Re: [PATCH 2/4] writeback: allow for dirty metadata accounting

[Josef Bacik <josef@toxicpanda.com>]

On Thu, Nov 09, 2017 at 11:32:46AM +0100, Jan Kara wrote:
> On Wed 08-11-17 14:00:58, Josef Bacik wrote:
> > From: Josef Bacik <jbacik@fb.com>
> > 
> > Provide a mechanism for file systems to indicate how much dirty metadata they
> > are holding.  This introduces a few things
> > 
> > 1) Zone stats for dirty metadata, which is the same as the NR_FILE_DIRTY.
> > 2) WB stat for dirty metadata.  This way we know if we need to try and call into
> > the file system to write out metadata.  This could potentially be used in the
> > future to make balancing of dirty pages smarter.
> > 
> > Signed-off-by: Josef Bacik <jbacik@fb.com>
> ...
> > diff --git a/mm/vmscan.c b/mm/vmscan.c
> > index 13d711dd8776..0281abd62e87 100644
> > --- a/mm/vmscan.c
> > +++ b/mm/vmscan.c
> > @@ -3827,7 +3827,8 @@ static unsigned long node_pagecache_reclaimable(struct pglist_data *pgdat)
> >  
> >  	/* If we can't clean pages, remove dirty pages from consideration */
> >  	if (!(node_reclaim_mode & RECLAIM_WRITE))
> > -		delta += node_page_state(pgdat, NR_FILE_DIRTY);
> > +		delta += node_page_state(pgdat, NR_FILE_DIRTY) +
> > +			node_page_state(pgdat, NR_METADATA_DIRTY);
> >  
> >  	/* Watch for any possible underflows due to delta */
> >  	if (unlikely(delta > nr_pagecache_reclaimable))
> 
> Do you expect your metadata pages to be accounted in NR_FILE_PAGES?
> Otherwise this doesn't make sense. And even if they would, this function is
> about kswapd / direct page reclaim and I don't think you've added smarts
> there to writeout metadata. So if your metadata pages are going to show up
> in NR_FILE_PAGES, you need to subtract NR_METADATA_DIRTY from reclaimable
> pages always. It would be good to see btrfs counterpart to these patches so
> that we can answer questions like this easily...
> 

Ah good point, this accounting doesn't belong here, I'll fix it up.  I haven't
been sending the btrfs patch because it's fucking huge, since untangling the
btree inode usage requires a lot of reworking all at once so it's actually
buildable, so it didn't seem useful for the larger non-btrfs audience.  You can
see it in my git tree here

https://git.kernel.org/pub/scm/linux/kernel/git/josef/btrfs-next.git/commit/?h=new-kill-btree-inode&id=5dfd4a0012c1253260da07bee3fa3d4c13aac616

I'll fix this up.  Thanks,

Josef

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>