Yohann

I think H(div) should already be supported as a non-tensor basis (you might have to put each vectorial component in a different scalar field), similarly to other basis the interpolation...

I don't think it's an issue that non-tensor H(div) basis are bigger than Hˆ1, the libCEED interface is more or less agnostic of this. Whereas for the continuity, I thought...

> This same change probably also goes in `backends/hip-gen`, right? Yes, if you're fine with the change I can spend some time doing it in `hip-gen`, it should be quick....

> This same change probably also goes in `backends/hip-gen`, right? @jeremylt Change propagated to `hip-gen`.

> The CUDA CI flaked, but the HIP CI is happy now. Should I run `make test` on Lassen?

> This has identical changes in CUDA and HIP - CUDA CI is borked but HIP CI is passing so I think we are ok to merge @jeremylt What would...

I would unify `hip/ kernels/*` and `cuda/kernels/*` into `gpu_common_kernels/*` (or something similar), currently the code is duplicated. In the future it could be different, Hip/Cuda could have different code, but...

This is a good point, if we gather code, then we have to document in the same place the reasons for the differences. My proposal above is not a "first...

As soon as mfem/mfem#866 gets approved I can update the MFEM examples in libCEED.

@jedbrown I would need to work a bit more than that to make this work, it should be doable, but it is not ready (even for cuda-gen).