From mboxrd@z Thu Jan 1 00:00:00 1970 From: Shakeel Butt Subject: [PATCH v2 0/3] memcg: optimize charge codepath Date: Thu, 25 Aug 2022 00:05:03 +0000 Message-ID: <20220825000506.239406-1-shakeelb@google.com> Mime-Version: 1.0 Return-path: DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20210112; h=cc:to:from:subject:mime-version:message-id:date:from:to:cc; bh=vgA35UGCl2OHwAbiBhc19zCa8HdxydG1D9TkQ3os1vo=; b=Ge7brMIDek9qbeisJgfHL6rmBn4FcJKYA9NOFZ0ssAyFWYzyQc4kXS30ZHP2em4Z1h P3TuTjpB+zzcwiH6oROwzlRsWero2l5QxfYmurrc9hJBSB8kUcwtU4kRAIqOTRDxWiI9 FnZosPt+RBbfPjyDFj14FpKbfQ3sV+YnH3jHUhIvfJSXyA0AUA63cdDg11e8xX4/Myfg EqjDHsQ+yqVJdJoUnYtFmb24rYbR/drHoWA/7uWQxfNfnukE7muUDFSfj7q7MxclK3cA TIrcuHrDD3wmMumrUkRreEIKBtZbjTxfTYXMiVtJ9KUxV0mLM3E70ZKhnC+Rx7IbBro5 vwJg== List-ID: Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: Johannes Weiner , Michal Hocko , Roman Gushchin , Muchun Song Cc: =?UTF-8?q?Michal=20Koutn=C3=BD?= , Eric Dumazet , Soheil Hassas Yeganeh , Feng Tang , Oliver Sang , Andrew Morton , 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 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