Hi!

On Fri, 1 Apr 2016 10:55:49 +0200, I wrote:
> On Thu, 24 Mar 2016 22:31:29 +0100, I wrote:
> > On Wed, 23 Mar 2016 19:57:50 +0100, Bernd Schmidt <bschmidt@redhat.com> wrote:
> > > Ok with [...].
> > 
> > Thanks for the review; committed in r234471:
> 
> >     Also test -O0 for OpenACC C, C++ offloading test cases
> 
> Merged into gomp-4_0-branch in r234664:

> --- libgomp/testsuite/libgomp.oacc-c-c++-common/host_data-1.c
> +++ libgomp/testsuite/libgomp.oacc-c-c++-common/host_data-1.c
> @@ -1,5 +1,6 @@
>  /* { dg-do run { target openacc_nvidia_accel_selected } } */
>  /* { dg-additional-options "-lcuda -lcublas -lcudart" } */
> +/* { dg-xfail-run-if "TODO" { *-*-* } { "-O0" } { "" } } */
>  
>  #include <stdlib.h>
>  #include <openacc.h>

Filed <https://gcc.gnu.org/PR70579>.

> --- libgomp/testsuite/libgomp.oacc-c-c++-common/if-1.c
> +++ libgomp/testsuite/libgomp.oacc-c-c++-common/if-1.c
> @@ -1,3 +1,5 @@
> +/* { dg-xfail-run-if "TODO" { *-*-* } { "-O0" } { "" } } */
> +
>  #include <openacc.h>
>  #include <stdlib.h>
>  #include <stdbool.h>

Filed <https://gcc.gnu.org/PR70580>.

The both PASS on trunk.

Currently unclear if it's the same underlying problem or not.


Grüße
 Thomas