-
-
Notifications
You must be signed in to change notification settings - Fork 18
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
Remove scotch/metis link #112
Conversation
rather than directly linking mumps dependencies metis and scotch
…nda-forge-pinning 2024.09.25.18.51.16
@@ -26,11 +26,10 @@ requirements: | |||
- cmake # [win] | |||
- make # [not win] | |||
- wget # [not win] | |||
- pkgconfig | |||
- pkg-config |
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.
btw, pkgconfig
is a Python package, pkg-config
is command-line tool
Hi! This is the friendly automated conda-forge-linting service. I just wanted to let you know that I linted all conda-recipes in your PR ( |
analysis seems good:
|
Co-authored-by: Silvio Traversaro <[email protected]>
@conda-forge-admin please rerender |
…nda-forge-pinning 2024.09.26.07.35.45
New output:
I think this is a good to go, what do you think @minrk ? |
Yup, looks right to me! |
For future reference and for the repodata patch, linking scotch without expressing the dependency started in 3.13.3 build 2 with #50 . I will probably only do the repodata patch for 3.14.16 because finding the scotch version used at build time will be a pain. Since mumps is pinned x.y.z transitively pins libscotch also to x.y.z, it's really only builds against a given version of mumps when that version of mumps is later rebuilt against newer scotch that are affected, which I don't think happens that often, and likely means the latest build of most ipopt versions is fine and can't break in the future because past versions of mumps aren't getting fresh builds. |
I am afraid there was a regression: #114 . |
#111 fixed linking because scotch was linked but not expressed as a dependency, resulting in dll load errors when there's a scotch update.
But I think this is the 'real' fix, because mumps dependencies shouldn't be linked transitively (that's the case for static mumps, not the shared libraries we have in conda-forge).
Let's wait to merge until we can check the DSO inspection output from conda-build to make sure a link isn't still there that's not expressed in the dependencies.