Mikael Simberg
Mikael Simberg
> Wow. I had no idea that syntax was possible. > > Out of curiosity, do you know of any other use case for writing a function name that way?...
Thanks @khuck for bringing this up. We've definitely not been active users of the Kokkos profiling hooks with hpx-kokkos so this is undertested. I think this is orthogonal to hpx-kokkos...
Ok, thanks. I suppose the fencing question was actually a bit off topic. As far as I can tell, this is in the end simply a feature request to extend...
> Correct, this is a request to extend the profiling interface. That way, I can use a process-level map to map the deep copy completion event to the begin event....
This looks good to me besides the formatting changes required by CI. However, I'm just waiting for us to merge https://github.com/pika-org/pika/pull/436 before actually approving this since it's required for this...
Is there anything required on this PR to get `waiting for style check to succeed` to finish?
It looks reasonable, but actually this doesn't fix the problem, possibly due to a bug in CMake? I tried two spack `build-env`s, one with `[email protected]` and one with `[email protected]`. With...
> Have either of you tested with the 15 rc by any chance? Leaving it completely open ended as a patch isn't something I like to do if we can...
> Have either of you tested with the 15 rc by any chance? Leaving it completely open ended as a patch isn't something I like to do if we can...
So, `llvm@main` does still need _some_ patch. The build doesn't segfault with a vanilla GCC, but fails with the same problem as 13 and 14 have while building OpenMP without...