linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
From: Vladimir Davydov <vdavydov@parallels.com>
To: Michal Hocko <mhocko@suse.cz>
Cc: linux-kernel@vger.kernel.org, linux-mm@kvack.org,
	cgroups@vger.kernel.org, devel@openvz.org,
	Johannes Weiner <hannes@cmpxchg.org>,
	Glauber Costa <glommer@gmail.com>,
	Christoph Lameter <cl@linux.com>,
	Pekka Enberg <penberg@kernel.org>,
	Andrew Morton <akpm@linux-foundation.org>
Subject: Re: [PATCH 2/6] memcg, slab: kmem_cache_create_memcg(): free memcg params on error
Date: Thu, 19 Dec 2013 13:01:28 +0400	[thread overview]
Message-ID: <52B2B5E8.6020307@parallels.com> (raw)
In-Reply-To: <20131219084845.GB9331@dhcp22.suse.cz>

On 12/19/2013 12:48 PM, Michal Hocko wrote:
> On Thu 19-12-13 10:32:29, Vladimir Davydov wrote:
>> On 12/18/2013 09:06 PM, Michal Hocko wrote:
>>> On Wed 18-12-13 17:16:53, Vladimir Davydov wrote:
>>>> Plus, rename memcg_register_cache() to memcg_init_cache_params(),
>>>> because it actually does not register the cache anywhere, but simply
>>>> initialize kmem_cache::memcg_params.
>>> I've almost missed this is a memory leak fix.
>> Yeah, the comment is poor, sorry about that. Will fix it.
>>
>>> I do not mind renaming and the name but wouldn't
>>> memcg_alloc_cache_params suit better?
>> As you wish. I don't have a strong preference for memcg_init_cache_params.
> I really hate naming... but it seems that alloc is a better fit. _init_
> would expect an already allocated object.
>
> Btw. memcg_free_cache_params is called only once which sounds
> suspicious. The regular destroy path should use it as well?
> [...]

The usual destroy path uses memcg_release_cache(), which does the trick.
Plus, it actually "unregisters" the cache. BTW, I forgot to substitute
kfree(s->memcg_params) with the new memcg_free_cache_params() there.
Although it currently does not break anything, better to fix it in case
new memcg_free_cache_params() will have to do something else.

And you're right about the naming is not good.

Currently we have:

  on create:
    memcg_register_cache()
    memcg_cache_list_add()
  on destroy:
    memcg_release_cache()

After this patch we would have:

  on create:
    memcg_alloc_cache_params()
    memcg_register_cache()
  on destroy:
    memcg_release_cache()

Still not perfect: "alloc" does not have corresponding "free", while
"register" does not have corresponding "unregister", everything is done
by "release".

What do you think about splitting memcg_release_cache() into two functions:

    memcg_unregister_cache()
    memcg_free_cache_params()

?

Thanks.

--
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>

  reply	other threads:[~2013-12-19  9:01 UTC|newest]

Thread overview: 42+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-12-18 13:16 [PATCH 1/6] slab: cleanup kmem_cache_create_memcg() Vladimir Davydov
2013-12-18 13:16 ` [PATCH 2/6] memcg, slab: kmem_cache_create_memcg(): free memcg params on error Vladimir Davydov
2013-12-18 17:06   ` Michal Hocko
2013-12-19  6:32     ` Vladimir Davydov
2013-12-19  8:48       ` Michal Hocko
2013-12-19  9:01         ` Vladimir Davydov [this message]
2013-12-19  9:19           ` Michal Hocko
2013-12-18 13:16 ` [PATCH 3/6] memcg, slab: cleanup barrier usage when accessing memcg_caches Vladimir Davydov
2013-12-18 17:14   ` Michal Hocko
2013-12-19  6:37     ` Vladimir Davydov
2013-12-19  9:10       ` Michal Hocko
2013-12-19  9:16         ` Vladimir Davydov
2013-12-19  9:21           ` Michal Hocko
2013-12-19  9:29             ` Vladimir Davydov
2013-12-19  9:36               ` Michal Hocko
2013-12-19  9:53                 ` Vladimir Davydov
2013-12-18 13:16 ` [PATCH 4/6] memcg, slab: check and init memcg_cahes under slab_mutex Vladimir Davydov
2013-12-18 17:41   ` Michal Hocko
2013-12-19  7:07     ` Vladimir Davydov
2013-12-19  8:00       ` Glauber Costa
2013-12-19  9:12         ` Michal Hocko
2013-12-19  9:17           ` Vladimir Davydov
2013-12-19  9:21         ` Vladimir Davydov
2013-12-18 13:16 ` [PATCH 5/6] memcg: clear memcg_params after removing cache from memcg_slab_caches list Vladimir Davydov
2013-12-18 13:16 ` [PATCH 6/6] memcg, slab: RCU protect memcg_params for root caches Vladimir Davydov
2013-12-19  9:28   ` Michal Hocko
2013-12-19  9:36     ` Vladimir Davydov
2013-12-19  9:43       ` Michal Hocko
2013-12-19  9:47         ` Vladimir Davydov
2013-12-19 10:06           ` Michal Hocko
2013-12-18 16:56 ` [PATCH 1/6] slab: cleanup kmem_cache_create_memcg() Michal Hocko
2013-12-19  6:31   ` Vladimir Davydov
2013-12-19  8:44     ` Michal Hocko
2013-12-19  8:51       ` Vladimir Davydov
2013-12-19  9:16         ` Michal Hocko
2013-12-19  7:27 ` Pekka Enberg
2013-12-19  8:17 ` [Devel] " Vasily Averin
2013-12-19  8:39   ` Vladimir Davydov
2013-12-19  9:26     ` Vasily Averin
2013-12-19  9:42       ` Vladimir Davydov
2013-12-19  9:45       ` Michal Hocko
2013-12-19 10:23       ` Pekka Enberg

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=52B2B5E8.6020307@parallels.com \
    --to=vdavydov@parallels.com \
    --cc=akpm@linux-foundation.org \
    --cc=cgroups@vger.kernel.org \
    --cc=cl@linux.com \
    --cc=devel@openvz.org \
    --cc=glommer@gmail.com \
    --cc=hannes@cmpxchg.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=mhocko@suse.cz \
    --cc=penberg@kernel.org \
    /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;
as well as URLs for NNTP newsgroup(s).