abrown commented on Issue #1318:
Yeah, I know! I'll keep looking at #1306 but I wanted to avoid having a bug in master for too long.
abrown commented on Issue #1318:
@alexcrichton, any idea what https://github.com/bytecodealliance/wasmtime/pull/1318/checks?check_run_id=506812657#step:9:4117 means?
thread '<unnamed>' panicked at 'trampoline compilation should not produce external symbol relocs', crates/jit/src/compiler.rs:434:9
alexcrichton commented on Issue #1318:
Oh dear that sounds like another bug! I haven't run into that, but I can try to dig in.
abrown commented on Issue #1318:
Good thing we have fuzzing...
alexcrichton commented on Issue #1318:
Hm ok, so it looks like that condition may still be necessary. It looks like the trampoline is so big it needs the probestack libcall as a relocation, but we assert trampolines don't need relocations. Mind adding the smaller 17-f32-argument test in and leaving that condition in? That way we should still assert this bug at least is fixed and I can get to the other when I get a chance.
abrown commented on Issue #1318:
Sure, I guess we should open another bug for that specific issue and change the FIXME text to point to that?
alexcrichton commented on Issue #1318:
Sounds good to me!
abrown commented on Issue #1318:
I think this is failing for the same reason as #1319, some
cc
link error.
abrown edited a comment on Issue #1318:
I think this is failing for the same reason as #1319--some
cc
link error--not an actual issue with this PR.
iximeow commented on Issue #1318:
Where do you see a
cc
error? I'm seeinggit checkout --progress --force ea9bfb97b4eb6b5dca08bc3de09a83de59245601 ##[error]fatal: reference is not a tree: ea9bfb97b4eb6b5dca08bc3de09a83de59245601 Removed matchers: 'checkout-git' ##[error]Git checkout failed with exit code: 128 ##[error]Exit code 1 returned from process: file name '/home/runner/runners/2.165.2/bin/Runner.PluginHost', arguments 'action "GitHub.Runner.Plugins.Repository.v1_0.CheckoutTask, Runner.Plugins"'.in all the logs, which make me wonder if GitHub is feeling okay? How would that commit be missing? But a
cc
error earlier on might result in the repo disappearing?
abrown commented on Issue #1318:
No, you're right... this is different. Hm...
abrown commented on Issue #1318:
Well, I guess the same thing happened on #1319; first I saw a
cc
error and after re-running all of the actions now I see a Git failure...
iximeow commented on Issue #1318:
FWIW I have since seen The
cc
Error: https://github.com/bytecodealliance/wasmtime/runs/507019811 - my intuition is that our build might hit a disk space limit wherecc
fails on writing out artifacts? I'm a bit sensitive to that specific cause from dealing with it locally.
github-actions[bot] commented on Issue #1318:
Subscribe to Label Action
This issue or pull request has been labeled: "cranelift", "cranelift:meta"
<details> <summary>Users Subscribed to "cranelift"</summary>
- @bnjbvr
</details>
To subscribe or unsubscribe from this label, edit the <code>.github/subscribe-to-label.json</code> configuration file.
Last updated: Dec 23 2024 at 12:05 UTC