Skip to content

Conversation

@tomsmeding
Copy link
Collaborator

@tomsmeding tomsmeding commented Sep 16, 2025

This PR builds on top of #83 (CUDA 13 support), merge that PR first!
Actual changes in addition to #83 can be seen in this compare view.

  • Updated github actions to newest versions.
  • Fixed some bugs in Setup.hs (most of my own making in Support new Cabal library in Setup.hs (3.12, 3.14, 3.16) #82) that only occurred with old GHC/Cabal.
  • Unlist GHC <8.4 from CI matrix because (the latest) stack doesn't support them any more
  • Add extra-deps for language-c to the stack.yamls before stackage LTS 23 in order to support CUDA 13
  • CI is green :)

The oldest 4 GHC versions that we have a stack.yaml for are now untested in CI, but I kept the files nevertheless (and added the extra-deps for language-c without testing). I figured that they might be useful for someone. Also happy to remove them if that is better.

EDIT: It's worth noting that, as before, this CI only tests whether the thing builds. While that's a useful test, it's a far cry from proper testing. Fortunately, some of this code is at least tested on recent GHC/Cabal/CUDA versions via integration with accelerate-llvm-ptx, but it's not all.

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.

1 participant