This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [patch, Fortran] Fix PR 50690
On Sat, Oct 15, 2011 at 07:19:43PM +0200, Jakub Jelinek wrote:
> On Sat, Oct 15, 2011 at 06:36:55PM +0200, Tobias Burnus wrote:
> > I think that's the wrong solution to the problem: I think it makes
> > more sense to fix the handling for OMP workshare to handle BLOCK
> > correctly rather than to move the variable to the function scope and
> > then to fiddle with the private attribute. (By the way, an
>
> I don't agree here. Until we know if OpenMP will allow or disallow
> BLOCK in workshare and what the semantics for it should be, we shouldn't
> allow it. OpenMP <= 3.1 has very tight rules on what is allowed in
> workshare, it is verified during resolving and during translation
> we rely on it still being present.
Though, what could be done is just special case OpenMP workshare regions,
insert everything into BLOCK local vars unless in OpenMP workshare, in that
case put the BLOCK with the temporary around the workshare rather than
inside of it. In the case of omp parallel workshare it would need
to go in between omp parallel and omp workshare.
Jakub