Move some linear algebra/matrix operations to internal API

[sebastiangrimberg]

Addresses #1156 (comment)

[jeremylt]

For some reason not all CI tests are running, but I think some of the t30* tests now need ceed/backend.h added

[sebastiangrimberg]

For some reason not all CI tests are running, but I think some of the t30* tests now need ceed/backend.h added

They aren't running because all of your Github Actions workflows are marked as on: push and so PRs from forks won't trigger jobs. You can either modify them to have a header which looks like:

on:
  push:
    branches:
      - main
  pull_request:

or, I'd love to develop right from CEED/libCEED in which case none of that is necessary. But I need write access to the repo to do that. Indeed I was waiting for tests to run to add the necessary includes.

[jeremylt]

Huh, I thought that on pull_request also covered forks?

I don't have perms for adding people I don't think

[sebastiangrimberg]

on: pull_request does cover forks but most tests in libCEED are only on: push which does not cover forks. See for example https://github.com/CEED/libCEED/blob/main/.github/workflows/c-fortran-test-linux-osx.yml#L3.

[jeremylt]

Ahh. I can fix that on Monday

[sebastiangrimberg]

No problem! Another question for this is should these functions get pulled out of the Python/Fortran APIs? Looks like they are already not in the Rust one.

[jeremylt]

Probably. Users have better options for that stuff in those languages other than libCEED

[sebastiangrimberg]

Awesome I agree! Thanks

[jedbrown]

I added you to the repo so you can supersede with an in-repo branch if you'd like.

[sebastiangrimberg]

Thank you!

…

On Fri, Feb 24, 2023 at 8:29 PM Jed Brown ***@***.***> wrote: I added you to the repo so you can supersede with an in-repo branch if you'd like. — Reply to this email directly, view it on GitHub <#1165 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AGLF4FE4I6B7324AM5YBEUTWZGDA5ANCNFSM6AAAAAAVHPQY2Q> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[jeremylt]

See #1166 for the CI fix.

I want to get CI running for this branch for sanity, but the overall changes look right to my eye (though I don't trust my eye as much as the CI)

[jeremylt]

#1166 merged, so rebasing should let CI run for this PR

[jeremylt]

Awesome, this looks good to me

[jeremylt]

I think this is ready for squash+merge but wanted to check if you had anything else to do on this branch first. It looks like there is a merge conflict?

[sebastiangrimberg]

All good from my end, do what you need to do. I'm happy for you to rebase, squash, and merge as you'd like.