[prev in list] [next in list] [prev in thread] [next in thread] 

List:       linux-mm
Subject:    Re: [PATCH v4 06/11] memcg: add dirty page accounting
From:       KAMEZAWA Hiroyuki <kamezawa.hiroyu () jp ! fujitsu ! com>
Date:       2010-10-29 11:17:26
Message-ID: 20101029201726.370e462b.kamezawa.hiroyu () jp ! fujitsu ! com
[Download RAW message or body]

On Fri, 29 Oct 2010 19:13:00 +0800
Wu Fengguang <fengguang.wu@intel.com> wrote:

> On Fri, Oct 29, 2010 at 03:09:09PM +0800, Greg Thelen wrote:
> 
> > +
> > +	case MEMCG_NR_FILE_DIRTY:
> > +		/* Use Test{Set,Clear} to only un/charge the memcg once. */
> > +		if (val > 0) {
> > +			if (TestSetPageCgroupFileDirty(pc))
> > +				val = 0;
> > +		} else {
> > +			if (!TestClearPageCgroupFileDirty(pc))
> > +				val = 0;
> > +		}
> 
> I'm wondering why TestSet/TestClear and even the cgroup page flags for
> dirty/writeback/unstable pages are necessary at all (it helps to
> document in changelog if there are any). For example, VFS will call
> TestSetPageDirty() before calling
> mem_cgroup_inc_page_stat(MEMCG_NR_FILE_DIRTY), so there should be no
> chance of false double counting.
> 

1. flag is necessary for moving accounting information between cgroups
   when account_move() occurs.

2. TestSet... is required because there are always race with page_cgroup_lock()'s
   lock bit.

Thanks,
-Kame

--
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>

[prev in list] [next in list] [prev in thread] [next in thread]