Boris Kaus

That is required if you want to run SuperLU_Dist on a GPU. Is that an immediate priority?

thanks - that worked smoothly. Seems ready now.

@eschnett we currently have a failure with the following error when using mpitrampoline:: ``` [14:22:13] /workspace/srcdir/petsc-3.16.6/src/mat/impls/scalapack/matscalapack.c:20:45: error: initializer element is not a compile-time constant [14:22:13] static PetscMPIInt Petsc_ScaLAPACK_keyval = MPI_KEYVAL_INVALID;...

@eschnett @giordano : this compiles fine now. Please have a look if you are happy with it.

I see no reason why this would be done. In the testing we do this but with a `finalize` at the end

yes good point; the current version of Perple_X does indeed not allow that (even when it is already distributed as open source code). Email to Jamie is on the way.

`Perple_X_jll` is now [available](https://github.com/JuliaBinaryWrappers/Perple_X_jll.jl) and appears to work on all available systems. It is build using this [binary builder script](https://github.com/JuliaPackaging/Yggdrasil/blob/master/P/Perple_X/build_tarballs.jl) from the [github repository](https://github.com/jadconnolly/Perple_X) that Jamie created to store the...

so windows compilation still fails as can be seen [here](https://buildkite.com/julialang/yggdrasil/builds/1199#0186a688-a053-4e05-9bf8-d34fae5d702f) with the following error: ```bash [08:18:04] /workspace/srcdir/gmt/src/gmt_api.c:952:10: fatal error: TlHelp32.h: No such file or directory --   | [08:18:04] #include   |...

thanks. The patch file suggests that we should also change to at a couple of places. See this [patch](https://github.com/JuliaPackaging/Yggdrasil/pull/5199/files#diff-242362bc935c2ff33934b28ba8d365498b7fc1d9877cbee903cf2caf07273a12)

Do you need the executable or the dynamic library? Same with gmt: should the dynamic library be shipped as well?