https://github.com/DeinAlptraum requested changes to this pull request.
I think I found the reason why the CI doesn't run. Unfortunately it made me realize that the changes here might require some deeper adaptions... So far the CI was a single-step workflow, that just calls the `llvm-project-tests` workflow with some parameters. But by design that workflow is meant to be run standalone (it is the only step in the pipeline) so it does one-off things, such as determining itself which runner image to use. But once we add a second step in our own workflow, we cannot reasonably delegate the runner image choice to `llvm-project-tests` anymore... Due to that, I'm not sure if the current approach can be made workable at all without larger changes to our CI setup. @rorth I know this is a bit much to figure out if you don't have experience with GHA. I will do some experiments to try and figure out if/how this can be solved, but I'm a bit busy atm, so not sure when I'll get around to that. Feel free to abuse the CI to try things out yourself, or to ask if you want any guidance. https://github.com/llvm/llvm-project/pull/142948 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
