diff options
| author | Thurston Dang <thurston@google.com> | 2025-07-08 20:51:19 -0700 |
|---|---|---|
| committer | GitHub <noreply@github.com> | 2025-07-08 20:51:19 -0700 |
| commit | cc95e4039b75c931f928464dd248e0d3712d7643 (patch) | |
| tree | d0bfbb9f833359bca23c18a8657e50c1043908fb /flang/lib/Frontend/CompilerInvocation.cpp | |
| parent | 0819fec61a4e6189bae11d1dbbd0a8794321c746 (diff) | |
| download | llvm-cc95e4039b75c931f928464dd248e0d3712d7643.zip llvm-cc95e4039b75c931f928464dd248e0d3712d7643.tar.gz llvm-cc95e4039b75c931f928464dd248e0d3712d7643.tar.bz2 | |
[msan] Handle AVX512 vector down convert (non-mem) intrinsics (#147606)
This handles `llvm.x86.avx512.mask.pmov{,s,us}.*.512` using
`handleIntrinsicByApplyingToShadow()` where possible, otherwise using a
customized slow-path handler, `handleAVX512VectorDownConvert()`.
Note that shadow propagation of `pmov{s,us}` (signed/unsigned
saturation) are approximated using truncation. Future work could extend
`handleAVX512VectorDownConvert()` to use `GetMinMaxUnsigned()` to handle
saturation precisely.
Diffstat (limited to 'flang/lib/Frontend/CompilerInvocation.cpp')
0 files changed, 0 insertions, 0 deletions