From: Nikolay Borisov <kernel@kyup.com>
To: js1304@gmail.com, Andrew Morton <akpm@linux-foundation.org>
Cc: Christoph Lameter <cl@linux.com>,
Pekka Enberg <penberg@kernel.org>,
David Rientjes <rientjes@google.com>,
Jesper Dangaard Brouer <brouer@redhat.com>,
linux-mm@kvack.org, linux-kernel@vger.kernel.org,
Joonsoo Kim <iamjoonsoo.kim@lge.com>
Subject: Re: [PATCH 01/11] mm/slab: hold a slab_mutex when calling __kmem_cache_shrink()
Date: Thu, 31 Mar 2016 13:53:14 +0300 [thread overview]
Message-ID: <56FD019A.10906@kyup.com> (raw)
In-Reply-To: <1459142821-20303-2-git-send-email-iamjoonsoo.kim@lge.com>
On 03/28/2016 08:26 AM, js1304@gmail.com wrote:
> From: Joonsoo Kim <iamjoonsoo.kim@lge.com>
>
> Major kmem_cache metadata in slab subsystem is synchronized with
> the slab_mutex. In SLAB, if some of them is changed, node's shared
> array cache would be freed and re-populated. If __kmem_cache_shrink()
> is called at the same time, it will call drain_array() with n->shared
> without holding node lock so problem can happen.
>
> We can fix this small theoretical race condition by holding node lock
> in drain_array(), but, holding a slab_mutex in kmem_cache_shrink()
> looks more appropriate solution because stable state would make things
> less error-prone and this is not performance critical path.
>
> In addtion, annotate on SLAB functions.
Just a nit but would it not be better instead of doing comment-style
annotation to use lockdep_assert_held/_once. In both cases for someone
to understand what locks have to be held will go and read the source. In
my mind it's easier to miss a comment line, rather than the
lockdep_assert. Furthermore in case lockdep is enabled a locking
violation would spew useful info to dmesg.
--
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>
WARNING: multiple messages have this Message-ID (diff)
From: Nikolay Borisov <kernel@kyup.com>
To: js1304@gmail.com, Andrew Morton <akpm@linux-foundation.org>
Cc: Christoph Lameter <cl@linux.com>,
Pekka Enberg <penberg@kernel.org>,
David Rientjes <rientjes@google.com>,
Jesper Dangaard Brouer <brouer@redhat.com>,
linux-mm@kvack.org, linux-kernel@vger.kernel.org,
Joonsoo Kim <iamjoonsoo.kim@lge.com>
Subject: Re: [PATCH 01/11] mm/slab: hold a slab_mutex when calling __kmem_cache_shrink()
Date: Thu, 31 Mar 2016 13:53:14 +0300 [thread overview]
Message-ID: <56FD019A.10906@kyup.com> (raw)
In-Reply-To: <1459142821-20303-2-git-send-email-iamjoonsoo.kim@lge.com>
On 03/28/2016 08:26 AM, js1304@gmail.com wrote:
> From: Joonsoo Kim <iamjoonsoo.kim@lge.com>
>
> Major kmem_cache metadata in slab subsystem is synchronized with
> the slab_mutex. In SLAB, if some of them is changed, node's shared
> array cache would be freed and re-populated. If __kmem_cache_shrink()
> is called at the same time, it will call drain_array() with n->shared
> without holding node lock so problem can happen.
>
> We can fix this small theoretical race condition by holding node lock
> in drain_array(), but, holding a slab_mutex in kmem_cache_shrink()
> looks more appropriate solution because stable state would make things
> less error-prone and this is not performance critical path.
>
> In addtion, annotate on SLAB functions.
Just a nit but would it not be better instead of doing comment-style
annotation to use lockdep_assert_held/_once. In both cases for someone
to understand what locks have to be held will go and read the source. In
my mind it's easier to miss a comment line, rather than the
lockdep_assert. Furthermore in case lockdep is enabled a locking
violation would spew useful info to dmesg.
next prev parent reply other threads:[~2016-03-31 10:53 UTC|newest]
Thread overview: 58+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-03-28 5:26 mm/slab: reduce lock contention in alloc path js1304
2016-03-28 5:26 ` js1304
2016-03-28 5:26 ` [PATCH 01/11] mm/slab: hold a slab_mutex when calling __kmem_cache_shrink() js1304
2016-03-28 5:26 ` js1304
2016-03-29 0:50 ` Christoph Lameter
2016-03-29 0:50 ` Christoph Lameter
2016-03-30 8:11 ` Joonsoo Kim
2016-03-30 8:11 ` Joonsoo Kim
2016-03-31 10:53 ` Nikolay Borisov [this message]
2016-03-31 10:53 ` Nikolay Borisov
2016-04-01 2:18 ` Joonsoo Kim
2016-04-01 2:18 ` Joonsoo Kim
2016-03-28 5:26 ` [PATCH 02/11] mm/slab: remove BAD_ALIEN_MAGIC again js1304
2016-03-28 5:26 ` js1304
2016-03-28 8:58 ` Geert Uytterhoeven
2016-03-28 8:58 ` Geert Uytterhoeven
2016-03-30 8:11 ` Joonsoo Kim
2016-03-30 8:11 ` Joonsoo Kim
2016-03-28 21:19 ` Andrew Morton
2016-03-28 21:19 ` Andrew Morton
2016-03-29 0:53 ` Christoph Lameter
2016-03-29 0:53 ` Christoph Lameter
2016-03-28 5:26 ` [PATCH 03/11] mm/slab: drain the free slab as much as possible js1304
2016-03-28 5:26 ` js1304
2016-03-29 0:54 ` Christoph Lameter
2016-03-29 0:54 ` Christoph Lameter
2016-03-28 5:26 ` [PATCH 04/11] mm/slab: factor out kmem_cache_node initialization code js1304
2016-03-28 5:26 ` js1304
2016-03-29 0:56 ` Christoph Lameter
2016-03-29 0:56 ` Christoph Lameter
2016-03-30 8:12 ` Joonsoo Kim
2016-03-30 8:12 ` Joonsoo Kim
2016-03-28 5:26 ` [PATCH 05/11] mm/slab: clean-up kmem_cache_node setup js1304
2016-03-28 5:26 ` js1304
2016-03-29 0:58 ` Christoph Lameter
2016-03-29 0:58 ` Christoph Lameter
2016-03-30 8:15 ` Joonsoo Kim
2016-03-30 8:15 ` Joonsoo Kim
2016-03-28 5:26 ` [PATCH 06/11] mm/slab: don't keep free slabs if free_objects exceeds free_limit js1304
2016-03-28 5:26 ` js1304
2016-03-29 1:03 ` Christoph Lameter
2016-03-29 1:03 ` Christoph Lameter
2016-03-30 8:25 ` Joonsoo Kim
2016-03-30 8:25 ` Joonsoo Kim
2016-03-28 5:26 ` [PATCH 07/11] mm/slab: racy access/modify the slab color js1304
2016-03-28 5:26 ` js1304
2016-03-29 1:05 ` Christoph Lameter
2016-03-29 1:05 ` Christoph Lameter
2016-03-30 8:25 ` Joonsoo Kim
2016-03-30 8:25 ` Joonsoo Kim
2016-03-28 5:26 ` [PATCH 08/11] mm/slab: make cache_grow() handle the page allocated on arbitrary node js1304
2016-03-28 5:26 ` js1304
2016-03-28 5:26 ` [PATCH 09/11] mm/slab: separate cache_grow() to two parts js1304
2016-03-28 5:26 ` js1304
2016-03-28 5:27 ` [PATCH 10/11] mm/slab: refill cpu cache through a new slab without holding a node lock js1304
2016-03-28 5:27 ` js1304
2016-03-28 5:27 ` [PATCH 11/11] mm/slab: lockless decision to grow cache js1304
2016-03-28 5:27 ` js1304
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=56FD019A.10906@kyup.com \
--to=kernel@kyup.com \
--cc=akpm@linux-foundation.org \
--cc=brouer@redhat.com \
--cc=cl@linux.com \
--cc=iamjoonsoo.kim@lge.com \
--cc=js1304@gmail.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=penberg@kernel.org \
--cc=rientjes@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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.