Re: [PATCH] Fix DIO EIO error caused by race between jbd_commit_transaction() and journal_try_to_drop_buffers()

[Mingming Cao <cmm@us.ibm.com>]

On Fri, 2008-05-16 at 10:17 -0700, Badari Pulavarty wrote:

> > @@ -1713,7 +1753,31 @@ int journal_try_to_free_buffers(journal_
> >  		if (buffer_jbd(bh))
> >  			goto busy;
> >  	} while ((bh = bh->b_this_page) != head);
> > +
> >  	ret = try_to_free_buffers(page);
> > +
> > + 	/*
> > +	 * In the case of concurrent direct IO and buffered IO,
> > +	 * There are a number of places where we
> > +	 * could race with journal_commit_transaction(), the later still
> > +	 * helds the reference to the buffers to free while processing them.
> > +	 * try_to_free_buffers() failed to free those buffers,
> > +	 * resulting in an unexpected EIO error
> > +	 * returns back to the generic_file_direct_IO()
> > +	 *
> > +	 * So let's wait for the current transaction finished flush
> > +	 * dirty data buffers before we try to free those buffers
> > +	 * again. This wait is needed by direct IO code path only,
> > +	 * gfp_mask __GFP_REPEAT is passed from the direct IO code
> > +	 * path to flag if we need to wait and retry free buffers.
> > +	 */
> > +	if (ret == 0 && dio) {
> 
> drop "dio" variable and compare here, like 
> 	  if (ret == 0 && (gfp_mask & __GFP_REPEAT)
> 

Okay, will do. Also will update the patch with other format changes you
suggested.

> > +        	spin_lock(&journal->j_state_lock);
> > +		journal_wait_for_transaction_sync_data(journal);
> > +		ret = try_to_free_buffers(page);
> > +		spin_unlock(&journal->j_state_lock);
> > +	}
> 


Also, This patch changed the struct journal_s to add a new wait queue,
so that journal_try_to_free_buffers() could only need to wait for data
to be commited, rather than using j_wait_done_commit queue and wait for
the whole transaction being committed. It might break other fs  that
uses journal_s, thus not worth it. Will update the patch.