maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity

LoadStoreVectorizer: if one element of a vector is integer, default to 

integer.

Fixes issues on some architectures where we use arithmetic ops to build
vectors, which can cause bad things to happen for loads/stores of mixed
types.

Patch by Fiona Glaser

llvm-svn: 274307
This commit is contained in:
Matt Arsenault 2016-07-01 00:37:01 +00:00
parent 8a4ab5e19f
commit d7e8898bdd
2 changed files with 19 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

20
llvm/lib/Transforms/Vectorize/LoadStoreVectorizer.cpp
View File

@ -579,7 +579,15 @@ bool Vectorizer::vectorizeInstructions(ArrayRef<Value *> Instrs) {
bool Vectorizer::vectorizeStoreChain(ArrayRef<Value *> Chain) {
StoreInst *S0 = cast<StoreInst>(Chain[0]);
Type *StoreTy = S0->getValueOperand()->getType();
// If the vector has an int element, default to int for the whole load.
Type *StoreTy;
for (const auto &V : Chain) {
StoreTy = cast<StoreInst>(V)->getValueOperand()->getType();
if (StoreTy->isIntOrIntVectorTy())
break;
}
unsigned Sz = DL.getTypeSizeInBits(StoreTy);
unsigned VF = VecRegSize / Sz;
unsigned ChainSize = Chain.size();
@ -700,7 +708,15 @@ bool Vectorizer::vectorizeStoreChain(ArrayRef<Value *> Chain) {
bool Vectorizer::vectorizeLoadChain(ArrayRef<Value *> Chain) {
LoadInst *L0 = cast<LoadInst>(Chain[0]);
Type *LoadTy = L0->getType();
// If the vector has an int element, default to int for the whole load.
Type *LoadTy;
for (const auto &V : Chain) {
LoadTy = cast<LoadInst>(V)->getType();
if (LoadTy->isIntOrIntVectorTy())
break;
}
unsigned Sz = DL.getTypeSizeInBits(LoadTy);
unsigned VF = VecRegSize / Sz;
unsigned ChainSize = Chain.size();

2
llvm/test/Transforms/LoadStoreVectorizer/AMDGPU/merge-stores.ll
View File

@ -140,7 +140,7 @@ define void @merge_global_store_4_constants_f32(float addrspace(1)* %out) #0 {
}
; CHECK-LABEL: @merge_global_store_4_constants_mixed_i32_f32
; CHECK: store <4 x float> <float 8.000000e+00, float 0x36D6000000000000, float 2.000000e+00, float 0x36E1000000000000>, <4 x float> addrspace(1)* %{{[0-9]+}}
; CHECK: store <4 x i32> <i32 1090519040, i32 11, i32 1073741824, i32 17>, <4 x i32> addrspace(1)* %{{[0-9]+$}}
define void @merge_global_store_4_constants_mixed_i32_f32(float addrspace(1)* %out) #0 {
%out.gep.1 = getelementptr float, float addrspace(1)* %out, i32 1
%out.gep.2 = getelementptr float, float addrspace(1)* %out, i32 2