lkml.org 
[lkml]   [2010]   [Nov]   [16]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: [PATCH] jbd2: avoid the concurrent data writeback
Hi Hellwig,

On Mon, 15 Nov 2010 19:27:32 +0800
Christoph Hellwig <hch@infradead.org> wrote:

> On Mon, Nov 15, 2010 at 05:59:43PM +0800, Feng Tang wrote:
> > + *
> > + * Sometimes when this get called, the host inode may be under data
> > + * syncing initiated by flush thread(especially for a large file),
> > and
> > + * in such situation, we should skip this path of writeback
> > */
> > static int journal_submit_inode_data_buffers(struct address_space
> > *mapping) {
> > @@ -181,6 +185,13 @@ static int
> > journal_submit_inode_data_buffers(struct address_space
> > *mapping) .range_end = i_size_read(mapping->host), };
> >
> > + spin_lock(&inode_lock);
> > + if (mapping->host->i_state & I_SYNC) {
> > + spin_unlock(&inode_lock);
> > + return 0;
> > + }
> > + spin_unlock(&inode_lock);
> > +
>
> inode_lock is not exported to modules, and that's for a pretty good
> reason. I think you want to change this code at a higher level to not
> compete with the flusher threads at all.
>
Good point. The alternative I can think of is to use writeback_in_progress(),
thus the check will be changed to:

if (writeback_in_progress(mapping->backing_dev_info))
return 0;
This have the same effect as the original patch.


Thanks,
Feng


\
 
 \ /
  Last update: 2010-11-16 08:43    [W:0.060 / U:0.588 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site