From: Shakeel Butt <shakeelb@google.com>
To: Johannes Weiner <hannes@cmpxchg.org>,
Michal Hocko <mhocko@kernel.org>,
Roman Gushchin <roman.gushchin@linux.dev>,
Muchun Song <songmuchun@bytedance.com>
Cc: "Michal Koutný" <mkoutny@suse.com>,
"Eric Dumazet" <edumazet@google.com>,
"Soheil Hassas Yeganeh" <soheil@google.com>,
"Feng Tang" <feng.tang@intel.com>,
"Oliver Sang" <oliver.sang@intel.com>,
"Andrew Morton" <akpm@linux-foundation.org>,
lkp@lists.01.org, cgroups@vger.kernel.org, linux-mm@kvack.org,
netdev@vger.kernel.org, linux-kernel@vger.kernel.org,
"Shakeel Butt" <shakeelb@google.com>
Subject: [PATCH v2 0/3] memcg: optimize charge codepath
Date: Thu, 25 Aug 2022 00:05:03 +0000 [thread overview]
Message-ID: <20220825000506.239406-1-shakeelb@google.com> (raw)
Recently Linux networking stack has moved from a very old per socket
pre-charge caching to per-cpu caching to avoid pre-charge fragmentation
and unwarranted OOMs. One impact of this change is that for network
traffic workloads, memcg charging codepath can become a bottleneck. The
kernel test robot has also reported this regression[1]. This patch
series tries to improve the memcg charging for such workloads.
This patch series implement three optimizations:
(A) Reduce atomic ops in page counter update path.
(B) Change layout of struct page_counter to eliminate false sharing
between usage and high.
(C) Increase the memcg charge batch to 64.
To evaluate the impact of these optimizations, on a 72 CPUs machine, we
ran the following workload in root memcg and then compared with scenario
where the workload is run in a three level of cgroup hierarchy with top
level having min and low setup appropriately.
$ netserver -6
# 36 instances of netperf with following params
$ netperf -6 -H ::1 -l 60 -t TCP_SENDFILE -- -m 10K
Results (average throughput of netperf):
1. root memcg 21694.8 Mbps
2. 6.0-rc1 10482.7 Mbps (-51.6%)
3. 6.0-rc1 + (A) 14542.5 Mbps (-32.9%)
4. 6.0-rc1 + (B) 12413.7 Mbps (-42.7%)
5. 6.0-rc1 + (C) 17063.7 Mbps (-21.3%)
6. 6.0-rc1 + (A+B+C) 20120.3 Mbps (-7.2%)
With all three optimizations, the memcg overhead of this workload has
been reduced from 51.6% to just 7.2%.
[1] https://lore.kernel.org/linux-mm/20220619150456.GB34471@xsang-OptiPlex-9020/
Changes since v1:
- Commit message updates
- Instead of explicit padding add align compiler option with struct
Shakeel Butt (3):
mm: page_counter: remove unneeded atomic ops for low/min
mm: page_counter: rearrange struct page_counter fields
memcg: increase MEMCG_CHARGE_BATCH to 64
include/linux/memcontrol.h | 7 ++++---
include/linux/page_counter.h | 34 +++++++++++++++++++++++-----------
mm/page_counter.c | 13 ++++++-------
3 files changed, 33 insertions(+), 21 deletions(-)
--
2.37.1.595.g718a3a8f04-goog
WARNING: multiple messages have this Message-ID (diff)
From: Shakeel Butt <shakeelb@google.com>
To: lkp@lists.01.org
Subject: [PATCH v2 0/3] memcg: optimize charge codepath
Date: Thu, 25 Aug 2022 00:05:03 +0000 [thread overview]
Message-ID: <20220825000506.239406-1-shakeelb@google.com> (raw)
[-- Attachment #1: Type: text/plain, Size: 2079 bytes --]
Recently Linux networking stack has moved from a very old per socket
pre-charge caching to per-cpu caching to avoid pre-charge fragmentation
and unwarranted OOMs. One impact of this change is that for network
traffic workloads, memcg charging codepath can become a bottleneck. The
kernel test robot has also reported this regression[1]. This patch
series tries to improve the memcg charging for such workloads.
This patch series implement three optimizations:
(A) Reduce atomic ops in page counter update path.
(B) Change layout of struct page_counter to eliminate false sharing
between usage and high.
(C) Increase the memcg charge batch to 64.
To evaluate the impact of these optimizations, on a 72 CPUs machine, we
ran the following workload in root memcg and then compared with scenario
where the workload is run in a three level of cgroup hierarchy with top
level having min and low setup appropriately.
$ netserver -6
# 36 instances of netperf with following params
$ netperf -6 -H ::1 -l 60 -t TCP_SENDFILE -- -m 10K
Results (average throughput of netperf):
1. root memcg 21694.8 Mbps
2. 6.0-rc1 10482.7 Mbps (-51.6%)
3. 6.0-rc1 + (A) 14542.5 Mbps (-32.9%)
4. 6.0-rc1 + (B) 12413.7 Mbps (-42.7%)
5. 6.0-rc1 + (C) 17063.7 Mbps (-21.3%)
6. 6.0-rc1 + (A+B+C) 20120.3 Mbps (-7.2%)
With all three optimizations, the memcg overhead of this workload has
been reduced from 51.6% to just 7.2%.
[1] https://lore.kernel.org/linux-mm/20220619150456.GB34471(a)xsang-OptiPlex-9020/
Changes since v1:
- Commit message updates
- Instead of explicit padding add align compiler option with struct
Shakeel Butt (3):
mm: page_counter: remove unneeded atomic ops for low/min
mm: page_counter: rearrange struct page_counter fields
memcg: increase MEMCG_CHARGE_BATCH to 64
include/linux/memcontrol.h | 7 ++++---
include/linux/page_counter.h | 34 +++++++++++++++++++++++-----------
mm/page_counter.c | 13 ++++++-------
3 files changed, 33 insertions(+), 21 deletions(-)
--
2.37.1.595.g718a3a8f04-goog
WARNING: multiple messages have this Message-ID (diff)
From: Shakeel Butt <shakeelb-hpIqsD4AKlfQT0dZR+AlfA@public.gmane.org>
To: Johannes Weiner <hannes-druUgvl0LCNAfugRpC6u6w@public.gmane.org>,
Michal Hocko <mhocko-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>,
Roman Gushchin
<roman.gushchin-fxUVXftIFDnyG1zEObXtfA@public.gmane.org>,
Muchun Song <songmuchun-EC8Uxl6Npydl57MIdRCFDg@public.gmane.org>
Cc: "Michal Koutný" <mkoutny-IBi9RG/b67k@public.gmane.org>,
"Eric Dumazet" <edumazet-hpIqsD4AKlfQT0dZR+AlfA@public.gmane.org>,
"Soheil Hassas Yeganeh"
<soheil-hpIqsD4AKlfQT0dZR+AlfA@public.gmane.org>,
"Feng Tang" <feng.tang-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>,
"Oliver Sang"
<oliver.sang-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>,
"Andrew Morton"
<akpm-de/tnXTf+JLsfHDXvbKv3WD2FQJk+8+b@public.gmane.org>,
lkp-hn68Rpc1hR1g9hUCZPvPmw@public.gmane.org,
cgroups-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
linux-mm-Bw31MaZKKs3YtjvyW6yDsg@public.gmane.org,
netdev-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
"Shakeel Butt" <shakeelb-hpIqsD4AKlfQT0dZR+AlfA@public.gmane.org>
Subject: [PATCH v2 0/3] memcg: optimize charge codepath
Date: Thu, 25 Aug 2022 00:05:03 +0000 [thread overview]
Message-ID: <20220825000506.239406-1-shakeelb@google.com> (raw)
Recently Linux networking stack has moved from a very old per socket
pre-charge caching to per-cpu caching to avoid pre-charge fragmentation
and unwarranted OOMs. One impact of this change is that for network
traffic workloads, memcg charging codepath can become a bottleneck. The
kernel test robot has also reported this regression[1]. This patch
series tries to improve the memcg charging for such workloads.
This patch series implement three optimizations:
(A) Reduce atomic ops in page counter update path.
(B) Change layout of struct page_counter to eliminate false sharing
between usage and high.
(C) Increase the memcg charge batch to 64.
To evaluate the impact of these optimizations, on a 72 CPUs machine, we
ran the following workload in root memcg and then compared with scenario
where the workload is run in a three level of cgroup hierarchy with top
level having min and low setup appropriately.
$ netserver -6
# 36 instances of netperf with following params
$ netperf -6 -H ::1 -l 60 -t TCP_SENDFILE -- -m 10K
Results (average throughput of netperf):
1. root memcg 21694.8 Mbps
2. 6.0-rc1 10482.7 Mbps (-51.6%)
3. 6.0-rc1 + (A) 14542.5 Mbps (-32.9%)
4. 6.0-rc1 + (B) 12413.7 Mbps (-42.7%)
5. 6.0-rc1 + (C) 17063.7 Mbps (-21.3%)
6. 6.0-rc1 + (A+B+C) 20120.3 Mbps (-7.2%)
With all three optimizations, the memcg overhead of this workload has
been reduced from 51.6% to just 7.2%.
[1] https://lore.kernel.org/linux-mm/20220619150456.GB34471@xsang-OptiPlex-9020/
Changes since v1:
- Commit message updates
- Instead of explicit padding add align compiler option with struct
Shakeel Butt (3):
mm: page_counter: remove unneeded atomic ops for low/min
mm: page_counter: rearrange struct page_counter fields
memcg: increase MEMCG_CHARGE_BATCH to 64
include/linux/memcontrol.h | 7 ++++---
include/linux/page_counter.h | 34 +++++++++++++++++++++++-----------
mm/page_counter.c | 13 ++++++-------
3 files changed, 33 insertions(+), 21 deletions(-)
--
2.37.1.595.g718a3a8f04-goog
next reply other threads:[~2022-08-25 0:05 UTC|newest]
Thread overview: 34+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-08-25 0:05 Shakeel Butt [this message]
2022-08-25 0:05 ` [PATCH v2 0/3] memcg: optimize charge codepath Shakeel Butt
2022-08-25 0:05 ` Shakeel Butt
2022-08-25 0:05 ` [PATCH v2 1/3] mm: page_counter: remove unneeded atomic ops for low/min Shakeel Butt
2022-08-25 0:05 ` Shakeel Butt
2022-08-25 6:43 ` Michal Hocko
2022-08-25 6:43 ` Michal Hocko
2022-08-25 6:43 ` Michal Hocko
2022-08-25 0:05 ` [PATCH v2 2/3] mm: page_counter: rearrange struct page_counter fields Shakeel Butt
2022-08-25 0:05 ` Shakeel Butt
2022-08-25 0:33 ` Andrew Morton
2022-08-25 0:33 ` Andrew Morton
2022-08-25 0:33 ` Andrew Morton
2022-08-25 4:41 ` Shakeel Butt
2022-08-25 4:41 ` Shakeel Butt
2022-08-25 4:41 ` Shakeel Butt
2022-08-25 5:21 ` Andrew Morton
2022-08-25 5:21 ` Andrew Morton
2022-08-25 5:21 ` Andrew Morton
2022-08-25 15:24 ` Shakeel Butt
2022-08-25 15:24 ` Shakeel Butt
2022-08-25 15:24 ` Shakeel Butt
2022-08-25 6:47 ` Michal Hocko
2022-08-25 6:47 ` Michal Hocko
2022-08-25 15:25 ` Shakeel Butt
2022-08-25 15:25 ` Shakeel Butt
2022-08-25 15:25 ` Shakeel Butt
2022-08-25 0:05 ` [PATCH v2 3/3] memcg: increase MEMCG_CHARGE_BATCH to 64 Shakeel Butt
2022-08-25 0:05 ` Shakeel Butt
2022-08-25 6:49 ` Michal Hocko
2022-08-25 6:49 ` Michal Hocko
2022-08-25 8:30 ` Muchun Song
2022-08-25 8:30 ` Muchun Song
2022-08-25 8:30 ` Muchun Song
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=20220825000506.239406-1-shakeelb@google.com \
--to=shakeelb@google.com \
--cc=akpm@linux-foundation.org \
--cc=cgroups@vger.kernel.org \
--cc=edumazet@google.com \
--cc=feng.tang@intel.com \
--cc=hannes@cmpxchg.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=lkp@lists.01.org \
--cc=mhocko@kernel.org \
--cc=mkoutny@suse.com \
--cc=netdev@vger.kernel.org \
--cc=oliver.sang@intel.com \
--cc=roman.gushchin@linux.dev \
--cc=soheil@google.com \
--cc=songmuchun@bytedance.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.