Bug#406902: kernel NFS data loss

2009-09-10 Thread psz
Dear Moritz, > Can you give us a status update please? Is this fixed upstream > in the mean time? If not, did you submit your patches and was > there any feedback? I believe this issue has been "fixed upstream", as per comment http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=406902#29 I guess the

Bug#406902: kernel NFS data loss

2009-09-08 Thread Moritz Muehlenhoff
On Wed, Sep 19, 2007 at 08:17:12AM +1000, Paul Szabo wrote: > On 24 Jul 07 I wrote: > > > The patch below (against 2.6.8-16sarge7) seems to solve the problem. > > ... [patch for fs/exportfs/expfs.c] ... > > Seems to me that this has been incorporated (or done) in > linux-source-2.6.18.dfsg.1-13et

Bug#406902: kernel NFS data loss

2007-09-18 Thread Paul Szabo
On 24 Jul 07 I wrote: > The patch below (against 2.6.8-16sarge7) seems to solve the problem. > ... [patch for fs/exportfs/expfs.c] ... Seems to me that this has been incorporated (or done) in linux-source-2.6.18.dfsg.1-13etch2 . Earlier, on 20 Jul 07 I wrote: > ... I noticed what I thought were

Bug#406902: [PATCH] Re: Bug#406902 kernel NFS data loss

2007-07-23 Thread Paul Szabo
The patch below (against 2.6.8-16sarge7) seems to solve the problem. I now run my machines with both this patch, and also the one I submitted on 20 Jul. Please include in future versions of the kernel. Thanks, Paul Szabo [EMAIL PROTECTED] http://www.maths.usyd.edu.au/u/psz/ School of Mathemat

Bug#406902: kernel NFS data loss

2007-07-21 Thread Paul Szabo
Have now tested the "patch" in my previous message: it does not solve the problem I reported. Cheers, Paul Szabo [EMAIL PROTECTED] http://www.maths.usyd.edu.au/u/psz/ School of Mathematics and Statistics University of SydneyAustralia -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with

Bug#406902: kernel NFS data loss

2007-07-19 Thread Paul Szabo
I ran some tests today, and it seemed (but not conclusive) that the problem only occurs when the client is a multi-CPU SMP machine. Looking at kernel source code, I noticed what I thought were oddities. Do you think the following "patch" against 2.6.8-16sarge7 code would be useful? I have not yet