https://gcc.gnu.org/bugzilla/show_bug.cgi?id=88460
--- Comment #4 from Tom de Vries <vries at gcc dot gnu.org> ---
(In reply to Jakub Jelinek from comment #3)
> That test is huge, can you narrow it to something smaller that still ICEs?
> Like, bisect to a single omp target in main and corresponding fN function +
> support?
Sure:
...
int results[2000];
#pragma omp declare target
void
baz (int i)
{
results[i]++;
}
void
f1 ()
{
#pragma omp distribute parallel for default(none)
for (int i = 0; i < 2000; i++)
baz (i);
}
#pragma omp end declare target
int
main ()
{
#pragma omp target teams map (tofrom: results)
f1 ();
}
...
