This is the mail archive of the libc-alpha@sourceware.org mailing list for the glibc project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [PATCH 5/6] nptl: Convert some rwlock tests to use libsupport


On Monday 08 April 2019 at 09:47:32 +0200, Florian Weimer wrote:
> * Mike Crowe:
> > However, I'm less sure about this code from tst-rwlock7.c(do_test):
> >
> >   size_t cnt;
> >   for (cnt = 0; cnt < sizeof (kind) / sizeof (kind[0]); ++cnt)
> >     {
> >       pthread_rwlock_t r;
> >       pthread_rwlockattr_t a;
> >       if (pthread_rwlockattr_init (&a) != 0)
> >         FAIL_EXIT1 ("round %Zu: rwlockattr_t failed\n", cnt);
> >
> >       if (pthread_rwlockattr_setkind_np (&a, kind[cnt]) != 0)
> >         FAIL_EXIT1 ("round %Zu: rwlockattr_setkind failed\n", cnt);
> >     ...
> >
> > If the round numbers are important then I need to keep the messages. If
> > not, then I can also replace these with TEST_COMPARE. Do you have an
> > opinion on that?
> 
> I don't think I've ever seen tst-rwlock7 fail, so I don't know if the
> round numbers are useful for diagnosing test failures.

The patch series I posted yesterday[1] leaves these messages showing the
round number (as above.) I'm happy to change them to use TEST_COMPARE
instead if required.

Thanks.

Mike.

[1] https://sourceware.org/ml/libc-alpha/2019-04/msg00161.html


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]