Hi Jim, sorry for not writing the patch myself as I promised, had some troubles with git clone before leaving office.
Jim Meyering wrote: > Based on your description, I wrote the following. > Does it solve the problem? Yep, the patch is almost the same as I have commited and built in Fedora Rawhide and additionally attached to RHBZ #449910 as https://bugzilla.redhat.com/attachment.cgi?id=308560 . The only difference is additional test case (which checks correct preserving of time in cp and touch) and a few differences in comments (will change that 0/success to -1/failure in next Rawhide build). I think as the problem is likely only very rare (and probably only Fedora building system specific), test case is not necessary and it could be kept as it is until it will get fixed in koji xen kernel. > > From 51aca03ab224ad1bf0a5c432403a0131001cda36 Mon Sep 17 00:00:00 2001 > From: Jim Meyering <[EMAIL PROTECTED]> > Date: Fri, 6 Jun 2008 18:20:14 +0200 > Subject: [PATCH] utimens.c: correct yesterday's kernel bug work-around > MIME-Version: 1.0 > Content-Type: text/plain; charset=utf-8 > Content-Transfer-Encoding: 8bit > > Ondřej Vašík found that the invalid return value of 280 indicates > failure, not success, and the kernel bug we're trying to work > around affects not just the utimensat call, but also the fallback > futimens call. > * lib/utimens.c (gl_futimens) [HAVE_UTIMENSAT]: Simulate failure, > not success. > [HAVE_FUTIMENS]: Use the same work-around, here. Ondřej Vašík
signature.asc
Description: Toto je digitálně podepsaná část zprávy
