From mboxrd@z Thu Jan 1 00:00:00 1970 From: Dan Carpenter Date: Mon, 14 Jun 2010 21:11:20 +0000 Subject: Re: [patch 11/11] btrfs: The file argument for fsync() is never Message-Id: <20100614211120.GB5483@bicker> List-Id: References: <20100529094907.GL5483@bicker> <201006142207.25930.johannes.hirte@fem.tu-ilmenau.de> <20100614205849.GA26828@lst.de> In-Reply-To: <20100614205849.GA26828@lst.de> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: Christoph Hellwig Cc: Johannes Hirte , linux-btrfs@vger.kernel.org, Yan Zheng , Josef Bacik , Chris Mason , kernel-janitors@vger.kernel.org On Mon, Jun 14, 2010 at 10:58:49PM +0200, Christoph Hellwig wrote: > On Mon, Jun 14, 2010 at 10:07:23PM +0200, Johannes Hirte wrote: > > I think you're wrong here. I've run into a kernel null pointer dereference at > > this point with a NFS exported btrfs filesystem: > > Looks like you've applied the patch to a far too old kernel. It can't > be NULL for quite a while already. > You're the expert, but it looks like it could be null in 2.6.34 like he says. I'm just looking at vfs_fsync_range() in "git show v2.6.34:fs/sync.c". > If you're testing patches posted to the mailinglists always make sure > you have a recent enough kernel. This actually was merged into the btrfs and the kernel.org trees already. I'm not sure how the btrfs backports work... regards, dan carpenter