Message ID | 1571135440-24313-1-git-send-email-xiangxia.m.yue@gmail.com |
---|---|
Headers | show |
Series | optimize openvswitch flow looking up | expand |
From: xiangxia.m.yue@gmail.com Date: Tue, 15 Oct 2019 18:30:30 +0800 > This series patch optimize openvswitch for performance or simplify > codes. ... Pravin and company, please review!
On 10/17/2019 12:22 PM, David Miller wrote: > From: xiangxia.m.yue@gmail.com > Date: Tue, 15 Oct 2019 18:30:30 +0800 > >> This series patch optimize openvswitch for performance or simplify >> codes. > ... > > Pravin and company, please review! Hi Dave, I've tested the patches and provided my 'Tested-by' tag. Pravin has had some suggestions for versions 1-3 of these patches so I'll let him provide the ack on this latest patch set. Thanks, - Greg
On Wed, Oct 16, 2019 at 5:50 AM <xiangxia.m.yue@gmail.com> wrote: > > From: Tonghao Zhang <xiangxia.m.yue@gmail.com> > > This series patch optimize openvswitch for performance or simplify > codes. > > Patch 1, 2, 4: Port Pravin B Shelar patches to > linux upstream with little changes. btw, should we keep Pravin as the author of the above three patches? Regards, William > > Patch 5, 6, 7: Optimize the flow looking up and > simplify the flow hash. > > Patch 8, 9: are bugfix. > > The performance test is on Intel Xeon E5-2630 v4. > The test topology is show as below: > > +-----------------------------------+ > | +---------------------------+ | > | | eth0 ovs-switch eth1 | | Host0 > | +---------------------------+ | > +-----------------------------------+ > ^ | > | | > | | > | | > | v > +-----+----+ +----+-----+ > | netperf | Host1 | netserver| Host2 > +----------+ +----------+ > > We use netperf send the 64B packets, and insert 255+ flow-mask: > $ ovs-dpctl add-flow ovs-switch "in_port(1),eth(dst=00:01:00:00:00:00/ff:ff:ff:ff:ff:01),eth_type(0x0800),ipv4(frag=no)" 2 > ... > $ ovs-dpctl add-flow ovs-switch "in_port(1),eth(dst=00:ff:00:00:00:00/ff:ff:ff:ff:ff:ff),eth_type(0x0800),ipv4(frag=no)" 2 > $ > $ netperf -t UDP_STREAM -H 2.2.2.200 -l 40 -- -m 18 > > * Without series patch, throughput 8.28Mbps > * With series patch, throughput 46.05Mbps > > v3->v4: > access ma->count with READ_ONCE/WRITE_ONCE API. More information, > see patch 5 comments. > > v2->v3: > update ma point when realloc mask_array in patch 5. > > v1->v2: > use kfree_rcu instead of call_rcu > > Tonghao Zhang (10): > net: openvswitch: add flow-mask cache for performance > net: openvswitch: convert mask list in mask array > net: openvswitch: shrink the mask array if necessary > net: openvswitch: optimize flow mask cache hash collision > net: openvswitch: optimize flow-mask looking up > net: openvswitch: simplify the flow_hash > net: openvswitch: add likely in flow_lookup > net: openvswitch: fix possible memleak on destroy flow-table > net: openvswitch: don't unlock mutex when changing the user_features > fails > net: openvswitch: simplify the ovs_dp_cmd_new > > net/openvswitch/datapath.c | 65 +++++---- > net/openvswitch/flow.h | 1 - > net/openvswitch/flow_table.c | 316 +++++++++++++++++++++++++++++++++++++------ > net/openvswitch/flow_table.h | 19 ++- > 4 files changed, 329 insertions(+), 72 deletions(-) > > -- > 1.8.3.1 > > _______________________________________________ > dev mailing list > dev@openvswitch.org > https://mail.openvswitch.org/mailman/listinfo/ovs-dev
On Tue, Oct 22, 2019 at 1:14 AM William Tu <u9012063@gmail.com> wrote: > > On Wed, Oct 16, 2019 at 5:50 AM <xiangxia.m.yue@gmail.com> wrote: > > > > From: Tonghao Zhang <xiangxia.m.yue@gmail.com> > > > > This series patch optimize openvswitch for performance or simplify > > codes. > > > > Patch 1, 2, 4: Port Pravin B Shelar patches to > > linux upstream with little changes. > > btw, should we keep Pravin as the author of the above three patches? Agree, but how i can to that, these patches should be sent by Pravin ? > Regards, > William > > > > > Patch 5, 6, 7: Optimize the flow looking up and > > simplify the flow hash. > > > > Patch 8, 9: are bugfix. > > > > The performance test is on Intel Xeon E5-2630 v4. > > The test topology is show as below: > > > > +-----------------------------------+ > > | +---------------------------+ | > > | | eth0 ovs-switch eth1 | | Host0 > > | +---------------------------+ | > > +-----------------------------------+ > > ^ | > > | | > > | | > > | | > > | v > > +-----+----+ +----+-----+ > > | netperf | Host1 | netserver| Host2 > > +----------+ +----------+ > > > > We use netperf send the 64B packets, and insert 255+ flow-mask: > > $ ovs-dpctl add-flow ovs-switch "in_port(1),eth(dst=00:01:00:00:00:00/ff:ff:ff:ff:ff:01),eth_type(0x0800),ipv4(frag=no)" 2 > > ... > > $ ovs-dpctl add-flow ovs-switch "in_port(1),eth(dst=00:ff:00:00:00:00/ff:ff:ff:ff:ff:ff),eth_type(0x0800),ipv4(frag=no)" 2 > > $ > > $ netperf -t UDP_STREAM -H 2.2.2.200 -l 40 -- -m 18 > > > > * Without series patch, throughput 8.28Mbps > > * With series patch, throughput 46.05Mbps > > > > v3->v4: > > access ma->count with READ_ONCE/WRITE_ONCE API. More information, > > see patch 5 comments. > > > > v2->v3: > > update ma point when realloc mask_array in patch 5. > > > > v1->v2: > > use kfree_rcu instead of call_rcu > > > > Tonghao Zhang (10): > > net: openvswitch: add flow-mask cache for performance > > net: openvswitch: convert mask list in mask array > > net: openvswitch: shrink the mask array if necessary > > net: openvswitch: optimize flow mask cache hash collision > > net: openvswitch: optimize flow-mask looking up > > net: openvswitch: simplify the flow_hash > > net: openvswitch: add likely in flow_lookup > > net: openvswitch: fix possible memleak on destroy flow-table > > net: openvswitch: don't unlock mutex when changing the user_features > > fails > > net: openvswitch: simplify the ovs_dp_cmd_new > > > > net/openvswitch/datapath.c | 65 +++++---- > > net/openvswitch/flow.h | 1 - > > net/openvswitch/flow_table.c | 316 +++++++++++++++++++++++++++++++++++++------ > > net/openvswitch/flow_table.h | 19 ++- > > 4 files changed, 329 insertions(+), 72 deletions(-) > > > > -- > > 1.8.3.1 > > > > _______________________________________________ > > dev mailing list > > dev@openvswitch.org > > https://mail.openvswitch.org/mailman/listinfo/ovs-dev
On Mon, Oct 21, 2019 at 6:16 PM Tonghao Zhang <xiangxia.m.yue@gmail.com> wrote: > > On Tue, Oct 22, 2019 at 1:14 AM William Tu <u9012063@gmail.com> wrote: > > > > On Wed, Oct 16, 2019 at 5:50 AM <xiangxia.m.yue@gmail.com> wrote: > > > > > > From: Tonghao Zhang <xiangxia.m.yue@gmail.com> > > > > > > This series patch optimize openvswitch for performance or simplify > > > codes. > > > > > > Patch 1, 2, 4: Port Pravin B Shelar patches to > > > linux upstream with little changes. > > > > btw, should we keep Pravin as the author of the above three patches? > Agree, but how i can to that, these patches should be sent by Pravin ? you can send the patch, and use git commit --amend --author="" to change author William
From: Tonghao Zhang <xiangxia.m.yue@gmail.com> This series patch optimize openvswitch for performance or simplify codes. Patch 1, 2, 4: Port Pravin B Shelar patches to linux upstream with little changes. Patch 5, 6, 7: Optimize the flow looking up and simplify the flow hash. Patch 8, 9: are bugfix. The performance test is on Intel Xeon E5-2630 v4. The test topology is show as below: +-----------------------------------+ | +---------------------------+ | | | eth0 ovs-switch eth1 | | Host0 | +---------------------------+ | +-----------------------------------+ ^ | | | | | | | | v +-----+----+ +----+-----+ | netperf | Host1 | netserver| Host2 +----------+ +----------+ We use netperf send the 64B packets, and insert 255+ flow-mask: $ ovs-dpctl add-flow ovs-switch "in_port(1),eth(dst=00:01:00:00:00:00/ff:ff:ff:ff:ff:01),eth_type(0x0800),ipv4(frag=no)" 2 ... $ ovs-dpctl add-flow ovs-switch "in_port(1),eth(dst=00:ff:00:00:00:00/ff:ff:ff:ff:ff:ff),eth_type(0x0800),ipv4(frag=no)" 2 $ $ netperf -t UDP_STREAM -H 2.2.2.200 -l 40 -- -m 18 * Without series patch, throughput 8.28Mbps * With series patch, throughput 46.05Mbps v3->v4: access ma->count with READ_ONCE/WRITE_ONCE API. More information, see patch 5 comments. v2->v3: update ma point when realloc mask_array in patch 5. v1->v2: use kfree_rcu instead of call_rcu Tonghao Zhang (10): net: openvswitch: add flow-mask cache for performance net: openvswitch: convert mask list in mask array net: openvswitch: shrink the mask array if necessary net: openvswitch: optimize flow mask cache hash collision net: openvswitch: optimize flow-mask looking up net: openvswitch: simplify the flow_hash net: openvswitch: add likely in flow_lookup net: openvswitch: fix possible memleak on destroy flow-table net: openvswitch: don't unlock mutex when changing the user_features fails net: openvswitch: simplify the ovs_dp_cmd_new net/openvswitch/datapath.c | 65 +++++---- net/openvswitch/flow.h | 1 - net/openvswitch/flow_table.c | 316 +++++++++++++++++++++++++++++++++++++------ net/openvswitch/flow_table.h | 19 ++- 4 files changed, 329 insertions(+), 72 deletions(-)