https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100916
Thomas Koenig <tkoenig at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Keywords| |needs-stdcheck
Blocks| |32630
CC| |tkoenig at gcc dot gnu.org
--- Comment #7 from Thomas Koenig <tkoenig at gcc dot gnu.org> ---
(In reply to Tobias Burnus from comment #6)
> (In reply to sandra from comment #5)
> > Jose's test case for this issue is still failing.
>
> I find the testcase very confusing. For
>
> type(c_other), intent(in) :: a(:)
> call check_tk_as(a, t, k, e, 1_c_size_t)
>
> it fails with:
>
> FAIL! type (A): 6 != 255
>
> which is printed by:
> printf ("FAIL! type (A): %d != %d\n", auxp->type, type);
>
> gfortran uses the following, which seems to be fine:
> #define CFI_type_struct 6
> I do not quite understand why the testcase expects 255 as type now how it
> tries to generate that number.
>
>
> Other fails:
> FAIL! type (B): 6 != -1
> FAIL! type (C): -1 != 6
> FAIL! type: -1 != 6
>
> Here the issue is that GCC uses CFI_type_struct (= 6)
> while the testcase expects CFI_type_other (= -1).
> I think one can argue for either.
Not clearly defined, then? Adding the relevant keyword, now we have it :-)
Referenced Bugs:
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=32630
[Bug 32630] [meta-bug] ISO C binding
