diff options
| author | Artem Gindinson <gindinson@roofline.ai> | 2025-06-02 13:27:44 +0200 |
|---|---|---|
| committer | GitHub <noreply@github.com> | 2025-06-02 12:27:44 +0100 |
| commit | af6e3c045b07ebc7ce09318c90048f407a15b391 (patch) | |
| tree | ffe0ba0f232b44f78c0e41b88653a1d520df5ac4 /clang/lib/Basic/SourceManager.cpp | |
| parent | 0ada5c7b1a52afb668bc42dd2d5573e5805433d1 (diff) | |
| download | llvm-af6e3c045b07ebc7ce09318c90048f407a15b391.zip llvm-af6e3c045b07ebc7ce09318c90048f407a15b391.tar.gz llvm-af6e3c045b07ebc7ce09318c90048f407a15b391.tar.bz2 | |
[mlir][math] Fix intrinsic conversions to LLVM for 0D-vector types (#141020)
`vector<t>` types are not compatible with the LLVM type system – with
the current approach employed within `LLVMTypeConverter`, they must be
explicitly converted into `vector<1xt>` when lowering. Employ this rule
within the conversion patterns for intrinsics that are handled directly
within `MathToLLVM`: `math.ctlz` `.cttz`, `.absi`, `.expm1`, `.log1p`,
`.rsqrt`, `.isnan`, `.isfinite`.
This change does not cover/test patterns that are based off
`VectorConvertToLLVMPattern` template from `LLVMCommon/VectorPattern.h`.
---------
Signed-off-by: Artem Gindinson <gindinson@roofline.ai>
Diffstat (limited to 'clang/lib/Basic/SourceManager.cpp')
0 files changed, 0 insertions, 0 deletions