Message ID | BLU0-SMTP416114389C569D135154F297690@phx.gbl |
---|---|
State | New |
Headers | show |
On Apr 6, 2014, at 9:23 AM, John David Anglin <dave.anglin@bell.net> wrote:
> This test is an expected fail on hppa*-*-hpux* because there is no lockless atomic support.
One can add “no lockless atomic” in there somewhere… The reason is that a secondary port that someone is doing can see a failure, look at the test case, and stop reading when they see your comment and think they are in the same boat. I’ve done that in many cases with some of the hp comments… so handy.
On 6-Apr-14, at 12:51 PM, Mike Stump wrote: > On Apr 6, 2014, at 9:23 AM, John David Anglin <dave.anglin@bell.net> > wrote: >> This test is an expected fail on hppa*-*-hpux* because there is no >> lockless atomic support. > > One can add “no lockless atomic” in there somewhere… The reason is > that a secondary port that someone is doing can see a failure, look > at the test case, and stop reading when they see your comment and > think they are in the same boat. I’ve done that in many cases with > some of the hp comments… so handy. I'll take a look. Dave -- John David Anglin dave.anglin@bell.net
Index: gcc.dg/atomic/stdatomic-flag.c =================================================================== --- gcc.dg/atomic/stdatomic-flag.c (revision 209163) +++ gcc.dg/atomic/stdatomic-flag.c (working copy) @@ -1,5 +1,5 @@ /* Test atomic_flag routines for existence and execution. */ -/* { dg-do run } */ +/* { dg-do run { xfail hppa*-*-hpux* } } */ /* { dg-options "-std=c11 -pedantic-errors" } */ #include <stdatomic.h>