[llvm] [AArch64] Consider streaming mode in TTI interfaces for vectorization. (PR #96305)
Sander de Smalen via llvm-commits
llvm-commits at lists.llvm.org
Mon Jun 24 03:04:37 PDT 2024
================
@@ -2162,19 +2162,20 @@ AArch64TTIImpl::getRegisterBitWidth(TargetTransformInfo::RegisterKind K) const {
case TargetTransformInfo::RGK_Scalar:
return TypeSize::getFixed(64);
case TargetTransformInfo::RGK_FixedWidthVector:
- if (!ST->isNeonAvailable() && !EnableFixedwidthAutovecInStreamingMode)
- return TypeSize::getFixed(0);
-
- if (ST->hasSVE())
+ if (ST->useSVEForFixedLengthVectors() &&
+ (ST->isSVEAvailable() || EnableFixedwidthAutovecInStreamingMode))
return TypeSize::getFixed(
std::max(ST->getMinSVEVectorSizeInBits(), 128u));
----------------
sdesmalen-arm wrote:
That's a good point, we'll need to fix this. To make this work for streaming-SVE functions, do we need a change to the ACLE itself as well?
https://github.com/llvm/llvm-project/pull/96305
More information about the llvm-commits
mailing list