbroadcastb %xmm2, %ymm2
 ; AVX2-NEXT:    vmovdqa {{.*#+}} xmm4 = [8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8]
 ; AVX2-NEXT:    vpsubb %xmm2, %xmm4, %xmm4
 ; AVX2-NEXT:    vpmovzxbq {{.*#+}} xmm4 = xmm4[0],zero,zero,zero,zero,zero,zero,zero,xmm4[1],zero,zero,zero,zero,zero,zero,zero
@@ -1467,7 +1468,6 @@ define <32 x i8> @splatvar_funnnel_v32i8(<32 x i8> %x, <32 x i8> %y, <32 x i8> %
 ; AVX2-NEXT:    vpbroadcastb %xmm4, %ymm4
 ; AVX2-NEXT:    vpand %ymm4, %ymm1, %ymm1
 ; AVX2-NEXT:    vpor %ymm1, %ymm3, %ymm1
-; AVX2-NEXT:    vpbroadcastb %xmm2, %ymm2
 ; AVX2-NEXT:    vpxor %xmm3, %xmm3, %xmm3
 ; AVX2-NEXT:    vpcmpeqb %ymm3, %ymm2, %ymm2
 ; AVX2-NEXT:    vpblendvb %ymm2, %ymm0, %ymm1, %ymm0
@@ -1482,6 +1482,7 @@ define <32 x i8> @splatvar_funnnel_v32i8(<32 x i8> %x, <32 x i8> %y, <32 x i8> %
 ; AVX512F-NEXT:    vpsllw %xmm3, %xmm5, %xmm3
 ; AVX512F-NEXT:    vpbroadcastb %xmm3, %ymm3
 ; AVX512F-NEXT:    vpand %ymm3, %ymm4, %ymm3
+; AVX512F-NEXT:    vpbroadcastb %xmm2, %ymm2
 ; AVX512F-NEXT:    vmovdqa {{.*#+}} xmm4 = [8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8]
 ; AVX512F-NEXT:    vpsubb %xmm2, %xmm4, %xmm4
 ; AVX512F-NEXT:    vpmovzxbq {{.*#+}} xmm4 = xmm4[0],zero,zero,zero,zero,zero,zero,zero,xmm4[1],zero,zero,zero,zero,zero,zero,zero
@@ -1491,7 +1492,6 @@ define <32 x i8> @splatvar_funnnel_v32i8(<32 x i8> %x, <32 x i8> %y, <32 x i8> %
 ; AVX512F-NEXT:    vpbroadcastb %xmm4, %ymm4
 ; AVX512F-NEXT:    vpand %ymm4, %ymm1, %ymm1
 ; AVX512F-NEXT:    vpor %ymm1, %ymm3, %ymm1
-; AVX512F-NEXT:    vpbroadcastb %xmm2, %ymm2
 ; AVX512F-NEXT:    vpxor %xmm3, %xmm3, %xmm3
 ; AVX512F-NEXT:    vpcmpeqb %ymm3, %ymm2, %ymm2
 ; AVX512F-NEXT:    vpblendvb %ymm2, %ymm0, %ymm1, %ymm0
@@ -1506,6 +1506,7 @@ define <32 x i8> @splatvar_funnnel_v32i8(<32 x i8> %x, <32 x i8> %y, <32 x i8> %
 ; AVX512VL-NEXT:    vpsllw %xmm3, %xmm5, %xmm3
 ; AVX512VL-NEXT:    vpbroadcastb %xmm3, %ymm3
 ; AVX512VL-NEXT:    vpand %ymm3, %ymm4, %ymm3
+; AVX512VL-NEXT:    vpbroadcastb %xmm2, %ymm2
 ; AVX512VL-NEXT:    vmovdqa {{.*#+}} xmm4 = [8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8]
 ; AVX512VL-NEXT:    vpsubb %xmm2, %xmm4, %xmm4
 ; AVX512VL-NEXT:    vpmovzxbq {{.*#+}} xmm4 = xmm4[0],zero,zero,zero,zero,zero,zero,zero,xmm4[1],zero,zero,zero,zero,zero,zero,zero
@@ -1514,9 +1515,8 @@ define <32 x i8> @splatvar_funnnel_v32i8(<32 x i8> %x, <32 x i8> %y, <32 x i8> %
 ; AVX512VL-NEXT:    vpsrlw $8, %xmm4, %xmm4
 ; AVX512VL-NEXT:    vpbroadcastb %xmm4, %ymm4
 ; AVX512VL-NEXT:    vpternlogq $236, %ymm1, %ymm3, %ymm4
-; AVX512VL-NEXT:    vpbroadcastb %xmm2, %ymm1
-; AVX512VL-NEXT:    vpxor %xmm2, %xmm2, %xmm2
-; AVX512VL-NEXT:    vpcmpeqb %ymm2, %ymm1, %ymm1
+; AVX512VL-NEXT:    vpxor %xmm1, %xmm1, %xmm1
+; AVX512VL-NEXT:    vpcmpeqb %ymm1, %ymm2, %ymm1
 ; AVX512VL-NEXT:    vpblendvb %ymm1, %ymm0, %ymm4, %ymm0
 ; AVX512VL-NEXT:    retq
 ;