From: Xunlei Pang <xlpang@linux.alibaba.com>
To: Vlastimil Babka <vbabka@suse.cz>,
xlpang@linux.alibaba.com, Christoph Lameter <cl@linux.com>,
Pekka Enberg <penberg@kernel.org>, Roman Gushchin <guro@fb.com>,
Konstantin Khlebnikov <khlebnikov@yandex-team.ru>,
David Rientjes <rientjes@google.com>,
Matthew Wilcox <willy@infradead.org>,
Shu Ming <sming56@gmail.com>,
Andrew Morton <akpm@linux-foundation.org>,
Christoph Lameter <cl@gentwo.de>
Cc: linux-kernel@vger.kernel.org, linux-mm@kvack.org,
Wen Yang <wenyang@linux.alibaba.com>,
James Wang <jnwang@linux.alibaba.com>,
Thomas Gleixner <tglx@linutronix.de>
Subject: Re: [PATCH v3 0/4] mm/slub: Fix count_partial() problem
Date: Tue, 16 Mar 2021 19:49:33 +0800 [thread overview]
Message-ID: <9ea6829a-bf10-4c24-bc8c-492862a76b54@linux.alibaba.com> (raw)
In-Reply-To: <c5028ee5-6407-14ef-d060-1c8755ca213a@suse.cz>
On 3/16/21 7:02 PM, Vlastimil Babka wrote:
> On 3/16/21 11:42 AM, Xunlei Pang wrote:
>> On 3/16/21 2:49 AM, Vlastimil Babka wrote:
>>> On 3/9/21 4:25 PM, Xunlei Pang wrote:
>>>> count_partial() can hold n->list_lock spinlock for quite long, which
>>>> makes much trouble to the system. This series eliminate this problem.
>>>
>>> Before I check the details, I have two high-level comments:
>>>
>>> - patch 1 introduces some counting scheme that patch 4 then changes, could we do
>>> this in one step to avoid the churn?
>>>
>>> - the series addresses the concern that spinlock is being held, but doesn't
>>> address the fact that counting partial per-node slabs is not nearly enough if we
>>> want accurate <active_objs> in /proc/slabinfo because there are also percpu
>>> slabs and per-cpu partial slabs, where we don't track the free objects at all.
>>> So after this series while the readers of /proc/slabinfo won't block the
>>> spinlock, they will get the same garbage data as before. So Christoph is not
>>> wrong to say that we can just report active_objs == num_objs and it won't
>>> actually break any ABI.
>>
>> If maintainers don't mind this inaccuracy which I also doubt its
>> importance, then it becomes easy. For fear that some people who really
>> cares, introducing an extra config(default-off) for it would be a good
>> option.
>
> Great.
>
>>> At the same time somebody might actually want accurate object statistics at the
>>> expense of peak performance, and it would be nice to give them such option in
>>> SLUB. Right now we don't provide this accuracy even with CONFIG_SLUB_STATS,
>>> although that option provides many additional tuning stats, with additional
>>> overhead.
>>> So my proposal would be a new config for "accurate active objects" (or just tie
>>> it to CONFIG_SLUB_DEBUG?) that would extend the approach of percpu counters in
>>> patch 4 to all alloc/free, so that it includes percpu slabs. Without this config
>>> enabled, let's just report active_objs == num_objs.
>> For percpu slabs, the numbers can be retrieved from the existing
>> slub_percpu_partial()->pobjects, looks no need extra work.
>
> Hm, unfortunately it's not that simple, the number there is a snapshot that can
> become wildly inacurate afterwards.
>
It's hard to make it absoultely accurate using percpu, the data can
change during you iterating all the cpus and total_objects, I can't
imagine its real-world usage, not to mention the percpu freelist cache.
I think sysfs slabs_cpu_partial should work enough for common debug purpose.
prev parent reply other threads:[~2021-03-16 11:50 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-03-09 15:25 [PATCH v3 0/4] mm/slub: Fix count_partial() problem Xunlei Pang
2021-03-09 15:25 ` [PATCH v3 1/4] mm/slub: Introduce two counters for partial objects Xunlei Pang
2021-03-09 15:25 ` [PATCH v3 2/4] mm/slub: Get rid of count_partial() Xunlei Pang
2021-03-09 15:25 ` [PATCH v3 3/4] percpu: Export per_cpu_sum() Xunlei Pang
2021-03-09 15:25 ` [PATCH v3 4/4] mm/slub: Use percpu partial free counter Xunlei Pang
2021-03-15 18:49 ` [PATCH v3 0/4] mm/slub: Fix count_partial() problem Vlastimil Babka
2021-03-15 19:05 ` Roman Gushchin
2021-03-15 19:22 ` Yang Shi
2021-03-16 10:07 ` Christoph Lameter
2021-03-16 10:32 ` Vlastimil Babka
2021-03-16 10:42 ` Xunlei Pang
2021-03-16 11:02 ` Vlastimil Babka
2021-03-16 11:49 ` Xunlei Pang [this message]
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=9ea6829a-bf10-4c24-bc8c-492862a76b54@linux.alibaba.com \
--to=xlpang@linux.alibaba.com \
--cc=akpm@linux-foundation.org \
--cc=cl@gentwo.de \
--cc=cl@linux.com \
--cc=guro@fb.com \
--cc=jnwang@linux.alibaba.com \
--cc=khlebnikov@yandex-team.ru \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=penberg@kernel.org \
--cc=rientjes@google.com \
--cc=sming56@gmail.com \
--cc=tglx@linutronix.de \
--cc=vbabka@suse.cz \
--cc=wenyang@linux.alibaba.com \
--cc=willy@infradead.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).