From: Feng Tang <feng.tang@intel.com>
To: Eric Dumazet <edumazet@google.com>
Cc: Jakub Kicinski <kuba@kernel.org>, Xin Long <lucien.xin@gmail.com>,
Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>,
kernel test robot <oliver.sang@intel.com>,
Shakeel Butt <shakeelb@google.com>,
Soheil Hassas Yeganeh <soheil@google.com>,
LKML <linux-kernel@vger.kernel.org>,
Linux Memory Management List <linux-mm@kvack.org>,
network dev <netdev@vger.kernel.org>,
linux-s390@vger.kernel.org,
MPTCP Upstream <mptcp@lists.linux.dev>,
"linux-sctp @ vger . kernel . org" <linux-sctp@vger.kernel.org>,
lkp@lists.01.org, kbuild test robot <lkp@intel.com>,
Huang Ying <ying.huang@intel.com>,
zhengjun.xing@linux.intel.com, fengwei.yin@intel.com,
Ying Xu <yinxu@redhat.com>
Subject: Re: [net] 4890b686f4: netperf.Throughput_Mbps -69.4% regression
Date: Fri, 24 Jun 2022 14:00:53 +0800 [thread overview]
Message-ID: <20220624060053.GD79500@shbuild999.sh.intel.com> (raw)
In-Reply-To: <CANn89iLwwN7hRsJD_skbcRNY9sBtPh1fhULKco5wosx_i4x6gg@mail.gmail.com>
On Fri, Jun 24, 2022 at 07:45:00AM +0200, Eric Dumazet wrote:
> On Fri, Jun 24, 2022 at 7:14 AM Feng Tang <feng.tang@intel.com> wrote:
> >
> > Hi Eric,
> >
> > On Fri, Jun 24, 2022 at 06:13:51AM +0200, Eric Dumazet wrote:
> > > On Fri, Jun 24, 2022 at 3:57 AM Jakub Kicinski <kuba@kernel.org> wrote:
> > > >
> > > > On Thu, 23 Jun 2022 18:50:07 -0400 Xin Long wrote:
> > > > > From the perf data, we can see __sk_mem_reduce_allocated() is the one
> > > > > using CPU the most more than before, and mem_cgroup APIs are also
> > > > > called in this function. It means the mem cgroup must be enabled in
> > > > > the test env, which may explain why I couldn't reproduce it.
> > > > >
> > > > > The Commit 4890b686f4 ("net: keep sk->sk_forward_alloc as small as
> > > > > possible") uses sk_mem_reclaim(checking reclaimable >= PAGE_SIZE) to
> > > > > reclaim the memory, which is *more frequent* to call
> > > > > __sk_mem_reduce_allocated() than before (checking reclaimable >=
> > > > > SK_RECLAIM_THRESHOLD). It might be cheap when
> > > > > mem_cgroup_sockets_enabled is false, but I'm not sure if it's still
> > > > > cheap when mem_cgroup_sockets_enabled is true.
> > > > >
> > > > > I think SCTP netperf could trigger this, as the CPU is the bottleneck
> > > > > for SCTP netperf testing, which is more sensitive to the extra
> > > > > function calls than TCP.
> > > > >
> > > > > Can we re-run this testing without mem cgroup enabled?
> > > >
> > > > FWIW I defer to Eric, thanks a lot for double checking the report
> > > > and digging in!
> > >
> > > I did tests with TCP + memcg and noticed a very small additional cost
> > > in memcg functions,
> > > because of suboptimal layout:
> > >
> > > Extract of an internal Google bug, update from June 9th:
> > >
> > > --------------------------------
> > > I have noticed a minor false sharing to fetch (struct
> > > mem_cgroup)->css.parent, at offset 0xc0,
> > > because it shares the cache line containing struct mem_cgroup.memory,
> > > at offset 0xd0
> > >
> > > Ideally, memcg->socket_pressure and memcg->parent should sit in a read
> > > mostly cache line.
> > > -----------------------
> > >
> > > But nothing that could explain a "-69.4% regression"
> >
> > We can double check that.
> >
> > > memcg has a very similar strategy of per-cpu reserves, with
> > > MEMCG_CHARGE_BATCH being 32 pages per cpu.
> >
> > We have proposed patch to increase the batch numer for stats
> > update, which was not accepted as it hurts the accuracy and
> > the data is used by many tools.
> >
> > > It is not clear why SCTP with 10K writes would overflow this reserve constantly.
> > >
> > > Presumably memcg experts will have to rework structure alignments to
> > > make sure they can cope better
> > > with more charge/uncharge operations, because we are not going back to
> > > gigantic per-socket reserves,
> > > this simply does not scale.
> >
> > Yes, the memcg statitics and charge/unchage update is very sensitive
> > with the data alignemnt layout, and can easily trigger peformance
> > changes, as we've seen quite some similar cases in the past several
> > years.
> >
> > One pattern we've seen is, even if a memcg stats updating or charge
> > function only takes about 2%~3% of the CPU cycles in perf-profile data,
> > once it got affected, the peformance change could be amplified to up to
> > 60% or more.
> >
>
> Reorganizing "struct mem_cgroup" to put "struct page_counter memory"
> in a separate cache line would be beneficial.
That may help.
And I also want to say the benchmarks(especially micro one) are very
sensitive to the layout of mem_cgroup. As the 'page_counter' is 112
bytes in size, I recently made a patch to make it cacheline aligned
(take 2 cachelines), which improved some hackbench/netperf test
cases, but caused huge (49%) drop for some vm-scalability tests.
> Many low hanging fruits, assuming nobody will use __randomize_layout on it ;)
>
> Also some fields are written even if their value is not changed.
>
> diff --git a/mm/memcontrol.c b/mm/memcontrol.c
> index abec50f31fe64100f4be5b029c7161b3a6077a74..53d9c1e581e78303ef73942e2b34338567987b74
> 100644
> --- a/mm/memcontrol.c
> +++ b/mm/memcontrol.c
> @@ -7037,10 +7037,12 @@ bool mem_cgroup_charge_skmem(struct mem_cgroup
> *memcg, unsigned int nr_pages,
> struct page_counter *fail;
>
> if (page_counter_try_charge(&memcg->tcpmem, nr_pages, &fail)) {
> - memcg->tcpmem_pressure = 0;
> + if (READ_ONCE(memcg->tcpmem_pressure))
> + WRITE_ONCE(memcg->tcpmem_pressure, 0);
> return true;
> }
> - memcg->tcpmem_pressure = 1;
> + if (!READ_ONCE(memcg->tcpmem_pressure))
> + WRITE_ONCE(memcg->tcpmem_pressure, 1);
> if (gfp_mask & __GFP_NOFAIL) {
> page_counter_charge(&memcg->tcpmem, nr_pages);
> return true;
I will also try this patch, which may take some time.
Thanks,
Feng
next prev parent reply other threads:[~2022-06-24 6:00 UTC|newest]
Thread overview: 35+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-06-19 15:04 [net] 4890b686f4: netperf.Throughput_Mbps -69.4% regression kernel test robot
2022-06-23 0:28 ` Jakub Kicinski
2022-06-23 3:08 ` Xin Long
2022-06-23 22:50 ` Xin Long
2022-06-24 1:57 ` Jakub Kicinski
2022-06-24 4:13 ` Eric Dumazet
2022-06-24 4:22 ` Eric Dumazet
2022-06-24 5:13 ` Feng Tang
2022-06-24 5:45 ` Eric Dumazet
2022-06-24 6:00 ` Feng Tang [this message]
2022-06-24 6:07 ` Eric Dumazet
2022-06-24 6:34 ` Shakeel Butt
2022-06-24 7:06 ` Feng Tang
2022-06-24 14:43 ` Shakeel Butt
2022-06-25 2:36 ` Feng Tang
2022-06-27 2:38 ` Feng Tang
2022-06-27 8:46 ` Eric Dumazet
2022-06-27 12:34 ` Feng Tang
2022-06-27 14:07 ` Eric Dumazet
2022-06-27 14:48 ` Feng Tang
2022-06-27 16:25 ` Eric Dumazet
2022-06-27 16:48 ` Shakeel Butt
2022-06-27 17:05 ` Eric Dumazet
2022-06-28 1:46 ` Roman Gushchin
2022-06-28 3:49 ` Feng Tang
2022-07-01 15:47 ` Shakeel Butt
2022-07-03 10:43 ` Feng Tang
2022-07-03 22:55 ` Roman Gushchin
2022-07-05 5:03 ` Feng Tang
2022-08-16 5:52 ` Oliver Sang
2022-08-16 15:55 ` Shakeel Butt
2022-06-27 14:52 ` Shakeel Butt
2022-06-27 14:56 ` Eric Dumazet
2022-06-27 15:12 ` Feng Tang
2022-06-27 16:25 ` Shakeel Butt
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=20220624060053.GD79500@shbuild999.sh.intel.com \
--to=feng.tang@intel.com \
--cc=edumazet@google.com \
--cc=fengwei.yin@intel.com \
--cc=kuba@kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=linux-s390@vger.kernel.org \
--cc=linux-sctp@vger.kernel.org \
--cc=lkp@intel.com \
--cc=lkp@lists.01.org \
--cc=lucien.xin@gmail.com \
--cc=marcelo.leitner@gmail.com \
--cc=mptcp@lists.linux.dev \
--cc=netdev@vger.kernel.org \
--cc=oliver.sang@intel.com \
--cc=shakeelb@google.com \
--cc=soheil@google.com \
--cc=ying.huang@intel.com \
--cc=yinxu@redhat.com \
--cc=zhengjun.xing@linux.intel.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).