| Issue |
242930
|
| Summary |
Potential error in _mm_mask_minmax_ph
|
| Labels |
new issue
|
| Assignees |
|
| Reporter |
Gadal
|
The current implementation is:
```
#define _mm_mask_minmax_ph(W, U, A, B, C) \
((__m128h)__builtin_ia32_vminmaxph128_mask( \
(__v8hf)(__m128h)(A), (__v8hf)(__m128h)(B), (int)(C), \
(__v8hf)(__m128h)(W), (__mmask16)-1))
```
I believe `(__mmask16)-1` is intended to be `(__mmask16)(U)`.
https://github.com/llvm/llvm-project/blob/c348d3b91c126414367f839465d9d67b7db38acf/clang/lib/Headers/avx10_2minmaxintrin.h#L91
_______________________________________________
llvm-bugs mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs