[llvm] e3bd011 - [X86][SSE] Combine combineLogicBlendIntoConditionalNegate for VSELECT nodes (PR43660)

Simon Pilgrim via llvm-commits llvm-commits at lists.llvm.org
Sun Jan 5 10:54:05 PST 2020


Author: Simon Pilgrim
Date: 2020-01-05T18:50:44Z
New Revision: e3bd0118903ccb545ca5cb2641646eb66ae8180e

URL: https://github.com/llvm/llvm-project/commit/e3bd0118903ccb545ca5cb2641646eb66ae8180e
DIFF: https://github.com/llvm/llvm-project/commit/e3bd0118903ccb545ca5cb2641646eb66ae8180e.diff

LOG: [X86][SSE] Combine combineLogicBlendIntoConditionalNegate for VSELECT nodes (PR43660)

Attempt to use combineLogicBlendIntoConditionalNegate for (select M, (sub 0, X), X) -> (sub (xor X, M), M)

We limit this to cases that can't easily replace the VSELECT with a shuffle (non-constant masks) or where a BLENDV is likely to occur (which tends to result in slower codegen).

Added: 
    

Modified: 
    llvm/lib/Target/X86/X86ISelLowering.cpp
    llvm/test/CodeGen/X86/combine-sdiv.ll
    llvm/test/CodeGen/X86/vector-blend.ll

Removed: 
    


################################################################################
diff  --git a/llvm/lib/Target/X86/X86ISelLowering.cpp b/llvm/lib/Target/X86/X86ISelLowering.cpp
index ba29d9e513d0..e41717c332a1 100644
--- a/llvm/lib/Target/X86/X86ISelLowering.cpp
+++ b/llvm/lib/Target/X86/X86ISelLowering.cpp
@@ -37697,10 +37697,21 @@ static SDValue combineSelect(SDNode *N, SelectionDAG &DAG,
   EVT VT = LHS.getValueType();
   EVT CondVT = Cond.getValueType();
   const TargetLowering &TLI = DAG.getTargetLoweringInfo();
+  bool CondConstantVector = ISD::isBuildVectorOfConstantSDNodes(Cond.getNode());
+
+  // Attempt to combine (select M, (sub 0, X), X) -> (sub (xor X, M), M).
+  // Limit this to cases of non-constant masks that createShuffleMaskFromVSELECT
+  // can't catch, plus vXi8 cases where we'd likely end up with BLENDV.
+  if (CondVT.isVector() && CondVT.isInteger() &&
+      CondVT.getScalarSizeInBits() == VT.getScalarSizeInBits() &&
+      (!CondConstantVector || CondVT.getScalarType() == MVT::i8) &&
+      DAG.ComputeNumSignBits(Cond) == CondVT.getScalarSizeInBits())
+    if (SDValue V = combineLogicBlendIntoConditionalNegate(VT, Cond, RHS, LHS,
+                                                           DL, DAG, Subtarget))
+      return V;
 
   // Convert vselects with constant condition into shuffles.
-  if (ISD::isBuildVectorOfConstantSDNodes(Cond.getNode()) &&
-      DCI.isBeforeLegalizeOps()) {
+  if (CondConstantVector && DCI.isBeforeLegalizeOps()) {
     SmallVector<int, 64> Mask;
     if (createShuffleMaskFromVSELECT(Mask, Cond))
       return DAG.getVectorShuffle(VT, DL, LHS, RHS, Mask);

diff  --git a/llvm/test/CodeGen/X86/combine-sdiv.ll b/llvm/test/CodeGen/X86/combine-sdiv.ll
index 137e03228c0f..235a6f0f3342 100644
--- a/llvm/test/CodeGen/X86/combine-sdiv.ll
+++ b/llvm/test/CodeGen/X86/combine-sdiv.ll
@@ -2104,46 +2104,32 @@ define <4 x i32> @combine_vec_sdiv_by_pow2b_undef3(<4 x i32> %x) {
 
 ; PR37119
 define <16 x i8> @non_splat_minus_one_divisor_0(<16 x i8> %A) {
-; SSE2-LABEL: non_splat_minus_one_divisor_0:
-; SSE2:       # %bb.0:
-; SSE2-NEXT:    movdqa {{.*#+}} xmm1 = [0,0,255,0,0,0,255,0,0,255,255,255,255,255,255,255]
-; SSE2-NEXT:    pxor %xmm1, %xmm0
-; SSE2-NEXT:    psubb %xmm0, %xmm1
-; SSE2-NEXT:    movdqa %xmm1, %xmm0
-; SSE2-NEXT:    retq
-;
-; SSE41-LABEL: non_splat_minus_one_divisor_0:
-; SSE41:       # %bb.0:
-; SSE41-NEXT:    movdqa %xmm0, %xmm1
-; SSE41-NEXT:    pxor %xmm2, %xmm2
-; SSE41-NEXT:    psubb %xmm0, %xmm2
-; SSE41-NEXT:    movaps {{.*#+}} xmm0 = [0,0,255,0,0,0,255,0,0,255,255,255,255,255,255,255]
-; SSE41-NEXT:    pblendvb %xmm0, %xmm1, %xmm2
-; SSE41-NEXT:    movdqa %xmm2, %xmm0
-; SSE41-NEXT:    retq
+; SSE-LABEL: non_splat_minus_one_divisor_0:
+; SSE:       # %bb.0:
+; SSE-NEXT:    movdqa {{.*#+}} xmm1 = [255,255,0,255,255,255,0,255,255,0,0,0,0,0,0,0]
+; SSE-NEXT:    pxor %xmm1, %xmm0
+; SSE-NEXT:    psubb %xmm1, %xmm0
+; SSE-NEXT:    retq
 ;
 ; AVX1-LABEL: non_splat_minus_one_divisor_0:
 ; AVX1:       # %bb.0:
-; AVX1-NEXT:    vpxor %xmm1, %xmm1, %xmm1
-; AVX1-NEXT:    vpsubb %xmm0, %xmm1, %xmm1
-; AVX1-NEXT:    vmovdqa {{.*#+}} xmm2 = [0,0,255,0,0,0,255,0,0,255,255,255,255,255,255,255]
-; AVX1-NEXT:    vpblendvb %xmm2, %xmm0, %xmm1, %xmm0
+; AVX1-NEXT:    vmovdqa {{.*#+}} xmm1 = [255,255,0,255,255,255,0,255,255,0,0,0,0,0,0,0]
+; AVX1-NEXT:    vpxor %xmm1, %xmm0, %xmm0
+; AVX1-NEXT:    vpsubb %xmm1, %xmm0, %xmm0
 ; AVX1-NEXT:    retq
 ;
 ; AVX2-LABEL: non_splat_minus_one_divisor_0:
 ; AVX2:       # %bb.0:
-; AVX2-NEXT:    vpxor %xmm1, %xmm1, %xmm1
-; AVX2-NEXT:    vpsubb %xmm0, %xmm1, %xmm1
-; AVX2-NEXT:    vmovdqa {{.*#+}} xmm2 = [0,0,255,0,0,0,255,0,0,255,255,255,255,255,255,255]
-; AVX2-NEXT:    vpblendvb %xmm2, %xmm0, %xmm1, %xmm0
+; AVX2-NEXT:    vmovdqa {{.*#+}} xmm1 = [255,255,0,255,255,255,0,255,255,0,0,0,0,0,0,0]
+; AVX2-NEXT:    vpxor %xmm1, %xmm0, %xmm0
+; AVX2-NEXT:    vpsubb %xmm1, %xmm0, %xmm0
 ; AVX2-NEXT:    retq
 ;
 ; AVX512F-LABEL: non_splat_minus_one_divisor_0:
 ; AVX512F:       # %bb.0:
-; AVX512F-NEXT:    vpxor %xmm1, %xmm1, %xmm1
-; AVX512F-NEXT:    vpsubb %xmm0, %xmm1, %xmm1
-; AVX512F-NEXT:    vmovdqa {{.*#+}} xmm2 = [0,0,255,0,0,0,255,0,0,255,255,255,255,255,255,255]
-; AVX512F-NEXT:    vpblendvb %xmm2, %xmm0, %xmm1, %xmm0
+; AVX512F-NEXT:    vmovdqa {{.*#+}} xmm1 = [255,255,0,255,255,255,0,255,255,0,0,0,0,0,0,0]
+; AVX512F-NEXT:    vpxor %xmm1, %xmm0, %xmm0
+; AVX512F-NEXT:    vpsubb %xmm1, %xmm0, %xmm0
 ; AVX512F-NEXT:    retq
 ;
 ; AVX512BW-LABEL: non_splat_minus_one_divisor_0:
@@ -2156,10 +2142,9 @@ define <16 x i8> @non_splat_minus_one_divisor_0(<16 x i8> %A) {
 ;
 ; XOP-LABEL: non_splat_minus_one_divisor_0:
 ; XOP:       # %bb.0:
-; XOP-NEXT:    vpxor %xmm1, %xmm1, %xmm1
-; XOP-NEXT:    vpsubb %xmm0, %xmm1, %xmm1
-; XOP-NEXT:    vmovdqa {{.*#+}} xmm2 = [0,0,255,0,0,0,255,0,0,255,255,255,255,255,255,255]
-; XOP-NEXT:    vpblendvb %xmm2, %xmm0, %xmm1, %xmm0
+; XOP-NEXT:    vmovdqa {{.*#+}} xmm1 = [255,255,0,255,255,255,0,255,255,0,0,0,0,0,0,0]
+; XOP-NEXT:    vpxor %xmm1, %xmm0, %xmm0
+; XOP-NEXT:    vpsubb %xmm1, %xmm0, %xmm0
 ; XOP-NEXT:    retq
   %div = sdiv <16 x i8> %A, <i8 -1, i8 -1, i8 1, i8 -1, i8 -1, i8 -1, i8 1, i8 -1, i8 -1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1>
   ret <16 x i8> %div
@@ -2203,38 +2188,38 @@ define <16 x i8> @non_splat_minus_one_divisor_1(<16 x i8> %A) {
 ; SSE41-LABEL: non_splat_minus_one_divisor_1:
 ; SSE41:       # %bb.0:
 ; SSE41-NEXT:    movdqa %xmm0, %xmm1
-; SSE41-NEXT:    pxor %xmm2, %xmm2
 ; SSE41-NEXT:    pxor %xmm0, %xmm0
-; SSE41-NEXT:    pcmpgtb %xmm1, %xmm0
+; SSE41-NEXT:    pxor %xmm3, %xmm3
+; SSE41-NEXT:    pcmpgtb %xmm1, %xmm3
 ; SSE41-NEXT:    pxor %xmm4, %xmm4
-; SSE41-NEXT:    punpcklbw {{.*#+}} xmm4 = xmm4[0],xmm0[0],xmm4[1],xmm0[1],xmm4[2],xmm0[2],xmm4[3],xmm0[3],xmm4[4],xmm0[4],xmm4[5],xmm0[5],xmm4[6],xmm0[6],xmm4[7],xmm0[7]
-; SSE41-NEXT:    pmovzxbw {{.*#+}} xmm3 = xmm0[0],zero,xmm0[1],zero,xmm0[2],zero,xmm0[3],zero,xmm0[4],zero,xmm0[5],zero,xmm0[6],zero,xmm0[7],zero
-; SSE41-NEXT:    psllw $1, %xmm3
-; SSE41-NEXT:    pblendw {{.*#+}} xmm3 = xmm4[0,1],xmm3[2],xmm4[3,4,5],xmm3[6],xmm4[7]
+; SSE41-NEXT:    punpcklbw {{.*#+}} xmm4 = xmm4[0],xmm3[0],xmm4[1],xmm3[1],xmm4[2],xmm3[2],xmm4[3],xmm3[3],xmm4[4],xmm3[4],xmm4[5],xmm3[5],xmm4[6],xmm3[6],xmm4[7],xmm3[7]
+; SSE41-NEXT:    pmovzxbw {{.*#+}} xmm2 = xmm3[0],zero,xmm3[1],zero,xmm3[2],zero,xmm3[3],zero,xmm3[4],zero,xmm3[5],zero,xmm3[6],zero,xmm3[7],zero
+; SSE41-NEXT:    psllw $1, %xmm2
+; SSE41-NEXT:    pblendw {{.*#+}} xmm2 = xmm4[0,1],xmm2[2],xmm4[3,4,5],xmm2[6],xmm4[7]
+; SSE41-NEXT:    psrlw $8, %xmm2
+; SSE41-NEXT:    punpckhbw {{.*#+}} xmm3 = xmm3[8],xmm0[8],xmm3[9],xmm0[9],xmm3[10],xmm0[10],xmm3[11],xmm0[11],xmm3[12],xmm0[12],xmm3[13],xmm0[13],xmm3[14],xmm0[14],xmm3[15],xmm0[15]
+; SSE41-NEXT:    pmullw {{.*}}(%rip), %xmm3
 ; SSE41-NEXT:    psrlw $8, %xmm3
+; SSE41-NEXT:    packuswb %xmm3, %xmm2
+; SSE41-NEXT:    paddb %xmm1, %xmm2
+; SSE41-NEXT:    movdqa %xmm2, %xmm0
 ; SSE41-NEXT:    punpckhbw {{.*#+}} xmm0 = xmm0[8],xmm2[8],xmm0[9],xmm2[9],xmm0[10],xmm2[10],xmm0[11],xmm2[11],xmm0[12],xmm2[12],xmm0[13],xmm2[13],xmm0[14],xmm2[14],xmm0[15],xmm2[15]
-; SSE41-NEXT:    pmullw {{.*}}(%rip), %xmm0
-; SSE41-NEXT:    psrlw $8, %xmm0
-; SSE41-NEXT:    packuswb %xmm0, %xmm3
-; SSE41-NEXT:    paddb %xmm1, %xmm3
-; SSE41-NEXT:    movdqa %xmm3, %xmm0
-; SSE41-NEXT:    punpckhbw {{.*#+}} xmm0 = xmm0[8],xmm3[8],xmm0[9],xmm3[9],xmm0[10],xmm3[10],xmm0[11],xmm3[11],xmm0[12],xmm3[12],xmm0[13],xmm3[13],xmm0[14],xmm3[14],xmm0[15],xmm3[15]
 ; SSE41-NEXT:    psraw $8, %xmm0
-; SSE41-NEXT:    movdqa %xmm0, %xmm4
-; SSE41-NEXT:    psllw $1, %xmm4
+; SSE41-NEXT:    movdqa %xmm0, %xmm3
+; SSE41-NEXT:    psllw $1, %xmm3
 ; SSE41-NEXT:    psllw $7, %xmm0
-; SSE41-NEXT:    pblendw {{.*#+}} xmm0 = xmm0[0,1,2,3,4],xmm4[5],xmm0[6],xmm4[7]
+; SSE41-NEXT:    pblendw {{.*#+}} xmm0 = xmm0[0,1,2,3,4],xmm3[5],xmm0[6],xmm3[7]
 ; SSE41-NEXT:    psrlw $8, %xmm0
-; SSE41-NEXT:    punpcklbw {{.*#+}} xmm3 = xmm3[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
-; SSE41-NEXT:    psraw $8, %xmm3
-; SSE41-NEXT:    psllw $7, %xmm3
-; SSE41-NEXT:    psrlw $8, %xmm3
-; SSE41-NEXT:    packuswb %xmm0, %xmm3
+; SSE41-NEXT:    punpcklbw {{.*#+}} xmm2 = xmm2[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; SSE41-NEXT:    psraw $8, %xmm2
+; SSE41-NEXT:    psllw $7, %xmm2
+; SSE41-NEXT:    psrlw $8, %xmm2
+; SSE41-NEXT:    packuswb %xmm0, %xmm2
 ; SSE41-NEXT:    movaps {{.*#+}} xmm0 = [0,0,255,0,0,0,255,0,0,255,255,255,255,255,255,255]
-; SSE41-NEXT:    pblendvb %xmm0, %xmm3, %xmm1
-; SSE41-NEXT:    psubb %xmm1, %xmm2
-; SSE41-NEXT:    movaps {{.*#+}} xmm0 = [255,255,0,255,255,255,0,255,255,0,0,0,0,255,0,255]
 ; SSE41-NEXT:    pblendvb %xmm0, %xmm2, %xmm1
+; SSE41-NEXT:    movdqa {{.*#+}} xmm0 = [255,255,0,255,255,255,0,255,255,0,0,0,0,255,0,255]
+; SSE41-NEXT:    pxor %xmm0, %xmm1
+; SSE41-NEXT:    psubb %xmm0, %xmm1
 ; SSE41-NEXT:    movdqa %xmm1, %xmm0
 ; SSE41-NEXT:    retq
 ;
@@ -2247,68 +2232,68 @@ define <16 x i8> @non_splat_minus_one_divisor_1(<16 x i8> %A) {
 ; AVX1-NEXT:    vpsllw $1, %xmm4, %xmm4
 ; AVX1-NEXT:    vpblendw {{.*#+}} xmm3 = xmm3[0,1],xmm4[2],xmm3[3,4,5],xmm4[6],xmm3[7]
 ; AVX1-NEXT:    vpsrlw $8, %xmm3, %xmm3
-; AVX1-NEXT:    vpunpckhbw {{.*#+}} xmm2 = xmm2[8],xmm1[8],xmm2[9],xmm1[9],xmm2[10],xmm1[10],xmm2[11],xmm1[11],xmm2[12],xmm1[12],xmm2[13],xmm1[13],xmm2[14],xmm1[14],xmm2[15],xmm1[15]
-; AVX1-NEXT:    vpmullw {{.*}}(%rip), %xmm2, %xmm2
-; AVX1-NEXT:    vpsrlw $8, %xmm2, %xmm2
-; AVX1-NEXT:    vpackuswb %xmm2, %xmm3, %xmm2
-; AVX1-NEXT:    vpaddb %xmm2, %xmm0, %xmm2
-; AVX1-NEXT:    vpunpckhbw {{.*#+}} xmm3 = xmm2[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15]
-; AVX1-NEXT:    vpsraw $8, %xmm3, %xmm3
-; AVX1-NEXT:    vpsllw $1, %xmm3, %xmm4
-; AVX1-NEXT:    vpsllw $7, %xmm3, %xmm3
-; AVX1-NEXT:    vpblendw {{.*#+}} xmm3 = xmm3[0,1,2,3,4],xmm4[5],xmm3[6],xmm4[7]
-; AVX1-NEXT:    vpsrlw $8, %xmm3, %xmm3
-; AVX1-NEXT:    vpunpcklbw {{.*#+}} xmm2 = xmm2[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; AVX1-NEXT:    vpunpckhbw {{.*#+}} xmm1 = xmm2[8],xmm1[8],xmm2[9],xmm1[9],xmm2[10],xmm1[10],xmm2[11],xmm1[11],xmm2[12],xmm1[12],xmm2[13],xmm1[13],xmm2[14],xmm1[14],xmm2[15],xmm1[15]
+; AVX1-NEXT:    vpmullw {{.*}}(%rip), %xmm1, %xmm1
+; AVX1-NEXT:    vpsrlw $8, %xmm1, %xmm1
+; AVX1-NEXT:    vpackuswb %xmm1, %xmm3, %xmm1
+; AVX1-NEXT:    vpaddb %xmm1, %xmm0, %xmm1
+; AVX1-NEXT:    vpunpckhbw {{.*#+}} xmm2 = xmm1[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15]
 ; AVX1-NEXT:    vpsraw $8, %xmm2, %xmm2
+; AVX1-NEXT:    vpsllw $1, %xmm2, %xmm3
 ; AVX1-NEXT:    vpsllw $7, %xmm2, %xmm2
+; AVX1-NEXT:    vpblendw {{.*#+}} xmm2 = xmm2[0,1,2,3,4],xmm3[5],xmm2[6],xmm3[7]
 ; AVX1-NEXT:    vpsrlw $8, %xmm2, %xmm2
-; AVX1-NEXT:    vpackuswb %xmm3, %xmm2, %xmm2
-; AVX1-NEXT:    vmovdqa {{.*#+}} xmm3 = [0,0,255,0,0,0,255,0,0,255,255,255,255,255,255,255]
-; AVX1-NEXT:    vpblendvb %xmm3, %xmm2, %xmm0, %xmm0
-; AVX1-NEXT:    vpsubb %xmm0, %xmm1, %xmm1
-; AVX1-NEXT:    vmovdqa {{.*#+}} xmm2 = [255,255,0,255,255,255,0,255,255,0,0,0,0,255,0,255]
+; AVX1-NEXT:    vpunpcklbw {{.*#+}} xmm1 = xmm1[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; AVX1-NEXT:    vpsraw $8, %xmm1, %xmm1
+; AVX1-NEXT:    vpsllw $7, %xmm1, %xmm1
+; AVX1-NEXT:    vpsrlw $8, %xmm1, %xmm1
+; AVX1-NEXT:    vpackuswb %xmm2, %xmm1, %xmm1
+; AVX1-NEXT:    vmovdqa {{.*#+}} xmm2 = [0,0,255,0,0,0,255,0,0,255,255,255,255,255,255,255]
 ; AVX1-NEXT:    vpblendvb %xmm2, %xmm1, %xmm0, %xmm0
+; AVX1-NEXT:    vmovdqa {{.*#+}} xmm1 = [255,255,0,255,255,255,0,255,255,0,0,0,0,255,0,255]
+; AVX1-NEXT:    vpxor %xmm1, %xmm0, %xmm0
+; AVX1-NEXT:    vpsubb %xmm1, %xmm0, %xmm0
 ; AVX1-NEXT:    retq
 ;
 ; AVX2-LABEL: non_splat_minus_one_divisor_1:
 ; AVX2:       # %bb.0:
 ; AVX2-NEXT:    vpxor %xmm1, %xmm1, %xmm1
-; AVX2-NEXT:    vpcmpgtb %xmm0, %xmm1, %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:    vpmullw {{.*}}(%rip), %ymm2, %ymm2
-; AVX2-NEXT:    vpsrlw $8, %ymm2, %ymm2
-; AVX2-NEXT:    vextracti128 $1, %ymm2, %xmm3
-; AVX2-NEXT:    vpackuswb %xmm3, %xmm2, %xmm2
-; AVX2-NEXT:    vpaddb %xmm2, %xmm0, %xmm2
-; AVX2-NEXT:    vpmovsxbw %xmm2, %ymm2
-; AVX2-NEXT:    vpmullw {{.*}}(%rip), %ymm2, %ymm2
-; AVX2-NEXT:    vpsrlw $8, %ymm2, %ymm2
-; AVX2-NEXT:    vextracti128 $1, %ymm2, %xmm3
-; AVX2-NEXT:    vpackuswb %xmm3, %xmm2, %xmm2
-; AVX2-NEXT:    vmovdqa {{.*#+}} xmm3 = [0,0,255,0,0,0,255,0,0,255,255,255,255,255,255,255]
-; AVX2-NEXT:    vpblendvb %xmm3, %xmm2, %xmm0, %xmm0
-; AVX2-NEXT:    vpsubb %xmm0, %xmm1, %xmm1
-; AVX2-NEXT:    vmovdqa {{.*#+}} xmm2 = [255,255,0,255,255,255,0,255,255,0,0,0,0,255,0,255]
+; AVX2-NEXT:    vpcmpgtb %xmm0, %xmm1, %xmm1
+; 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:    vpmullw {{.*}}(%rip), %ymm1, %ymm1
+; AVX2-NEXT:    vpsrlw $8, %ymm1, %ymm1
+; AVX2-NEXT:    vextracti128 $1, %ymm1, %xmm2
+; AVX2-NEXT:    vpackuswb %xmm2, %xmm1, %xmm1
+; AVX2-NEXT:    vpaddb %xmm1, %xmm0, %xmm1
+; AVX2-NEXT:    vpmovsxbw %xmm1, %ymm1
+; AVX2-NEXT:    vpmullw {{.*}}(%rip), %ymm1, %ymm1
+; AVX2-NEXT:    vpsrlw $8, %ymm1, %ymm1
+; AVX2-NEXT:    vextracti128 $1, %ymm1, %xmm2
+; AVX2-NEXT:    vpackuswb %xmm2, %xmm1, %xmm1
+; AVX2-NEXT:    vmovdqa {{.*#+}} xmm2 = [0,0,255,0,0,0,255,0,0,255,255,255,255,255,255,255]
 ; AVX2-NEXT:    vpblendvb %xmm2, %xmm1, %xmm0, %xmm0
+; AVX2-NEXT:    vmovdqa {{.*#+}} xmm1 = [255,255,0,255,255,255,0,255,255,0,0,0,0,255,0,255]
+; AVX2-NEXT:    vpxor %xmm1, %xmm0, %xmm0
+; AVX2-NEXT:    vpsubb %xmm1, %xmm0, %xmm0
 ; AVX2-NEXT:    vzeroupper
 ; AVX2-NEXT:    retq
 ;
 ; AVX512F-LABEL: non_splat_minus_one_divisor_1:
 ; AVX512F:       # %bb.0:
 ; AVX512F-NEXT:    vpxor %xmm1, %xmm1, %xmm1
-; AVX512F-NEXT:    vpcmpgtb %xmm0, %xmm1, %xmm2
-; AVX512F-NEXT:    vpmovzxbd {{.*#+}} zmm2 = xmm2[0],zero,zero,zero,xmm2[1],zero,zero,zero,xmm2[2],zero,zero,zero,xmm2[3],zero,zero,zero,xmm2[4],zero,zero,zero,xmm2[5],zero,zero,zero,xmm2[6],zero,zero,zero,xmm2[7],zero,zero,zero,xmm2[8],zero,zero,zero,xmm2[9],zero,zero,zero,xmm2[10],zero,zero,zero,xmm2[11],zero,zero,zero,xmm2[12],zero,zero,zero,xmm2[13],zero,zero,zero,xmm2[14],zero,zero,zero,xmm2[15],zero,zero,zero
-; AVX512F-NEXT:    vpsrlvd {{.*}}(%rip), %zmm2, %zmm2
-; AVX512F-NEXT:    vpmovdb %zmm2, %xmm2
-; AVX512F-NEXT:    vpaddb %xmm2, %xmm0, %xmm2
-; AVX512F-NEXT:    vpmovsxbd %xmm2, %zmm2
-; AVX512F-NEXT:    vpsravd {{.*}}(%rip), %zmm2, %zmm2
-; AVX512F-NEXT:    vpmovdb %zmm2, %xmm2
-; AVX512F-NEXT:    vmovdqa {{.*#+}} xmm3 = [0,0,255,0,0,0,255,0,0,255,255,255,255,255,255,255]
-; AVX512F-NEXT:    vpblendvb %xmm3, %xmm2, %xmm0, %xmm0
-; AVX512F-NEXT:    vpsubb %xmm0, %xmm1, %xmm1
-; AVX512F-NEXT:    vmovdqa {{.*#+}} xmm2 = [255,255,0,255,255,255,0,255,255,0,0,0,0,255,0,255]
+; AVX512F-NEXT:    vpcmpgtb %xmm0, %xmm1, %xmm1
+; AVX512F-NEXT:    vpmovzxbd {{.*#+}} zmm1 = xmm1[0],zero,zero,zero,xmm1[1],zero,zero,zero,xmm1[2],zero,zero,zero,xmm1[3],zero,zero,zero,xmm1[4],zero,zero,zero,xmm1[5],zero,zero,zero,xmm1[6],zero,zero,zero,xmm1[7],zero,zero,zero,xmm1[8],zero,zero,zero,xmm1[9],zero,zero,zero,xmm1[10],zero,zero,zero,xmm1[11],zero,zero,zero,xmm1[12],zero,zero,zero,xmm1[13],zero,zero,zero,xmm1[14],zero,zero,zero,xmm1[15],zero,zero,zero
+; AVX512F-NEXT:    vpsrlvd {{.*}}(%rip), %zmm1, %zmm1
+; AVX512F-NEXT:    vpmovdb %zmm1, %xmm1
+; AVX512F-NEXT:    vpaddb %xmm1, %xmm0, %xmm1
+; AVX512F-NEXT:    vpmovsxbd %xmm1, %zmm1
+; AVX512F-NEXT:    vpsravd {{.*}}(%rip), %zmm1, %zmm1
+; AVX512F-NEXT:    vpmovdb %zmm1, %xmm1
+; AVX512F-NEXT:    vmovdqa {{.*#+}} xmm2 = [0,0,255,0,0,0,255,0,0,255,255,255,255,255,255,255]
 ; AVX512F-NEXT:    vpblendvb %xmm2, %xmm1, %xmm0, %xmm0
+; AVX512F-NEXT:    vmovdqa {{.*#+}} xmm1 = [255,255,0,255,255,255,0,255,255,0,0,0,0,255,0,255]
+; AVX512F-NEXT:    vpxor %xmm1, %xmm0, %xmm0
+; AVX512F-NEXT:    vpsubb %xmm1, %xmm0, %xmm0
 ; AVX512F-NEXT:    vzeroupper
 ; AVX512F-NEXT:    retq
 ;
@@ -2336,15 +2321,15 @@ define <16 x i8> @non_splat_minus_one_divisor_1(<16 x i8> %A) {
 ; XOP-LABEL: non_splat_minus_one_divisor_1:
 ; XOP:       # %bb.0:
 ; XOP-NEXT:    vpxor %xmm1, %xmm1, %xmm1
-; XOP-NEXT:    vpcmpgtb %xmm0, %xmm1, %xmm2
-; XOP-NEXT:    vpshlb {{.*}}(%rip), %xmm2, %xmm2
-; XOP-NEXT:    vpaddb %xmm2, %xmm0, %xmm2
-; XOP-NEXT:    vpshab {{.*}}(%rip), %xmm2, %xmm2
-; XOP-NEXT:    vmovdqa {{.*#+}} xmm3 = [0,0,255,0,0,0,255,0,0,255,255,255,255,255,255,255]
-; XOP-NEXT:    vpblendvb %xmm3, %xmm2, %xmm0, %xmm0
-; XOP-NEXT:    vpsubb %xmm0, %xmm1, %xmm1
-; XOP-NEXT:    vmovdqa {{.*#+}} xmm2 = [255,255,0,255,255,255,0,255,255,0,0,0,0,255,0,255]
+; XOP-NEXT:    vpcmpgtb %xmm0, %xmm1, %xmm1
+; XOP-NEXT:    vpshlb {{.*}}(%rip), %xmm1, %xmm1
+; XOP-NEXT:    vpaddb %xmm1, %xmm0, %xmm1
+; XOP-NEXT:    vpshab {{.*}}(%rip), %xmm1, %xmm1
+; XOP-NEXT:    vmovdqa {{.*#+}} xmm2 = [0,0,255,0,0,0,255,0,0,255,255,255,255,255,255,255]
 ; XOP-NEXT:    vpblendvb %xmm2, %xmm1, %xmm0, %xmm0
+; XOP-NEXT:    vmovdqa {{.*#+}} xmm1 = [255,255,0,255,255,255,0,255,255,0,0,0,0,255,0,255]
+; XOP-NEXT:    vpxor %xmm1, %xmm0, %xmm0
+; XOP-NEXT:    vpsubb %xmm1, %xmm0, %xmm0
 ; XOP-NEXT:    retq
   %div = sdiv <16 x i8> %A, <i8 -1, i8 -1, i8 2, i8 -1, i8 -1, i8 -1, i8 2, i8 -1, i8 -1, i8 2, i8 2, i8 2, i8 2, i8 -128, i8 2, i8 -128>
   ret <16 x i8> %div

diff  --git a/llvm/test/CodeGen/X86/vector-blend.ll b/llvm/test/CodeGen/X86/vector-blend.ll
index 6d46064469de..ce97fec08415 100644
--- a/llvm/test/CodeGen/X86/vector-blend.ll
+++ b/llvm/test/CodeGen/X86/vector-blend.ll
@@ -922,37 +922,19 @@ entry:
 }
 
 define <4 x i32> @blend_neg_logic_v4i32_2(<4 x i32> %v, <4 x i32> %c) {
-; SSE2-LABEL: blend_neg_logic_v4i32_2:
-; SSE2:       # %bb.0: # %entry
-; SSE2-NEXT:    psrad $31, %xmm1
-; SSE2-NEXT:    pxor %xmm1, %xmm0
-; SSE2-NEXT:    psubd %xmm0, %xmm1
-; SSE2-NEXT:    movdqa %xmm1, %xmm0
-; SSE2-NEXT:    retq
-;
-; SSSE3-LABEL: blend_neg_logic_v4i32_2:
-; SSSE3:       # %bb.0: # %entry
-; SSSE3-NEXT:    psrad $31, %xmm1
-; SSSE3-NEXT:    pxor %xmm1, %xmm0
-; SSSE3-NEXT:    psubd %xmm0, %xmm1
-; SSSE3-NEXT:    movdqa %xmm1, %xmm0
-; SSSE3-NEXT:    retq
-;
-; SSE41-LABEL: blend_neg_logic_v4i32_2:
-; SSE41:       # %bb.0: # %entry
-; SSE41-NEXT:    movdqa %xmm0, %xmm2
-; SSE41-NEXT:    pxor %xmm3, %xmm3
-; SSE41-NEXT:    psubd %xmm0, %xmm3
-; SSE41-NEXT:    movaps %xmm1, %xmm0
-; SSE41-NEXT:    blendvps %xmm0, %xmm2, %xmm3
-; SSE41-NEXT:    movaps %xmm3, %xmm0
-; SSE41-NEXT:    retq
+; SSE-LABEL: blend_neg_logic_v4i32_2:
+; SSE:       # %bb.0: # %entry
+; SSE-NEXT:    psrad $31, %xmm1
+; SSE-NEXT:    pxor %xmm1, %xmm0
+; SSE-NEXT:    psubd %xmm0, %xmm1
+; SSE-NEXT:    movdqa %xmm1, %xmm0
+; SSE-NEXT:    retq
 ;
 ; AVX-LABEL: blend_neg_logic_v4i32_2:
 ; AVX:       # %bb.0: # %entry
-; AVX-NEXT:    vpxor %xmm2, %xmm2, %xmm2
-; AVX-NEXT:    vpsubd %xmm0, %xmm2, %xmm2
-; AVX-NEXT:    vblendvps %xmm1, %xmm0, %xmm2, %xmm0
+; AVX-NEXT:    vpsrad $31, %xmm1, %xmm1
+; AVX-NEXT:    vpxor %xmm1, %xmm0, %xmm0
+; AVX-NEXT:    vpsubd %xmm0, %xmm1, %xmm0
 ; AVX-NEXT:    retq
 entry:
   %0 = ashr <4 x i32> %c, <i32 31, i32 31, i32 31, i32 31>


        


More information about the llvm-commits mailing list