Skip to content

Conversation

@kvpanch
Copy link
Contributor

@kvpanch kvpanch commented Oct 31, 2025

This changeset is based on #346

@kvpanch kvpanch requested a review from xermicus October 31, 2025 15:22
@kvpanch kvpanch force-pushed the kvpanch/llvm_submodule branch from 6c7f201 to ddfb47b Compare October 31, 2025 16:14
@kvpanch
Copy link
Contributor Author

kvpanch commented Nov 3, 2025

@xermicus now it's good to go

Copy link
Member

@xermicus xermicus left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks, very nice! Two things stick out:

  • The submodule should only be required in LLVM test and release workflows?
  • The Test LLVM Builder action doesn't seem being run in this PR. But it should surely have been triggered based on this change set?

branches: ["main"]
types: [opened, synchronize]
paths:
- 'LLVM.lock'
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Maybe this should run when the .gitmodules has changed?

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The file no longer exists. The way it suppose to work in future is we update new llvm file with new commit to pull from LLVM

Comment on lines 25 to 26
with:
submodules: true
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Same here, shouldn't be needed?

Copy link
Contributor Author

@kvpanch kvpanch Nov 4, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

let me try to remove it

Copy link
Member

@xermicus xermicus left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Oh no - the LLVM builder workflow not showing up is due to an invalid spec: https://github.com/paritytech/revive/actions/runs/19039292406

This isn't this PR's fault but would you mind fixing it here at the same time as drive-by?

@kvpanch
Copy link
Contributor Author

kvpanch commented Nov 4, 2025

Oh no - the LLVM builder workflow not showing up is due to an invalid spec: https://github.com/paritytech/revive/actions/runs/19039292406

This isn't this PR's fault but would you mind fixing it here at the same time as drive-by?

interesting. The error complain about path and path-ignore, but the structure there didn't change, not sure what happens

@kvpanch kvpanch force-pushed the kvpanch/llvm_submodule branch from b81b7f3 to 64ceedf Compare November 4, 2025 17:05
@xermicus
Copy link
Member

xermicus commented Nov 4, 2025

The problem is that we can only have either paths or paths-ignore defined.

From my side there is no need to bikeshed this. Just remove paths-ignore to fix it. It was intended as a minor convenience tweak but it's not the end of the world if we run the workflow for markdown changes.

I didn't catch this during reviewing the PR introducing the regression. The docs would have explained that but of course I didn't read them careful enough. It's really unfortunate (and also very surprising to me) that in the PR itself there seems to be no indicator whatsoever if there are invalid workflow yml files found 😒

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants