On Sat, Oct 27, 2012 at 12:16 PM, Volker Armin Hemmann <volkerar...@googlemail.com> wrote: > Am Freitag, 26. Oktober 2012, 01:34:35 schrieb Neil Bothwick: >> On Thu, 25 Oct 2012 15:10:17 -0500, Dale wrote: >> > Or just don't update the kernel until it is fixed. After all, if the >> > kernel you are using works and is not a serious security problem, just >> > use it for a while. They will fix it pretty soon I'm sure. >> >> Several versions are affected, it's not a new issue, just a newly >> discovered one. That in itself indicates the likelihood of being affected >> by it, considering how many boxes use ext4 with these kernel versions. > > since only two persons ever hit it, it is a pretty save bet that this bug is > so uncommon that you should not waste any sleep over it. > > -- > #163933 >
For fear of leading anyone down a path that eventually causes them problems I won't attempt to communicate the root cause or fixes but there is new information, and a potential patch verified by the only person to see this problem, about what causes the corruption. Read the thread if this makes any difference in your life. Don't worry about it if it doesn't. Basically, it's apparently driven by mount options and something the OP (of the lkml thread) did which most users are unlikely IMO to ever try. - Mark
