linux-stable/net/openvswitch
Tonghao Zhang c4b2bf6b4a openvswitch: Optimize operations for OvS flow_stats.
When calling the flow_free() to free the flow, we call many times
(cpu_possible_mask, eg. 128 as default) cpumask_next(). That will
take up our CPU usage if we call the flow_free() frequently.
When we put all packets to userspace via upcall, and OvS will send
them back via netlink to ovs_packet_cmd_execute(will call flow_free).

The test topo is shown as below. VM01 sends TCP packets to VM02,
and OvS forward packtets. When testing, we use perf to report the
system performance.

VM01 --- OvS-VM --- VM02

Without this patch, perf-top show as below: The flow_free() is
3.02% CPU usage.

	4.23%  [kernel]            [k] _raw_spin_unlock_irqrestore
	3.62%  [kernel]            [k] __do_softirq
	3.16%  [kernel]            [k] __memcpy
	3.02%  [kernel]            [k] flow_free
	2.42%  libc-2.17.so        [.] __memcpy_ssse3_back
	2.18%  [kernel]            [k] copy_user_generic_unrolled
	2.17%  [kernel]            [k] find_next_bit

When applied this patch, perf-top show as below: Not shown on
the list anymore.

	4.11%  [kernel]            [k] _raw_spin_unlock_irqrestore
	3.79%  [kernel]            [k] __do_softirq
	3.46%  [kernel]            [k] __memcpy
	2.73%  libc-2.17.so        [.] __memcpy_ssse3_back
	2.25%  [kernel]            [k] copy_user_generic_unrolled
	1.89%  libc-2.17.so        [.] _int_malloc
	1.53%  ovs-vswitchd        [.] xlate_actions

With this patch, the TCP throughput(we dont use Megaflow Cache
+ Microflow Cache) between VMs is 1.18Gbs/sec up to 1.30Gbs/sec
(maybe ~10% performance imporve).

This patch adds cpumask struct, the cpu_used_mask stores the cpu_id
that the flow used. And we only check the flow_stats on the cpu we
used, and it is unncessary to check all possible cpu when getting,
cleaning, and updating the flow_stats. Adding the cpu_used_mask to
sw_flow struct does’t increase the cacheline number.

Signed-off-by: Tonghao Zhang <xiangxia.m.yue@gmail.com>
Acked-by: Pravin B Shelar <pshelar@ovn.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2017-07-19 13:49:39 -07:00
..
actions.c Openvswitch: Refactor sample and recirc actions implementation 2017-03-22 11:28:35 -07:00
conntrack.c netfilter: introduce nf_conntrack_helper_put helper function 2017-05-15 12:42:29 +02:00
conntrack.h openvswitch: Pack struct sw_flow_key. 2017-02-09 22:59:34 -05:00
datapath.c net: Remove all references to SKB_GSO_UDP. 2017-07-17 09:52:58 -07:00
datapath.h openvswitch: fix mis-ordered comment lines for ovs_skb_cb 2017-07-03 05:54:00 -07:00
dp_notify.c openvswitch: fix hangup on vxlan/gre/geneve device deletion 2015-12-03 14:29:25 -05:00
flow.c openvswitch: Optimize operations for OvS flow_stats. 2017-07-19 13:49:39 -07:00
flow.h openvswitch: Optimize operations for OvS flow_stats. 2017-07-19 13:49:39 -07:00
flow_netlink.c net: store port/representator id in metadata_dst 2017-06-25 11:42:01 -04:00
flow_netlink.h openvswitch: Add original direction conntrack tuple to sw_flow_key. 2017-02-09 22:59:34 -05:00
flow_table.c openvswitch: Optimize operations for OvS flow_stats. 2017-07-19 13:49:39 -07:00
flow_table.h openvswitch: Zero flows on allocation. 2015-09-22 17:33:41 -07:00
Kconfig openvswitch: call only into reachable nf-nat code 2016-03-28 17:58:59 +02:00
Makefile openvswitch: Remove conntrack Kconfig option. 2015-09-06 23:48:33 -07:00
vport-geneve.c openvswitch: do not ignore netdev errors when creating tunnel vports 2016-08-10 23:13:23 -07:00
vport-gre.c openvswitch: do not ignore netdev errors when creating tunnel vports 2016-08-10 23:13:23 -07:00
vport-internal_dev.c net: Fix inconsistent teardown and release of private netdev state. 2017-06-07 15:53:24 -04:00
vport-internal_dev.h openvswitch: introduce rtnl ops stub 2014-07-01 14:40:17 -07:00
vport-netdev.c openvswitch: allow L3 netdev ports 2016-11-13 00:51:02 -05:00
vport-netdev.h openvswitch: Use dev_queue_xmit for vport send. 2015-10-22 06:46:16 -07:00
vport-vxlan.c vxlan: get rid of redundant vxlan_dev.flags 2017-06-20 13:37:02 -04:00
vport.c openvswitch: add processing of L3 packets 2016-11-13 00:51:02 -05:00
vport.h openvswitch: pass mac_proto to ovs_vport_send 2016-11-13 00:51:02 -05:00