[PATCH] fix PR target/54407

Jack Howarth howarth@bromo.med.uc.edu
Tue Oct 16 20:24:00 GMT 2012


On Tue, Oct 16, 2012 at 12:26:39PM -0700, Mike Stump wrote:
> On Oct 16, 2012, at 10:28 AM, Dominique Dhumieres <dominiq@lps.ens.fr> wrote:
> >> I'm not a big fan of just hiding the bug; why not just xfail it of darwin10 and darwin[5-9]?
> > 
> > I am viewing xfail for bugs that have a chance to be fixed some day.
> > Since the bug is fixed for darwin11 and 12, I doubt that Apple will backport it.
> > I would prefer a dg-skip-if may be with a comment about the bug.
> 
> Not a huge fan of skip either, but, if you guys really want to, that's ok by me.  Please do include a comment and a pointer to the Apple bug report number if available else a pointer to the gcc PR number where we talk about why we are skipping it.  Something like: OS bug, see gcc PR/x.  Short, easy to understand and one can read the PR for details, if one is really, really bored.

Mike,
   I can't absolutely swear that the pthread issue fixed in radr://7905357. "testcase 20_util/shared_ptr/thread/default_weaktoshared.cc thread handling bug",
is the same bug. Certainly Apple gave pthreads in libSystem some TLC in Lion which will never be backported to earlier OS releases.
     Jack



More information about the Libstdc++ mailing list