obeis opened PR #13854 from obeis:arm32-isa to bytecodealliance:main:
some features aren't implemented yet.
I added file tests for my code.
obeis requested alexcrichton for a review on PR #13854.
obeis requested wasmtime-compiler-reviewers for a review on PR #13854.
obeis requested wasmtime-default-reviewers for a review on PR #13854.
obeis updated PR #13854.
obeis updated PR #13854.
alexcrichton commented on PR #13854:
Before going too too much further on this, have you seen this discussion on Zulip and/or https://github.com/bytecodealliance/wasmtime/pull/13815?
I did not see them.
:memo: programmerjake submitted PR review.
:speech_balloon: programmerjake created PR review comment:
shouldn't this be release, not acquire?
:memo: obeis submitted PR review.
:speech_balloon: obeis created PR review comment:
oops, should be
:speech_balloon: obeis edited PR review comment.
obeis updated PR #13854.
:speech_balloon: obeis deleted PR review comment.
obeis updated PR #13854.
github-actions[bot] added the label cranelift:area:machinst on PR #13854.
github-actions[bot] added the label cranelift on PR #13854.
github-actions[bot] added the label isle on PR #13854.
github-actions[bot] added the label cranelift:meta on PR #13854.
github-actions[bot] commented on PR #13854:
Subscribe to Label Action
cc @cfallin, @fitzgen
<details>
This issue or pull request has been labeled: "cranelift", "cranelift:area:machinst", "cranelift:meta", "isle"Thus the following users have been cc'd because of the following labels:
- cfallin: isle
- fitzgen: isle
To subscribe or unsubscribe from this label, edit the <code>.github/subscribe-to-label.json</code> configuration file.
Learn more.
</details>
alexcrichton commented on PR #13854:
Apologies @obeis I forgot this was assigned to me after this week's Cranelift meeting. For review of this I'm going to mostly defer to @cfallin and @fitzgen who are the primary maintainers of Cranelift's backends and probably have more insight into various pieces here than I do.
One thing I can provide some review on though is the CI infrastructure. Currently this doesn't add anything new which means that the filetests are executed but there aren't, for example, any runtests nor QEMU emulation. This means that we're never actually executing any of the generated code here which is a gap we'll either want to close in this PR or very shortly thereafter. Would you be interested in integrating the QEMU/execution side into this PR as well? For example enabling any
runtests that this PR enables for arm32? I believe @cpetig's work could be a good starting point here.
alexcrichton requested fitzgen for a review on PR #13854.
alexcrichton requested cfallin for a review on PR #13854.
alexcrichton unassigned alexcrichton from PR #13854 cranelift: add arm32 backend.
No problem, I will add the QEMU execution.
obeis updated PR #13854.
obeis updated PR #13854.
obeis updated PR #13854.
alexcrichton commented on PR #13854:
@cpetig I'm a bit confused by the latest commit, could you expand on it a bit? For example it seems like it duplicates the preexisting logic at
crates/wastandcargo test --test wastfor this repo. Is the intention to get some pieces running while everything's not stood up yet? If that's the case thecrates/test-utils/src/wast.rsfile has a bunch of filters/hooks to assert tests are either passing or not.
cpetig commented on PR #13854:
Hi @alexcrichton, your doubts are valid. I was suspecting that I can use a lot more existing infrastructure for the qemu tests, but I didn't find it at first try.
So I wanted to make sure that @obeis is able to use it before I disappeared for two weeks. Thus, I created a merge request to his branch which now got merged here, which wasn't my original intention. Probably motivated by a reading of your last message.
I plan to further refine the patch a week from now. Thank you for the hints on where to find the other parts of the logic.
cpetig edited a comment on PR #13854:
Hi @alexcrichton, your doubts are valid. I was suspecting that I can use a lot more existing infrastructure for the qemu tests, but I didn't find it at first try.
So I wanted to make sure that @obeis is able to use it before I disappeared busy for two weeks. Thus, I created a merge request to his branch which now got merged here, which wasn't my original intention. Probably motivated by a reading of your last message.
I plan to further refine the patch a week from now. Thank you for the hints on where to find the other parts of the logic.
Last updated: Aug 30 2026 at 09:07 UTC