-
Notifications
You must be signed in to change notification settings - Fork 697
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
CI: add GHC 9.10 (backport #9914) #10058
Conversation
Two dots vs three dots strikes again. I remember we fixed it, but I don't remember on which branch and if PR commit that contained it (might have been a big PR about something else) got backported:
Would somebody try to find this? |
there was a really old PR (last summer) to that effect: I think this may be the effect of combing this one and #9671 (use in-tree Cabal library for cabal-install tests with custom setup) but since both are supposedly on the 3.12 branch, I don't understand why this failure doesn't happen on And there's another failure, which seems to be somewhat unrelated but also exhibiting some other deviation from the reference output:
|
e62a2ea
to
5c6f670
Compare
wat 😖 ETA: nemmind, the test actually passes here, |
Now I'm stuck on
Is this #9917? |
I doubt it. More likely #8133 but also looks like you could bump the bound in
|
#8133 would be especially bad news since the next thing I do is build the test suite, so I can run the failing test. |
Okay, got it. I hit this once before and forgot about it until I read the |
I don't even see what the failure is here. |
Oh, got it. Except that I don't see why it's looking for a ghc-9.6.4 inplace build of cabal when I told it to use ghc-9.10.1. ETA: gotta specify Reproduced test failure locally. Remaining questions:
|
Potentially I need 3 different patch files to make this work…. |
@geekosaur are you seeing the same failure that's currently shown in CI? ( |
Yes. |
Can you think of any reasons why this dots issue would only show up with 9.10.1? |
Also, as I feared, it just ran out of commits it could apply the 9.10 patch to (it can only apply to commits that have |
|
Yeh, so bisect finds nothing. All commits fail. I'm at something of a loss as to how to test for the included |
A simple sanity check that would be great to do, I think, is to check that changes from #9671 are still present on the 3.12 branch and weren't lost due to any quick conflict resolutions. |
Trying to repro it locally gives me a jarring exception:
Also, I dug a little into #9671, and at least the CI-related part of it (the |
I hit that one as well; it's because
|
my default is 9.10 :-) I never use |
A quick look suggests that the path contains |
* CI: add GHC 9.10 * GHC 9.10 compat in testsuite: CPP symbols don't get passed to CC as eagearly Discussion: #9914 (comment) Related GHC issue: https://gitlab.haskell.org/ghc/ghc/-/issues/21291 * GHC 9.10 compat in tests: disable tests regressing due to #9940 (cherry picked from commit 3a8c69c)
5c6f670
to
604c68a
Compare
All right, Linux has succeeded, so I assume we're good here. |
Correction: backports don't need a second review. |
Please read Github PR Conventions and then fill in one of these two templates.
Template Β: This PR does not modify
cabal
behaviour (documentation, tests, refactoring, etc.)Include the following checklist in your PR:
This is an automatic backport of pull request #9914 done by [Mergify](https://mergify.com).