Repository navigation
Fix ARM (Apple Silicon) build of the fast pair HMM - #12
Merged
Merged
Conversation
cactus defines __AVX2__ on ARM so abPOA can use SIMDe, which made the fast pair HMM include immintrin.h there and fail to build. The intrinsics are now gated on the target as well as the feature macros. On aarch64, FPCR.FZ is set for the same span as the x86 FTZ/DAZ bits so results match x86. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
cactus adds
-D__AVX2__on ARM so abPOA can use SIMDe, and since #9 that makespairwiseAligner.cincludeimmintrin.hthere, which breaks the build on Apple Silicon (and presumably aarch64 Linux).PAIR_HMM_X86_SSE2/PAIR_HMM_X86_AVX2macros. The scalar path is used everywhere else.FPCR.FZis set and restored around the same block as the x86 FTZ/DAZ bits, so denormals are flushed the same way and results should match x86.Tested by building cactus master on an M-series Mac (macOS 15, Apple clang 17) and running
examples/evolverMammals.txtend to end.🤖 Generated with Claude Code