On Jul 18, 2020, at 8:19 PM, Hans-Peter Nilsson <h...@bitrange.com> wrote:
>
> Long-standing FAIL remedied; committed. Maybe better to list
> the targets that *do* support arbitrary frame access?
Yes, it would be better if test cases that fall way to low in portability are
listed instead against what platforms the test is expected to work on, instead
of having everyone else join a long list in the exceptions clause.
Another way to handle this is to expose the ability to do such a thing in the
first place and expose that via a .h or a preprocessing condition, then the
test itself will only test on machines that claim it should work in the first
place.
In this case, and with this feature specifically, it isn't portable and
reliable enough.
> gcc/testsuite:
> * gcc.dg/Wno-frame-address.c: Skip for cris and mmix.
>
> --- gcc/gcc/testsuite/gcc.dg/Wno-frame-address.c.orig Mon Jan 13 22:30:47 2020
> +++ gcc/gcc/testsuite/gcc.dg/Wno-frame-address.c Sun Jul 19 05:05:49 2020
> @@ -1,5 +1,5 @@
> /* { dg-do compile } */
> -/* { dg-skip-if "Cannot access arbitrary stack frames" { arm*-*-* amdgpu-*-*
> avr-*-* hppa*-*-* ia64-*-* visium-*-* csky-*-* msp430-*-* } } */
> +/* { dg-skip-if "Cannot access arbitrary stack frames" { arm*-*-* amdgpu-*-*
> avr-*-* hppa*-*-* ia64-*-* visium-*-* csky-*-* msp430-*-* cris-*-* mmix-*-* }
> } */
> /* { dg-options "-Werror" } */
> /* { dg-additional-options "-mbackchain" { target { s390*-*-* } } } */
