Allow for native Apple M1 ARM builds! #1298
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
The version of the macOS sdk we're using doesn't implement it, see more detailed in-code comment included in this commit
First (pretty obviously), don't set the preprocessor define "CPU_X86_64" if we're not actually targeting x86_64 :) Secondly, don't set some linker flags that seem to cause problems on aarch64 builds -- with them present, launching the binary under lldb exits with "Malformed Mach-O file" and regular launches fail with mysterious error codes. -- Having tested myself, it seems to work just fine without them set; I believe that even for x86_64 they were added by accident! (See in-code comment)
If we're targeting a native M1 aarch64 build, obviously we don't want to tell FFMpeg to build for x86_64 :)
It doesn't make set a deployment target earlier than the first OS that actually supports the CPU :)
We only actually use cpuid.h on Linux, and including it causes a problem for M1 builds because it doesn't exist outside x86_64!
The calc code uses x86_64 SIMD instrinsics for speed. However, those are unavailable on ARM. SSE2Neon provides ARM equivalents of each missing instruction using ARM's Neon SIMD instructions, providing equivalent speed for the majority of x86_64 SIMD instructions :)
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Outdated suggestions cannot be applied.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.
In 2020, Apple introduced their own ARM chip, the M1, which replaced the Intel x86_64 chips they had been using.
Consequently, all modern macs run on ARM processors and run ARM code.
Etterna has only been able to work on them because Apple provided Rosetta v2, a translation layer allowing our x86_64 mac builds to be dynamically JIT-d into ARM code, which can be run.
However, this JIT process has some overhead and produces less efficient code than code compiled for the target processor in the first place, and more importantly Rosetta v2 will have support dropped for it at some point in the future.
This PR allows Etterna to be built for Apple Silicon targets, and is an important first step in the process of automatically generating Apple-Silicon-compatible releases.
The other two major components are telling GHA to build for Apple Silicion targets, and notarizing our executable so it can actually be distributed (see: #1292)
Depends-on: #1297 , as without it Etterna will not compile for macOS at all, even for a x86_64 target.