Refactor avx512bw: reduction operations

This commit is contained in:
Tobias Decking 2024-07-01 00:07:09 +02:00 committed by Amanieu d'Antras
parent fe0a378499
commit bbb2ba5424
2 changed files with 1184 additions and 74 deletions

View file

@ -50,74 +50,6 @@
* [ ] [`_tile_zero`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_tile_zero)
</p></details>
<details><summary>["AVX512BW", "AVX512VL"]</summary><p>
* [ ] [`_mm256_mask_reduce_add_epi16`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm256_mask_reduce_add_epi16)
* [ ] [`_mm256_mask_reduce_add_epi8`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm256_mask_reduce_add_epi8)
* [ ] [`_mm256_mask_reduce_and_epi16`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm256_mask_reduce_and_epi16)
* [ ] [`_mm256_mask_reduce_and_epi8`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm256_mask_reduce_and_epi8)
* [ ] [`_mm256_mask_reduce_max_epi16`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm256_mask_reduce_max_epi16)
* [ ] [`_mm256_mask_reduce_max_epi8`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm256_mask_reduce_max_epi8)
* [ ] [`_mm256_mask_reduce_max_epu16`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm256_mask_reduce_max_epu16)
* [ ] [`_mm256_mask_reduce_max_epu8`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm256_mask_reduce_max_epu8)
* [ ] [`_mm256_mask_reduce_min_epi16`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm256_mask_reduce_min_epi16)
* [ ] [`_mm256_mask_reduce_min_epi8`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm256_mask_reduce_min_epi8)
* [ ] [`_mm256_mask_reduce_min_epu16`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm256_mask_reduce_min_epu16)
* [ ] [`_mm256_mask_reduce_min_epu8`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm256_mask_reduce_min_epu8)
* [ ] [`_mm256_mask_reduce_mul_epi16`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm256_mask_reduce_mul_epi16)
* [ ] [`_mm256_mask_reduce_mul_epi8`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm256_mask_reduce_mul_epi8)
* [ ] [`_mm256_mask_reduce_or_epi16`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm256_mask_reduce_or_epi16)
* [ ] [`_mm256_mask_reduce_or_epi8`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm256_mask_reduce_or_epi8)
* [ ] [`_mm256_reduce_add_epi16`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm256_reduce_add_epi16)
* [ ] [`_mm256_reduce_add_epi8`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm256_reduce_add_epi8)
* [ ] [`_mm256_reduce_and_epi16`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm256_reduce_and_epi16)
* [ ] [`_mm256_reduce_and_epi8`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm256_reduce_and_epi8)
* [ ] [`_mm256_reduce_max_epi16`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm256_reduce_max_epi16)
* [ ] [`_mm256_reduce_max_epi8`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm256_reduce_max_epi8)
* [ ] [`_mm256_reduce_max_epu16`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm256_reduce_max_epu16)
* [ ] [`_mm256_reduce_max_epu8`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm256_reduce_max_epu8)
* [ ] [`_mm256_reduce_min_epi16`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm256_reduce_min_epi16)
* [ ] [`_mm256_reduce_min_epi8`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm256_reduce_min_epi8)
* [ ] [`_mm256_reduce_min_epu16`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm256_reduce_min_epu16)
* [ ] [`_mm256_reduce_min_epu8`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm256_reduce_min_epu8)
* [ ] [`_mm256_reduce_mul_epi16`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm256_reduce_mul_epi16)
* [ ] [`_mm256_reduce_mul_epi8`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm256_reduce_mul_epi8)
* [ ] [`_mm256_reduce_or_epi16`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm256_reduce_or_epi16)
* [ ] [`_mm256_reduce_or_epi8`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm256_reduce_or_epi8)
* [ ] [`_mm_mask_reduce_add_epi16`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm_mask_reduce_add_epi16)
* [ ] [`_mm_mask_reduce_add_epi8`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm_mask_reduce_add_epi8)
* [ ] [`_mm_mask_reduce_and_epi16`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm_mask_reduce_and_epi16)
* [ ] [`_mm_mask_reduce_and_epi8`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm_mask_reduce_and_epi8)
* [ ] [`_mm_mask_reduce_max_epi16`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm_mask_reduce_max_epi16)
* [ ] [`_mm_mask_reduce_max_epi8`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm_mask_reduce_max_epi8)
* [ ] [`_mm_mask_reduce_max_epu16`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm_mask_reduce_max_epu16)
* [ ] [`_mm_mask_reduce_max_epu8`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm_mask_reduce_max_epu8)
* [ ] [`_mm_mask_reduce_min_epi16`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm_mask_reduce_min_epi16)
* [ ] [`_mm_mask_reduce_min_epi8`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm_mask_reduce_min_epi8)
* [ ] [`_mm_mask_reduce_min_epu16`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm_mask_reduce_min_epu16)
* [ ] [`_mm_mask_reduce_min_epu8`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm_mask_reduce_min_epu8)
* [ ] [`_mm_mask_reduce_mul_epi16`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm_mask_reduce_mul_epi16)
* [ ] [`_mm_mask_reduce_mul_epi8`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm_mask_reduce_mul_epi8)
* [ ] [`_mm_mask_reduce_or_epi16`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm_mask_reduce_or_epi16)
* [ ] [`_mm_mask_reduce_or_epi8`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm_mask_reduce_or_epi8)
* [ ] [`_mm_reduce_add_epi16`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm_reduce_add_epi16)
* [ ] [`_mm_reduce_add_epi8`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm_reduce_add_epi8)
* [ ] [`_mm_reduce_and_epi16`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm_reduce_and_epi16)
* [ ] [`_mm_reduce_and_epi8`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm_reduce_and_epi8)
* [ ] [`_mm_reduce_max_epi16`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm_reduce_max_epi16)
* [ ] [`_mm_reduce_max_epi8`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm_reduce_max_epi8)
* [ ] [`_mm_reduce_max_epu16`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm_reduce_max_epu16)
* [ ] [`_mm_reduce_max_epu8`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm_reduce_max_epu8)
* [ ] [`_mm_reduce_min_epi16`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm_reduce_min_epi16)
* [ ] [`_mm_reduce_min_epi8`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm_reduce_min_epi8)
* [ ] [`_mm_reduce_min_epu16`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm_reduce_min_epu16)
* [ ] [`_mm_reduce_min_epu8`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm_reduce_min_epu8)
* [ ] [`_mm_reduce_mul_epi16`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm_reduce_mul_epi16)
* [ ] [`_mm_reduce_mul_epi8`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm_reduce_mul_epi8)
* [ ] [`_mm_reduce_or_epi16`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm_reduce_or_epi16)
* [ ] [`_mm_reduce_or_epi8`](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm_reduce_or_epi8)
</p></details>
<details><summary>["AVX512_FP16"]</summary><p>

File diff suppressed because it is too large Load diff