Add test coroutine

[Pavel-Durov]

Related Issue: #58

[Pavel-Durov]

We still can't run CI validation with the latest YK
Failed buildbot job - https://ci.soft-dev.org/#/builders/1/builds/2527

[vext01]

The issue is still open, so how comes we are re-enabling this test?

[Pavel-Durov]

The issue is still open because I don't want to close it before I can validate tests with buildbot.
I've tested coroutine.lua myself and it works with try_repeat 1000.

[vext01]

Was there any change that you'd expect to have fixed that test?

[Pavel-Durov]

Yes, this commit: b931338

[vext01]

bors r+

[bors]

Build failed:

• buildbot/buildbot-build-script

[vext01]

What should we do with this?

[Pavel-Durov]

good question :)
I don't think we can merge any PRs in yklua since breaking changes in YK.
Last working version: 4a955668d5a6647c5de8f043314983f057a5039e (Sep 15, 2023)

[Pavel-Durov]

bors try

[bors]

try

Build succeeded!

The publicly hosted instance of bors-ng is deprecated and will go away soon.

If you want to self-host your own instance, instructions are here.
For more help, visit the forum.

If you want to switch to GitHub's built-in merge queue, visit their help page.

• buildbot/buildbot-build-script

[Pavel-Durov]

Bors CI succeeded I guess we can merge it then!? @vext01

[vext01]

I think so.

Any squashing to do?

[Pavel-Durov]

No squashing needed :)

[vext01]

Great!

bors r+

[bors]

Build failed:

• buildbot/buildbot-build-script

[vext01]

16:19:38 thread '<unnamed>' panicked at ykrt/src/mt.rs:469:21:
16:19:38 assertion failed: sidetrace.is_none() ||
16:19:38     matches!(hl_arc.lock().kind, HotLocationKind :: Compiled(_))

One for @ptersilie ?

[Pavel-Durov]

Oh, that's new :)

[ptersilie]

16:19:38 thread '<unnamed>' panicked at ykrt/src/mt.rs:469:21:
16:19:38 assertion failed: sidetrace.is_none() ||
16:19:38     matches!(hl_arc.lock().kind, HotLocationKind :: Compiled(_))

One for @ptersilie ?

Hm, this error is caused by https://github.com/ykjit/yk/blob/master/ykrt/src/mt.rs#L470. This essentially says "if the current location doesn't have a compiled trace, then there also can't be a side trace". Since the only way sidetrace can be set here is via StopSideTracing (https://github.com/ykjit/yk/blob/master/ykrt/src/mt.rs#L279), I don't quite see how we can end up here.

It can't be a coincidence that this is caused by adding coroutines.lua. Maybe @ltratt is more suited for this?

[Pavel-Durov]

Closing this PR as coroutine.lua test is currently failing.

[ltratt]

@Pavel-Durov I don't think we should give up quite that quickly :) Can you investigate what part of the assert is failing? Might be worth inserting something like db!hl_arc.lock().kind); just before the assert.

[Pavel-Durov]

I'm not giving up :)
I'm just closing the PR until we fix the issue #58

[ltratt]

The bug in that issues seems different than the failing assert here?

[Pavel-Durov]

It does look different but I can't reproduce the same error as in buildbot with the latest YK version c6a0b9ca333408693a0c904bd502b1cdf0dc8d3f

[ltratt]

I wonder if you might have a slightly stale build or similar. If you rm -rf target in yk and make clean in yklua, and rebuild both, does that change anything?

[Pavel-Durov]

I will try doing clean build and run this test again

[Pavel-Durov]

bors try

[bors]

try

Build succeeded!

The publicly hosted instance of bors-ng is deprecated and will go away soon.

If you want to self-host your own instance, instructions are here.
For more help, visit the forum.

If you want to switch to GitHub's built-in merge queue, visit their help page.

• buildbot/buildbot-build-script

[vext01]

Is this one ready for squashing then? If so, please go ahead.

[Pavel-Durov]

not ready yet, tests are failing again.

[Pavel-Durov]

bors try

[bors]

try

Build failed:

• buildbot/buildbot-build-script

[Pavel-Durov]

closing in favour of #79