strange koji build failures...?

[Matthias Clasen]

On Wed, 2008-06-04 at 19:25 +0200, Jim Meyering wrote:
> Ondřej Vašík <ovasik at redhat.com> wrote:
> > Hi Jim,
> >
> > Jim Meyering wrote:
> >> Eric Blake fixed that with this change to gnulib's utimens.c:
> >>
> >>   http://git.savannah.gnu.org/gitweb/?p=gnulib.git;a=commitdiff;h=93f08406537
> >
> > I know about the fix (was causing troubles with test-suite) - therefore
> > that patch is already part of the coreutils-6.12-1.fc10 build - so it
> > will not solve the troubles with koji. Problem seems to be in return
> > code 280 from futimensat() call. So futimensat is available, but failing
> > - the same problem as I had with the coreutils test suite few weeks ago
> > (if you remember touch/create-no-missing issue). I would say koji xen
> > kernel issue, but could be workarounded by Fedora coreutils.
> 
> Hi Ondřej,
> 
> I think you're right that it's a kernel problem.
> I've proposed a patch attached to http://bugzilla.redhat.com/449910
> as https://bugzilla.redhat.com/attachment.cgi?id=308374
> that might suffice to work around the problem.
> 
> Can you test it?
> 

Hmm, I see that coreutils 6.12 is back in rawhide, and the build
failures are back, too:
http://koji.fedoraproject.org/koji/buildinfo?buildID=51679