I compiled 3.0.1 with the 3.0.2-rc1 patch and your patch:
diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c
index ccc1afa..e0ea721 100644
--- a/fs/cifs/connect.c
+++ b/fs/cifs/connect.c
@@ -2838,7 +2838,8 @@ cleanup_volume_info_contents(struct smb_vol *volume_info)
kfree(volume_info->username);
kzfree(volume_info->password);
kfree(volume_info->UNC);
- kfree(volume_info->UNCip);
+ if (volume_info->UNCip != volume_info->UNC + 2)
+ kfree(volume_info->UNCip);
kfree(volume_info->domainname);
kfree(volume_info->iocharset);
kfree(volume_info->prepath);
No oops! I'm using the same version of AUFS that I used when 3.0.1 came out, so
definitely not a AUFS problem! Thanks again.
--- On Mon, 8/15/11, kirk w <kirkpu...@yahoo.com> wrote:
> From: kirk w <kirkpu...@yahoo.com>
> Subject: Re: Kernel oops with cifs
> To: sf...@users.sourceforge.net
> Cc: aufs-users@lists.sourceforge.net
> Date: Monday, August 15, 2011, 2:01 PM
> Hi Junjiro,
>
> I'm don't think it's a AUFS problem. See:
>
> https://lkml.org/lkml/2011/8/15/121
>
> And during 3.0rcX there were cifs problems that would cause
> a oops. It's very difficult to get any information when it
> crashes, only once could I save some logs. I'll wait for the
> 3.0.2 kernel and test with that. Sorry for the noise.
>
> Thanks for your work. Hopefully Al and the other kernel
> folks will merge AUFS.
>
> Regards,
>
> Kirk
>
> --- On Mon, 8/15/11, sf...@users.sourceforge.net
> <sf...@users.sourceforge.net>
> wrote:
>
> > From: sf...@users.sourceforge.net
> <sf...@users.sourceforge.net>
> > Subject: Re: Kernel oops with cifs
> > To: "kirk w" <kirkpu...@yahoo.com>,
> aufs-users@lists.sourceforge.net
> > Date: Monday, August 15, 2011, 10:04 AM
> >
> > sf...@users.sourceforge.net:
> > > kirk w:
> > > > I think I ran into a bug with the 3.0 kernel
> and
> > AUFS. All seems fine until you try to do a cifs mount,
> then
> > the kernel crashes:
> > :::
> > > Thank you for report.
> > > I'll try by myself and also read cifs code when I
> have
> > time.
> >
> > Hi,
> >
> > I have just tried it, and both of ls -l and getfattr
> > worked.
> > Will you give me more infromation? such as
> > - /proc/mounts (instead of the output of mount(8))
> > - /sys/module/aufs/*
> > - /sys/fs/aufs/* (if you have them)
> > - /debug/aufs/* (if you have them)
> >
> > And I don't know about ROX-Filer at all. Additionally
> your
> > stack trace
> > is unreliable, and it doesn't tell me the trigger is
> > listxattr(2) or
> > lstat(2). It may be lstat(2), I guess. But which file
> did
> > ROX-Filer
> > issues lstat(2) for? Was it a file on cifs (via
> aufs)?
> >
> > By the way, I found a bug in cifs in linux-3.0 which
> may
> > cause memory
> > corruption. But I am not sure it is related to your
> > problem.
> > I will post a patch if it is not fixed yet.
> >
> > Could you separate or identify the problem comes from
> aufs
> > or cifs?
> >
> >
> > J. R. Okajima
> >
>
------------------------------------------------------------------------------
uberSVN's rich system and user administration capabilities and model
configuration take the hassle out of deploying and managing Subversion and
the tools developers use with it. Learn more about uberSVN and get a free
download at: http://p.sf.net/sfu/wandisco-dev2dev
