Hartmut Kaiser
Hartmut Kaiser
@diehlpk Yes, this is expected. We have disabled the schedulers that depend on 128bit atomics on platforms where those are either not available or are not lockfree. We have seen...
@Pansysk75 what's the status here? Can this be merged?
> @hkaiser Should we set the [CMP0144](https://cmake.org/cmake/help/latest/policy/CMP0144.html) CMake policy to NEW on a project-wise level? Now, because it is unset, the existence of env variables like BOOST_ROOT, HWLOC_ROOT etc. cause...
> However, I also noticed that you recently modified all `_ROOT` to `_ROOT` across the codebase. Could you point me to what motivated that change? Edit: Never mind, I now...
@Pansysk75 we still have the circular dependency between the concurrency and allocator modules. Also, some of the header tests are failing (most likely caused by the circular dependency issue)
> > @Pansysk75 we still have the circular dependency between the concurrency and allocator modules. Also, some of the header tests are failing (most likely caused by the circular dependency...
@Pansysk75 I think this is good to go as a stop-gap measure circumventing the underlying problem from happening. What do you think?
@Pansysk75 the original issues are resolved here. Let's merge it now.
> Should we add feature tests for these, or just disable them for Clang/GCC? I suppose these were added with the Intel compiler in mind? They were created for gcc...
@opensdh I will update this PR as I'm making progress on documenting the HPX backend.