https://gcc.gnu.org/bugzilla/show_bug.cgi?id=120958
--- Comment #24 from rguenther at suse dot de <rguenther at suse dot de> --- On Thu, 17 Jul 2025, anlauf at gcc dot gnu.org wrote: > https://gcc.gnu.org/bugzilla/show_bug.cgi?id=120958 > > --- Comment #23 from anlauf at gcc dot gnu.org --- > (In reply to Richard Biener from comment #22) > > (In reply to Richard Biener from comment #21) > > > A pragmatic solution might be to pattern-match (by name) some of the > > > affected APIs and exclude them from the fnspec processing that causes the > > > issue. > > > > Main reasoning the body of code using Fortran Standards version not capable > > of correctly annotating said APIs isn't (or shouldn't...) be increasing, > > so the chance we can catch everything important is high (not to suggest > > a -fasync-api-pattern=mpi_*,foo* way of user control to adjust the patterns > > - yet). > > So if FOO is a wrapper subroutine that calls MPI_I*, one needs to find all > of them in the code? Technically yes. You already have to be unlucky to see SRA do sth given with Fortran the chances the address of something is taken is quite high (everything is passed by reference).
