Benedict

Thanks for the suggestions! Here is a small update: When I put `MPI_Barriers` **after** `if n_boundaries > 0 ... end` they are passed by all ranks. I see the corresponding...

Ready for review @trixi-framework/developers . This PR involves places I have not been before (visualization, packages extensions, ...). Any feedback or hints would be highly appreciated.

I got stuck at some point because I was unsure about how to handle Requires.jl. However, we removed Requires.jl in the meantime, so this PR is now ready to be...

Thanks for picking this up! For a start, running `@profile` on 1000 iterations of `rhs!` and using `PProf` gives these flamegraphs (top: 1.10.4, bottom: 1.11-rc1)  We...

Yes! For 1.11 we have `checkbounds` beneath these calls, which make up a fair amount of samples. For 1.10 they are not captured.

Confirming @sloede 's observations: With the alternative implementation of `get_node_vars` we see an improvement both with 1.10.4 and 1.11.0-rc1. For `get_contravariant_vector` we see an improvement with 1.11.0-rc1 but a slowdown...

When working on the visualization callback (#2225 ) I only recently realized that `PlotData2D` is specialized for TreeMesh. It produces a `PlotData2DCartesian` object in this case, while there will be...

> It may be related to this issue of workflow runs can't be triggered by processes authenticated with the repository's GITHUB_TOKEN: https://github.com/orgs/community/discussions/57484 Yes, this makes sense. But isn't that the...

I had a look at the advection diffusion problems. All of them (elixir_advection_diffusion_*) show some slight deviations. Only one exceeds the tolerance, however.

So far I can only confirm what you observed, @DanielDoehring. I see almost the exact numbers when running with threads or MPI (MPI.jl and system OpenMPI).