| To: | Justin Piszcz <jpiszcz@xxxxxxxxxxxxxxx> |
|---|---|
| Subject: | Re: [GIT PULL] XFS update for 2.6.23 - revert a commit |
| From: | Timothy Shimmin <tes@xxxxxxx> |
| Date: | Wed, 03 Oct 2007 11:49:11 +1000 |
| Cc: | Lachlan McIlroy <lachlan@xxxxxxx>, Eric Sandeen <sandeen@xxxxxxxxxxx>, xfs@xxxxxxxxxxx |
| In-reply-to: | <Pine.LNX.4.64.0710020442540.24427@xxxxxxxxxxxxxxxx> |
| References: | <20071001072350.DF61C58C4C0A@xxxxxxxxxxxxxxxxxxxxxxx> <4700EE2A.1020304@xxxxxxxxxxx> <4701A1D0.5010709@xxxxxxx> <4701ED51.8050706@xxxxxxx> <Pine.LNX.4.64.0710020442540.24427@xxxxxxxxxxxxxxxx> |
| Sender: | xfs-bounce@xxxxxxxxxxx |
| User-agent: | Thunderbird 2.0.0.6 (Macintosh/20070728) |
Hi Justin, Justin Piszcz wrote: On Tue, 2 Oct 2007, Lachlan McIlroy wrote:Yeah that about sums it up. In an attempt to prevent log replay of inodes in cases when we shouldn't replay we also prevented log replay of inodes in cases when we should replay. We end up with directories that refer to inodesthat were not replayed and we read existing data off disk. That existingdata is usually previous instances of inodes. We had cases of regular filesturning into directories and inode version mismatches. LachlanIn 2.6.23-rc8? The bad change went into rc7 and was still there in rc8. It was backed out for rc9. --Tim |
| Previous by Date: | Re: REVIEW: xfs_reno, Timothy Shimmin |
|---|---|
| Next by Date: | Re: REVIEW: xfs_reno, David Chinner |
| Previous by Thread: | Re: [GIT PULL] XFS update for 2.6.23 - revert a commit, Justin Piszcz |
| Next by Thread: | Re: [GIT PULL] XFS update for 2.6.23 - revert a commit, Justin Piszcz |
| Indexes: | [Date] [Thread] [Top] [All Lists] |