Re: [PATCH v3 3/4] limit nr_dentries per superblock

From: Pavel Emelyanov <xemul-bzQdu9zFT3WakBO8gow8eQ@public.gmane.org>
To: Pekka Enberg <penberg-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>
Cc: Andrea Arcangeli
	<aarcange-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>,
	Rik van Riel <riel-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>,
	Nick Piggin <npiggin-tSWWG44O7X1aa/9Udqfwiw@public.gmane.org>,
	Eric Dumazet
	<eric.dumazet-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>,
	Hugh Dickins <hughd-hpIqsD4AKlfQT0dZR+AlfA@public.gmane.org>,
	"containers-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org"
	<containers-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org>,
	Dave Chinner <david-FqsqvQoI3Ljby3iVrkZq2A@public.gmane.org>,
	"linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org"
	<linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org>,
	James Bottomley
	<jbottomley-bzQdu9zFT3WakBO8gow8eQ@public.gmane.org>,
	Al Viro <viro-RmSDqhL/yNMiFSDQTTA3OLVCufUGDwFn@public.gmane.org>,
	David Rientjes <rientjes-hpIqsD4AKlfQT0dZR+AlfA@public.gmane.org>,
	"linux-fsdevel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org"
	<linux-fsdevel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org>,
	Christoph Lameter <cl-vYTEC60ixJUAvxtiuMwx3w@public.gmane.org>,
	Dave Hansen
	<dave-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8@public.gmane.org>
Subject: Re: [PATCH v3 3/4] limit nr_dentries per superblock
Date: Mon, 15 Aug 2011 16:12:55 +0400	[thread overview]
Message-ID: <4E490D47.8050105__24260.0837759257$1313410445$gmane$org@parallels.com> (raw)
In-Reply-To: <CAOJsxLErFcxuuqnWkRbOAHEFbeGrKp3YMZ-144=m16oBXCHJ2g-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>

On 08/15/2011 03:55 PM, Pekka Enberg wrote:
> On Mon, Aug 15, 2011 at 2:32 PM, Pavel Emelyanov <xemul-bzQdu9zFT3WakBO8gow8eQ@public.gmane.org> wrote:
>>> Couldn't you simply do per-container "struct kmem_accounted_cache" in struct superblock?
>>
>> If by this you mean "account for all the kmem associated with particular superblock" then
>> this is OK for us, but this can't be done in a simple
>>
>>        if (used + size > limit)
>>                return -ENOMEM
>>        else {
>>                used += size;
>>                return 0;
>>        }
>>
>> manner, since once we hit the limit we should shrink the unused dentries. And most of the
>> patches are about this.
> 
> So you want to shrink dentries that were allocated in the container
> that hit the dcache limit? How does this patch set deal with that? Or
> are you referring to some other patches?

Yes, this patchset does only the API part. The full idea is implemented (with poor API though) here:
http://lwn.net/Articles/441164/

>                          Pekka
> .
> 