[llvm] f5a26cc - [X86][SSE] combineMulToPMADDWD - enable sext(v8i16) -> zext(v8i16) fold on pre-SSE41 targets

Simon Pilgrim via llvm-commits llvm-commits at lists.llvm.org
Sat Sep 25 06:35:48 PDT 2021


Author: Simon Pilgrim
Date: 2021-09-25T14:35:31+01:00
New Revision: f5a26ccae24caa21f7adafe0aab704fd0bc5a618

URL: https://github.com/llvm/llvm-project/commit/f5a26ccae24caa21f7adafe0aab704fd0bc5a618
DIFF: https://github.com/llvm/llvm-project/commit/f5a26ccae24caa21f7adafe0aab704fd0bc5a618.diff

LOG: [X86][SSE] combineMulToPMADDWD - enable sext(v8i16) -> zext(v8i16) fold on pre-SSE41 targets

We already do this on SSE41 targets where we have sext/zext instructions, now that combineShiftToPMULH handles SSE2 targets, we can enable this here as well.

Added: 
    

Modified: 
    llvm/lib/Target/X86/X86ISelLowering.cpp
    llvm/test/CodeGen/X86/madd.ll

Removed: 
    


################################################################################
diff  --git a/llvm/lib/Target/X86/X86ISelLowering.cpp b/llvm/lib/Target/X86/X86ISelLowering.cpp
index f3a4501a20b7..a3b5640c1a62 100644
--- a/llvm/lib/Target/X86/X86ISelLowering.cpp
+++ b/llvm/lib/Target/X86/X86ISelLowering.cpp
@@ -44265,10 +44265,9 @@ static SDValue combineMulToPMADDWD(SDNode *N, SelectionDAG &DAG,
     if (DAG.MaskedValueIsZero(Op, Mask17))
       return Op;
     // Convert sext(vXi16) to zext(vXi16).
-    // TODO: Enable pre-SSE41 once we can prefer MULHU/MULHS first.
     // TODO: Handle sext from smaller types as well?
     if (Op.getOpcode() == ISD::SIGN_EXTEND && VT.is128BitVector() &&
-        Subtarget.hasSSE41() && N->isOnlyUserOf(Op.getNode())) {
+        N->isOnlyUserOf(Op.getNode())) {
       SDValue Src = Op.getOperand(0);
       if (Src.getScalarValueSizeInBits() == 16)
         return DAG.getNode(ISD::ZERO_EXTEND, SDLoc(N), VT, Src);

diff  --git a/llvm/test/CodeGen/X86/madd.ll b/llvm/test/CodeGen/X86/madd.ll
index 330ca9061867..5dd1d6cb4c61 100644
--- a/llvm/test/CodeGen/X86/madd.ll
+++ b/llvm/test/CodeGen/X86/madd.ll
@@ -11,25 +11,25 @@ define i32 @_Z10test_shortPsS_i_128(i16* nocapture readonly, i16* nocapture read
 ; SSE2-NEXT:    movl %edx, %eax
 ; SSE2-NEXT:    pxor %xmm0, %xmm0
 ; SSE2-NEXT:    xorl %ecx, %ecx
+; SSE2-NEXT:    pxor %xmm1, %xmm1
 ; SSE2-NEXT:    .p2align 4, 0x90
 ; SSE2-NEXT:  .LBB0_1: # %vector.body
 ; SSE2-NEXT:    # =>This Inner Loop Header: Depth=1
-; SSE2-NEXT:    movq {{.*#+}} xmm1 = mem[0],zero
 ; SSE2-NEXT:    movq {{.*#+}} xmm2 = mem[0],zero
-; SSE2-NEXT:    movdqa %xmm2, %xmm3
-; SSE2-NEXT:    pmulhw %xmm1, %xmm3
-; SSE2-NEXT:    pmullw %xmm1, %xmm2
-; SSE2-NEXT:    punpcklwd {{.*#+}} xmm2 = xmm2[0],xmm3[0],xmm2[1],xmm3[1],xmm2[2],xmm3[2],xmm2[3],xmm3[3]
-; SSE2-NEXT:    paddd %xmm2, %xmm0
+; SSE2-NEXT:    movq {{.*#+}} xmm3 = mem[0],zero
+; SSE2-NEXT:    punpcklwd {{.*#+}} xmm2 = xmm2[0],xmm0[0],xmm2[1],xmm0[1],xmm2[2],xmm0[2],xmm2[3],xmm0[3]
+; SSE2-NEXT:    punpcklwd {{.*#+}} xmm3 = xmm3[0],xmm0[0],xmm3[1],xmm0[1],xmm3[2],xmm0[2],xmm3[3],xmm0[3]
+; SSE2-NEXT:    pmaddwd %xmm2, %xmm3
+; SSE2-NEXT:    paddd %xmm3, %xmm1
 ; SSE2-NEXT:    addq $8, %rcx
 ; SSE2-NEXT:    cmpq %rcx, %rax
 ; SSE2-NEXT:    jne .LBB0_1
 ; SSE2-NEXT:  # %bb.2: # %middle.block
-; SSE2-NEXT:    pshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
-; SSE2-NEXT:    paddd %xmm0, %xmm1
-; SSE2-NEXT:    pshufd {{.*#+}} xmm0 = xmm1[1,1,1,1]
+; SSE2-NEXT:    pshufd {{.*#+}} xmm0 = xmm1[2,3,2,3]
 ; SSE2-NEXT:    paddd %xmm1, %xmm0
-; SSE2-NEXT:    movd %xmm0, %eax
+; SSE2-NEXT:    pshufd {{.*#+}} xmm1 = xmm0[1,1,1,1]
+; SSE2-NEXT:    paddd %xmm0, %xmm1
+; SSE2-NEXT:    movd %xmm1, %eax
 ; SSE2-NEXT:    retq
 ;
 ; AVX-LABEL: _Z10test_shortPsS_i_128:
@@ -2568,35 +2568,28 @@ define <4 x i32> @pmaddwd_swapped_indices(<8 x i16>* %Aptr, <8 x i16>* %Bptr) {
   ret <4 x i32> %add
 }
 
-; Negative test were indices aren't paired properly
+; Negative test where indices aren't paired properly
 define <4 x i32> @pmaddwd_bad_indices(<8 x i16>* %Aptr, <8 x i16>* %Bptr) {
 ; SSE2-LABEL: pmaddwd_bad_indices:
 ; SSE2:       # %bb.0:
-; SSE2-NEXT:    movdqa (%rdi), %xmm0
-; SSE2-NEXT:    movdqa (%rsi), %xmm1
-; SSE2-NEXT:    pshuflw {{.*#+}} xmm2 = xmm0[2,1,2,3,4,5,6,7]
+; SSE2-NEXT:    movdqa (%rdi), %xmm1
+; SSE2-NEXT:    movdqa (%rsi), %xmm0
+; SSE2-NEXT:    pshuflw {{.*#+}} xmm2 = xmm1[2,1,2,3,4,5,6,7]
 ; SSE2-NEXT:    pshufhw {{.*#+}} xmm2 = xmm2[0,1,2,3,6,5,6,7]
 ; SSE2-NEXT:    pshufd {{.*#+}} xmm2 = xmm2[0,2,2,3]
 ; SSE2-NEXT:    pshuflw {{.*#+}} xmm2 = xmm2[1,0,3,2,4,5,6,7]
-; SSE2-NEXT:    pshuflw {{.*#+}} xmm0 = xmm0[0,3,2,3,4,5,6,7]
-; SSE2-NEXT:    pshufhw {{.*#+}} xmm0 = xmm0[0,1,2,3,4,7,6,7]
-; SSE2-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[0,2,2,3]
-; SSE2-NEXT:    pshuflw {{.*#+}} xmm3 = xmm1[0,2,2,3,4,5,6,7]
-; SSE2-NEXT:    pshufhw {{.*#+}} xmm3 = xmm3[0,1,2,3,4,6,6,7]
-; SSE2-NEXT:    pshufd {{.*#+}} xmm3 = xmm3[0,2,2,3]
-; SSE2-NEXT:    pshuflw {{.*#+}} xmm1 = xmm1[3,1,2,3,4,5,6,7]
-; SSE2-NEXT:    pshufhw {{.*#+}} xmm1 = xmm1[0,1,2,3,7,5,6,7]
+; SSE2-NEXT:    pxor %xmm3, %xmm3
+; SSE2-NEXT:    punpcklwd {{.*#+}} xmm2 = xmm2[0],xmm3[0],xmm2[1],xmm3[1],xmm2[2],xmm3[2],xmm2[3],xmm3[3]
+; SSE2-NEXT:    movdqa {{.*#+}} xmm4 = [65535,0,65535,0,65535,0,65535,0]
+; SSE2-NEXT:    pand %xmm0, %xmm4
+; SSE2-NEXT:    pmaddwd %xmm2, %xmm4
+; SSE2-NEXT:    pshuflw {{.*#+}} xmm1 = xmm1[0,3,2,3,4,5,6,7]
+; SSE2-NEXT:    pshufhw {{.*#+}} xmm1 = xmm1[0,1,2,3,4,7,6,7]
 ; SSE2-NEXT:    pshufd {{.*#+}} xmm1 = xmm1[0,2,2,3]
-; SSE2-NEXT:    pshuflw {{.*#+}} xmm1 = xmm1[1,0,3,2,4,5,6,7]
-; SSE2-NEXT:    movdqa %xmm2, %xmm4
-; SSE2-NEXT:    pmulhw %xmm3, %xmm4
-; SSE2-NEXT:    pmullw %xmm3, %xmm2
-; SSE2-NEXT:    punpcklwd {{.*#+}} xmm2 = xmm2[0],xmm4[0],xmm2[1],xmm4[1],xmm2[2],xmm4[2],xmm2[3],xmm4[3]
-; SSE2-NEXT:    movdqa %xmm0, %xmm3
-; SSE2-NEXT:    pmulhw %xmm1, %xmm3
-; SSE2-NEXT:    pmullw %xmm1, %xmm0
-; SSE2-NEXT:    punpcklwd {{.*#+}} xmm0 = xmm0[0],xmm3[0],xmm0[1],xmm3[1],xmm0[2],xmm3[2],xmm0[3],xmm3[3]
-; SSE2-NEXT:    paddd %xmm2, %xmm0
+; SSE2-NEXT:    punpcklwd {{.*#+}} xmm1 = xmm1[0],xmm3[0],xmm1[1],xmm3[1],xmm1[2],xmm3[2],xmm1[3],xmm3[3]
+; SSE2-NEXT:    psrld $16, %xmm0
+; SSE2-NEXT:    pmaddwd %xmm1, %xmm0
+; SSE2-NEXT:    paddd %xmm4, %xmm0
 ; SSE2-NEXT:    retq
 ;
 ; AVX-LABEL: pmaddwd_bad_indices:


        


More information about the llvm-commits mailing list