avx512bw,vbmi: Optimize shift_left and shift_right. #1147
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This patch picks the instructions from avx512vbmi for the fast path. Masking is faster than an additional AND instruction.
Instead of loading from an aligned array on the stack. So we yield the `set` instead of `load` intrinsic, which makes it easier for the compiler to constant fold this parts. Sadly, MSVC needs this....
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
_maskz_permutexvarinstead ofpermute + andon avx512bw, tooxsimd::make_batch_constantand so_mm512_setinstead of_mm512_load(constexpr std::array)The first one is just a tiny optimization already done in the AVX512VBMI implementation.
The second patch makes it easier for the compiler to move this constants into the .text section. GCC is not affected, MSVC sadly is affected a lot. It used to generated 8 times
mov [rbp + N], constwith a singlevmovdqu32 zmm, [rdp]afterwards - poor MSVC.... And worse for the store_forwarding unit....