-
Notifications
You must be signed in to change notification settings - Fork 4.8k
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
Fix zlib-ng build with cmake 3.31.0 #110359
Conversation
Tagging subscribers to this area: @dotnet/area-infrastructure-libraries |
@jakobbotsch I'll revert the dummy change in JIT once windows format leg is green. |
@jakobbotsch, a side note: in https://github.com/dotnet/runtime/actions/runs/12145206891/job/33866308923?pr=110359 we can see repeated downloads:
first in |
I did not see the jit-format Windows run succeeding with this PR, it just seemed to be permanently stuck when I looked. Did you see it succeed? |
Probably, I'm not really familiar with the setup of this, cc @dotnet/jit-contrib
As long as it doesn't come with a large amount of churn to the JIT code base... |
The comment above it has a link to green run? |
Windows jit-format action has been broken on main for like 2 weeks, it's failing on every single JIT PR. |
Co-authored-by: Jeremy Koritzinsky <jkoritzinsky@gmail.com>
Ah, ok, looks good then. For some reason when I look at the "Checks" page it shows me a still running jit-format job.
Yes, @am11 is aiming to fix the failure with this PR. |
Yup, I had to update a file under coreclr/nativeaot and one eng/native to cleanup few other warnings. The problem is that the python script fails when there is even a single config warning. Then GitHub running gets stuck for five hours in 'spawned process exiting with non-zero line' part of the pythong script. Maybe we should try to make the script robust against warnings. |
It's fine to only download the tools once. It's not clear which download is actually getting used. I'm not familiar with the current job setup; @jkoritzinsky created that. As for the version change, we shouldn't update unless we are required to. Or it provides large benefits of some kind. At one point there was an idea to share the tools between JIT and non-JIT runtime users, but I don't think that happened (it would make it much harder to update if it did). |
I specifically wanted to build We shouldn't be downloading jitutils again in the |
Now it's downloading once.
The python script clones it under $TEMP and does the "right thing". It won't clutter the diffs. |
You can remove the Checkout jitutils step as well then. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This looks good to me but I'm not very familiar with any of this, so I will defer to @jkoritzinsky for another approval.
Thanks! |
We are seeing warnings like these (with cmake 3.31 in github actions):
When we started src/native/external projects, the idea was that we will keep our configs outside the vendor directory and we won't be using their cmake configs if they are providing one. That made version updates simple and "our" configs were just a flat list of files we want to build in the product. Now we are adding new dependencies with "fetch plan", that uses the cmake config from the vendor dir and we have to keep up with the config issues..
fyi @jkoritzinsky one downside of fetch plan.