On Thu, Mar 01, 2012 at 04:52:41PM +0000, Colin King wrote:
> From: Colin Ian King <firstname.lastname@example.org>
> SRU Justification:
> When failing to read the lower file's crypto metadata during a lookup,
> eCryptfs must continue on without throwing an error. For example, there
> may be a plaintext file in the lower mount point that the user wants to
> delete through the eCryptfs mount.
> If an error is encountered while reading the metadata in lookup(), the
> eCryptfs inode's size could be incorrect. We must be sure to reread the
> plaintext inode size from the metadata when performing an open() or
> setattr(). The metadata is already being read in those paths, so this
> adds minimal performance overhead.
> BugLink: http://bugs.launchpad.net/bugs/509180
> This patch is a backport by Tim Gardner that was SRU'd back in 20 Jul 2011
> but wasn't tested and so was reverted as part of the SRU process. I'm
> sending back to the list again as I think it should be included.
> See https://bugs.launchpad.net/ecryptfs/+bug/509180/comments/50,
> without this patch, the file as garbage at the end of the, file. With
> the fix, the file contains just the expected data with no garbage at
> the end.
> I've also got a suitable test working for the ecryptfs test suite
> which I hope will be included fairly shortly.
> Tim Gardner (1):
> eCryptfs: Handle failed metadata read in lookup
Ack, I guess the testcase got lost on the bug noise originally, other
issues were mixed up, and people are still complaining of problems in
Oneiric/Precise which has the fix already, which should be in new bug