Alexander Sinn
Alexander Sinn
Good point, maybe at the next WarpX+AMReX development meeting?
Now that ParallelFor is N dimensional, I think a base level of functionality for N dimensions is provided. Below, I listed some additional classes that could be converted to N...
Yes. There is still a lot left to do. I am waiting for Axel to return and give feedback first
This PR depends on #4529 btw
I found a `Gpu::inLaunchRegion()` in `meanParticleVelocity`. With how that's written, maybe It's fine to keep it, but I noticed that `meanParticleVelocity` does not take into account particle weight.
I changed it so now it is `getInvalidRandomEngine()` to make it more clear.
It is indeed super complicated when used with multiple threads. In that specific example, since the kernel launch from thread 0 happens before cudaStreamSynchronize is called from thread 1, the...
Yes I am still working on this. Next I will try to give each stream an array with one bool per omp thread to store if the stream is synced.
This makes it easier to use `PolymorphicArenaAllocator`: https://github.com/AMReX-Codes/amrex/pull/4603