Hartmut Kaiser
Hartmut Kaiser
> I don't like renaming `interface` to an `ugly_` name, since it is in fact exposed to specialization authors (albeit not to application developers). I know we've been down a...
@ipdemes I have removed the WIP tag from this as I think the PR has reached a minimal viable state. The tests from the `run` and `exec` submodules compile and...
@ipdemes I have updated the PR with recent changes from https://github.com/flecsi/flecsi/tree/2. All tests pass for me now. Please let me know if there is anything else you would like me...
> @ipdemes I have updated the PR with recent changes from https://github.com/flecsi/flecsi/tree/2. All tests pass for me now. Please let me know if there is anything else you would like...
> @hkaiser > > Do you mean when building HPX or building FleCSI? I think it should work without any additional settings. For both, HPX and FleCSI the corresponding options...
> @hkaiser > > FleCSI it self does not use the ENABLE_MPI_THREAD_MULTIPLE macro. Does it impact some of the HPX header files? No, the HPX backend uses multi-threaded MPI in...
> We do need release notes and an update to `options.rst` even if no other documentation is required because it works just like the other backends. I updated options.rst (added...
> Shouldn't we have a Spack variant for HPX? We do have HPX support in Spack.
@opensdh this PR is now ready for final review. Thanks!
> @hkaiser > > Current version breaks the MPI backend. Did you make the needed change to it? Will check, thanks!