aaron/ispc
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

Tweak to not oversize short vec types for 64 bit values

Browse Source
This commit is contained in:
james.brodman
2013-01-17 15:45:51 -05:00
parent 44e6be7914
commit a4e94a26ba
1 changed files with 16 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

17
expr.cpp
View File

@@ -3803,6 +3803,7 @@ ExprList::GetConstant(const Type *type) const {
// uniform short vector type
AssertPos(pos, type->IsUniformType() &&
CastType<VectorType>(type) != NULL);
llvm::VectorType *lvt =
llvm::dyn_cast<llvm::VectorType>(lt);
AssertPos(pos, lvt != NULL);
@@ -3810,8 +3811,22 @@ ExprList::GetConstant(const Type *type) const {
// Uniform short vectors are stored as vectors of length
// rounded up to the native vector width. So we add additional
// undef values here until we get the right size.
while ((cv.size() % g->target.nativeVectorWidth) != 0)
int vectorWidth = g->target.nativeVectorWidth;
const VectorType *vt = CastType<VectorType>(type);
const AtomicType *bt = vt->GetElementType();
if (Type::Equal(bt->GetAsUniformType(), AtomicType::UniformInt64) ||
Type::Equal(bt->GetAsUniformType(), AtomicType::UniformUInt64) ||
Type::Equal(bt->GetAsUniformType(), AtomicType::UniformDouble)) {
// target.nativeVectorWidth should be in terms of 32-bit
// values, so for the 64-bit guys, it takes half as many of
// them to fill the native width
vectorWidth /= 2;
}
while ((cv.size() % vectorWidth) != 0) {
cv.push_back(llvm::UndefValue::get(lvt->getElementType()));
}
return llvm::ConstantVector::get(cv);
}