Re: EXT3 filesystem corruption
On Wed, May 21, 2008 at 02:32:52AM -0600, dann frazier wrote:
> On Wed, May 21, 2008 at 12:29:33PM +0800, Steven Wilton wrote:
> > Hi,
> > I added a response to the following bug report about 6 months ago, and
> > was wondering if it can be resolved:
> > http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=425534
> > The bug is a documented race condition in ext3, with a patch that has
> > been applied upstream, which I thought might be considered important
> > enough to fix. I would imagine that it woud not cause problems for a
> > lot of users, who would either not hit the race condition, or it would
> > be corrected (and possibly ignored) when the machine runs a fsck on
> > reboot.
> Corruption fixes are definitely reasonable for stable updates, I'll
> take a look at porting this. Thanks for tracking it down.
I read through the bug report and the lkml threads, but I don't see a
connection between #425534 (which causes filesystem corruption) and
the issue in http://lkml.org/lkml/2006/12/29/26 (which appears to
cause internal file corruption)- can you elaborate on why you believe
they are connected?