diff options
| author | Allen <zhongyunde@huawei.com> | 2024-04-16 09:00:03 +0800 |
|---|---|---|
| committer | GitHub <noreply@github.com> | 2024-04-16 09:00:03 +0800 |
| commit | 8aa7e378dee27ec81959ef6750a7dd07cefdc77d (patch) | |
| tree | 0fdfc46674c2d6727e3469a9bd422bcd4f1694ba /clang/lib/CodeGen/CodeGenModule.cpp | |
| parent | 2ac562ab784c6bf04e6d3026c567d3552a735668 (diff) | |
| download | llvm-8aa7e378dee27ec81959ef6750a7dd07cefdc77d.zip llvm-8aa7e378dee27ec81959ef6750a7dd07cefdc77d.tar.gz llvm-8aa7e378dee27ec81959ef6750a7dd07cefdc77d.tar.bz2 | |
[InterleavedAccessPass] Get round the unsupported large scalarize vectors (#88643)
When build with option -msve-vector-bits=512, the return vaule of
Subtarget->getMinSVEVectorSizeInBits() is 512;
While the MinElts is still 4 for <vscale x 4 x double> in
getNumInterleavedAccesses, so it creates invalid
llvm.aarch64.sve.ld2.sret.nxv4f64, which need be splited.
Unlikely, the related custom spilting is not supported now.
Fix https://github.com/llvm/llvm-project/issues/88247
Diffstat (limited to 'clang/lib/CodeGen/CodeGenModule.cpp')
0 files changed, 0 insertions, 0 deletions