Hans Pabst
LGTM. However, I wonder how this interferes with the pending decision to drop GCC8? Given this change, GCC8 can be kept.
@dmishura Awesome - thank you, Dmitry!
I explored an approach to use CP2K's convention checker and essentially ban nested parallelism or more specifically a parallel region inside of a master section. However, the GFortran AST cannot...
> Also please note that MASTER construct is deprecated in latest spec. This questions the whole MPI-like programming model in OpenMP based on `get_get_thread_num`. This is about "OpenMP tasks everywhere"...
The current thesis is, nested parallelism like PARALLEL WORKSHARE even inside inside of a MASTER section (which in turn is inside of another PARALLEL region) should never hang. GNU Fortran...
> Thanks @hfp I will take a look next week... Might be easier now (less work, only three files). I separated some cleanup work (#860).
@mkrack this single PR is apparently resolving all runtime issues with Intel Fortran Compiler (IFX) in CP2K/DBCSR. There is still the question why `PARALLEL WORKSHARE DEFAULT(none) SHARED(...) IF(spawn)` with `spawn=.NOT....
@alazzaro any feedback welcome - so far, no issues with this code on my side, I am actively using it since sending the PR.
Objective of this implementation is to exercise a way to preserve the (parallel) workshare. It does not judge (or confirms) whether the workshare is beneficial or not.
Hold on merging this PR. I pulled in more expertise. I will reply on the original issue https://github.com/cp2k/dbcsr/issues/857, and revise/close this PR accordingly.