xfs
[Top] [All Lists]

Re: TAKE 893451 - Fix for the nfs corruption bug

To: Russell Cattelan <cattelan@xxxxxxx>
Subject: Re: TAKE 893451 - Fix for the nfs corruption bug
From: Feizhou <feizhou@xxxxxxxxxxxxx>
Date: Wed, 16 Jun 2004 15:32:19 +0800
Cc: linux-xfs@xxxxxxxxxxx, Kent Ho <kent@xxxxxxxxxxxx>, Timothy Webster <timothyw@xxxxxxxxxxxx>
In-reply-to: <20040615180549.B1259812C64@naboo.americas.sgi.com>
References: <20040615180549.B1259812C64@naboo.americas.sgi.com>
Sender: linux-xfs-bounce@xxxxxxxxxxx
User-agent: Mozilla Thunderbird 0.6 (X11/20040502)
Hi,

Is this in 2.6.7 and if not, is there a way to migitate it?

What does it take to hit it?

Would recommend a NFS server running XFS filesystems with 2.6.7?

Thanks,

Christopher

Russell Cattelan wrote:

Partial fix for long standing nfs+xfs corruption.

See:
http://oss.sgi.com/bugzilla/show_bug.cgi?id=198
for full details

Date:  Tue Jun 15 11:04:55 PDT 2004
Workarea:  naboo.americas.sgi.com:/go/space/XFS/xfs-linux-p
Inspected by:  felixb@xxxxxxx

The following file(s) were checked into:
 bonnie.engr.sgi.com:/isms/xfs-kern/xfs-linux


Modid: xfs-linux:xfs-kern:173598a
linux-2.6/xfs_aops.c - 1.75
linux-2.4/xfs_aops.c - 1.82
- Masanori TSUDA fix for xfs+xfs corruption
Basically do not flush pages beyond end of file (i_size)
There still needs to be some investigation done to see
if this is just fixing a symptom of i_size != di_size / valid page handling.
But for the moment this fixes the problem








<Prev in Thread] Current Thread [Next in Thread>