Detect all Apple M* CPUs and enable the wide multiplier assembly implementations #1901
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Outdated suggestions cannot be applied.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.
Description of changes:
Looking at the benchmarks on https://github.com/ctz/graviola/ I noticed AWS-LC was slower than expected. Those benchmarks were performed on an M2 Mac and previously AWS-LC was only checking for M1 CPUs. This change opts all Apple M* CPUs in the alt/wide multiplier implementations. A future Apple M CPU might benefit from the non-alt implementation, but for now it seems like all M CPUs benefit from the alt implementation and this is a sane default.
Call-outs:
I think this is safe because both implementations use the same instructions, it's just a reordering. So if a future M CPU doesn't support the wide multiplier it will still work it will just be slower than if it was using the non-alt implementation.
Testing:
Tested on an M1 and M3 that the alt implementation is picked and is faster. M3 before:
M3 After
As expected P-224 and P-256 are unaffected, while 384/521/X25519 which do use s2n-bignum are much faster.
By submitting this pull request, I confirm that my contribution is made under the terms of the Apache 2.0 license and the ISC license.