jimmyhmiller opened PR #6468 from jimmyhmiller:aarch64-sub-and-mul
to bytecodealliance:main
:
I just filled in some
todo!
for sub and mul on aarch64. I copied the file tests from x86. It should be fairly straight forward, but of course there is a chance I missed something obvious.@saulecabrera You mentioned fuzz testing being your main approach now. I wasn't sure how to do that, but would happily go run some if you can point me in the right direction.
jimmyhmiller requested abrown for a review on PR #6468.
jimmyhmiller requested wasmtime-compiler-reviewers for a review on PR #6468.
jimmyhmiller created PR review comment:
So this test on x86 looks like this. It does not have sub but add instead. I assumed this was a mistake, but just copied them for right now. Happy to change in both if that's correct.
saulecabrera created PR review comment:
Thanks for pointing this out -- it should definitely be
sub
.
jimmyhmiller updated PR #6468.
saulecabrera submitted PR review:
Looks good to me; I left a comment regarding one file test that includes x64 output, but the specified architecture is aarch64.
Would you also be willing to fix the x64/aarch64 subtraction test that you pointed out here https://github.com/bytecodealliance/wasmtime/pull/6468/files#r1207028404?
I think that after those fixes, we can land this change.
saulecabrera submitted PR review:
Looks good to me; I left a comment regarding one file test that includes x64 output, but the specified architecture is aarch64.
Would you also be willing to fix the x64/aarch64 subtraction test that you pointed out here https://github.com/bytecodealliance/wasmtime/pull/6468/files#r1207028404?
I think that after those fixes, we can land this change.
saulecabrera created PR review comment:
This test has x64 output, although the specified target is aarch64. Perhaps a bug in the filetest crate? Or maybe a copy paste mistake?
jimmyhmiller updated PR #6468.
jimmyhmiller updated PR #6468.
saulecabrera merged PR #6468.
Last updated: Dec 23 2024 at 13:07 UTC