Re: [PATCH 4.4 13/16] ocfs2: should wait dio before inode lock in ocfs2_setattr()

[Ben Hutchings <ben.hutchings@codethink.co.uk>]

On Fri, 2017-12-08 at 12:03 +0800, alex chen wrote:
> 
> On 2017/12/8 10:26, Ben Hutchings wrote:
> > On Fri, 2017-12-08 at 08:39 +0800, alex chen wrote:
> > > 
> > > On 2017/12/8 2:25, Ben Hutchings wrote:
> > > > On Wed, 2017-12-06 at 09:02 +0800, alex chen wrote:
> > > > > Hi Ben,
> > > > > 
> > > > > Thanks for your reply.
> > > > > 
> > > > > On 2017/12/5 23:49, Ben Hutchings wrote:
> > > > > > On Wed, 2017-11-22 at 11:12 +0100, Greg Kroah-Hartman wrote:
> > > > > > > 4.4-stable review patch.  If anyone has any objections,
> > > > > > > please let me know.
> > > > > > > 
> > > > > > > ------------------
> > > > > > > 
> > > > > > > From: alex chen <alex.chen@huawei.com>
> > > > > > > 
> > > > > > > commit 28f5a8a7c033cbf3e32277f4cc9c6afd74f05300 upstream.
> > > > > > > 
> > > > > > > we should wait dio requests to finish before inode lock in
> > > > > > > ocfs2_setattr(), otherwise the following deadlock will
> > > > > > > happen:
> > > > > > 
> > > > > > [...]
> > > > > > 
> > > > > > I looked at the kernel-doc for inode_dio_wait():
> > > > > > 
> > > > > > /**
> > > > > >  * inode_dio_wait - wait for outstanding DIO requests to finish
> > > > > >  * @inode: inode to wait for
> > > > > >  *
> > > > > >  * Waits for all pending direct I/O requests to finish so that we can
> > > > > >  * proceed with a truncate or equivalent operation.
> > > > > >  *
> > > > > >  * Must be called under a lock that serializes taking new references
> > > > > >  * to i_dio_count, usually by inode->i_mutex.
> > > > > >  */
> > > > > > 
> > > > > > Now that ocfs2_setattr() calls this outside of the inode locked region,
> > > > > > what prevents another task adding a new dio request immediately
> > > > > > afterward?
> > > > > > 
> > > > > 
> > > > > In the kernel 4.6, firstly, we use the inode_lock() in do_truncate() to
> > > > > prevent another bio to be issued from this node.
> > > > 
> > > > [...]
> > > > 
> > > > Yes but there seems to be a race condition - after the call to
> > > > inode_dio_wait() and before the call to inode_lock(), another dio
> > > > request can be added.
> > 
> > Sorry, I've been mixing up inode_lock() and ocfs2_inode_lock(). 
> > However:
> > 
> > > In the truncating file situation, the lock order is as follow:
> > > do_truncate()
> > >  inode_lock()
> > >  notify_change()
> > >   ocfs2_setattr()
> > >    inode_dio_wait()
> > >     --here it is under the protect of inode_lock(), so another dio requests
> > >       from another process will not be added.
> > 
> > only DIO reads seem to take the inode lock.
> > 
> 
> I do not clearly understand what you mean.
> The inode_lock() will be called in ocfs2_file_write_iter().

Oh I see.  I didn't realise that was part of the call chain.

> You mean only DIO writes seem to take the inode_lock()?

I did mean reads, as do_blockdev_direct_IO() may call inode_lock() for
reads - but ocfs2 doesn't set the flag for that.  Maybe that's OK?

> BTW, in this patch, I just adjusted the inode_dio_wait() to the front of the ocfs2_rw_lock()
> and didn't adjust the order of inode_lock() and inode_dio_wait().

Right.  I think you've convinced me to stop worrying about this.

Ben.

-- 
Ben Hutchings
Software Developer, Codethink Ltd.