| To: | Barry Naujok <bnaujok@xxxxxxx> |
|---|---|
| Subject: | Re: REVIEW: Zero rest of superblock sector always |
| From: | Dave Chinner <david@xxxxxxxxxxxxx> |
| Date: | Thu, 4 Sep 2008 09:12:20 +1000 |
| Cc: | "xfs@xxxxxxxxxxx" <xfs@xxxxxxxxxxx> |
| In-reply-to: | <op.ugv7zt013jf8g2@xxxxxxxxxxxxxxxxxxxxxxxxxxxx> |
| Mail-followup-to: | Barry Naujok <bnaujok@xxxxxxx>, "xfs@xxxxxxxxxxx" <xfs@xxxxxxxxxxx> |
| References: | <op.ugv7zt013jf8g2@xxxxxxxxxxxxxxxxxxxxxxxxxxxx> |
| User-agent: | Mutt/1.5.18 (2008-05-17) |
On Wed, Sep 03, 2008 at 04:51:19PM +1000, Barry Naujok wrote: > I found that zeroing the "garbage" beyond the end of the superblock in > the first sector of each AG rather inconsistant. It depended on some > obscure combination of version bits to be set. Just for filling out hte history, that isfixing a bug in an old, old mkfs.xfs on Irix where a certain version failed to zero the superblock(s) before initialising them. Every XFS filesystem made in the last 10 years should have properly initialised superblocks. But still, doing it unconditionally is fine as it will fix stuff like fsfuzzer damage.... Cheers, Dave. -- Dave Chinner david@xxxxxxxxxxxxx |
| <Prev in Thread] | Current Thread | [Next in Thread> |
|---|---|---|
| ||
| Previous by Date: | Re: xfs_growfs fix backport for 2.6.16.y, Ed Cashin |
|---|---|
| Next by Date: | Re: REVIEW: Improve caching in libxfs, Christoph Hellwig |
| Previous by Thread: | Re: REVIEW: Zero rest of superblock sector always, Christoph Hellwig |
| Next by Thread: | REVIEW: Change mkfs.xfs to set primary superblock inodes in ALL secondaries, Barry Naujok |
| Indexes: | [Date] [Thread] [Top] [All Lists] |