[llvm] [X86] Add widenBuildVector to create a wider build vector if the scalars are mergeable (PR #167667)
via llvm-commits
llvm-commits at lists.llvm.org
Wed Nov 12 02:30:42 PST 2025
llvmbot wrote:
<!--LLVM PR SUMMARY COMMENT-->
@llvm/pr-subscribers-backend-x86
Author: Simon Pilgrim (RKSimon)
<details>
<summary>Changes</summary>
See if each pair of scalar operands of a build vector can be freely merged together - typically if they've been split for some reason by legalization.
If we can create a new build vector node with double the scalar size, but half the element count - reducing codegen complexity and potentially allowing further optimization.
I did look at performing this generically in DAGCombine, but we don't have as much control over when a legal build vector can be folded - another generic fold would be to handle this on insert_vector_elt pairs, but again legality checks could be limiting.
Fixes #<!-- -->167498
---
Patch is 27.66 KiB, truncated to 20.00 KiB below, full version: https://github.com/llvm/llvm-project/pull/167667.diff
3 Files Affected:
- (modified) llvm/lib/Target/X86/X86ISelLowering.cpp (+54)
- (modified) llvm/test/CodeGen/X86/build-vector-128.ll (+73-311)
- (modified) llvm/test/CodeGen/X86/setcc-wide-types.ll (+16-28)
``````````diff
diff --git a/llvm/lib/Target/X86/X86ISelLowering.cpp b/llvm/lib/Target/X86/X86ISelLowering.cpp
index fa3dce256046f..190324a9f7474 100644
--- a/llvm/lib/Target/X86/X86ISelLowering.cpp
+++ b/llvm/lib/Target/X86/X86ISelLowering.cpp
@@ -8863,6 +8863,58 @@ static SDValue lowerBuildVectorAsBlend(BuildVectorSDNode *BVOp, SDLoc const &DL,
return SDValue();
}
+/// Widen a BUILD_VECTOR if the scalar operands are freely mergeable.
+static SDValue widenBuildVector(BuildVectorSDNode *BVOp, SDLoc const &DL,
+ X86Subtarget const &Subtarget,
+ SelectionDAG &DAG) {
+ using namespace SDPatternMatch;
+ MVT VT = BVOp->getSimpleValueType(0);
+ MVT SVT = VT.getScalarType();
+ unsigned NumElts = VT.getVectorNumElements();
+ unsigned EltBits = SVT.getSizeInBits();
+
+ if (SVT != MVT::i8 && SVT != MVT::i16 && SVT != MVT::i32)
+ return SDValue();
+
+ unsigned WideBits = 2 * EltBits;
+ MVT WideSVT = MVT::getIntegerVT(WideBits);
+ MVT WideVT = MVT::getVectorVT(WideSVT, NumElts / 2);
+ if (!DAG.getTargetLoweringInfo().isTypeLegal(WideSVT))
+ return SDValue();
+
+ SmallVector<SDValue, 8> WideOps;
+ for (unsigned I = 0; I != NumElts; I += 2) {
+ SDValue Op0 = BVOp->getOperand(I + 0);
+ SDValue Op1 = BVOp->getOperand(I + 1);
+
+ if (Op0.isUndef() && Op1.isUndef()) {
+ WideOps.push_back(DAG.getUNDEF(WideSVT));
+ continue;
+ }
+
+ // TODO: Constant repacking?
+
+ // Merge scalars that have been split from the same source.
+ SDValue X, Y;
+ if (sd_match(Op0, m_Trunc(m_Value(X))) &&
+ sd_match(Op1, m_Trunc(m_Srl(m_Value(Y), m_SpecificInt(EltBits)))) &&
+ peekThroughTruncates(X) == peekThroughTruncates(Y) &&
+ X.getValueType().bitsGE(WideSVT)) {
+ if (X.getValueType().bitsGT(WideSVT))
+ X = DAG.getNode(ISD::TRUNCATE, DL, WideSVT, X);
+ WideOps.push_back(X);
+ continue;
+ }
+
+ break;
+ }
+
+ if (WideOps.size() == (NumElts / 2))
+ return DAG.getBitcast(VT, DAG.getBuildVector(WideVT, DL, WideOps));
+
+ return SDValue();
+}
+
/// Create a vector constant without a load. SSE/AVX provide the bare minimum
/// functionality to do this, so it's all zeros, all ones, or some derivation
/// that is cheap to calculate.
@@ -9333,6 +9385,8 @@ X86TargetLowering::LowerBUILD_VECTOR(SDValue Op, SelectionDAG &DAG) const {
return BitOp;
if (SDValue Blend = lowerBuildVectorAsBlend(BV, dl, Subtarget, DAG))
return Blend;
+ if (SDValue WideBV = widenBuildVector(BV, dl, Subtarget, DAG))
+ return WideBV;
unsigned NumZero = ZeroMask.popcount();
unsigned NumNonZero = NonZeroMask.popcount();
diff --git a/llvm/test/CodeGen/X86/build-vector-128.ll b/llvm/test/CodeGen/X86/build-vector-128.ll
index b8bb417e1860c..59eb776ac365e 100644
--- a/llvm/test/CodeGen/X86/build-vector-128.ll
+++ b/llvm/test/CodeGen/X86/build-vector-128.ll
@@ -418,32 +418,12 @@ define <4 x i32> @test_buildvector_v2i64_split_v4i32(i64 %a0, i64 %a1) nounwind
; SSE-32-NEXT: movups {{[0-9]+}}(%esp), %xmm0
; SSE-32-NEXT: retl
;
-; SSE2-64-LABEL: test_buildvector_v2i64_split_v4i32:
-; SSE2-64: # %bb.0:
-; SSE2-64-NEXT: movl %edi, %eax
-; SSE2-64-NEXT: movl %esi, %ecx
-; SSE2-64-NEXT: shrq $32, %rdi
-; SSE2-64-NEXT: shrq $32, %rsi
-; SSE2-64-NEXT: movd %ecx, %xmm1
-; SSE2-64-NEXT: movd %esi, %xmm0
-; SSE2-64-NEXT: punpckldq {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1]
-; SSE2-64-NEXT: movd %eax, %xmm0
-; SSE2-64-NEXT: movd %edi, %xmm2
-; SSE2-64-NEXT: punpckldq {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1]
-; SSE2-64-NEXT: punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
-; SSE2-64-NEXT: retq
-;
-; SSE41-64-LABEL: test_buildvector_v2i64_split_v4i32:
-; SSE41-64: # %bb.0:
-; SSE41-64-NEXT: movl %edi, %eax
-; SSE41-64-NEXT: movl %esi, %ecx
-; SSE41-64-NEXT: shrq $32, %rdi
-; SSE41-64-NEXT: shrq $32, %rsi
-; SSE41-64-NEXT: movd %eax, %xmm0
-; SSE41-64-NEXT: pinsrd $1, %edi, %xmm0
-; SSE41-64-NEXT: pinsrd $2, %ecx, %xmm0
-; SSE41-64-NEXT: pinsrd $3, %esi, %xmm0
-; SSE41-64-NEXT: retq
+; SSE-64-LABEL: test_buildvector_v2i64_split_v4i32:
+; SSE-64: # %bb.0:
+; SSE-64-NEXT: movq %rsi, %xmm1
+; SSE-64-NEXT: movq %rdi, %xmm0
+; SSE-64-NEXT: punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
+; SSE-64-NEXT: retq
;
; AVX-32-LABEL: test_buildvector_v2i64_split_v4i32:
; AVX-32: # %bb.0:
@@ -452,14 +432,9 @@ define <4 x i32> @test_buildvector_v2i64_split_v4i32(i64 %a0, i64 %a1) nounwind
;
; AVX-64-LABEL: test_buildvector_v2i64_split_v4i32:
; AVX-64: # %bb.0:
-; AVX-64-NEXT: movl %edi, %eax
-; AVX-64-NEXT: movl %esi, %ecx
-; AVX-64-NEXT: shrq $32, %rdi
-; AVX-64-NEXT: shrq $32, %rsi
-; AVX-64-NEXT: vmovd %eax, %xmm0
-; AVX-64-NEXT: vpinsrd $1, %edi, %xmm0, %xmm0
-; AVX-64-NEXT: vpinsrd $2, %ecx, %xmm0, %xmm0
-; AVX-64-NEXT: vpinsrd $3, %esi, %xmm0, %xmm0
+; AVX-64-NEXT: vmovq %rsi, %xmm0
+; AVX-64-NEXT: vmovq %rdi, %xmm1
+; AVX-64-NEXT: vpunpcklqdq {{.*#+}} xmm0 = xmm1[0],xmm0[0]
; AVX-64-NEXT: retq
%a0.lo = trunc i64 %a0 to i32
%a1.lo = trunc i64 %a1 to i32
@@ -475,12 +450,10 @@ define <4 x i32> @test_buildvector_v2i64_split_v4i32(i64 %a0, i64 %a1) nounwind
}
define <8 x i16> @test_buildvector_v4i32_split_v8i16(i32 %a0, i32 %a1, i32 %a2, i32 %a3) nounwind {
-; SSE2-32-LABEL: test_buildvector_v4i32_split_v8i16:
-; SSE2-32: # %bb.0:
-; SSE2-32-NEXT: movsd {{.*#+}} xmm1 = mem[0],zero
-; SSE2-32-NEXT: movsd {{.*#+}} xmm0 = mem[0],zero
-; SSE2-32-NEXT: movlhps {{.*#+}} xmm0 = xmm0[0],xmm1[0]
-; SSE2-32-NEXT: retl
+; SSE-32-LABEL: test_buildvector_v4i32_split_v8i16:
+; SSE-32: # %bb.0:
+; SSE-32-NEXT: movups {{[0-9]+}}(%esp), %xmm0
+; SSE-32-NEXT: retl
;
; SSE2-64-LABEL: test_buildvector_v4i32_split_v8i16:
; SSE2-64: # %bb.0:
@@ -493,80 +466,25 @@ define <8 x i16> @test_buildvector_v4i32_split_v8i16(i32 %a0, i32 %a1, i32 %a2,
; SSE2-64-NEXT: punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; SSE2-64-NEXT: retq
;
-; SSE41-32-LABEL: test_buildvector_v4i32_split_v8i16:
-; SSE41-32: # %bb.0:
-; SSE41-32-NEXT: pushl %esi
-; SSE41-32-NEXT: movl {{[0-9]+}}(%esp), %eax
-; SSE41-32-NEXT: movl {{[0-9]+}}(%esp), %ecx
-; SSE41-32-NEXT: movl {{[0-9]+}}(%esp), %edx
-; SSE41-32-NEXT: movl {{[0-9]+}}(%esp), %esi
-; SSE41-32-NEXT: movd %esi, %xmm0
-; SSE41-32-NEXT: shrl $16, %esi
-; SSE41-32-NEXT: pinsrw $1, %esi, %xmm0
-; SSE41-32-NEXT: pinsrw $2, %edx, %xmm0
-; SSE41-32-NEXT: shrl $16, %edx
-; SSE41-32-NEXT: pinsrw $3, %edx, %xmm0
-; SSE41-32-NEXT: pinsrw $4, %ecx, %xmm0
-; SSE41-32-NEXT: shrl $16, %ecx
-; SSE41-32-NEXT: pinsrw $5, %ecx, %xmm0
-; SSE41-32-NEXT: pinsrw $6, %eax, %xmm0
-; SSE41-32-NEXT: shrl $16, %eax
-; SSE41-32-NEXT: pinsrw $7, %eax, %xmm0
-; SSE41-32-NEXT: popl %esi
-; SSE41-32-NEXT: retl
-;
; SSE41-64-LABEL: test_buildvector_v4i32_split_v8i16:
; SSE41-64: # %bb.0:
; SSE41-64-NEXT: movd %edi, %xmm0
-; SSE41-64-NEXT: shrl $16, %edi
-; SSE41-64-NEXT: pinsrw $1, %edi, %xmm0
-; SSE41-64-NEXT: pinsrw $2, %esi, %xmm0
-; SSE41-64-NEXT: shrl $16, %esi
-; SSE41-64-NEXT: pinsrw $3, %esi, %xmm0
-; SSE41-64-NEXT: pinsrw $4, %edx, %xmm0
-; SSE41-64-NEXT: shrl $16, %edx
-; SSE41-64-NEXT: pinsrw $5, %edx, %xmm0
-; SSE41-64-NEXT: pinsrw $6, %ecx, %xmm0
-; SSE41-64-NEXT: shrl $16, %ecx
-; SSE41-64-NEXT: pinsrw $7, %ecx, %xmm0
+; SSE41-64-NEXT: pinsrd $1, %esi, %xmm0
+; SSE41-64-NEXT: pinsrd $2, %edx, %xmm0
+; SSE41-64-NEXT: pinsrd $3, %ecx, %xmm0
; SSE41-64-NEXT: retq
;
; AVX-32-LABEL: test_buildvector_v4i32_split_v8i16:
; AVX-32: # %bb.0:
-; AVX-32-NEXT: pushl %esi
-; AVX-32-NEXT: movl {{[0-9]+}}(%esp), %eax
-; AVX-32-NEXT: movl {{[0-9]+}}(%esp), %ecx
-; AVX-32-NEXT: movl {{[0-9]+}}(%esp), %edx
-; AVX-32-NEXT: movl {{[0-9]+}}(%esp), %esi
-; AVX-32-NEXT: vmovd %esi, %xmm0
-; AVX-32-NEXT: shrl $16, %esi
-; AVX-32-NEXT: vpinsrw $1, %esi, %xmm0, %xmm0
-; AVX-32-NEXT: vpinsrw $2, %edx, %xmm0, %xmm0
-; AVX-32-NEXT: shrl $16, %edx
-; AVX-32-NEXT: vpinsrw $3, %edx, %xmm0, %xmm0
-; AVX-32-NEXT: vpinsrw $4, %ecx, %xmm0, %xmm0
-; AVX-32-NEXT: shrl $16, %ecx
-; AVX-32-NEXT: vpinsrw $5, %ecx, %xmm0, %xmm0
-; AVX-32-NEXT: vpinsrw $6, %eax, %xmm0, %xmm0
-; AVX-32-NEXT: shrl $16, %eax
-; AVX-32-NEXT: vpinsrw $7, %eax, %xmm0, %xmm0
-; AVX-32-NEXT: popl %esi
+; AVX-32-NEXT: vmovups {{[0-9]+}}(%esp), %xmm0
; AVX-32-NEXT: retl
;
; AVX-64-LABEL: test_buildvector_v4i32_split_v8i16:
; AVX-64: # %bb.0:
; AVX-64-NEXT: vmovd %edi, %xmm0
-; AVX-64-NEXT: shrl $16, %edi
-; AVX-64-NEXT: vpinsrw $1, %edi, %xmm0, %xmm0
-; AVX-64-NEXT: vpinsrw $2, %esi, %xmm0, %xmm0
-; AVX-64-NEXT: shrl $16, %esi
-; AVX-64-NEXT: vpinsrw $3, %esi, %xmm0, %xmm0
-; AVX-64-NEXT: vpinsrw $4, %edx, %xmm0, %xmm0
-; AVX-64-NEXT: shrl $16, %edx
-; AVX-64-NEXT: vpinsrw $5, %edx, %xmm0, %xmm0
-; AVX-64-NEXT: vpinsrw $6, %ecx, %xmm0, %xmm0
-; AVX-64-NEXT: shrl $16, %ecx
-; AVX-64-NEXT: vpinsrw $7, %ecx, %xmm0, %xmm0
+; AVX-64-NEXT: vpinsrd $1, %esi, %xmm0, %xmm0
+; AVX-64-NEXT: vpinsrd $2, %edx, %xmm0, %xmm0
+; AVX-64-NEXT: vpinsrd $3, %ecx, %xmm0, %xmm0
; AVX-64-NEXT: retq
%a0.lo = trunc i32 %a0 to i16
%a1.lo = trunc i32 %a1 to i16
@@ -594,244 +512,88 @@ define <8 x i16> @test_buildvector_v4i32_split_v8i16(i32 %a0, i32 %a1, i32 %a2,
define <16 x i8> @test_buildvector_v8i16_split_v16i8(i16 %a0, i16 %a1, i16 %a2, i16 %a3, i16 %a4, i16 %a5, i16 %a6, i16 %a7) nounwind {
; SSE2-32-LABEL: test_buildvector_v8i16_split_v16i8:
; SSE2-32: # %bb.0:
-; SSE2-32-NEXT: movzwl {{[0-9]+}}(%esp), %eax
-; SSE2-32-NEXT: movd %eax, %xmm1
-; SSE2-32-NEXT: movdqa %xmm1, %xmm0
-; SSE2-32-NEXT: psrld $8, %xmm0
-; SSE2-32-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]
-; SSE2-32-NEXT: movzwl {{[0-9]+}}(%esp), %eax
-; SSE2-32-NEXT: movd %eax, %xmm0
-; SSE2-32-NEXT: movdqa %xmm0, %xmm2
-; SSE2-32-NEXT: psrld $8, %xmm2
-; SSE2-32-NEXT: punpcklbw {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1],xmm0[2],xmm2[2],xmm0[3],xmm2[3],xmm0[4],xmm2[4],xmm0[5],xmm2[5],xmm0[6],xmm2[6],xmm0[7],xmm2[7]
-; SSE2-32-NEXT: punpcklwd {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
-; SSE2-32-NEXT: movzwl {{[0-9]+}}(%esp), %eax
-; SSE2-32-NEXT: movd %eax, %xmm2
-; SSE2-32-NEXT: movdqa %xmm2, %xmm1
-; SSE2-32-NEXT: psrld $8, %xmm1
-; SSE2-32-NEXT: punpcklbw {{.*#+}} xmm2 = xmm2[0],xmm1[0],xmm2[1],xmm1[1],xmm2[2],xmm1[2],xmm2[3],xmm1[3],xmm2[4],xmm1[4],xmm2[5],xmm1[5],xmm2[6],xmm1[6],xmm2[7],xmm1[7]
-; SSE2-32-NEXT: movzwl {{[0-9]+}}(%esp), %eax
-; SSE2-32-NEXT: movd %eax, %xmm1
-; SSE2-32-NEXT: movdqa %xmm1, %xmm3
-; SSE2-32-NEXT: psrld $8, %xmm3
-; SSE2-32-NEXT: punpcklbw {{.*#+}} xmm1 = xmm1[0],xmm3[0],xmm1[1],xmm3[1],xmm1[2],xmm3[2],xmm1[3],xmm3[3],xmm1[4],xmm3[4],xmm1[5],xmm3[5],xmm1[6],xmm3[6],xmm1[7],xmm3[7]
-; SSE2-32-NEXT: punpcklwd {{.*#+}} xmm1 = xmm1[0],xmm2[0],xmm1[1],xmm2[1],xmm1[2],xmm2[2],xmm1[3],xmm2[3]
-; SSE2-32-NEXT: punpckldq {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1]
-; SSE2-32-NEXT: movzwl {{[0-9]+}}(%esp), %eax
-; SSE2-32-NEXT: movd %eax, %xmm0
-; SSE2-32-NEXT: movdqa %xmm0, %xmm2
-; SSE2-32-NEXT: psrld $8, %xmm2
-; SSE2-32-NEXT: punpcklbw {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1],xmm0[2],xmm2[2],xmm0[3],xmm2[3],xmm0[4],xmm2[4],xmm0[5],xmm2[5],xmm0[6],xmm2[6],xmm0[7],xmm2[7]
-; SSE2-32-NEXT: movzwl {{[0-9]+}}(%esp), %eax
-; SSE2-32-NEXT: movd %eax, %xmm2
-; SSE2-32-NEXT: movdqa %xmm2, %xmm3
-; SSE2-32-NEXT: psrld $8, %xmm3
-; SSE2-32-NEXT: punpcklbw {{.*#+}} xmm2 = xmm2[0],xmm3[0],xmm2[1],xmm3[1],xmm2[2],xmm3[2],xmm2[3],xmm3[3],xmm2[4],xmm3[4],xmm2[5],xmm3[5],xmm2[6],xmm3[6],xmm2[7],xmm3[7]
+; SSE2-32-NEXT: movd {{.*#+}} xmm0 = mem[0],zero,zero,zero
+; SSE2-32-NEXT: movd {{.*#+}} xmm1 = mem[0],zero,zero,zero
+; SSE2-32-NEXT: punpcklwd {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1],xmm1[2],xmm0[2],xmm1[3],xmm0[3]
+; SSE2-32-NEXT: movd {{.*#+}} xmm0 = mem[0],zero,zero,zero
+; SSE2-32-NEXT: movd {{.*#+}} xmm2 = mem[0],zero,zero,zero
; SSE2-32-NEXT: punpcklwd {{.*#+}} xmm2 = xmm2[0],xmm0[0],xmm2[1],xmm0[1],xmm2[2],xmm0[2],xmm2[3],xmm0[3]
-; SSE2-32-NEXT: movzwl {{[0-9]+}}(%esp), %eax
-; SSE2-32-NEXT: movd %eax, %xmm3
-; SSE2-32-NEXT: movdqa %xmm3, %xmm0
-; SSE2-32-NEXT: psrld $8, %xmm0
-; SSE2-32-NEXT: punpcklbw {{.*#+}} xmm3 = xmm3[0],xmm0[0],xmm3[1],xmm0[1],xmm3[2],xmm0[2],xmm3[3],xmm0[3],xmm3[4],xmm0[4],xmm3[5],xmm0[5],xmm3[6],xmm0[6],xmm3[7],xmm0[7]
-; SSE2-32-NEXT: movzwl {{[0-9]+}}(%esp), %eax
-; SSE2-32-NEXT: movd %eax, %xmm0
-; SSE2-32-NEXT: movdqa %xmm0, %xmm4
-; SSE2-32-NEXT: psrld $8, %xmm4
-; SSE2-32-NEXT: punpcklbw {{.*#+}} xmm0 = xmm0[0],xmm4[0],xmm0[1],xmm4[1],xmm0[2],xmm4[2],xmm0[3],xmm4[3],xmm0[4],xmm4[4],xmm0[5],xmm4[5],xmm0[6],xmm4[6],xmm0[7],xmm4[7]
+; SSE2-32-NEXT: punpckldq {{.*#+}} xmm2 = xmm2[0],xmm1[0],xmm2[1],xmm1[1]
+; SSE2-32-NEXT: movd {{.*#+}} xmm0 = mem[0],zero,zero,zero
+; SSE2-32-NEXT: movd {{.*#+}} xmm1 = mem[0],zero,zero,zero
+; SSE2-32-NEXT: punpcklwd {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1],xmm1[2],xmm0[2],xmm1[3],xmm0[3]
+; SSE2-32-NEXT: movd {{.*#+}} xmm3 = mem[0],zero,zero,zero
+; SSE2-32-NEXT: movd {{.*#+}} xmm0 = mem[0],zero,zero,zero
; SSE2-32-NEXT: punpcklwd {{.*#+}} xmm0 = xmm0[0],xmm3[0],xmm0[1],xmm3[1],xmm0[2],xmm3[2],xmm0[3],xmm3[3]
-; SSE2-32-NEXT: punpckldq {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1]
-; SSE2-32-NEXT: punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
+; SSE2-32-NEXT: punpckldq {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; SSE2-32-NEXT: punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm2[0]
; SSE2-32-NEXT: retl
;
; SSE2-64-LABEL: test_buildvector_v8i16_split_v16i8:
; SSE2-64: # %bb.0:
-; SSE2-64-NEXT: pushq %rbp
-; SSE2-64-NEXT: pushq %r15
-; SSE2-64-NEXT: pushq %r14
-; SSE2-64-NEXT: pushq %rbx
-; SSE2-64-NEXT: movzwl %di, %eax
-; SSE2-64-NEXT: movzwl %si, %r10d
-; SSE2-64-NEXT: movzwl %dx, %r11d
-; SSE2-64-NEXT: movzwl %cx, %ebx
-; SSE2-64-NEXT: movzwl %r8w, %ebp
-; SSE2-64-NEXT: movzwl %r9w, %r14d
-; SSE2-64-NEXT: movzwl {{[0-9]+}}(%rsp), %r15d
-; SSE2-64-NEXT: movd %r15d, %xmm0
-; SSE2-64-NEXT: movdqa %xmm0, %xmm1
-; SSE2-64-NEXT: psrld $8, %xmm1
-; SSE2-64-NEXT: punpcklbw {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3],xmm0[4],xmm1[4],xmm0[5],xmm1[5],xmm0[6],xmm1[6],xmm0[7],xmm1[7]
-; SSE2-64-NEXT: movzwl {{[0-9]+}}(%rsp), %r15d
-; SSE2-64-NEXT: movd %r15d, %xmm2
-; SSE2-64-NEXT: movdqa %xmm2, %xmm1
-; SSE2-64-NEXT: psrld $8, %xmm1
-; SSE2-64-NEXT: punpcklbw {{.*#+}} xmm2 = xmm2[0],xmm1[0],xmm2[1],xmm1[1],xmm2[2],xmm1[2],xmm2[3],xmm1[3],xmm2[4],xmm1[4],xmm2[5],xmm1[5],xmm2[6],xmm1[6],xmm2[7],xmm1[7]
-; SSE2-64-NEXT: punpcklwd {{.*#+}} xmm2 = xmm2[0],xmm0[0],xmm2[1],xmm0[1],xmm2[2],xmm0[2],xmm2[3],xmm0[3]
-; SSE2-64-NEXT: movd %r9d, %xmm0
-; SSE2-64-NEXT: movd %r14d, %xmm1
-; SSE2-64-NEXT: psrld $8, %xmm1
-; SSE2-64-NEXT: punpcklbw {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3],xmm0[4],xmm1[4],xmm0[5],xmm1[5],xmm0[6],xmm1[6],xmm0[7],xmm1[7]
-; SSE2-64-NEXT: movd %r8d, %xmm1
-; SSE2-64-NEXT: movd %ebp, %xmm3
-; SSE2-64-NEXT: psrld $8, %xmm3
-; SSE2-64-NEXT: punpcklbw {{.*#+}} xmm1 = xmm1[0],xmm3[0],xmm1[1],xmm3[1],xmm1[2],xmm3[2],xmm1[3],xmm3[3],xmm1[4],xmm3[4],xmm1[5],xmm3[5],xmm1[6],xmm3[6],xmm1[7],xmm3[7]
+; SSE2-64-NEXT: movd {{.*#+}} xmm0 = mem[0],zero,zero,zero
+; SSE2-64-NEXT: movd {{.*#+}} xmm1 = mem[0],zero,zero,zero
; SSE2-64-NEXT: punpcklwd {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1],xmm1[2],xmm0[2],xmm1[3],xmm0[3]
-; SSE2-64-NEXT: punpckldq {{.*#+}} xmm1 = xmm1[0],xmm2[0],xmm1[1],xmm2[1]
-; SSE2-64-NEXT: movd %ecx, %xmm0
-; SSE2-64-NEXT: movd %ebx, %xmm2
-; SSE2-64-NEXT: psrld $8, %xmm2
-; SSE2-64-NEXT: punpcklbw {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1],xmm0[2],xmm2[2],xmm0[3],xmm2[3],xmm0[4],xmm2[4],xmm0[5],xmm2[5],xmm0[6],xmm2[6],xmm0[7],xmm2[7]
-; SSE2-64-NEXT: movd %edx, %xmm2
-; SSE2-64-NEXT: movd %r11d, %xmm3
-; SSE2-64-NEXT: psrld $8, %xmm3
-; SSE2-64-NEXT: punpcklbw {{.*#+}} xmm2 = xmm2[0],xmm3[0],xmm2[1],xmm3[1],xmm2[2],xmm3[2],xmm2[3],xmm3[3],xmm2[4],xmm3[4],xmm2[5],xmm3[5],xmm2[6],xmm3[6],xmm2[7],xmm3[7]
+; SSE2-64-NEXT: movd %r9d, %xmm0
+; SSE2-64-NEXT: movd %r8d, %xmm2
; SSE2-64-NEXT: punpcklwd {{.*#+}} xmm2 = xmm2[0],xmm0[0],xmm2[1],xmm0[1],xmm2[2],xmm0[2],xmm2[3],xmm0[3]
+; SSE2-64-NEXT: punpckldq {{.*#+}} xmm2 = xmm2[0],xmm1[0],xmm2[1],xmm1[1]
+; SSE2-64-NEXT: movd %ecx, %xmm0
+; SSE2-64-NEXT: movd %edx, %xmm1
+; SSE2-64-NEXT: punpcklwd {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1],xmm1[2],xmm0[2],xmm1[3],xmm0[3]
; SSE2-64-NEXT: movd %esi, %xmm3
-; SSE2-64-NEXT: movd %r10d, %xmm0
-; SSE2-64-NEXT: psrld $8, %xmm0
-; SSE2-64-NEXT: punpcklbw {{.*#+}} xmm3 = xmm3[0],xmm0[0],xmm3[1],xmm0[1],xmm3[2],xmm0[2],xmm3[3],xmm0[3],xmm3[4],xmm0[4],xmm3[5],xmm0[5],xmm3[6],xmm0[6],xmm3[7],xmm0[7]
; SSE2-64-NEXT: movd %edi, %xmm0
-; SSE2-64-NEXT: movd %eax, %xmm4
-; SSE2-64-NEXT: psrld $8, %xmm4
-; SSE2-64-NEXT: punpcklbw {{.*#+}} xmm0 = xmm0[0],xmm4[0],xmm0[1],xmm4[1],xmm0[2],xmm4[2],xmm0[3],xmm4[3],xmm0[4],xmm4[4],xmm0[5],xmm4[5],xmm0[6],xmm4[6],xmm0[7],xmm4[7]
; SSE2-64-NEXT: punpcklwd {{.*#+}} xmm0 = xmm0[0],xmm3[0],xmm0[1],xmm3[1],xmm0[2],xmm3[2],xmm0[3],xmm3[3]
-; SSE2-64-NEXT: punpckldq {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1]
-; SSE2-64-NEXT: punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
-; SSE2-64-NEXT: popq %rbx
-; SSE2-64-NEXT: popq %r14
-; SSE2-64-NEXT: popq %r15
-; SSE2-64-NEXT: popq %rbp
+; SSE2-64-NEXT: punpckldq {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; SSE2-64-NEXT: punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm2[0]
; SSE2-64-NEXT: retq
;
; SSE41-32-LABEL: test_buildvector_v8i16_split_v16i8:
; SSE41-32: # %bb.0:
-; SSE41-32-NEXT: movzwl {{[0-9]+}}(%esp), %eax
-; SSE41-32-NEXT: movd %eax, %xmm0
-; SSE41-32-NEXT: shrl $8, %eax
-; SSE41-32-NEXT: pinsrb $1, %eax, %xmm0
-; SSE41-32-NEXT: movzwl {{[0-9]+}}(%esp), %eax
-; SSE41-32-NEXT: pinsrb $2, %eax, %xmm0
-; SSE41-32-NEXT: shrl $8, %eax
-; SSE41-32-NEXT: pinsrb $3, %eax, %xmm0
-; SSE41-32-NEXT: movzwl {{[0-9]+}}(%esp), %eax
-; SSE41-32-NEXT: pinsrb $4, %eax, %xmm0
-; SSE41-32-NEXT: shrl $8, %eax
-; SSE41-32-NEXT: pinsrb $5, %eax, %xmm0
-; SSE41-32-NEXT: movzwl {{[0-9]+}}(%esp), %eax
-; SSE41-32-NEXT: pinsrb $6, %eax, %xmm0
-; SSE41-32-NEXT: shrl $8, %eax
-; SSE41-32-NEXT: pinsrb $7, %eax, %xmm0
-; SSE41-32-NEXT: movzwl {{[0-9]+}}(%esp), %eax
-; SSE41-32-NEXT: pinsrb $8, %eax, %xmm0
-; SSE41-32-NEXT: shrl $8, %eax
-; SSE41-32-NEXT: pinsrb $9, %eax, %xmm0
-; SSE41-32-NEXT: movzwl {{[0-9]+}}(%esp), %eax
-; SSE41-32-NEXT: pinsrb $10, %eax, %xmm0
-; SSE41-32-NEXT: shrl $8, %eax
-; SSE41-32-NEXT: pinsrb $11, %eax, %xmm0
-; SSE41-32-NEXT: movzwl {{[0-9]+}}(%esp), %eax
-; SSE41-32-NEXT: pinsrb $12, %eax, %xmm0
-; SSE41-32-NEXT: shrl $8, %eax
-; SSE41-32-NEXT: pinsrb $13, %eax, %xmm0
-; SSE41-32-NEXT: movzwl {{[0-9]+}}(%esp), %eax
-; SSE41-32-NEXT: pinsrb $14, %eax, %xmm0
-; SSE41-32-NEXT: shrl $8, %eax
-; SSE41-32-NEXT: pinsrb $15, %eax, %xmm0
+; SSE41-32-NEXT: movd {{.*#+}} xmm0 = mem[0],zero,zero,zero
+; SSE41-32-NEXT: pinsrw $1, {{[0-9]+}}(%esp), %xmm0
+; SSE41-32-NEXT: pinsrw $2, {{[0-9]+}}(%esp), %xmm0
+; SSE41-32-NEXT: pinsrw $3, {{[0-9]+}}(%esp), %xmm0
+; SSE41-32-NEXT: pinsrw $4, {{[0-9]+}}(%esp), %xmm0
+; SSE41-32-NEXT: pinsrw $5, {{[0-9]+}}(%esp), %xmm0
+; SSE41-32-NEXT: pinsrw $6, {{[0-9]+}}(%esp), %xmm0
+; SSE41-32-NEXT: pinsrw $7, {{[0-9]+}}(%esp), %xmm0
; SSE41-32-NEXT: retl
;
; SSE41-64-LABEL: test_buildvector_v8i16_split_v16i8:
; SSE41-64: # %bb.0:
; SSE41-64-NEXT: movd %edi, %xmm0
-; SSE41-64-NEXT: shrl $8, %edi
-; SSE41-64-NEXT: pinsrb...
[truncated]
``````````
</details>
https://github.com/llvm/llvm-project/pull/167667
More information about the llvm-commits
mailing list