Hi Tobias, ok for the Fortran part.
- Andre On Tue, 3 Dec 2024 10:34:34 +0100 Tobias Burnus <tbur...@baylibre.com> wrote: > In C, [[omp::decl]] wasn't supported and when discussing 'alignof' with Jakub, > he pointed out that there are corner cases (see commit log for one), which > imply that it is better that 'omp allocate align(…) doesn't affect 'alignof'. > The patch now removes the alignment handling from the FE (C and Fortran, but > only required for C); for static variables it is now handled in > varpool_node::finalize_decl (and, as before, for stack variables in > gimplify_bind_expr). [Stack variables are omp_alloc'ed, for static > variables only the alignment is honored, albeit the backends could do > something special. While they don't, I could imagine that the GPU > backends might eventually handle them in a special way.] > > Comments suggestions? > > Tobias > > PS: C++ only has a very basic support (fails with sorry during parsing). > However, a patch to add proper support will by shortly be submitted by my > colleague. It is based on a patch that was posted end of last year, but fixes > to Jakub's review comments have/had still to be done. > The attached patch is kind of a fallout of discussions that were started > during the C++ revision work. -- Andre Vehreschild * Email: vehre ad gmx dot de
