From: Glauber Costa <glommer@parallels.com>
To: David Rientjes <rientjes@google.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>,
cgroups@vger.kernel.org, linux-mm@kvack.org,
linux-kernel@vger.kernel.org, devel@openvz.org,
KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>,
Michal Hocko <mhocko@suse.cz>,
Johannes Weiner <hannes@cmpxchg.org>,
Greg Thelen <gthelen@google.com>,
Suleiman Souhlal <suleiman@google.com>,
Christoph Lameter <cl@linux.com>,
Pekka Enberg <penberg@cs.helsinki.fi>
Subject: Re: [PATCH 17/23] kmem controller charge/uncharge infrastructure
Date: Wed, 25 Apr 2012 11:43:28 -0300 [thread overview]
Message-ID: <4F980D90.8050109@parallels.com> (raw)
In-Reply-To: <alpine.DEB.2.00.1204241550110.2537@chino.kir.corp.google.com>
On 04/24/2012 07:54 PM, David Rientjes wrote:
> On Tue, 24 Apr 2012, Glauber Costa wrote:
>
>>> Yes, for user memory, I see charging to p->mm->owner as allowing that
>>> process to eventually move and be charged to a different memcg and there's
>>> no way to do proper accounting if the charge is split amongst different
>>> memcgs because of thread membership to a set of memcgs. This is
>>> consistent with charges for shared memory being moved when a thread
>>> mapping it moves to a new memcg, as well.
>>
>> But that's the problem.
>>
>> When we are dealing with kernel memory, we are allocating a whole slab page.
>> It is essentially impossible to track, given a page, which task allocated
>> which object.
>>
>
> Right, so you have to make the distinction that slab charges cannot be
> migrated by memory.move_charge_at_immigrate (and it's not even specified
> to do anything beyond user pages in Documentation/cgroups/memory.txt),
Never intended to.
> but
> it would be consistent to charge the same memcg for a process's slab
> allocations as the process's user allocations.
>
> My response was why we shouldn't be charging user pages to
> mem_cgroup_from_task(current) rather than
> mem_cgroup_from_task(current->mm->owner) which is what is currently
> implemented.
Ah, all right. Well, for user memory I agree with you. My point was
exactly that user memory can always be pinpointed to a specific address
space, while kernel memory can't.
>
> If that can't be changed so that we can still migrate user memory amongst
> memcgs for memory.move_charge_at_immigrate, then it seems consistent to
> have all allocations done by a task to be charged to the same memcg.
> Hence, I suggested current->mm->owner for slab charging as well.
All right. This can be done. Although I don't see this as a must for
slab as already explained, I certainly don't oppose doing so as well.
--
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/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
next prev parent reply other threads:[~2012-04-25 14:45 UTC|newest]
Thread overview: 70+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-04-20 21:57 [PATCH 00/23] slab+slub accounting for memcg Glauber Costa
2012-04-20 21:57 ` [PATCH 01/23] slub: don't create a copy of the name string in kmem_cache_create Glauber Costa
2012-04-20 21:57 ` [PATCH 02/23] slub: always get the cache from its page in kfree Glauber Costa
2012-04-20 21:57 ` [PATCH 03/23] slab: rename gfpflags to allocflags Glauber Costa
2012-04-20 21:57 ` [PATCH 04/23] memcg: Make it possible to use the stock for more than one page Glauber Costa
2012-04-25 0:59 ` KAMEZAWA Hiroyuki
2012-04-20 21:57 ` [PATCH 05/23] memcg: Reclaim when more than one page needed Glauber Costa
2012-04-25 1:16 ` KAMEZAWA Hiroyuki
2012-04-20 21:57 ` [PATCH 06/23] slab: use obj_size field of struct kmem_cache when not debugging Glauber Costa
2012-04-20 21:57 ` [PATCH 07/23] change defines to an enum Glauber Costa
2012-04-25 1:18 ` KAMEZAWA Hiroyuki
2012-04-20 21:57 ` [PATCH 08/23] don't force return value checking in res_counter_charge_nofail Glauber Costa
2012-04-25 1:28 ` KAMEZAWA Hiroyuki
2012-04-20 21:57 ` [PATCH 09/23] kmem slab accounting basic infrastructure Glauber Costa
2012-04-25 1:32 ` KAMEZAWA Hiroyuki
2012-04-25 14:38 ` Glauber Costa
2012-04-26 0:08 ` KAMEZAWA Hiroyuki
2012-04-30 19:33 ` Suleiman Souhlal
2012-05-02 15:15 ` Glauber Costa
2012-04-20 21:57 ` [PATCH 10/23] slab/slub: struct memcg_params Glauber Costa
2012-04-30 19:42 ` Suleiman Souhlal
2012-04-20 21:57 ` [PATCH 11/23] slub: consider a memcg parameter in kmem_create_cache Glauber Costa
2012-04-24 14:03 ` Frederic Weisbecker
2012-04-24 14:27 ` Glauber Costa
2012-04-25 1:38 ` KAMEZAWA Hiroyuki
2012-04-25 14:37 ` Glauber Costa
2012-04-30 19:51 ` Suleiman Souhlal
2012-05-02 15:18 ` Glauber Costa
2012-04-22 23:53 ` [PATCH 12/23] slab: pass memcg parameter to kmem_cache_create Glauber Costa
2012-04-30 19:54 ` Suleiman Souhlal
2012-04-22 23:53 ` [PATCH 13/23] slub: create duplicate cache Glauber Costa
2012-04-24 14:18 ` Frederic Weisbecker
2012-04-24 14:37 ` Glauber Costa
2012-04-26 13:10 ` Frederic Weisbecker
2012-04-30 20:15 ` Suleiman Souhlal
2012-04-22 23:53 ` [PATCH 14/23] slub: provide kmalloc_no_account Glauber Costa
2012-04-22 23:53 ` [PATCH 15/23] slab: create duplicate cache Glauber Costa
2012-04-22 23:53 ` [PATCH 16/23] slab: provide kmalloc_no_account Glauber Costa
2012-04-25 1:44 ` KAMEZAWA Hiroyuki
2012-04-25 14:29 ` Glauber Costa
2012-04-26 0:13 ` KAMEZAWA Hiroyuki
2012-04-22 23:53 ` [PATCH 17/23] kmem controller charge/uncharge infrastructure Glauber Costa
2012-04-23 22:25 ` David Rientjes
2012-04-24 14:22 ` Frederic Weisbecker
2012-04-24 14:40 ` Glauber Costa
2012-04-24 20:25 ` David Rientjes
2012-04-24 21:36 ` Glauber Costa
2012-04-24 22:54 ` David Rientjes
2012-04-25 14:43 ` Glauber Costa [this message]
2012-04-24 20:21 ` David Rientjes
2012-04-27 11:38 ` Frederic Weisbecker
2012-04-27 18:13 ` David Rientjes
2012-04-25 1:56 ` KAMEZAWA Hiroyuki
2012-04-25 14:44 ` Glauber Costa
2012-04-27 12:22 ` Frederic Weisbecker
2012-04-30 20:56 ` Suleiman Souhlal
2012-05-02 15:34 ` Glauber Costa
2012-04-22 23:53 ` [PATCH 18/23] slub: charge allocation to a memcg Glauber Costa
2012-04-22 23:53 ` [PATCH 19/23] slab: per-memcg accounting of slab caches Glauber Costa
2012-04-30 21:25 ` Suleiman Souhlal
2012-05-02 15:40 ` Glauber Costa
2012-04-22 23:53 ` [PATCH 20/23] memcg: disable kmem code when not in use Glauber Costa
2012-04-22 23:53 ` [PATCH 21/23] memcg: Track all the memcg children of a kmem_cache Glauber Costa
2012-04-22 23:53 ` [PATCH 22/23] memcg: Per-memcg memory.kmem.slabinfo file Glauber Costa
2012-04-22 23:53 ` [PATCH 23/23] slub: create slabinfo file for memcg Glauber Costa
2012-04-22 23:59 ` [PATCH 00/23] slab+slub accounting " Glauber Costa
2012-04-30 9:59 ` [PATCH 0/3] A few fixes for '[PATCH 00/23] slab+slub accounting for memcg' series Anton Vorontsov
2012-04-30 10:01 ` [PATCH 1/3] slab: Proper off-slabs handling when duplicating caches Anton Vorontsov
2012-04-30 10:01 ` [PATCH 2/3] slab: Fix imbalanced rcu locking Anton Vorontsov
2012-04-30 10:02 ` [PATCH 3/3] slab: Get rid of mem_cgroup_put_kmem_cache() Anton Vorontsov
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=4F980D90.8050109@parallels.com \
--to=glommer@parallels.com \
--cc=cgroups@vger.kernel.org \
--cc=cl@linux.com \
--cc=devel@openvz.org \
--cc=fweisbec@gmail.com \
--cc=gthelen@google.com \
--cc=hannes@cmpxchg.org \
--cc=kamezawa.hiroyu@jp.fujitsu.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=mhocko@suse.cz \
--cc=penberg@cs.helsinki.fi \
--cc=rientjes@google.com \
--cc=suleiman@google.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox