[llvm] r320669 - [SelectionDAG] When legalizing the result type of CONCAT_VECTORS, take into account whether the input type also needs to be promoted.
Craig Topper via llvm-commits
llvm-commits at lists.llvm.org
Wed Dec 13 22:49:07 PST 2017
Author: ctopper
Date: Wed Dec 13 22:49:07 2017
New Revision: 320669
URL: http://llvm.org/viewvc/llvm-project?rev=320669&view=rev
Log:
[SelectionDAG] When legalizing the result type of CONCAT_VECTORS, take into account whether the input type also needs to be promoted.
If so go ahead and get the promoted input vector to extract from. Previously, we would create a bunch of any_extends of extract_vector_elts with illegal input type that needs to be promoted. The legalization of those extract_vector_elts would then potentially introduce a truncate. So now we have a bunch of any_extends of truncates. By legalizing both parts together we avoid creating these extra nodes.
The test changes seem to be because we were previously combining the build_vector with the any_extend before the any_extend got combined with the truncate.
Modified:
llvm/trunk/lib/CodeGen/SelectionDAG/LegalizeIntegerTypes.cpp
llvm/trunk/test/CodeGen/X86/bitcast-and-setcc-512.ll
llvm/trunk/test/CodeGen/X86/vector-compare-results.ll
Modified: llvm/trunk/lib/CodeGen/SelectionDAG/LegalizeIntegerTypes.cpp
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/CodeGen/SelectionDAG/LegalizeIntegerTypes.cpp?rev=320669&r1=320668&r2=320669&view=diff
==============================================================================
--- llvm/trunk/lib/CodeGen/SelectionDAG/LegalizeIntegerTypes.cpp (original)
+++ llvm/trunk/lib/CodeGen/SelectionDAG/LegalizeIntegerTypes.cpp Wed Dec 13 22:49:07 2017
@@ -3489,7 +3489,6 @@ SDValue DAGTypeLegalizer::PromoteIntRes_
EVT NOutVT = TLI.getTypeToTransformTo(*DAG.getContext(), OutVT);
assert(NOutVT.isVector() && "This type must be promoted to a vector type");
- EVT InElemTy = OutVT.getVectorElementType();
EVT OutElemTy = NOutVT.getVectorElementType();
unsigned NumElem = N->getOperand(0).getValueType().getVectorNumElements();
@@ -3502,11 +3501,17 @@ SDValue DAGTypeLegalizer::PromoteIntRes_
SmallVector<SDValue, 8> Ops(NumOutElem);
for (unsigned i = 0; i < NumOperands; ++i) {
SDValue Op = N->getOperand(i);
+ if (getTypeAction(Op.getValueType()) == TargetLowering::TypePromoteInteger)
+ Op = GetPromotedInteger(Op);
+ EVT SclrTy = Op.getValueType().getVectorElementType();
+ assert(NumElem == Op.getValueType().getVectorNumElements() &&
+ "Unexpected number of elements");
+
for (unsigned j = 0; j < NumElem; ++j) {
SDValue Ext = DAG.getNode(
- ISD::EXTRACT_VECTOR_ELT, dl, InElemTy, Op,
+ ISD::EXTRACT_VECTOR_ELT, dl, SclrTy, Op,
DAG.getConstant(j, dl, TLI.getVectorIdxTy(DAG.getDataLayout())));
- Ops[i * NumElem + j] = DAG.getNode(ISD::ANY_EXTEND, dl, OutElemTy, Ext);
+ Ops[i * NumElem + j] = DAG.getAnyExtOrTrunc(Ext, dl, OutElemTy);
}
}
Modified: llvm/trunk/test/CodeGen/X86/bitcast-and-setcc-512.ll
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/X86/bitcast-and-setcc-512.ll?rev=320669&r1=320668&r2=320669&view=diff
==============================================================================
--- llvm/trunk/test/CodeGen/X86/bitcast-and-setcc-512.ll (original)
+++ llvm/trunk/test/CodeGen/X86/bitcast-and-setcc-512.ll Wed Dec 13 22:49:07 2017
@@ -12,33 +12,21 @@ define i8 @v8i64(<8 x i64> %a, <8 x i64>
; SSE-NEXT: movdqa {{[0-9]+}}(%rsp), %xmm9
; SSE-NEXT: movdqa {{[0-9]+}}(%rsp), %xmm10
; SSE-NEXT: movdqa {{[0-9]+}}(%rsp), %xmm11
-; SSE-NEXT: pcmpgtq %xmm5, %xmm1
-; SSE-NEXT: pshufd {{.*#+}} xmm1 = xmm1[0,2,2,3]
-; SSE-NEXT: pshuflw {{.*#+}} xmm1 = xmm1[0,2,2,3,4,5,6,7]
-; SSE-NEXT: pcmpgtq %xmm4, %xmm0
-; SSE-NEXT: pshufd {{.*#+}} xmm0 = xmm0[0,2,2,3]
-; SSE-NEXT: pshuflw {{.*#+}} xmm0 = xmm0[0,2,2,3,4,5,6,7]
-; SSE-NEXT: punpckldq {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
; SSE-NEXT: pcmpgtq %xmm7, %xmm3
; SSE-NEXT: pcmpgtq %xmm6, %xmm2
; SSE-NEXT: shufps {{.*#+}} xmm2 = xmm2[0,2],xmm3[0,2]
-; SSE-NEXT: packssdw %xmm2, %xmm2
-; SSE-NEXT: pblendw {{.*#+}} xmm2 = xmm0[0,1,2,3],xmm2[4,5,6,7]
+; SSE-NEXT: pcmpgtq %xmm5, %xmm1
+; SSE-NEXT: pcmpgtq %xmm4, %xmm0
+; SSE-NEXT: shufps {{.*#+}} xmm0 = xmm0[0,2],xmm1[0,2]
+; SSE-NEXT: packssdw %xmm2, %xmm0
; SSE-NEXT: pcmpgtq {{[0-9]+}}(%rsp), %xmm11
-; SSE-NEXT: pshufd {{.*#+}} xmm0 = xmm11[0,2,2,3]
-; SSE-NEXT: pshuflw {{.*#+}} xmm0 = xmm0[0,2,2,3,4,5,6,7]
; SSE-NEXT: pcmpgtq {{[0-9]+}}(%rsp), %xmm10
-; SSE-NEXT: pshufd {{.*#+}} xmm1 = xmm10[0,2,2,3]
-; SSE-NEXT: pshuflw {{.*#+}} xmm1 = xmm1[0,2,2,3,4,5,6,7]
-; SSE-NEXT: punpckldq {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1]
+; SSE-NEXT: shufps {{.*#+}} xmm10 = xmm10[0,2],xmm11[0,2]
; SSE-NEXT: pcmpgtq {{[0-9]+}}(%rsp), %xmm9
; SSE-NEXT: pcmpgtq {{[0-9]+}}(%rsp), %xmm8
; SSE-NEXT: shufps {{.*#+}} xmm8 = xmm8[0,2],xmm9[0,2]
-; SSE-NEXT: packssdw %xmm8, %xmm8
-; SSE-NEXT: pblendw {{.*#+}} xmm8 = xmm1[0,1,2,3],xmm8[4,5,6,7]
-; SSE-NEXT: pand %xmm2, %xmm8
-; SSE-NEXT: psllw $15, %xmm8
-; SSE-NEXT: psraw $15, %xmm8
+; SSE-NEXT: packssdw %xmm10, %xmm8
+; SSE-NEXT: pand %xmm0, %xmm8
; SSE-NEXT: packsswb %xmm0, %xmm8
; SSE-NEXT: pmovmskb %xmm8, %eax
; SSE-NEXT: # kill: def %al killed %al killed %eax
@@ -139,33 +127,21 @@ define i8 @v8f64(<8 x double> %a, <8 x d
; SSE-NEXT: movapd {{[0-9]+}}(%rsp), %xmm9
; SSE-NEXT: movapd {{[0-9]+}}(%rsp), %xmm10
; SSE-NEXT: movapd {{[0-9]+}}(%rsp), %xmm11
-; SSE-NEXT: cmpltpd %xmm1, %xmm5
-; SSE-NEXT: shufps {{.*#+}} xmm5 = xmm5[0,2,2,3]
-; SSE-NEXT: pshuflw {{.*#+}} xmm1 = xmm5[0,2,2,3,4,5,6,7]
-; SSE-NEXT: cmpltpd %xmm0, %xmm4
-; SSE-NEXT: shufps {{.*#+}} xmm4 = xmm4[0,2,2,3]
-; SSE-NEXT: pshuflw {{.*#+}} xmm0 = xmm4[0,2,2,3,4,5,6,7]
-; SSE-NEXT: punpckldq {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
; SSE-NEXT: cmpltpd %xmm3, %xmm7
; SSE-NEXT: cmpltpd %xmm2, %xmm6
; SSE-NEXT: shufps {{.*#+}} xmm6 = xmm6[0,2],xmm7[0,2]
-; SSE-NEXT: packssdw %xmm6, %xmm6
-; SSE-NEXT: pblendw {{.*#+}} xmm6 = xmm0[0,1,2,3],xmm6[4,5,6,7]
+; SSE-NEXT: cmpltpd %xmm1, %xmm5
+; SSE-NEXT: cmpltpd %xmm0, %xmm4
+; SSE-NEXT: shufps {{.*#+}} xmm4 = xmm4[0,2],xmm5[0,2]
+; SSE-NEXT: packssdw %xmm6, %xmm4
; SSE-NEXT: cmpltpd {{[0-9]+}}(%rsp), %xmm11
-; SSE-NEXT: shufps {{.*#+}} xmm11 = xmm11[0,2,2,3]
-; SSE-NEXT: pshuflw {{.*#+}} xmm0 = xmm11[0,2,2,3,4,5,6,7]
; SSE-NEXT: cmpltpd {{[0-9]+}}(%rsp), %xmm10
-; SSE-NEXT: shufps {{.*#+}} xmm10 = xmm10[0,2,2,3]
-; SSE-NEXT: pshuflw {{.*#+}} xmm1 = xmm10[0,2,2,3,4,5,6,7]
-; SSE-NEXT: punpckldq {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1]
+; SSE-NEXT: shufps {{.*#+}} xmm10 = xmm10[0,2],xmm11[0,2]
; SSE-NEXT: cmpltpd {{[0-9]+}}(%rsp), %xmm9
; SSE-NEXT: cmpltpd {{[0-9]+}}(%rsp), %xmm8
; SSE-NEXT: shufps {{.*#+}} xmm8 = xmm8[0,2],xmm9[0,2]
-; SSE-NEXT: packssdw %xmm8, %xmm8
-; SSE-NEXT: pblendw {{.*#+}} xmm8 = xmm1[0,1,2,3],xmm8[4,5,6,7]
-; SSE-NEXT: pand %xmm6, %xmm8
-; SSE-NEXT: psllw $15, %xmm8
-; SSE-NEXT: psraw $15, %xmm8
+; SSE-NEXT: packssdw %xmm10, %xmm8
+; SSE-NEXT: pand %xmm4, %xmm8
; SSE-NEXT: packsswb %xmm0, %xmm8
; SSE-NEXT: pmovmskb %xmm8, %eax
; SSE-NEXT: # kill: def %al killed %al killed %eax
Modified: llvm/trunk/test/CodeGen/X86/vector-compare-results.ll
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/X86/vector-compare-results.ll?rev=320669&r1=320668&r2=320669&view=diff
==============================================================================
--- llvm/trunk/test/CodeGen/X86/vector-compare-results.ll (original)
+++ llvm/trunk/test/CodeGen/X86/vector-compare-results.ll Wed Dec 13 22:49:07 2017
@@ -6562,72 +6562,52 @@ define <32 x i1> @test_cmp_v32f64(<32 x
; SSE2-LABEL: test_cmp_v32f64:
; SSE2: # %bb.0:
; SSE2-NEXT: movapd {{[0-9]+}}(%rsp), %xmm8
-; SSE2-NEXT: cmpltpd %xmm5, %xmm8
-; SSE2-NEXT: movapd {{[0-9]+}}(%rsp), %xmm5
-; SSE2-NEXT: cmpltpd %xmm4, %xmm5
-; SSE2-NEXT: shufps {{.*#+}} xmm8 = xmm8[0,2,2,3]
-; SSE2-NEXT: pshuflw {{.*#+}} xmm4 = xmm8[0,2,2,3,4,5,6,7]
-; SSE2-NEXT: shufps {{.*#+}} xmm5 = xmm5[0,2,2,3]
-; SSE2-NEXT: pshuflw {{.*#+}} xmm5 = xmm5[0,2,2,3,4,5,6,7]
-; SSE2-NEXT: punpckldq {{.*#+}} xmm5 = xmm5[0],xmm4[0],xmm5[1],xmm4[1]
-; SSE2-NEXT: movapd {{[0-9]+}}(%rsp), %xmm4
-; SSE2-NEXT: cmpltpd %xmm7, %xmm4
+; SSE2-NEXT: cmpltpd %xmm7, %xmm8
; SSE2-NEXT: movapd {{[0-9]+}}(%rsp), %xmm7
; SSE2-NEXT: cmpltpd %xmm6, %xmm7
-; SSE2-NEXT: shufps {{.*#+}} xmm7 = xmm7[0,2],xmm4[0,2]
-; SSE2-NEXT: packssdw %xmm7, %xmm7
-; SSE2-NEXT: movsd {{.*#+}} xmm7 = xmm5[0],xmm7[1]
+; SSE2-NEXT: shufps {{.*#+}} xmm7 = xmm7[0,2],xmm8[0,2]
+; SSE2-NEXT: movapd {{[0-9]+}}(%rsp), %xmm6
+; SSE2-NEXT: cmpltpd %xmm5, %xmm6
+; SSE2-NEXT: movapd {{[0-9]+}}(%rsp), %xmm5
+; SSE2-NEXT: cmpltpd %xmm4, %xmm5
+; SSE2-NEXT: shufps {{.*#+}} xmm5 = xmm5[0,2],xmm6[0,2]
+; SSE2-NEXT: packssdw %xmm7, %xmm5
; SSE2-NEXT: movapd {{[0-9]+}}(%rsp), %xmm4
-; SSE2-NEXT: cmpltpd %xmm1, %xmm4
+; SSE2-NEXT: cmpltpd %xmm3, %xmm4
+; SSE2-NEXT: movapd {{[0-9]+}}(%rsp), %xmm3
+; SSE2-NEXT: cmpltpd %xmm2, %xmm3
+; SSE2-NEXT: shufps {{.*#+}} xmm3 = xmm3[0,2],xmm4[0,2]
+; SSE2-NEXT: movapd {{[0-9]+}}(%rsp), %xmm2
+; SSE2-NEXT: cmpltpd %xmm1, %xmm2
; SSE2-NEXT: movapd {{[0-9]+}}(%rsp), %xmm1
; SSE2-NEXT: cmpltpd %xmm0, %xmm1
-; SSE2-NEXT: shufps {{.*#+}} xmm4 = xmm4[0,2,2,3]
-; SSE2-NEXT: pshuflw {{.*#+}} xmm0 = xmm4[0,2,2,3,4,5,6,7]
-; SSE2-NEXT: shufps {{.*#+}} xmm1 = xmm1[0,2,2,3]
-; SSE2-NEXT: pshuflw {{.*#+}} xmm4 = xmm1[0,2,2,3,4,5,6,7]
-; SSE2-NEXT: punpckldq {{.*#+}} xmm4 = xmm4[0],xmm0[0],xmm4[1],xmm0[1]
-; SSE2-NEXT: movapd {{[0-9]+}}(%rsp), %xmm5
-; SSE2-NEXT: cmpltpd %xmm3, %xmm5
; SSE2-NEXT: movapd {{[0-9]+}}(%rsp), %xmm0
-; SSE2-NEXT: cmpltpd %xmm2, %xmm0
-; SSE2-NEXT: movapd {{[0-9]+}}(%rsp), %xmm1
-; SSE2-NEXT: shufps {{.*#+}} xmm0 = xmm0[0,2],xmm5[0,2]
+; SSE2-NEXT: shufps {{.*#+}} xmm1 = xmm1[0,2],xmm2[0,2]
; SSE2-NEXT: movapd {{[0-9]+}}(%rsp), %xmm2
-; SSE2-NEXT: packssdw %xmm0, %xmm0
-; SSE2-NEXT: movsd {{.*#+}} xmm0 = xmm4[0],xmm0[1]
+; SSE2-NEXT: packssdw %xmm3, %xmm1
; SSE2-NEXT: movapd {{[0-9]+}}(%rsp), %xmm3
-; SSE2-NEXT: packsswb %xmm7, %xmm0
+; SSE2-NEXT: packsswb %xmm5, %xmm1
; SSE2-NEXT: movapd {{[0-9]+}}(%rsp), %xmm4
; SSE2-NEXT: cmpltpd {{[0-9]+}}(%rsp), %xmm4
-; SSE2-NEXT: shufps {{.*#+}} xmm4 = xmm4[0,2,2,3]
-; SSE2-NEXT: pshuflw {{.*#+}} xmm4 = xmm4[0,2,2,3,4,5,6,7]
; SSE2-NEXT: cmpltpd {{[0-9]+}}(%rsp), %xmm3
-; SSE2-NEXT: shufps {{.*#+}} xmm3 = xmm3[0,2,2,3]
-; SSE2-NEXT: pshuflw {{.*#+}} xmm3 = xmm3[0,2,2,3,4,5,6,7]
-; SSE2-NEXT: punpckldq {{.*#+}} xmm3 = xmm3[0],xmm4[0],xmm3[1],xmm4[1]
+; SSE2-NEXT: shufps {{.*#+}} xmm3 = xmm3[0,2],xmm4[0,2]
; SSE2-NEXT: movapd {{[0-9]+}}(%rsp), %xmm4
; SSE2-NEXT: cmpltpd {{[0-9]+}}(%rsp), %xmm4
; SSE2-NEXT: cmpltpd {{[0-9]+}}(%rsp), %xmm2
; SSE2-NEXT: shufps {{.*#+}} xmm2 = xmm2[0,2],xmm4[0,2]
; SSE2-NEXT: movapd {{[0-9]+}}(%rsp), %xmm4
-; SSE2-NEXT: packssdw %xmm2, %xmm2
-; SSE2-NEXT: movsd {{.*#+}} xmm2 = xmm3[0],xmm2[1]
+; SSE2-NEXT: packssdw %xmm3, %xmm2
; SSE2-NEXT: movapd {{[0-9]+}}(%rsp), %xmm3
; SSE2-NEXT: cmpltpd {{[0-9]+}}(%rsp), %xmm3
-; SSE2-NEXT: shufps {{.*#+}} xmm3 = xmm3[0,2,2,3]
-; SSE2-NEXT: pshuflw {{.*#+}} xmm3 = xmm3[0,2,2,3,4,5,6,7]
; SSE2-NEXT: cmpltpd {{[0-9]+}}(%rsp), %xmm4
-; SSE2-NEXT: shufps {{.*#+}} xmm4 = xmm4[0,2,2,3]
-; SSE2-NEXT: pshuflw {{.*#+}} xmm4 = xmm4[0,2,2,3,4,5,6,7]
-; SSE2-NEXT: punpckldq {{.*#+}} xmm4 = xmm4[0],xmm3[0],xmm4[1],xmm3[1]
+; SSE2-NEXT: shufps {{.*#+}} xmm4 = xmm4[0,2],xmm3[0,2]
; SSE2-NEXT: movapd {{[0-9]+}}(%rsp), %xmm3
; SSE2-NEXT: cmpltpd {{[0-9]+}}(%rsp), %xmm3
-; SSE2-NEXT: cmpltpd {{[0-9]+}}(%rsp), %xmm1
-; SSE2-NEXT: shufps {{.*#+}} xmm1 = xmm1[0,2],xmm3[0,2]
-; SSE2-NEXT: packssdw %xmm1, %xmm1
-; SSE2-NEXT: movsd {{.*#+}} xmm1 = xmm4[0],xmm1[1]
-; SSE2-NEXT: packsswb %xmm2, %xmm1
-; SSE2-NEXT: movdqa %xmm1, -{{[0-9]+}}(%rsp)
+; SSE2-NEXT: cmpltpd {{[0-9]+}}(%rsp), %xmm0
+; SSE2-NEXT: shufps {{.*#+}} xmm0 = xmm0[0,2],xmm3[0,2]
+; SSE2-NEXT: packssdw %xmm4, %xmm0
+; SSE2-NEXT: packsswb %xmm2, %xmm0
+; SSE2-NEXT: movdqa %xmm0, -{{[0-9]+}}(%rsp)
; SSE2-NEXT: movb -{{[0-9]+}}(%rsp), %al
; SSE2-NEXT: andb $1, %al
; SSE2-NEXT: movb %al, 2(%rdi)
@@ -6676,7 +6656,7 @@ define <32 x i1> @test_cmp_v32f64(<32 x
; SSE2-NEXT: movb %cl, 2(%rdi)
; SSE2-NEXT: andb $1, %al
; SSE2-NEXT: movb %al, 2(%rdi)
-; SSE2-NEXT: movdqa %xmm0, -{{[0-9]+}}(%rsp)
+; SSE2-NEXT: movdqa %xmm1, -{{[0-9]+}}(%rsp)
; SSE2-NEXT: movb -{{[0-9]+}}(%rsp), %al
; SSE2-NEXT: andb $1, %al
; SSE2-NEXT: movb %al, (%rdi)
@@ -6737,91 +6717,71 @@ define <32 x i1> @test_cmp_v32f64(<32 x
; SSE42-NEXT: pushq %r12
; SSE42-NEXT: pushq %rbx
; SSE42-NEXT: movapd {{[0-9]+}}(%rsp), %xmm8
-; SSE42-NEXT: cmpltpd %xmm5, %xmm8
-; SSE42-NEXT: movapd {{[0-9]+}}(%rsp), %xmm5
-; SSE42-NEXT: cmpltpd %xmm4, %xmm5
-; SSE42-NEXT: shufps {{.*#+}} xmm8 = xmm8[0,2,2,3]
-; SSE42-NEXT: pshuflw {{.*#+}} xmm4 = xmm8[0,2,2,3,4,5,6,7]
-; SSE42-NEXT: shufps {{.*#+}} xmm5 = xmm5[0,2,2,3]
-; SSE42-NEXT: pshuflw {{.*#+}} xmm5 = xmm5[0,2,2,3,4,5,6,7]
-; SSE42-NEXT: punpckldq {{.*#+}} xmm5 = xmm5[0],xmm4[0],xmm5[1],xmm4[1]
-; SSE42-NEXT: movapd {{[0-9]+}}(%rsp), %xmm4
-; SSE42-NEXT: cmpltpd %xmm7, %xmm4
+; SSE42-NEXT: cmpltpd %xmm7, %xmm8
; SSE42-NEXT: movapd {{[0-9]+}}(%rsp), %xmm7
; SSE42-NEXT: cmpltpd %xmm6, %xmm7
-; SSE42-NEXT: shufps {{.*#+}} xmm7 = xmm7[0,2],xmm4[0,2]
-; SSE42-NEXT: packssdw %xmm7, %xmm7
-; SSE42-NEXT: pblendw {{.*#+}} xmm7 = xmm5[0,1,2,3],xmm7[4,5,6,7]
+; SSE42-NEXT: shufps {{.*#+}} xmm7 = xmm7[0,2],xmm8[0,2]
+; SSE42-NEXT: movapd {{[0-9]+}}(%rsp), %xmm6
+; SSE42-NEXT: cmpltpd %xmm5, %xmm6
+; SSE42-NEXT: movapd {{[0-9]+}}(%rsp), %xmm5
+; SSE42-NEXT: cmpltpd %xmm4, %xmm5
+; SSE42-NEXT: shufps {{.*#+}} xmm5 = xmm5[0,2],xmm6[0,2]
+; SSE42-NEXT: packssdw %xmm7, %xmm5
; SSE42-NEXT: movapd {{[0-9]+}}(%rsp), %xmm4
-; SSE42-NEXT: cmpltpd %xmm1, %xmm4
+; SSE42-NEXT: cmpltpd %xmm3, %xmm4
+; SSE42-NEXT: movapd {{[0-9]+}}(%rsp), %xmm3
+; SSE42-NEXT: cmpltpd %xmm2, %xmm3
+; SSE42-NEXT: shufps {{.*#+}} xmm3 = xmm3[0,2],xmm4[0,2]
+; SSE42-NEXT: movapd {{[0-9]+}}(%rsp), %xmm2
+; SSE42-NEXT: cmpltpd %xmm1, %xmm2
; SSE42-NEXT: movapd {{[0-9]+}}(%rsp), %xmm1
; SSE42-NEXT: cmpltpd %xmm0, %xmm1
-; SSE42-NEXT: shufps {{.*#+}} xmm4 = xmm4[0,2,2,3]
-; SSE42-NEXT: pshuflw {{.*#+}} xmm0 = xmm4[0,2,2,3,4,5,6,7]
-; SSE42-NEXT: shufps {{.*#+}} xmm1 = xmm1[0,2,2,3]
-; SSE42-NEXT: pshuflw {{.*#+}} xmm4 = xmm1[0,2,2,3,4,5,6,7]
-; SSE42-NEXT: punpckldq {{.*#+}} xmm4 = xmm4[0],xmm0[0],xmm4[1],xmm0[1]
-; SSE42-NEXT: movapd {{[0-9]+}}(%rsp), %xmm5
-; SSE42-NEXT: cmpltpd %xmm3, %xmm5
; SSE42-NEXT: movapd {{[0-9]+}}(%rsp), %xmm0
-; SSE42-NEXT: cmpltpd %xmm2, %xmm0
-; SSE42-NEXT: movapd {{[0-9]+}}(%rsp), %xmm1
-; SSE42-NEXT: shufps {{.*#+}} xmm0 = xmm0[0,2],xmm5[0,2]
+; SSE42-NEXT: shufps {{.*#+}} xmm1 = xmm1[0,2],xmm2[0,2]
; SSE42-NEXT: movapd {{[0-9]+}}(%rsp), %xmm2
-; SSE42-NEXT: packssdw %xmm0, %xmm0
-; SSE42-NEXT: pblendw {{.*#+}} xmm0 = xmm4[0,1,2,3],xmm0[4,5,6,7]
+; SSE42-NEXT: packssdw %xmm3, %xmm1
; SSE42-NEXT: movapd {{[0-9]+}}(%rsp), %xmm3
-; SSE42-NEXT: packsswb %xmm7, %xmm0
+; SSE42-NEXT: packsswb %xmm5, %xmm1
; SSE42-NEXT: movapd {{[0-9]+}}(%rsp), %xmm4
; SSE42-NEXT: cmpltpd {{[0-9]+}}(%rsp), %xmm4
-; SSE42-NEXT: shufps {{.*#+}} xmm4 = xmm4[0,2,2,3]
-; SSE42-NEXT: pshuflw {{.*#+}} xmm4 = xmm4[0,2,2,3,4,5,6,7]
; SSE42-NEXT: cmpltpd {{[0-9]+}}(%rsp), %xmm3
-; SSE42-NEXT: shufps {{.*#+}} xmm3 = xmm3[0,2,2,3]
-; SSE42-NEXT: pshuflw {{.*#+}} xmm3 = xmm3[0,2,2,3,4,5,6,7]
-; SSE42-NEXT: punpckldq {{.*#+}} xmm3 = xmm3[0],xmm4[0],xmm3[1],xmm4[1]
+; SSE42-NEXT: shufps {{.*#+}} xmm3 = xmm3[0,2],xmm4[0,2]
; SSE42-NEXT: movapd {{[0-9]+}}(%rsp), %xmm4
; SSE42-NEXT: cmpltpd {{[0-9]+}}(%rsp), %xmm4
; SSE42-NEXT: cmpltpd {{[0-9]+}}(%rsp), %xmm2
; SSE42-NEXT: shufps {{.*#+}} xmm2 = xmm2[0,2],xmm4[0,2]
; SSE42-NEXT: movapd {{[0-9]+}}(%rsp), %xmm4
-; SSE42-NEXT: packssdw %xmm2, %xmm2
-; SSE42-NEXT: pblendw {{.*#+}} xmm2 = xmm3[0,1,2,3],xmm2[4,5,6,7]
+; SSE42-NEXT: packssdw %xmm3, %xmm2
; SSE42-NEXT: movapd {{[0-9]+}}(%rsp), %xmm3
; SSE42-NEXT: cmpltpd {{[0-9]+}}(%rsp), %xmm3
-; SSE42-NEXT: shufps {{.*#+}} xmm3 = xmm3[0,2,2,3]
-; SSE42-NEXT: pshuflw {{.*#+}} xmm3 = xmm3[0,2,2,3,4,5,6,7]
; SSE42-NEXT: cmpltpd {{[0-9]+}}(%rsp), %xmm4
-; SSE42-NEXT: shufps {{.*#+}} xmm4 = xmm4[0,2,2,3]
-; SSE42-NEXT: pshuflw {{.*#+}} xmm4 = xmm4[0,2,2,3,4,5,6,7]
-; SSE42-NEXT: punpckldq {{.*#+}} xmm4 = xmm4[0],xmm3[0],xmm4[1],xmm3[1]
+; SSE42-NEXT: shufps {{.*#+}} xmm4 = xmm4[0,2],xmm3[0,2]
; SSE42-NEXT: movapd {{[0-9]+}}(%rsp), %xmm3
; SSE42-NEXT: cmpltpd {{[0-9]+}}(%rsp), %xmm3
-; SSE42-NEXT: cmpltpd {{[0-9]+}}(%rsp), %xmm1
-; SSE42-NEXT: shufps {{.*#+}} xmm1 = xmm1[0,2],xmm3[0,2]
-; SSE42-NEXT: packssdw %xmm1, %xmm1
-; SSE42-NEXT: pblendw {{.*#+}} xmm1 = xmm4[0,1,2,3],xmm1[4,5,6,7]
-; SSE42-NEXT: packsswb %xmm2, %xmm1
-; SSE42-NEXT: pextrb $15, %xmm1, %eax
+; SSE42-NEXT: cmpltpd {{[0-9]+}}(%rsp), %xmm0
+; SSE42-NEXT: shufps {{.*#+}} xmm0 = xmm0[0,2],xmm3[0,2]
+; SSE42-NEXT: packssdw %xmm4, %xmm0
+; SSE42-NEXT: packsswb %xmm2, %xmm0
+; SSE42-NEXT: pextrb $15, %xmm0, %eax
; SSE42-NEXT: andb $1, %al
; SSE42-NEXT: movb %al, 2(%rdi)
-; SSE42-NEXT: pextrb $14, %xmm1, %eax
+; SSE42-NEXT: pextrb $14, %xmm0, %eax
; SSE42-NEXT: andb $1, %al
; SSE42-NEXT: movb %al, 2(%rdi)
-; SSE42-NEXT: pextrb $13, %xmm1, %r8d
-; SSE42-NEXT: pextrb $12, %xmm1, %r9d
-; SSE42-NEXT: pextrb $11, %xmm1, %r10d
-; SSE42-NEXT: pextrb $10, %xmm1, %r11d
-; SSE42-NEXT: pextrb $9, %xmm1, %r14d
-; SSE42-NEXT: pextrb $8, %xmm1, %r15d
-; SSE42-NEXT: pextrb $7, %xmm1, %r12d
-; SSE42-NEXT: pextrb $6, %xmm1, %r13d
-; SSE42-NEXT: pextrb $5, %xmm1, %ebx
-; SSE42-NEXT: pextrb $4, %xmm1, %ebp
-; SSE42-NEXT: pextrb $3, %xmm1, %eax
-; SSE42-NEXT: pextrb $2, %xmm1, %ecx
-; SSE42-NEXT: pextrb $1, %xmm1, %edx
-; SSE42-NEXT: pextrb $0, %xmm1, %esi
+; SSE42-NEXT: pextrb $13, %xmm0, %r8d
+; SSE42-NEXT: pextrb $12, %xmm0, %r9d
+; SSE42-NEXT: pextrb $11, %xmm0, %r10d
+; SSE42-NEXT: pextrb $10, %xmm0, %r11d
+; SSE42-NEXT: pextrb $9, %xmm0, %r14d
+; SSE42-NEXT: pextrb $8, %xmm0, %r15d
+; SSE42-NEXT: pextrb $7, %xmm0, %r12d
+; SSE42-NEXT: pextrb $6, %xmm0, %r13d
+; SSE42-NEXT: pextrb $5, %xmm0, %ebx
+; SSE42-NEXT: pextrb $4, %xmm0, %ebp
+; SSE42-NEXT: pextrb $3, %xmm0, %eax
+; SSE42-NEXT: pextrb $2, %xmm0, %ecx
+; SSE42-NEXT: pextrb $1, %xmm0, %edx
+; SSE42-NEXT: pextrb $0, %xmm0, %esi
; SSE42-NEXT: andb $1, %r8b
; SSE42-NEXT: movb %r8b, 2(%rdi)
; SSE42-NEXT: andb $1, %r9b
@@ -6850,26 +6810,26 @@ define <32 x i1> @test_cmp_v32f64(<32 x
; SSE42-NEXT: movb %dl, 2(%rdi)
; SSE42-NEXT: andb $1, %sil
; SSE42-NEXT: movb %sil, 2(%rdi)
-; SSE42-NEXT: pextrb $15, %xmm0, %eax
+; SSE42-NEXT: pextrb $15, %xmm1, %eax
; SSE42-NEXT: andb $1, %al
; SSE42-NEXT: movb %al, (%rdi)
-; SSE42-NEXT: pextrb $14, %xmm0, %eax
+; SSE42-NEXT: pextrb $14, %xmm1, %eax
; SSE42-NEXT: andb $1, %al
; SSE42-NEXT: movb %al, (%rdi)
-; SSE42-NEXT: pextrb $13, %xmm0, %r8d
-; SSE42-NEXT: pextrb $12, %xmm0, %r9d
-; SSE42-NEXT: pextrb $11, %xmm0, %r10d
-; SSE42-NEXT: pextrb $10, %xmm0, %r11d
-; SSE42-NEXT: pextrb $9, %xmm0, %r14d
-; SSE42-NEXT: pextrb $8, %xmm0, %r15d
-; SSE42-NEXT: pextrb $7, %xmm0, %r12d
-; SSE42-NEXT: pextrb $6, %xmm0, %r13d
-; SSE42-NEXT: pextrb $5, %xmm0, %ebx
-; SSE42-NEXT: pextrb $4, %xmm0, %ebp
-; SSE42-NEXT: pextrb $3, %xmm0, %eax
-; SSE42-NEXT: pextrb $2, %xmm0, %ecx
-; SSE42-NEXT: pextrb $1, %xmm0, %edx
-; SSE42-NEXT: pextrb $0, %xmm0, %esi
+; SSE42-NEXT: pextrb $13, %xmm1, %r8d
+; SSE42-NEXT: pextrb $12, %xmm1, %r9d
+; SSE42-NEXT: pextrb $11, %xmm1, %r10d
+; SSE42-NEXT: pextrb $10, %xmm1, %r11d
+; SSE42-NEXT: pextrb $9, %xmm1, %r14d
+; SSE42-NEXT: pextrb $8, %xmm1, %r15d
+; SSE42-NEXT: pextrb $7, %xmm1, %r12d
+; SSE42-NEXT: pextrb $6, %xmm1, %r13d
+; SSE42-NEXT: pextrb $5, %xmm1, %ebx
+; SSE42-NEXT: pextrb $4, %xmm1, %ebp
+; SSE42-NEXT: pextrb $3, %xmm1, %eax
+; SSE42-NEXT: pextrb $2, %xmm1, %ecx
+; SSE42-NEXT: pextrb $1, %xmm1, %edx
+; SSE42-NEXT: pextrb $0, %xmm1, %esi
; SSE42-NEXT: andb $1, %r8b
; SSE42-NEXT: movb %r8b, (%rdi)
; SSE42-NEXT: andb $1, %r9b
@@ -7294,117 +7254,115 @@ define <32 x i1> @test_cmp_v32i64(<32 x
; SSE2-LABEL: test_cmp_v32i64:
; SSE2: # %bb.0:
; SSE2-NEXT: movdqa {{.*#+}} xmm8 = [2147483648,0,2147483648,0]
-; SSE2-NEXT: pxor %xmm8, %xmm5
+; SSE2-NEXT: pxor %xmm8, %xmm7
; SSE2-NEXT: movdqa {{[0-9]+}}(%rsp), %xmm9
; SSE2-NEXT: pxor %xmm8, %xmm9
-; SSE2-NEXT: movdqa %xmm5, %xmm10
+; SSE2-NEXT: movdqa %xmm7, %xmm10
; SSE2-NEXT: pcmpgtd %xmm9, %xmm10
; SSE2-NEXT: pshufd {{.*#+}} xmm11 = xmm10[0,0,2,2]
-; SSE2-NEXT: pcmpeqd %xmm5, %xmm9
-; SSE2-NEXT: pshufd {{.*#+}} xmm9 = xmm9[1,1,3,3]
-; SSE2-NEXT: pand %xmm11, %xmm9
-; SSE2-NEXT: pshufd {{.*#+}} xmm5 = xmm10[1,1,3,3]
-; SSE2-NEXT: por %xmm9, %xmm5
-; SSE2-NEXT: pshufd {{.*#+}} xmm5 = xmm5[0,2,2,3]
-; SSE2-NEXT: pshuflw {{.*#+}} xmm9 = xmm5[0,2,2,3,4,5,6,7]
+; SSE2-NEXT: pcmpeqd %xmm7, %xmm9
+; SSE2-NEXT: pshufd {{.*#+}} xmm7 = xmm9[1,1,3,3]
+; SSE2-NEXT: pand %xmm11, %xmm7
+; SSE2-NEXT: pshufd {{.*#+}} xmm9 = xmm10[1,1,3,3]
+; SSE2-NEXT: por %xmm7, %xmm9
+; SSE2-NEXT: pxor %xmm8, %xmm6
+; SSE2-NEXT: movdqa {{[0-9]+}}(%rsp), %xmm7
+; SSE2-NEXT: pxor %xmm8, %xmm7
+; SSE2-NEXT: movdqa %xmm6, %xmm10
+; SSE2-NEXT: pcmpgtd %xmm7, %xmm10
+; SSE2-NEXT: pshufd {{.*#+}} xmm11 = xmm10[0,0,2,2]
+; SSE2-NEXT: pcmpeqd %xmm6, %xmm7
+; SSE2-NEXT: pshufd {{.*#+}} xmm6 = xmm7[1,1,3,3]
+; SSE2-NEXT: pand %xmm11, %xmm6
+; SSE2-NEXT: pshufd {{.*#+}} xmm7 = xmm10[1,1,3,3]
+; SSE2-NEXT: por %xmm6, %xmm7
+; SSE2-NEXT: shufps {{.*#+}} xmm7 = xmm7[0,2],xmm9[0,2]
+; SSE2-NEXT: pshuflw {{.*#+}} xmm6 = xmm7[0,2,2,3,4,5,6,7]
+; SSE2-NEXT: pshufhw {{.*#+}} xmm6 = xmm6[0,1,2,3,4,6,6,7]
+; SSE2-NEXT: pshufd {{.*#+}} xmm9 = xmm6[0,2,2,3]
+; SSE2-NEXT: pxor %xmm8, %xmm5
+; SSE2-NEXT: movdqa {{[0-9]+}}(%rsp), %xmm7
+; SSE2-NEXT: pxor %xmm8, %xmm7
+; SSE2-NEXT: movdqa %xmm5, %xmm6
+; SSE2-NEXT: pcmpgtd %xmm7, %xmm6
+; SSE2-NEXT: pshufd {{.*#+}} xmm10 = xmm6[0,0,2,2]
+; SSE2-NEXT: pcmpeqd %xmm5, %xmm7
+; SSE2-NEXT: pshufd {{.*#+}} xmm5 = xmm7[1,1,3,3]
+; SSE2-NEXT: pand %xmm10, %xmm5
+; SSE2-NEXT: pshufd {{.*#+}} xmm6 = xmm6[1,1,3,3]
+; SSE2-NEXT: por %xmm5, %xmm6
; SSE2-NEXT: pxor %xmm8, %xmm4
; SSE2-NEXT: movdqa {{[0-9]+}}(%rsp), %xmm5
; SSE2-NEXT: pxor %xmm8, %xmm5
-; SSE2-NEXT: movdqa %xmm4, %xmm10
-; SSE2-NEXT: pcmpgtd %xmm5, %xmm10
-; SSE2-NEXT: pshufd {{.*#+}} xmm11 = xmm10[0,0,2,2]
+; SSE2-NEXT: movdqa %xmm4, %xmm7
+; SSE2-NEXT: pcmpgtd %xmm5, %xmm7
+; SSE2-NEXT: pshufd {{.*#+}} xmm10 = xmm7[0,0,2,2]
; SSE2-NEXT: pcmpeqd %xmm4, %xmm5
; SSE2-NEXT: pshufd {{.*#+}} xmm4 = xmm5[1,1,3,3]
-; SSE2-NEXT: pand %xmm11, %xmm4
-; SSE2-NEXT: pshufd {{.*#+}} xmm5 = xmm10[1,1,3,3]
+; SSE2-NEXT: pand %xmm10, %xmm4
+; SSE2-NEXT: pshufd {{.*#+}} xmm5 = xmm7[1,1,3,3]
; SSE2-NEXT: por %xmm4, %xmm5
-; SSE2-NEXT: pshufd {{.*#+}} xmm4 = xmm5[0,2,2,3]
-; SSE2-NEXT: pshuflw {{.*#+}} xmm10 = xmm4[0,2,2,3,4,5,6,7]
-; SSE2-NEXT: punpckldq {{.*#+}} xmm10 = xmm10[0],xmm9[0],xmm10[1],xmm9[1]
-; SSE2-NEXT: pxor %xmm8, %xmm7
-; SSE2-NEXT: movdqa {{[0-9]+}}(%rsp), %xmm5
-; SSE2-NEXT: pxor %xmm8, %xmm5
-; SSE2-NEXT: movdqa %xmm7, %xmm4
-; SSE2-NEXT: pcmpgtd %xmm5, %xmm4
-; SSE2-NEXT: pshufd {{.*#+}} xmm9 = xmm4[0,0,2,2]
-; SSE2-NEXT: pcmpeqd %xmm7, %xmm5
-; SSE2-NEXT: pshufd {{.*#+}} xmm5 = xmm5[1,1,3,3]
-; SSE2-NEXT: pand %xmm9, %xmm5
-; SSE2-NEXT: pshufd {{.*#+}} xmm4 = xmm4[1,1,3,3]
-; SSE2-NEXT: por %xmm5, %xmm4
-; SSE2-NEXT: pxor %xmm8, %xmm6
-; SSE2-NEXT: movdqa {{[0-9]+}}(%rsp), %xmm5
-; SSE2-NEXT: pxor %xmm8, %xmm5
-; SSE2-NEXT: movdqa %xmm6, %xmm7
-; SSE2-NEXT: pcmpgtd %xmm5, %xmm7
-; SSE2-NEXT: pshufd {{.*#+}} xmm9 = xmm7[0,0,2,2]
-; SSE2-NEXT: pcmpeqd %xmm6, %xmm5
-; SSE2-NEXT: pshufd {{.*#+}} xmm5 = xmm5[1,1,3,3]
-; SSE2-NEXT: pand %xmm9, %xmm5
-; SSE2-NEXT: pshufd {{.*#+}} xmm6 = xmm7[1,1,3,3]
-; SSE2-NEXT: por %xmm5, %xmm6
-; SSE2-NEXT: shufps {{.*#+}} xmm6 = xmm6[0,2],xmm4[0,2]
-; SSE2-NEXT: pshuflw {{.*#+}} xmm4 = xmm6[0,2,2,3,4,5,6,7]
+; SSE2-NEXT: shufps {{.*#+}} xmm5 = xmm5[0,2],xmm6[0,2]
+; SSE2-NEXT: pshuflw {{.*#+}} xmm4 = xmm5[0,2,2,3,4,5,6,7]
; SSE2-NEXT: pshufhw {{.*#+}} xmm4 = xmm4[0,1,2,3,4,6,6,7]
-; SSE2-NEXT: pshufd {{.*#+}} xmm5 = xmm4[0,1,0,2]
-; SSE2-NEXT: movsd {{.*#+}} xmm5 = xmm10[0],xmm5[1]
-; SSE2-NEXT: movapd {{.*#+}} xmm9 = [255,255,255,255,255,255,255,255]
-; SSE2-NEXT: andpd %xmm9, %xmm5
-; SSE2-NEXT: pxor %xmm8, %xmm1
+; SSE2-NEXT: pshufd {{.*#+}} xmm5 = xmm4[0,2,2,3]
+; SSE2-NEXT: punpcklqdq {{.*#+}} xmm5 = xmm5[0],xmm9[0]
+; SSE2-NEXT: movdqa {{.*#+}} xmm9 = [255,255,255,255,255,255,255,255]
+; SSE2-NEXT: pand %xmm9, %xmm5
+; SSE2-NEXT: pxor %xmm8, %xmm3
; SSE2-NEXT: movdqa {{[0-9]+}}(%rsp), %xmm6
; SSE2-NEXT: pxor %xmm8, %xmm6
-; SSE2-NEXT: movdqa %xmm1, %xmm7
+; SSE2-NEXT: movdqa %xmm3, %xmm7
; SSE2-NEXT: pcmpgtd %xmm6, %xmm7
; SSE2-NEXT: pshufd {{.*#+}} xmm4 = xmm7[0,0,2,2]
-; SSE2-NEXT: pcmpeqd %xmm1, %xmm6
-; SSE2-NEXT: pshufd {{.*#+}} xmm1 = xmm6[1,1,3,3]
-; SSE2-NEXT: pand %xmm4, %xmm1
+; SSE2-NEXT: pcmpeqd %xmm3, %xmm6
+; SSE2-NEXT: pshufd {{.*#+}} xmm3 = xmm6[1,1,3,3]
+; SSE2-NEXT: pand %xmm4, %xmm3
; SSE2-NEXT: pshufd {{.*#+}} xmm4 = xmm7[1,1,3,3]
-; SSE2-NEXT: por %xmm1, %xmm4
-; SSE2-NEXT: pshufd {{.*#+}} xmm1 = xmm4[0,2,2,3]
-; SSE2-NEXT: pshuflw {{.*#+}} xmm4 = xmm1[0,2,2,3,4,5,6,7]
-; SSE2-NEXT: pxor %xmm8, %xmm0
-; SSE2-NEXT: movdqa {{[0-9]+}}(%rsp), %xmm1
-; SSE2-NEXT: pxor %xmm8, %xmm1
-; SSE2-NEXT: movdqa %xmm0, %xmm6
-; SSE2-NEXT: pcmpgtd %xmm1, %xmm6
+; SSE2-NEXT: por %xmm3, %xmm4
+; SSE2-NEXT: pxor %xmm8, %xmm2
+; SSE2-NEXT: movdqa {{[0-9]+}}(%rsp), %xmm3
+; SSE2-NEXT: pxor %xmm8, %xmm3
+; SSE2-NEXT: movdqa %xmm2, %xmm6
+; SSE2-NEXT: pcmpgtd %xmm3, %xmm6
; SSE2-NEXT: pshufd {{.*#+}} xmm7 = xmm6[0,0,2,2]
-; SSE2-NEXT: pcmpeqd %xmm0, %xmm1
-; SSE2-NEXT: pshufd {{.*#+}} xmm0 = xmm1[1,1,3,3]
-; SSE2-NEXT: pand %xmm7, %xmm0
-; SSE2-NEXT: pshufd {{.*#+}} xmm1 = xmm6[1,1,3,3]
-; SSE2-NEXT: por %xmm0, %xmm1
-; SSE2-NEXT: pshufd {{.*#+}} xmm0 = xmm1[0,2,2,3]
-; SSE2-NEXT: pshuflw {{.*#+}} xmm1 = xmm0[0,2,2,3,4,5,6,7]
-; SSE2-NEXT: punpckldq {{.*#+}} xmm1 = xmm1[0],xmm4[0],xmm1[1],xmm4[1]
+; SSE2-NEXT: pcmpeqd %xmm2, %xmm3
+; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm3[1,1,3,3]
+; SSE2-NEXT: pand %xmm7, %xmm2
+; SSE2-NEXT: pshufd {{.*#+}} xmm3 = xmm6[1,1,3,3]
+; SSE2-NEXT: por %xmm2, %xmm3
+; SSE2-NEXT: shufps {{.*#+}} xmm3 = xmm3[0,2],xmm4[0,2]
+; SSE2-NEXT: pshuflw {{.*#+}} xmm2 = xmm3[0,2,2,3,4,5,6,7]
+; SSE2-NEXT: pshufhw {{.*#+}} xmm2 = xmm2[0,1,2,3,4,6,6,7]
+; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm2[0,2,2,3]
+; SSE2-NEXT: pxor %xmm8, %xmm1
+; SSE2-NEXT: movdqa {{[0-9]+}}(%rsp), %xmm3
; SSE2-NEXT: pxor %xmm8, %xmm3
-; SSE2-NEXT: movdqa {{[0-9]+}}(%rsp), %xmm0
-; SSE2-NEXT: pxor %xmm8, %xmm0
-; SSE2-NEXT: movdqa %xmm3, %xmm4
-; SSE2-NEXT: pcmpgtd %xmm0, %xmm4
+; SSE2-NEXT: movdqa %xmm1, %xmm4
+; SSE2-NEXT: pcmpgtd %xmm3, %xmm4
; SSE2-NEXT: pshufd {{.*#+}} xmm6 = xmm4[0,0,2,2]
-; SSE2-NEXT: pcmpeqd %xmm3, %xmm0
-; SSE2-NEXT: pshufd {{.*#+}} xmm0 = xmm0[1,1,3,3]
-; SSE2-NEXT: pand %xmm6, %xmm0
+; SSE2-NEXT: pcmpeqd %xmm1, %xmm3
+; SSE2-NEXT: pshufd {{.*#+}} xmm1 = xmm3[1,1,3,3]
+; SSE2-NEXT: pand %xmm6, %xmm1
; SSE2-NEXT: pshufd {{.*#+}} xmm3 = xmm4[1,1,3,3]
-; SSE2-NEXT: por %xmm0, %xmm3
-; SSE2-NEXT: pxor %xmm8, %xmm2
-; SSE2-NEXT: movdqa {{[0-9]+}}(%rsp), %xmm0
+; SSE2-NEXT: por %xmm1, %xmm3
; SSE2-NEXT: pxor %xmm8, %xmm0
-; SSE2-NEXT: movdqa %xmm2, %xmm4
-; SSE2-NEXT: pcmpgtd %xmm0, %xmm4
+; SSE2-NEXT: movdqa {{[0-9]+}}(%rsp), %xmm1
+; SSE2-NEXT: pxor %xmm8, %xmm1
+; SSE2-NEXT: movdqa %xmm0, %xmm4
+; SSE2-NEXT: pcmpgtd %xmm1, %xmm4
; SSE2-NEXT: pshufd {{.*#+}} xmm6 = xmm4[0,0,2,2]
-; SSE2-NEXT: pcmpeqd %xmm2, %xmm0
-; SSE2-NEXT: pshufd {{.*#+}} xmm0 = xmm0[1,1,3,3]
+; SSE2-NEXT: pcmpeqd %xmm0, %xmm1
+; SSE2-NEXT: pshufd {{.*#+}} xmm0 = xmm1[1,1,3,3]
; SSE2-NEXT: pand %xmm6, %xmm0
-; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm4[1,1,3,3]
-; SSE2-NEXT: por %xmm0, %xmm2
-; SSE2-NEXT: shufps {{.*#+}} xmm2 = xmm2[0,2],xmm3[0,2]
-; SSE2-NEXT: pshuflw {{.*#+}} xmm0 = xmm2[0,2,2,3,4,5,6,7]
+; SSE2-NEXT: pshufd {{.*#+}} xmm1 = xmm4[1,1,3,3]
+; SSE2-NEXT: por %xmm0, %xmm1
+; SSE2-NEXT: shufps {{.*#+}} xmm1 = xmm1[0,2],xmm3[0,2]
+; SSE2-NEXT: pshuflw {{.*#+}} xmm0 = xmm1[0,2,2,3,4,5,6,7]
; SSE2-NEXT: pshufhw {{.*#+}} xmm0 = xmm0[0,1,2,3,4,6,6,7]
-; SSE2-NEXT: pshufd {{.*#+}} xmm0 = xmm0[0,1,0,2]
-; SSE2-NEXT: movsd {{.*#+}} xmm0 = xmm1[0],xmm0[1]
-; SSE2-NEXT: andpd %xmm9, %xmm0
+; SSE2-NEXT: pshufd {{.*#+}} xmm0 = xmm0[0,2,2,3]
+; SSE2-NEXT: punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm2[0]
+; SSE2-NEXT: pand %xmm9, %xmm0
; SSE2-NEXT: packuswb %xmm5, %xmm0
; SSE2-NEXT: movdqa {{[0-9]+}}(%rsp), %xmm1
; SSE2-NEXT: pxor %xmm8, %xmm1
@@ -7418,23 +7376,22 @@ define <32 x i1> @test_cmp_v32i64(<32 x
; SSE2-NEXT: pand %xmm4, %xmm1
; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm3[1,1,3,3]
; SSE2-NEXT: por %xmm1, %xmm2
-; SSE2-NEXT: pshufd {{.*#+}} xmm1 = xmm2[0,2,2,3]
-; SSE2-NEXT: pshuflw {{.*#+}} xmm1 = xmm1[0,2,2,3,4,5,6,7]
-; SSE2-NEXT: movdqa {{[0-9]+}}(%rsp), %xmm2
-; SSE2-NEXT: pxor %xmm8, %xmm2
+; SSE2-NEXT: movdqa {{[0-9]+}}(%rsp), %xmm1
+; SSE2-NEXT: pxor %xmm8, %xmm1
; SSE2-NEXT: movdqa {{[0-9]+}}(%rsp), %xmm3
; SSE2-NEXT: pxor %xmm8, %xmm3
; SSE2-NEXT: movdqa %xmm3, %xmm4
-; SSE2-NEXT: pcmpgtd %xmm2, %xmm4
+; SSE2-NEXT: pcmpgtd %xmm1, %xmm4
; SSE2-NEXT: pshufd {{.*#+}} xmm5 = xmm4[0,0,2,2]
-; SSE2-NEXT: pcmpeqd %xmm2, %xmm3
-; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm3[1,1,3,3]
-; SSE2-NEXT: pand %xmm5, %xmm2
+; SSE2-NEXT: pcmpeqd %xmm1, %xmm3
+; SSE2-NEXT: pshufd {{.*#+}} xmm1 = xmm3[1,1,3,3]
+; SSE2-NEXT: pand %xmm5, %xmm1
; SSE2-NEXT: pshufd {{.*#+}} xmm3 = xmm4[1,1,3,3]
-; SSE2-NEXT: por %xmm2, %xmm3
-; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm3[0,2,2,3]
-; SSE2-NEXT: pshuflw {{.*#+}} xmm2 = xmm2[0,2,2,3,4,5,6,7]
-; SSE2-NEXT: punpckldq {{.*#+}} xmm2 = xmm2[0],xmm1[0],xmm2[1],xmm1[1]
+; SSE2-NEXT: por %xmm1, %xmm3
+; SSE2-NEXT: shufps {{.*#+}} xmm3 = xmm3[0,2],xmm2[0,2]
+; SSE2-NEXT: pshuflw {{.*#+}} xmm1 = xmm3[0,2,2,3,4,5,6,7]
+; SSE2-NEXT: pshufhw {{.*#+}} xmm1 = xmm1[0,1,2,3,4,6,6,7]
+; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm1[0,2,2,3]
; SSE2-NEXT: movdqa {{[0-9]+}}(%rsp), %xmm1
; SSE2-NEXT: pxor %xmm8, %xmm1
; SSE2-NEXT: movdqa {{[0-9]+}}(%rsp), %xmm3
@@ -7462,9 +7419,9 @@ define <32 x i1> @test_cmp_v32i64(<32 x
; SSE2-NEXT: shufps {{.*#+}} xmm4 = xmm4[0,2],xmm3[0,2]
; SSE2-NEXT: pshuflw {{.*#+}} xmm1 = xmm4[0,2,2,3,4,5,6,7]
; SSE2-NEXT: pshufhw {{.*#+}} xmm1 = xmm1[0,1,2,3,4,6,6,7]
-; SSE2-NEXT: pshufd {{.*#+}} xmm1 = xmm1[0,1,0,2]
-; SSE2-NEXT: movsd {{.*#+}} xmm1 = xmm2[0],xmm1[1]
-; SSE2-NEXT: andpd %xmm9, %xmm1
+; SSE2-NEXT: pshufd {{.*#+}} xmm1 = xmm1[0,2,2,3]
+; SSE2-NEXT: punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm2[0]
+; SSE2-NEXT: pand %xmm9, %xmm1
; SSE2-NEXT: movdqa {{[0-9]+}}(%rsp), %xmm2
; SSE2-NEXT: pxor %xmm8, %xmm2
; SSE2-NEXT: movdqa {{[0-9]+}}(%rsp), %xmm3
@@ -7477,8 +7434,6 @@ define <32 x i1> @test_cmp_v32i64(<32 x
; SSE2-NEXT: pand %xmm5, %xmm2
; SSE2-NEXT: pshufd {{.*#+}} xmm3 = xmm4[1,1,3,3]
; SSE2-NEXT: por %xmm2, %xmm3
-; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm3[0,2,2,3]
-; SSE2-NEXT: pshuflw {{.*#+}} xmm3 = xmm2[0,2,2,3,4,5,6,7]
; SSE2-NEXT: movdqa {{[0-9]+}}(%rsp), %xmm2
; SSE2-NEXT: pxor %xmm8, %xmm2
; SSE2-NEXT: movdqa {{[0-9]+}}(%rsp), %xmm4
@@ -7491,9 +7446,10 @@ define <32 x i1> @test_cmp_v32i64(<32 x
; SSE2-NEXT: pand %xmm6, %xmm2
; SSE2-NEXT: pshufd {{.*#+}} xmm4 = xmm5[1,1,3,3]
; SSE2-NEXT: por %xmm2, %xmm4
-; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm4[0,2,2,3]
-; SSE2-NEXT: pshuflw {{.*#+}} xmm2 = xmm2[0,2,2,3,4,5,6,7]
-; SSE2-NEXT: punpckldq {{.*#+}} xmm2 = xmm2[0],xmm3[0],xmm2[1],xmm3[1]
+; SSE2-NEXT: shufps {{.*#+}} xmm4 = xmm4[0,2],xmm3[0,2]
+; SSE2-NEXT: pshuflw {{.*#+}} xmm2 = xmm4[0,2,2,3,4,5,6,7]
+; SSE2-NEXT: pshufhw {{.*#+}} xmm2 = xmm2[0,1,2,3,4,6,6,7]
+; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm2[0,2,2,3]
; SSE2-NEXT: movdqa {{[0-9]+}}(%rsp), %xmm3
; SSE2-NEXT: pxor %xmm8, %xmm3
; SSE2-NEXT: movdqa {{[0-9]+}}(%rsp), %xmm4
@@ -7520,9 +7476,9 @@ define <32 x i1> @test_cmp_v32i64(<32 x
; SSE2-NEXT: shufps {{.*#+}} xmm5 = xmm5[0,2],xmm4[0,2]
; SSE2-NEXT: pshuflw {{.*#+}} xmm3 = xmm5[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[0,1,0,2]
-; SSE2-NEXT: movsd {{.*#+}} xmm3 = xmm2[0],xmm3[1]
-; SSE2-NEXT: andpd %xmm9, %xmm3
+; SSE2-NEXT: pshufd {{.*#+}} xmm3 = xmm3[0,2,2,3]
+; SSE2-NEXT: punpcklqdq {{.*#+}} xmm3 = xmm3[0],xmm2[0]
+; SSE2-NEXT: pand %xmm9, %xmm3
; SSE2-NEXT: packuswb %xmm1, %xmm3
; SSE2-NEXT: movdqa %xmm3, -{{[0-9]+}}(%rsp)
; SSE2-NEXT: movb -{{[0-9]+}}(%rsp), %al
@@ -7628,62 +7584,42 @@ define <32 x i1> @test_cmp_v32i64(<32 x
; SSE42-LABEL: test_cmp_v32i64:
; SSE42: # %bb.0:
; SSE42-NEXT: movdqa {{[0-9]+}}(%rsp), %xmm8
-; SSE42-NEXT: movdqa {{[0-9]+}}(%rsp), %xmm10
; SSE42-NEXT: movdqa {{[0-9]+}}(%rsp), %xmm11
+; SSE42-NEXT: movdqa {{[0-9]+}}(%rsp), %xmm10
; SSE42-NEXT: movdqa {{[0-9]+}}(%rsp), %xmm12
; SSE42-NEXT: movdqa {{[0-9]+}}(%rsp), %xmm9
-; SSE42-NEXT: movdqa {{[0-9]+}}(%rsp), %xmm13
; SSE42-NEXT: movdqa {{[0-9]+}}(%rsp), %xmm14
+; SSE42-NEXT: movdqa {{[0-9]+}}(%rsp), %xmm13
; SSE42-NEXT: movdqa {{[0-9]+}}(%rsp), %xmm15
-; SSE42-NEXT: pcmpgtq {{[0-9]+}}(%rsp), %xmm5
-; SSE42-NEXT: pshufd {{.*#+}} xmm5 = xmm5[0,2,2,3]
-; SSE42-NEXT: pshuflw {{.*#+}} xmm5 = xmm5[0,2,2,3,4,5,6,7]
-; SSE42-NEXT: pcmpgtq {{[0-9]+}}(%rsp), %xmm4
-; SSE42-NEXT: pshufd {{.*#+}} xmm4 = xmm4[0,2,2,3]
-; SSE42-NEXT: pshuflw {{.*#+}} xmm4 = xmm4[0,2,2,3,4,5,6,7]
-; SSE42-NEXT: punpckldq {{.*#+}} xmm4 = xmm4[0],xmm5[0],xmm4[1],xmm5[1]
; SSE42-NEXT: pcmpgtq {{[0-9]+}}(%rsp), %xmm7
; SSE42-NEXT: pcmpgtq {{[0-9]+}}(%rsp), %xmm6
; SSE42-NEXT: shufps {{.*#+}} xmm6 = xmm6[0,2],xmm7[0,2]
-; SSE42-NEXT: packssdw %xmm6, %xmm6
-; SSE42-NEXT: pblendw {{.*#+}} xmm6 = xmm4[0,1,2,3],xmm6[4,5,6,7]
-; SSE42-NEXT: pcmpgtq {{[0-9]+}}(%rsp), %xmm1
-; SSE42-NEXT: pshufd {{.*#+}} xmm1 = xmm1[0,2,2,3]
-; SSE42-NEXT: pshuflw {{.*#+}} xmm1 = xmm1[0,2,2,3,4,5,6,7]
-; SSE42-NEXT: pcmpgtq {{[0-9]+}}(%rsp), %xmm0
-; SSE42-NEXT: pshufd {{.*#+}} xmm0 = xmm0[0,2,2,3]
-; SSE42-NEXT: pshuflw {{.*#+}} xmm0 = xmm0[0,2,2,3,4,5,6,7]
-; SSE42-NEXT: punpckldq {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; SSE42-NEXT: pcmpgtq {{[0-9]+}}(%rsp), %xmm5
+; SSE42-NEXT: pcmpgtq {{[0-9]+}}(%rsp), %xmm4
+; SSE42-NEXT: shufps {{.*#+}} xmm4 = xmm4[0,2],xmm5[0,2]
+; SSE42-NEXT: packssdw %xmm6, %xmm4
; SSE42-NEXT: pcmpgtq {{[0-9]+}}(%rsp), %xmm3
; SSE42-NEXT: pcmpgtq {{[0-9]+}}(%rsp), %xmm2
; SSE42-NEXT: shufps {{.*#+}} xmm2 = xmm2[0,2],xmm3[0,2]
-; SSE42-NEXT: packssdw %xmm2, %xmm2
-; SSE42-NEXT: pblendw {{.*#+}} xmm2 = xmm0[0,1,2,3],xmm2[4,5,6,7]
-; SSE42-NEXT: packsswb %xmm6, %xmm2
+; SSE42-NEXT: pcmpgtq {{[0-9]+}}(%rsp), %xmm1
+; SSE42-NEXT: pcmpgtq {{[0-9]+}}(%rsp), %xmm0
+; SSE42-NEXT: shufps {{.*#+}} xmm0 = xmm0[0,2],xmm1[0,2]
+; SSE42-NEXT: packssdw %xmm2, %xmm0
+; SSE42-NEXT: packsswb %xmm4, %xmm0
; SSE42-NEXT: pcmpgtq {{[0-9]+}}(%rsp), %xmm15
-; SSE42-NEXT: pshufd {{.*#+}} xmm0 = xmm15[0,2,2,3]
-; SSE42-NEXT: pshuflw {{.*#+}} xmm0 = xmm0[0,2,2,3,4,5,6,7]
-; SSE42-NEXT: pcmpgtq {{[0-9]+}}(%rsp), %xmm14
-; SSE42-NEXT: pshufd {{.*#+}} xmm1 = xmm14[0,2,2,3]
-; SSE42-NEXT: pshuflw {{.*#+}} xmm1 = xmm1[0,2,2,3,4,5,6,7]
-; SSE42-NEXT: punpckldq {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1]
; SSE42-NEXT: pcmpgtq {{[0-9]+}}(%rsp), %xmm13
+; SSE42-NEXT: shufps {{.*#+}} xmm13 = xmm13[0,2],xmm15[0,2]
+; SSE42-NEXT: pcmpgtq {{[0-9]+}}(%rsp), %xmm14
; SSE42-NEXT: pcmpgtq {{[0-9]+}}(%rsp), %xmm9
-; SSE42-NEXT: shufps {{.*#+}} xmm9 = xmm9[0,2],xmm13[0,2]
-; SSE42-NEXT: packssdw %xmm9, %xmm9
-; SSE42-NEXT: pblendw {{.*#+}} xmm9 = xmm1[0,1,2,3],xmm9[4,5,6,7]
+; SSE42-NEXT: shufps {{.*#+}} xmm9 = xmm9[0,2],xmm14[0,2]
+; SSE42-NEXT: packssdw %xmm13, %xmm9
; SSE42-NEXT: pcmpgtq {{[0-9]+}}(%rsp), %xmm12
-; SSE42-NEXT: pshufd {{.*#+}} xmm0 = xmm12[0,2,2,3]
-; SSE42-NEXT: pshuflw {{.*#+}} xmm0 = xmm0[0,2,2,3,4,5,6,7]
-; SSE42-NEXT: pcmpgtq {{[0-9]+}}(%rsp), %xmm11
-; SSE42-NEXT: pshufd {{.*#+}} xmm1 = xmm11[0,2,2,3]
-; SSE42-NEXT: pshuflw {{.*#+}} xmm1 = xmm1[0,2,2,3,4,5,6,7]
-; SSE42-NEXT: punpckldq {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1]
; SSE42-NEXT: pcmpgtq {{[0-9]+}}(%rsp), %xmm10
+; SSE42-NEXT: shufps {{.*#+}} xmm10 = xmm10[0,2],xmm12[0,2]
+; SSE42-NEXT: pcmpgtq {{[0-9]+}}(%rsp), %xmm11
; SSE42-NEXT: pcmpgtq {{[0-9]+}}(%rsp), %xmm8
-; SSE42-NEXT: shufps {{.*#+}} xmm8 = xmm8[0,2],xmm10[0,2]
-; SSE42-NEXT: packssdw %xmm8, %xmm8
-; SSE42-NEXT: pblendw {{.*#+}} xmm8 = xmm1[0,1,2,3],xmm8[4,5,6,7]
+; SSE42-NEXT: shufps {{.*#+}} xmm8 = xmm8[0,2],xmm11[0,2]
+; SSE42-NEXT: packssdw %xmm10, %xmm8
; SSE42-NEXT: packsswb %xmm9, %xmm8
; SSE42-NEXT: pextrb $15, %xmm8, %eax
; SSE42-NEXT: andb $1, %al
@@ -7733,52 +7669,52 @@ define <32 x i1> @test_cmp_v32i64(<32 x
; SSE42-NEXT: pextrb $0, %xmm8, %eax
; SSE42-NEXT: andb $1, %al
; SSE42-NEXT: movb %al, 2(%rdi)
-; SSE42-NEXT: pextrb $15, %xmm2, %eax
+; SSE42-NEXT: pextrb $15, %xmm0, %eax
; SSE42-NEXT: andb $1, %al
; SSE42-NEXT: movb %al, (%rdi)
-; SSE42-NEXT: pextrb $14, %xmm2, %eax
+; SSE42-NEXT: pextrb $14, %xmm0, %eax
; SSE42-NEXT: andb $1, %al
; SSE42-NEXT: movb %al, (%rdi)
-; SSE42-NEXT: pextrb $13, %xmm2, %eax
+; SSE42-NEXT: pextrb $13, %xmm0, %eax
; SSE42-NEXT: andb $1, %al
; SSE42-NEXT: movb %al, (%rdi)
-; SSE42-NEXT: pextrb $12, %xmm2, %eax
+; SSE42-NEXT: pextrb $12, %xmm0, %eax
; SSE42-NEXT: andb $1, %al
; SSE42-NEXT: movb %al, (%rdi)
-; SSE42-NEXT: pextrb $11, %xmm2, %eax
+; SSE42-NEXT: pextrb $11, %xmm0, %eax
; SSE42-NEXT: andb $1, %al
; SSE42-NEXT: movb %al, (%rdi)
-; SSE42-NEXT: pextrb $10, %xmm2, %eax
+; SSE42-NEXT: pextrb $10, %xmm0, %eax
; SSE42-NEXT: andb $1, %al
; SSE42-NEXT: movb %al, (%rdi)
-; SSE42-NEXT: pextrb $9, %xmm2, %eax
+; SSE42-NEXT: pextrb $9, %xmm0, %eax
; SSE42-NEXT: andb $1, %al
; SSE42-NEXT: movb %al, (%rdi)
-; SSE42-NEXT: pextrb $8, %xmm2, %eax
+; SSE42-NEXT: pextrb $8, %xmm0, %eax
; SSE42-NEXT: andb $1, %al
; SSE42-NEXT: movb %al, (%rdi)
-; SSE42-NEXT: pextrb $7, %xmm2, %eax
+; SSE42-NEXT: pextrb $7, %xmm0, %eax
; SSE42-NEXT: andb $1, %al
; SSE42-NEXT: movb %al, (%rdi)
-; SSE42-NEXT: pextrb $6, %xmm2, %eax
+; SSE42-NEXT: pextrb $6, %xmm0, %eax
; SSE42-NEXT: andb $1, %al
; SSE42-NEXT: movb %al, (%rdi)
-; SSE42-NEXT: pextrb $5, %xmm2, %eax
+; SSE42-NEXT: pextrb $5, %xmm0, %eax
; SSE42-NEXT: andb $1, %al
; SSE42-NEXT: movb %al, (%rdi)
-; SSE42-NEXT: pextrb $4, %xmm2, %eax
+; SSE42-NEXT: pextrb $4, %xmm0, %eax
; SSE42-NEXT: andb $1, %al
; SSE42-NEXT: movb %al, (%rdi)
-; SSE42-NEXT: pextrb $3, %xmm2, %eax
+; SSE42-NEXT: pextrb $3, %xmm0, %eax
; SSE42-NEXT: andb $1, %al
; SSE42-NEXT: movb %al, (%rdi)
-; SSE42-NEXT: pextrb $2, %xmm2, %eax
+; SSE42-NEXT: pextrb $2, %xmm0, %eax
; SSE42-NEXT: andb $1, %al
; SSE42-NEXT: movb %al, (%rdi)
-; SSE42-NEXT: pextrb $1, %xmm2, %eax
+; SSE42-NEXT: pextrb $1, %xmm0, %eax
; SSE42-NEXT: andb $1, %al
; SSE42-NEXT: movb %al, (%rdi)
-; SSE42-NEXT: pextrb $0, %xmm2, %eax
+; SSE42-NEXT: pextrb $0, %xmm0, %eax
; SSE42-NEXT: andb $1, %al
; SSE42-NEXT: movb %al, (%rdi)
; SSE42-NEXT: movq %rdi, %rax
More information about the llvm-commits
mailing list