Fix a bug in the AVX 256-bit shuffle code in cases where the splat element is on the boundary of two 128-bit vectors.

The attached testcase was stuck in an endless loop.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@148027 91177308-0d34-0410-b5e6-96231b3b80d8
This commit is contained in:
Nadav Rotem 2012-01-12 15:31:55 +00:00
parent fb418bab97
commit d2070b00ef
2 changed files with 13 additions and 1 deletions

View File

@ -4413,7 +4413,7 @@ static SDValue PromoteSplat(ShuffleVectorSDNode *SV, SelectionDAG &DAG) {
// Extract the 128-bit part containing the splat element and update
// the splat element index when it refers to the higher register.
if (Size == 256) {
unsigned Idx = (EltNo > NumElems/2) ? NumElems/2 : 0;
unsigned Idx = (EltNo >= NumElems/2) ? NumElems/2 : 0;
V1 = Extract128BitVector(V1, DAG.getConstant(Idx, MVT::i32), DAG, dl);
if (Idx > 0)
EltNo -= NumElems/2;

View File

@ -0,0 +1,12 @@
; RUN: llc < %s -march=x86 -mcpu=corei7-avx -mattr=+avx -mtriple=i686-pc-win32 | FileCheck %s
; CHECK: endless_loop
define void @endless_loop() {
entry:
%0 = load <8 x i32> addrspace(1)* undef, align 32
%1 = shufflevector <8 x i32> %0, <8 x i32> undef, <16 x i32> <i32 4, i32 4, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
%2 = shufflevector <16 x i32> <i32 undef, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 undef>, <16 x i32> %1, <16 x i32> <i32 16, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 17>
store <16 x i32> %2, <16 x i32> addrspace(1)* undef, align 64
ret void
; CHECK: ret
}