From: Michal Hocko <mhocko@kernel.org>
To: Feng Tang <feng.tang@intel.com>
Cc: Andrew Morton <akpm@linux-foundation.org>,
Johannes Weiner <hannes@cmpxchg.org>,
Matthew Wilcox <willy@infradead.org>,
Mel Gorman <mgorman@suse.de>, Kees Cook <keescook@chromium.org>,
Luis Chamberlain <mcgrof@kernel.org>,
Iurii Zaikin <yzaikin@google.com>,
"Kleen, Andi" <andi.kleen@intel.com>,
"Chen, Tim C" <tim.c.chen@intel.com>,
"Hansen, Dave" <dave.hansen@intel.com>,
"Huang, Ying" <ying.huang@intel.com>,
"linux-mm@kvack.org" <linux-mm@kvack.org>,
"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH 3/3] mm: adjust vm_committed_as_batch according to vm overcommit policy
Date: Fri, 15 May 2020 11:08:21 +0200 [thread overview]
Message-ID: <20200515090821.GO29153@dhcp22.suse.cz> (raw)
In-Reply-To: <20200515080210.GC69177@shbuild999.sh.intel.com>
On Fri 15-05-20 16:02:10, Feng Tang wrote:
> Hi Michal,
>
> Thanks for the thorough reviews for these 3 patches!
>
> On Fri, May 15, 2020 at 03:41:25PM +0800, Michal Hocko wrote:
> > On Fri 08-05-20 15:25:17, Feng Tang wrote:
> > > When checking a performance change for will-it-scale scalability
> > > mmap test [1], we found very high lock contention for spinlock of
> > > percpu counter 'vm_committed_as':
> > >
> > > 94.14% 0.35% [kernel.kallsyms] [k] _raw_spin_lock_irqsave
> > > 48.21% _raw_spin_lock_irqsave;percpu_counter_add_batch;__vm_enough_memory;mmap_region;do_mmap;
> > > 45.91% _raw_spin_lock_irqsave;percpu_counter_add_batch;__do_munmap;
> > >
> > > Actually this heavy lock contention is not always necessary. The
> > > 'vm_committed_as' needs to be very precise when the strict
> > > OVERCOMMIT_NEVER policy is set, which requires a rather small batch
> > > number for the percpu counter.
> > >
> > > So lift the batch number to 16X for OVERCOMMIT_ALWAYS and
> > > OVERCOMMIT_GUESS policies, and add a sysctl handler to adjust it
> > > when the policy is reconfigured.
> >
> > Increasing the batch size for weaker overcommit modes makes sense. But
> > your patch is also tuning OVERCOMMIT_NEVER without any explanation why
> > that is still "small enough to be precise".
>
> Actually, it keeps the batch algorithm for "OVERCOMMIT_NEVER", but
> change the other 2 policies, which I should set it clear in the
> commit log.
Yeah, I have misread that part. Sorry about that.
[...]
> > > +void mm_compute_batch(void)
> > > {
> > > u64 memsized_batch;
> > > s32 nr = num_present_cpus();
> > > s32 batch = max_t(s32, nr*2, 32);
> > > -
> > > - /* batch size set to 0.4% of (total memory/#cpus), or max int32 */
> > > - memsized_batch = min_t(u64, (totalram_pages()/nr)/256, 0x7fffffff);
> > > + unsigned long ram_pages = totalram_pages();
> > > +
> > > + /*
> > > + * For policy of OVERCOMMIT_NEVER, set batch size to 0.4%
> > > + * of (total memory/#cpus), and lift it to 6.25% for other
> > > + * policies to easy the possible lock contention for percpu_counter
> > > + * vm_committed_as, while the max limit is INT_MAX
> > > + */
> > > + if (sysctl_overcommit_memory == OVERCOMMIT_NEVER)
> > > + memsized_batch = min_t(u64, ram_pages/nr/256, INT_MAX);
> > > + else
> > > + memsized_batch = min_t(u64, ram_pages/nr/16, INT_MAX);
>
> Also as you mentioned there are real-world work loads with big mmap
> size and multi-threading, can we lift it even further ?
> memsized_batch = min_t(u64, ram_pages/nr/4, INT_MAX)
Try to measure those and see what numbers look like.
--
Michal Hocko
SUSE Labs
next prev parent reply other threads:[~2020-05-15 9:08 UTC|newest]
Thread overview: 34+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-05-08 7:25 [PATCH 0/3] make vm_committed_as_batch aware of vm overcommit policy Feng Tang
2020-05-08 7:25 ` [PATCH 1/3] proc/meminfo: avoid open coded reading of vm_committed_as Feng Tang
2020-05-15 7:22 ` Michal Hocko
2020-05-08 7:25 ` [PATCH 2/3] mm/util.c: make vm_memory_committed() more accurate Feng Tang
2020-05-15 7:23 ` Michal Hocko
2020-05-15 8:11 ` Feng Tang
2020-05-15 9:04 ` Michal Hocko
2020-05-15 13:01 ` Feng Tang
2020-05-08 7:25 ` [PATCH 3/3] mm: adjust vm_committed_as_batch according to vm overcommit policy Feng Tang
2020-05-08 11:24 ` Matthew Wilcox
2020-05-08 12:33 ` Feng Tang
2020-05-15 7:41 ` Michal Hocko
2020-05-15 8:02 ` Feng Tang
2020-05-15 9:08 ` Michal Hocko [this message]
2020-05-15 11:26 ` Feng Tang
2020-05-15 7:44 ` Michal Hocko
2020-05-15 8:38 ` Feng Tang
2020-05-21 21:27 ` [PATCH 0/3] make vm_committed_as_batch aware of " Qian Cai
2020-05-26 18:14 ` Qian Cai
2020-05-27 1:14 ` Andi Kleen
2020-05-27 1:43 ` Feng Tang
2020-05-27 1:46 ` Feng Tang
2020-05-27 2:25 ` Qian Cai
2020-05-27 10:46 ` Feng Tang
2020-05-27 12:05 ` Qian Cai
2020-05-27 13:33 ` Feng Tang
2020-05-27 15:42 ` Qian Cai
2020-05-28 14:18 ` Qian Cai
2020-05-28 15:10 ` Feng Tang
2020-05-28 15:21 ` Kleen, Andi
2020-05-29 15:43 ` Feng Tang
2020-05-29 15:50 ` Andi Kleen
2020-05-29 16:04 ` Feng Tang
2020-05-28 15:48 ` Michal Hocko
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=20200515090821.GO29153@dhcp22.suse.cz \
--to=mhocko@kernel.org \
--cc=akpm@linux-foundation.org \
--cc=andi.kleen@intel.com \
--cc=dave.hansen@intel.com \
--cc=feng.tang@intel.com \
--cc=hannes@cmpxchg.org \
--cc=keescook@chromium.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=mcgrof@kernel.org \
--cc=mgorman@suse.de \
--cc=tim.c.chen@intel.com \
--cc=willy@infradead.org \
--cc=ying.huang@intel.com \
--cc=yzaikin@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;
as well as URLs for NNTP newsgroup(s).