Hi Jakub,
thanks a lot for fixing this. Since it's a regression and the fix is
so trivial, I hope you're willing to backport it to the 6 and 7
branches as well? (If not, I can also take care of that.)
Cheers,
Janus
2017-11-24 22:42 GMT+01:00 Jakub Jelinek <ja...@redhat.com>:
> Hi!
>
> gfc_trans_omp_array_reduction_or_udr creates artificial intrinsics symbol
> just for the purpose of expansion of array min/max reductions.
> That is something we certainly don't want a -Wsurprising warning on.
>
> Fixed thusly, bootstrapped/regtested on x86_64-linux and i686-linux,
> committed to trunk.
>
> 2017-11-24 Jakub Jelinek <ja...@redhat.com>
>
> PR fortran/81304
> * trans-openmp.c (gfc_trans_omp_array_reduction_or_udr): Set
> attr.implicit_type in intrinsic_sym to avoid undesirable warning.
>
> * testsuite/libgomp.fortran/pr81304.f90: New test.
>
> --- gcc/fortran/trans-openmp.c.jj 2017-11-08 16:19:36.000000000 +0100
> +++ gcc/fortran/trans-openmp.c 2017-11-24 13:41:41.995610897 +0100
> @@ -1623,6 +1623,7 @@ gfc_trans_omp_array_reduction_or_udr (tr
> intrinsic_sym.attr.referenced = 1;
> intrinsic_sym.attr.intrinsic = 1;
> intrinsic_sym.attr.function = 1;
> + intrinsic_sym.attr.implicit_type = 1;
> intrinsic_sym.result = &intrinsic_sym;
> intrinsic_sym.declared_at = where;
>
> --- libgomp/testsuite/libgomp.fortran/pr81304.f90.jj 2017-11-24
> 13:50:39.668043753 +0100
> +++ libgomp/testsuite/libgomp.fortran/pr81304.f90 2017-11-24
> 13:50:11.000000000 +0100
> @@ -0,0 +1,17 @@
> +! PR fortran/81304
> +! { dg-do run }
> +! { dg-options "-Wsurprising" }
> +
> +program pr81304
> + integer :: i
> + real, dimension(1:3) :: a, b, c
> + a = 128
> + b = 0
> +!$omp parallel do reduction(min: a) reduction(max: b) private (c) ! {
> dg-bogus "Type specified for intrinsic function" }
> + do i = 1, 16
> + c = (/ i, i - 5, i + 5 /)
> + a = min (a, c)
> + b = max (b, c)
> + end do
> + if (any (a /= (/ 1, -4, 6 /)) .or. any (b /= (/ 16, 11, 21 /))) call abort
> +end
>
> Jakub
