Build VPERM At Runtime (Again?) #11385
Merged
Conversation
This file contains 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
Nekotekina
force-pushed
the
master2
branch
3 times, most recently
from
867d001
to
b8776b6
Compare
Notably, runtime-built SSSE3 version of VPERM. Some other instructions are refactored and vectorized. Aarch64 impl of multiple instructions including VPERM.
Fix: nearbyint -> roundeven
Implement more sensible check for EVEX prefix and SSE
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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Outdated suggestions cannot be applied.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.
Some other refactoring.
Use efficient runtime dispatch for SSSE3 version of VPERM (built at runtime) and fallback. The main point is that the final function doesn't contain branches for runtime dispatch, it's only done at program startup. While it may seem silly, it completely bypasses compiler shenanigans regarding different target options (for example, if target differs for two functions, one cannot inline another in GCC/clang; MSVC has some problems as well).
ARM version of VPERM instruction is also written.