xfs
[Top] [All Lists]

Re: xfs tries to write on readonly device on unmount

To: Christoph Hellwig <hch@xxxxxxxxxxxxx>
Subject: Re: xfs tries to write on readonly device on unmount
From: Michael Prokop <mika@xxxxxxxx>
Date: Wed, 22 Jul 2009 13:27:25 +0200
Cc: xfs@xxxxxxxxxxx
In-reply-to: <20090721111909.GA29399@xxxxxxxxxxxxx>
References: <2009-07-21T12-28-50@xxxxxxxxxxxxxxxxxxxxxxxxx> <20090721111909.GA29399@xxxxxxxxxxxxx>
User-agent: Mutt/1.5.13 (2006-08-11)
* Christoph Hellwig <hch@xxxxxxxxxxxxx> [20090721 13:19]:
> On Tue, Jul 21, 2009 at 12:51:20PM +0200, Michael Prokop wrote:

> > This issue might be related to
> > http://oss.sgi.com/bugzilla/show_bug.cgi?id=413

> > A XFS partition is set to read-only mode (using 'blockdev --setro')
> > and mounted. As soon as I *un*mount the filesystem there seem to
> > happen write requests.

> > I've a Tableau Forensic Bridge? which avoids the write requests and
> > attached it to a Linux live-system running kernel 2.6.28 inside
> > VMware and Virtualbox. I can easily reproduce the problem.

> Block 0 is the superblock, so it appears like there is a superblock
> write somewhere not guarded by an read-only check.  Can you still
> reproduce this with a current kernel?  We now do have a testcase in
> xfsqa (test 200) that checks exactly that scenario, but it doesn't
> show issues on recent kernels.

I reran the test procedure with kernel 2.6.31-rc3 and could not
reproduce the issue.

Thanks.

regards,
-mika-
-- 
http://michael-prokop.at/  || http://adminzen.org/
http://grml-solutions.com/ || http://grml.org/

Attachment: signature.asc
Description: Digital signature

<Prev in Thread] Current Thread [Next in Thread>