[committed] fortran/dump-parse-tree.c: Use proper enum type (was: Re: [Patch ]Fortran/OpenMP: Extend defaultmap clause for OpenMP 5 [PR92568])

2021-06-23 Thread Tobias Burnus
Indeed, I somehow managed to use the wrong of the two enums in the cast ... Fixed in r12-1760-gcac3527793b38164e2a83c7ccbfe0cfcf5ac95b8 On 23.06.21 16:13, Martin Liška wrote: I noticed the patch causes the following clang warnings: gcc/fortran/dump-parse-tree.c:1786:11: warning: comparison of

Re: [Patch, fortran V3] PR fortran/100683 - Array initialization refuses valid (list of pending patches)

2021-06-23 Thread dhumieres.dominique--- via Fortran
Hi José, > Thus: Do you have a list of patches pending review? https://gcc.gnu.org/pipermail/fortran/2021-April/055924.html PRs 100029 and 100040. I have the patch in my working tree for a long time. It works as expected. OK to commit. https://gcc.gnu.org/pipermail/fortran/2021-April/0559

Re: [Patch ]Fortran/OpenMP: Extend defaultmap clause for OpenMP 5 [PR92568]

2021-06-23 Thread Martin Liška
Hello. I noticed the patch causes the following clang warnings: gcc/fortran/dump-parse-tree.c:1786:11: warning: comparison of different enumeration types in switch statement ('enum gfc_omp_defaultmap' and 'gfc_omp_defaultmap_category') [-Wenum-compare-switch] gcc/fortran/dump-parse-tree.c:1787

Re: [Ping^2, Patch, Fortran] PR100337 Should be able to pass non-present optional arguments to CO_BROADCAST

2021-06-23 Thread Tobias Burnus
On 23.06.21 10:23, Andre Vehreschild wrote: Will wait two weeks for any errors introduced by this patch before backporting to gcc-11, ok? Fine with me. Thanks again for the patch. Tobias - Mentor Graphics (Deutschland) GmbH, Arnulfstrasse 201, 80634 München Registergericht

Re: [Ping^2, Patch, Fortran] PR100337 Should be able to pass non-present optional arguments to CO_BROADCAST

2021-06-23 Thread Andre Vehreschild via Fortran
Hi all, the fix for PR100337 was submitted as da13e4ebebb07a47d5fb50eab8893f8fe38683df. Thanks for the review Tobias. @Tobias: You are right, caf_single does not get much testing. But this part (not providing a stat) is tested multiple times, because of the laziness. Nearly none of the tests in