Re: [ovs-dev] [PATCH v3 0/2] add port-based ingress policing based packet-per-second rate-limiting

On Thu, Jul 01, 2021 at 09:47:35PM +0200, Ilya Maximets wrote: > On 7/1/21 9:17 PM, Simon Horman wrote: > > On Tue, Jun 29, 2021 at 08:17:04AM -0700, Marcelo Ricardo Leitner wrote: > >> On Wed, Jun 23, 2021 at 03:47:45PM +0200, Simon Horman wrote: > >>> On Wed, Jun 09, 2021 at 11:52:07AM +0200, Sim

Re: [ovs-dev] ovn-northd-ddlog scale issues

On Thu, Jul 01, 2021 at 12:18:12AM -0700, Ben Pfaff wrote: > On Tue, Jun 29, 2021 at 03:32:26PM -0700, Ben Pfaff wrote: > > On Mon, Jun 28, 2021 at 05:40:53PM +0200, Dumitru Ceara wrote: > > > On 5/20/21 5:50 PM, Ben Pfaff wrote: > > > > On Thu, May 20, 2021 at 05:06:26PM +0200, Dumitru Ceara wrote

[ovs-dev] [PATCH ovn] ovn-northd-ddlog: Avoid N**2 blowup for N connected logical routers.

It's easy to implement "connected components" in raw DDlog, but it takes N**2 time and space in the number of elements in a component. This was a huge waste for a test case supplied by Dumitru Ceara that had over 8000 logical routers. This commit solves the problem by using the "graph" transformer

[ovs-dev] [PATCH ovn 4/4] tests: Improve debuggability in gratuitous ARP test.

Checking everything and printing messages as the test proceeds makes it a lot easier to see what's going on. Signed-off-by: Ben Pfaff --- tests/ovn.at | 128 ++- 1 file changed, 66 insertions(+), 62 deletions(-) diff --git a/tests/ovn.at b/tests/o

[ovs-dev] [PATCH ovn 2/4] tests: Fix call to dpctl/del-flows.

This command name was misspelled and didn't work. Nothing checked the exit status, so the problem went unnoticed. Signed-off-by: Ben Pfaff --- tests/ovn.at | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tests/ovn.at b/tests/ovn.at index 722659739627..6b62e34bc0b3 100644 ---

[ovs-dev] [PATCH ovn 3/4] tests: Allow duplicate packets in gratuitous ARP test.

These packets are sent periodically so it's possible that more than one arrives if the test runs slowly. Signed-off-by: Ben Pfaff --- tests/ovn.at | 10 +- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/tests/ovn.at b/tests/ovn.at index 6b62e34bc0b3..cbdea3cabe55 100644 --

[ovs-dev] [PATCH ovn 1/4] tests: Fix "vlan traffic for external network with distributed..."

This fixes the test case "ovn -- vlan traffic for external network with distributed router gateway port", which often failed, especially under ddlog. The problem was that the test case was setting a column that only ovn-northd is supposed to set. The "tag" column in Logical_Switch_Port is for ovn

[ovs-dev] [PATCH ovn 0/4] Fix test "ovn -- vlan traffic for external network with distributed router gateway port"

The first commit is the one-liner that fixes most of the problems. The remaining commits are bonuses. With this and Numan's earlier patch "northd-ddlog: Add proxy arp flows for configured addresses in lsp router port.", I no longer get test failures on my machine. Ben Pfaff (4): tests: Fix "vl

[ovs-dev] [PATCH v2 7/7] tests: Add new test for cross-numa pmd rxq assignments.

Add some tests to ensure that if there are numa local PMDs they are used for polling an rxq. Also check that if there are only numa non-local PMDs they will be used and but the user will be warned. Signed-off-by: Kevin Traynor --- tests/pmd.at | 154 +

[ovs-dev] [PATCH v2 6/7] dpif-netdev: Allow pin rxq and non-isolate PMD.

Pinning an rxq to a PMD with pmd-rxq-affinity may be done for various reasons such as reserving a full PMD for an rxq, or to ensure that multiple rxqs from a port are handled on different PMDs. Previously pmd-rxq-affinity always isolated the PMD so no other rxqs could be assigned to it by OVS. The

[ovs-dev] [PATCH v2 0/7] Rxq scheduling updates.

A series to: - cleanup rxq scheduling and auto load balance - add some new rxq scheduling features NOTE: asan is complaining in some unit tests, I need to investigate this: https://github.com/kevintraynor/ovs/runs/2966525625?check_suite_focus=true#step:13:10256 General v2: - added unit tests - re

[ovs-dev] [PATCH v2 5/7] dpif-netdev: Add group rxq scheduling assignment type.

Add an rxq scheduling option that allows rxqs to be grouped on a pmd based purely on their load. The current default 'cycles' assignment sorts rxqs by measured processing load and then assigns them to a list of round robin PMDs. This helps to keep the rxqs that require most processing on different

[ovs-dev] [PATCH v2 4/7] dpif-netdev: Assign PMD for failed pinned rxqs.

Previously, if pmd-rxq-affinity was used to pin an rxq to a core that was not in pmd-cpu-mask the rxq was not polled and the user received a warning. Now that pinned and non-pinned rxqs are assigned to PMDs in a common call to rxq scheduling, if an invalid core is selected in pmd-rxq-affinity the

[ovs-dev] [PATCH v2 3/7] dpif-netdev: Sort PMD list by core id for rxq scheduling.

The list of PMDs is round robined through for the selection when assigning an rxq to a PMD. The list is based on a hash map, so there is no defined order. It means that in cases where there is no traffic or traffic patterns do not change, the same set of PMDs may get assigned different rxqs for no

[ovs-dev] [PATCH v2 2/7] dpif-netdev: Make PMD auto load balance use common rxq scheduling.

PMD auto load balance had its own separate implementation of the rxq scheduling that it used for dry runs. This was done because previously the rxq scheduling was not reusable for a dry run. Apart from the code duplication (which is a good enough reason to replace it alone) this meant that if any

[ovs-dev] [PATCH v2 1/7] dpif-netdev: Rework rxq scheduling code.

This reworks the current rxq scheduling code to break it into more generic and reusable pieces. The behaviour does not change from a user perspective, except the logs are updated to be more consistent. >From an implementation view, there are some changes with mind to adding functionality and reus

Re: [ovs-dev] [PATCH 0/5] Rxq scheduling updates.

On 24/06/2021 16:42, David Marchand wrote: > On Fri, Jun 4, 2021 at 11:19 PM Kevin Traynor wrote: >> >> The first two patches do not provide new functionality for the user >> (except the logs are reworked). They are reworking to make the >> rxq scheduling and PMD auto load balance code more modula

Re: [ovs-dev] [PATCH 5/5] dpif-netdev: Allow pin rxq and non-isolate PMD.

On 24/06/2021 16:24, David Marchand wrote: > On Fri, Jun 4, 2021 at 11:19 PM Kevin Traynor wrote: >> >> Pinning an rxq to a PMD with pmd-rxq-affinity may be done for >> various reasons such as reserving a full PMD for an rxq, or to >> ensure that multiple rxqs from a port are handled on different

Re: [ovs-dev] [PATCH 4/5] dpif-netdev: Assign PMD for failed pinned rxqs.

On 24/06/2021 16:04, David Marchand wrote: > On Fri, Jun 4, 2021 at 11:19 PM Kevin Traynor wrote: >> >> Previously, if pmd-rxq-affinity was used to pin an rxq to >> a core that was not in pmd-cpu-mask the rxq was not polled >> and the user received a warning. >> >> Now that pinned and non-pinned r

Re: [ovs-dev] [PATCH 4/5] dpif-netdev: Assign PMD for failed pinned rxqs.

On 24/06/2021 08:51, Pai G, Sunil wrote: > Hey Kevin , > > Patch looks good to me. > Builds fine , all test cases here > http://patchwork.ozlabs.org/project/openvswitch/patch/20210316154532.127858-1-ktray...@redhat.com/ > pass as well. > > > # ovs-appctl dpif-netdev/pmd-rxq-show > pmd threa

Re: [ovs-dev] [PATCH 3/5] dpif-netdev: Add group rxq scheduling assignment type.

On 24/06/2021 15:52, David Marchand wrote: > On Fri, Jun 4, 2021 at 11:19 PM Kevin Traynor wrote: >> >> Add an rxq scheduling option that allows rxqs to be grouped >> on a pmd based purely on their load. >> >> The current default 'cycles' assignment sorts rxqs by measured >> processing load and th

Re: [ovs-dev] [PATCH 3/5] dpif-netdev: Add group rxq scheduling assignment type.

On 23/06/2021 16:29, Pai G, Sunil wrote: > Hey Kevin , > > Patch looks good to me. > Builds fine , all test cases here > http://patchwork.ozlabs.org/project/openvswitch/patch/20210316154532.127858-1-ktray...@redhat.com/ > pass as well. > The groups assignment works fine too. > > From vswitchd

Re: [ovs-dev] [PATCH 2/5] dpif-netdev: Make PMD auto load balance use common rxq scheduling.

On 23/06/2021 12:21, Pai G, Sunil wrote: > Hey Kevin , > > Patch looks good to me. > Builds fine , all test cases here > http://patchwork.ozlabs.org/project/openvswitch/patch/20210316154532.127858-1-ktray...@redhat.com/ > pass as well. > > Some minor nits inline : > > > >> +static bool >> +

Re: [ovs-dev] [PATCH 2/5] dpif-netdev: Make PMD auto load balance use common rxq scheduling.

On 24/06/2021 15:30, David Marchand wrote: > On Fri, Jun 4, 2021 at 11:19 PM Kevin Traynor wrote: >> >> PMD auto load balance had its own separate implementation of the >> rxq scheduling that it used for dry runs. This was done because >> previously the rxq scheduling was not reusable for a dry ru

Re: [ovs-dev] [PATCH 1/5] dpif-netdev: Rework rxq scheduling code.

On 22/06/2021 20:02, Pai G, Sunil wrote: > Hey Kevin , > > Patch looks good to me: > > Builds fine , all test cases here > http://patchwork.ozlabs.org/project/openvswitch/patch/20210316154532.127858-1-ktray...@redhat.com/ > pass as well. > > Some minor nits inline : > > > >> diff --git a/

Re: [ovs-dev] [PATCH 1/5] dpif-netdev: Rework rxq scheduling code.

On 23/06/2021 14:04, David Marchand wrote: > On Fri, Jun 4, 2021 at 11:19 PM Kevin Traynor wrote: >> >> This reworks the current rxq scheduling code to break it into more >> generic and reusable pieces. >> >> The behaviour does not change from a user perspective, except the logs >> are updated to

Re: [ovs-dev] [PATCH 3/3] dpif-netlink: Introduce per-cpu upcall dispatch

Hi Mark, One more thing, this seems a relevant change to mention in the NEWS file. Thanks, fbl On Wed, Jun 30, 2021 at 05:56:11AM -0400, Mark Gray wrote: > The Open vSwitch kernel module uses the upcall mechanism to send > packets from kernel space to user space when it misses in the kernel >

Re: [ovs-dev] [PATCH net-next] openvswitch: Introduce per-cpu upcall dispatch

Hi Mark, Thanks for addressing the comments. The patch looks good to me and I plan to test it tomorrow. fbl On Wed, Jun 30, 2021 at 05:53:49AM -0400, Mark Gray wrote: > The Open vSwitch kernel module uses the upcall mechanism to send > packets from kernel space to user space when it misses in

Re: [ovs-dev] [PATCH 3/3] dpif-netlink: Introduce per-cpu upcall dispatch

Hi Mark, I've not tested this yet. See some comments below. On Wed, Jun 30, 2021 at 05:56:11AM -0400, Mark Gray wrote: > The Open vSwitch kernel module uses the upcall mechanism to send > packets from kernel space to user space when it misses in the kernel > space flow table. The upcall sends p

Re: [ovs-dev] [PATCH 2/3] dpif-netlink: fix report_loss() message

On Wed, Jun 30, 2021 at 05:56:10AM -0400, Mark Gray wrote: > Fixes: 1579cf677fcb ("dpif-linux: Implement the API functions to allow > multiple handler threads read upcall.") > Signed-off-by: Mark Gray > --- Acked-by: Flavio Leitner Thanks Mark! fbl __

Re: [ovs-dev] [PATCH 1/3] ofproto: change type of n_handlers and n_revalidators

On Wed, Jun 30, 2021 at 05:56:09AM -0400, Mark Gray wrote: > 'n_handlers' and 'n_revalidators' are declared as type 'size_t'. > However, dpif_handlers_set() requires parameter 'n_handlers' as > type 'uint32_t'. This patch fixes this type mismatch. > > Signed-off-by: Mark Gray > --- Acked-by: Fla

Re: [ovs-dev] [PATCH branch-2.15 0/2] Release patches for v2.15.1.

On 7/1/21 8:41 PM, Aaron Conole wrote: > Ilya Maximets writes: > >> We didn't have stable releases for a while. It's a good time >> to do them now before starting release process for 2.16. >> >> Ilya Maximets (2): >> Set release date for 2.15.1. >> Prepare for 2.15.2. >> >> NEWS

Re: [ovs-dev] [PATCH branch-2.13 0/2] Release patches for v2.13.4.

On 7/1/21 8:41 PM, Aaron Conole wrote: > Ilya Maximets writes: > >> We didn't have stable releases for a while. It's a good time >> to do them now before starting release process for 2.16. >> >> Ilya Maximets (2): >> Set release date for 2.13.4. >> Prepare for 2.13.5. >> >> NEWS

Re: [ovs-dev] [PATCH v3 0/2] add port-based ingress policing based packet-per-second rate-limiting

On 7/1/21 9:17 PM, Simon Horman wrote: > On Tue, Jun 29, 2021 at 08:17:04AM -0700, Marcelo Ricardo Leitner wrote: >> On Wed, Jun 23, 2021 at 03:47:45PM +0200, Simon Horman wrote: >>> On Wed, Jun 09, 2021 at 11:52:07AM +0200, Simon Horman wrote: Hi, this short test adds support for ad

Re: [ovs-dev] [PATCH ovn v2] ovn-controller: Fix port group I-P when they contain non-vif ports.

On Thu, Jul 1, 2021 at 10:18 AM Dumitru Ceara wrote: > > On 7/1/21 6:39 PM, Han Zhou wrote: > > On Thu, Jul 1, 2021 at 6:05 AM Dumitru Ceara wrote: > >> > >> On 7/1/21 2:53 PM, Numan Siddique wrote: > Thanks Dumitru! > Acked-by: Han Zhou > > Not sure if Numan would like to ta

Re: [ovs-dev] [PATCH v3 0/2] add port-based ingress policing based packet-per-second rate-limiting

On Tue, Jun 29, 2021 at 08:17:04AM -0700, Marcelo Ricardo Leitner wrote: > On Wed, Jun 23, 2021 at 03:47:45PM +0200, Simon Horman wrote: > > On Wed, Jun 09, 2021 at 11:52:07AM +0200, Simon Horman wrote: > > > Hi, > > > > > > this short test adds support for add port-based ingress policing based > >

Re: [ovs-dev] [PATCH 0/2] Stop configuring '--socket-mem'/'--socket-limit' by default for DPDK if not requested.

On 01/07/2021 18:46, Ilya Maximets wrote: > On 6/30/21 10:48 PM, Rosemarie O'Riorden wrote: >> From: Rosemarie O'Riorden >> >> Currently, there is a default value of 1024 for socket-mem if not >> configured. socket-limit automatically takes on the value of socket-mem >> unless otherwise specified.

Re: [ovs-dev] [PATCH ovn] add ovn-kubernetes job to github actions

Bleep bloop. Greetings Numan Siddique, I am a robot and I have tried out your patch. Thanks for your contribution. I encountered some error that I wasn't expecting. See the details below. checkpatch: WARNING: Line is 83 characters long (recommended limit is 79) #34 FILE: .ci/ovn-kubernetes/Do

Re: [ovs-dev] [PATCH branch-2.15 0/2] Release patches for v2.15.1.

Ilya Maximets writes: > We didn't have stable releases for a while. It's a good time > to do them now before starting release process for 2.16. > > Ilya Maximets (2): > Set release date for 2.15.1. > Prepare for 2.15.2. > > NEWS | 6 +- > configure.ac | 2 +- > debian/ch

Re: [ovs-dev] [PATCH branch-2.13 0/2] Release patches for v2.13.4.

Ilya Maximets writes: > We didn't have stable releases for a while. It's a good time > to do them now before starting release process for 2.16. > > Ilya Maximets (2): > Set release date for 2.13.4. > Prepare for 2.13.5. > > NEWS | 6 +- > configure.ac | 2 +- > debian/ch

[ovs-dev] [PATCH ovn] add ovn-kubernetes job to github actions

From: Antonio Ojea It allows to specify the ovn-kubernetes commits to test against. Submitted-at: https://github.com/ovn-org/ovn/pull/61 Signed-off-by: Antonio Ojea --- .ci/ovn-kubernetes/Dockerfile| 81 + .github/workflows/ovn-kubernetes.yml | 170

Re: [ovs-dev] [PATCH v4] netlink: removed incorrect optimization

On 6/16/21 11:03 AM, Ilya Maximets wrote: > On 6/16/21 1:30 AM, Ansis wrote: >> On Tue, Jun 15, 2021 at 3:02 PM William Tu wrote: >>> >>> On Tue, Jun 15, 2021 at 11:43 AM Ansis wrote: On Mon, Jun 14, 2021 at 10:22 PM Ansis wrote: > > On Mon, Jun 7, 2021 at 1:31 PM Toms Atteka

Re: [ovs-dev] [PATCH ovn v2] northd-ddlog: Add proxy arp flows for configured addresses in lsp router port.

On Thu, Jul 1, 2021 at 1:45 PM Ben Pfaff wrote: > > On Thu, Jul 01, 2021 at 08:45:21AM -0400, num...@ovn.org wrote: > > From: Numan Siddique > > > > The commit [1] didn't add the ddlog part. > > > > [1] - 8087cbc7462("ovn-northd.c: Add proxy ARP support to OVN") > > > > Signed-off-by: Numan Siddi

[ovs-dev] [PATCH branch-2.15 2/2] Prepare for 2.15.2.

Signed-off-by: Ilya Maximets --- NEWS | 3 +++ configure.ac | 2 +- debian/changelog | 6 ++ 3 files changed, 10 insertions(+), 1 deletion(-) diff --git a/NEWS b/NEWS index 328a50ce3..23a64c155 100644 --- a/NEWS +++ b/NEWS @@ -1,3 +1,6 @@ +v2.15.2 - xx xxx +-

[ovs-dev] [PATCH branch-2.15 0/2] Release patches for v2.15.1.

We didn't have stable releases for a while. It's a good time to do them now before starting release process for 2.16. Ilya Maximets (2): Set release date for 2.15.1. Prepare for 2.15.2. NEWS | 6 +- configure.ac | 2 +- debian/changelog | 8 +++- 3 files changed, 13

[ovs-dev] [PATCH branch-2.15 1/2] Set release date for 2.15.1.

Signed-off-by: Ilya Maximets --- NEWS | 3 ++- debian/changelog | 2 +- 2 files changed, 3 insertions(+), 2 deletions(-) diff --git a/NEWS b/NEWS index d6a9395be..328a50ce3 100644 --- a/NEWS +++ b/NEWS @@ -1,5 +1,6 @@ -v2.15.1 - xx xxx +v2.15.1 - 01 Jul 2021 ---

[ovs-dev] [PATCH branch-2.13 2/2] Prepare for 2.13.5.

Signed-off-by: Ilya Maximets --- NEWS | 3 +++ configure.ac | 2 +- debian/changelog | 6 ++ 3 files changed, 10 insertions(+), 1 deletion(-) diff --git a/NEWS b/NEWS index fea854d78..a1206adda 100644 --- a/NEWS +++ b/NEWS @@ -1,3 +1,6 @@ +v2.13.5 - xx xxx +-

[ovs-dev] [PATCH branch-2.13 0/2] Release patches for v2.13.4.

We didn't have stable releases for a while. It's a good time to do them now before starting release process for 2.16. Ilya Maximets (2): Set release date for 2.13.4. Prepare for 2.13.5. NEWS | 6 +- configure.ac | 2 +- debian/changelog | 8 +++- 3 files changed, 13

[ovs-dev] [PATCH branch-2.13 1/2] Set release date for 2.13.4.

Signed-off-by: Ilya Maximets --- NEWS | 3 ++- debian/changelog | 2 +- 2 files changed, 3 insertions(+), 2 deletions(-) diff --git a/NEWS b/NEWS index 733c785f0..fea854d78 100644 --- a/NEWS +++ b/NEWS @@ -1,5 +1,6 @@ -v2.13.4 - xx xxx +v2.13.4 - 01 Jul 2021 ---

[ovs-dev] [PATCH 1/1] [python] Add cooperative_yield() API method to Idl

When using eventlet monkey_patch()'d code, greenthreads can be blocked on connection for several seconds while the database contents are parsed. Eventlet recommends adding a sleep(0) call to cooperatively yield in cpu-bound code. asyncio code has asyncio.sleep(0). This patch adds an API method tha

[ovs-dev] [PATCH 0/1] [python] Cooperatively yielding to eventlet/gevent

When using python-ovs with eventlet/gevent and presumably asyncio, the code in __do_parse_update() can block due to being cpu-bound and not doing any I/O which would cooperatively yield to other greenthreads. The code below can show the issue. You'll notice when using the idl.Idl class, that after

Re: [ovs-dev] [PATCH 0/2] Stop configuring '--socket-mem'/'--socket-limit' by default for DPDK if not requested.

On 6/30/21 10:48 PM, Rosemarie O'Riorden wrote: > From: Rosemarie O'Riorden > > Currently, there is a default value of 1024 for socket-mem if not > configured. socket-limit automatically takes on the value of socket-mem > unless otherwise specified. With these changes, memory allocation will > be

Re: [ovs-dev] [PATCH ovn v2] northd-ddlog: Add proxy arp flows for configured addresses in lsp router port.

On Thu, Jul 01, 2021 at 08:45:21AM -0400, num...@ovn.org wrote: > From: Numan Siddique > > The commit [1] didn't add the ddlog part. > > [1] - 8087cbc7462("ovn-northd.c: Add proxy ARP support to OVN") > > Signed-off-by: Numan Siddique This fixes test failures for me. Thank you! Acked-by: Be

Re: [ovs-dev] [v4 02/12] dpif-netdev: Add auto validation function for miniflow extract

> -Original Message- > From: Eelco Chaudron > Sent: Thursday, July 1, 2021 5:19 PM > To: Van Haaren, Harry > Cc: Amber, Kumar ; d...@openvswitch.org; > i.maxim...@ovn.org; Flavio Leitner ; Stokes, Ian > > Subject: Re: [ovs-dev] [v4 02/12] dpif-netdev: Add auto validation function > for

Re: [ovs-dev] [PATCH ovn v2] ovn-controller: Fix port group I-P when they contain non-vif ports.

On 7/1/21 6:39 PM, Han Zhou wrote: > On Thu, Jul 1, 2021 at 6:05 AM Dumitru Ceara wrote: >> >> On 7/1/21 2:53 PM, Numan Siddique wrote: Thanks Dumitru! Acked-by: Han Zhou Not sure if Numan would like to take a second look as well, so let's > wait for one or two days befor

Re: [ovs-dev] [PATCH V7 00/13] Netdev vxlan-decap offload

> -Original Message- > From: Ilya Maximets > Sent: Thursday, July 1, 2021 11:32 AM > To: Van Haaren, Harry ; Ilya Maximets > > Cc: Eli Britstein ; ovs dev ; Ivan > Malov > ; Majd Dibbiny ; Stokes, Ian > ; Ferriter, Cian ; Ben Pfaff > ; Balazs Nemeth ; Sriharsha Basavapatna > > Subject:

Re: [ovs-dev] [PATCH] dpif-netdev: Report overhead busy cycles per pmd.

On Fri, Jun 11, 2021 at 2:02 PM Kevin Traynor wrote: > > Hi David, > > A few comments below, Thanks for the review and comments, all seemed fine to me. Retested with my setup and I just posted v2 on the ml. -- David Marchand ___ dev mailing list d..

[ovs-dev] [PATCH v2] dpif-netdev: Report overhead busy cycles per pmd.

Users complained that per rxq pmd usage was confusing: summing those values per pmd would never reach 100% even if increasing traffic load beyond pmd capacity. This is because the dpif-netdev/pmd-rxq-show command only reports "pure" rxq cycles while some cycles are used in the pmd mainloop and add

Re: [ovs-dev] [PATCH ovn v2] ovn-controller: Fix port group I-P when they contain non-vif ports.

On Thu, Jul 1, 2021 at 6:05 AM Dumitru Ceara wrote: > > On 7/1/21 2:53 PM, Numan Siddique wrote: > >> Thanks Dumitru! > >> Acked-by: Han Zhou > >> > >> Not sure if Numan would like to take a second look as well, so let's wait > >> for one or two days before merging. > > Thanks Dumitru and Han. >

Re: [ovs-dev] [PATCH V7 13/13] netdev-dpdk-offload: Add vxlan pattern matching function.

> -Original Message- > From: dev On Behalf Of Eli Britstein > Sent: Wednesday, June 23, 2021 4:53 PM > To: d...@openvswitch.org; Ilya Maximets > Cc: Eli Britstein ; Ivan Malov ; > Majd > Dibbiny > Subject: [ovs-dev] [PATCH V7 13/13] netdev-dpdk-offload: Add vxlan pattern > matching func

Re: [ovs-dev] [v4 02/12] dpif-netdev: Add auto validation function for miniflow extract

On 29 Jun 2021, at 13:05, Van Haaren, Harry wrote: > Hi Eelco, > > Would you describe the actual test being run below? > > I'm having a hard time figuring out what the actual datapath packet flow is. > It seems strange > that MFEX optimizations are affected by flow-count, that doesn't really >

Re: [ovs-dev] [v14 04/11] dpif-avx512: Add ISA implementation of dpif.

Bleep bloop. Greetings Cian Ferriter, I am a robot and I have tried out your patch. Thanks for your contribution. I encountered some error that I wasn't expecting. See the details below. checkpatch: WARNING: Line lacks whitespace around operator #224 FILE: lib/dpif-netdev-avx512.c:146:

Re: [ovs-dev] [v14 01/11] dpif-netdev: Refactor to multiple header files.

Bleep bloop. Greetings Cian Ferriter, I am a robot and I have tried out your patch. Thanks for your contribution. I encountered some error that I wasn't expecting. See the details below. checkpatch: ERROR: Inappropriate bracing around statement #530 FILE: lib/dpif-netdev-private-dpcls.h:97:

Re: [ovs-dev] [PATCH v2] conntrack: add coverage counters for L3 bad checksum

On 4/30/21 7:12 PM, Paolo Valerio wrote: > similarly to what already exists for L4, add conntrack_l3csum_err > and ipf_l3csum_err for L3. > > Received packets with L3 bad checksum will increase respectively > ipf_l3csum_err if they are fragments and conntrack_l3csum_err > otherwise. > > Although

[ovs-dev] [v14 11/11] dpcls-avx512: Enable avx512 vector popcount instruction.

From: Harry van Haaren This commit enables the AVX512-VPOPCNTDQ Vector Popcount instruction. This instruction is not available on every CPU that supports the AVX512-F Foundation ISA, hence it is enabled only when the additional VPOPCNTDQ ISA check is passed. The vector popcount instruction is us

[ovs-dev] [v14 10/11] dpdk: Cache result of CPU ISA checks.

From: Harry van Haaren As a small optimization, this patch caches the result of a CPU ISA check from DPDK. Particularly in the case of running the DPCLS autovalidator (which repeatedly probes subtables) this reduces the amount of CPU ISA lookups from the DPDK level. By caching them at the OVS/dp

[ovs-dev] [v14 09/11] dpif-netdev/dpcls: Specialize more subtable signatures.

From: Harry van Haaren This commit adds more subtables to be specialized. The traffic pattern here being matched is VXLAN traffic subtables, which commonly have (5,3), (9,1) and (9,4) subtable fingerprints. Signed-off-by: Harry van Haaren Acked-by: Flavio Leitner --- v14: - Added Flavio's Ac

[ovs-dev] [v14 08/11] dpif-netdev/dpcls-avx512: Enable 16 block processing.

From: Harry van Haaren This commit implements larger subtable searches in avx512. A limitation of the previous implementation was that up to 8 blocks of miniflow data could be matched on (so a subtable with 8 blocks was handled in avx, but 9 blocks or more would fall back to scalar/generic). This

[ovs-dev] [v14 07/11] dpif-netdev: Add a partial HWOL PMD statistic.

It is possible for packets traversing the userspace datapath to match a flow before hitting on EMC by using a mark ID provided by a NIC. Add a PMD statistic for this hit. Signed-off-by: Cian Ferriter Acked-by: Flavio Leitner --- Cc: Gaetan Rivet Cc: Sriharsha Basavapatna v14: - Added Flavio

[ovs-dev] [v14 06/11] dpif-netdev: Add command to get dpif implementations.

From: Harry van Haaren This commit adds a new command to retrieve the list of available DPIF implementations. This can be used by to check what implementations of the DPIF are available in any given OVS binary. It also returns which implementations are in use by the OVS PMD threads. Usage: $ ov

[ovs-dev] [v14 05/11] dpif-netdev: Add command to switch dpif implementation.

From: Harry van Haaren This commit adds a new command to allow the user to switch the active DPIF implementation at runtime. A probe function is executed before switching the DPIF implementation, to ensure the CPU is capable of running the ISA required. For example, the below code will switch to

[ovs-dev] [v14 04/11] dpif-avx512: Add ISA implementation of dpif.

From: Harry van Haaren This commit adds the AVX512 implementation of DPIF functionality, specifically the dp_netdev_input_outer_avx512 function. This function only handles outer (no re-circulations), and is optimized to use the AVX512 ISA for packet batching and other DPIF work. Sparse is not ab

[ovs-dev] [v14 02/11] dpif-netdev: Split HWOL out to own header file.

From: Harry van Haaren This commit moves the datapath lookup functions required for hardware offload to a separate file. This allows other DPIF implementations to access the lookup functions, encouraging code reuse. Signed-off-by: Harry van Haaren --- Cc: Gaetan Rivet Cc: Sriharsha Basavapat

[ovs-dev] [v14 03/11] dpif-netdev: Add function pointer for netdev input.

From: Harry van Haaren This commit adds a function pointer to the pmd thread data structure, giving the pmd thread flexibility in its dpif-input function choice. This allows choosing of the implementation based on ISA capabilities of the runtime CPU, leading to optimizations and higher performanc

[ovs-dev] [v14 01/11] dpif-netdev: Refactor to multiple header files.

From: Harry van Haaren Split the very large file dpif-netdev.c and the datastructures it contains into multiple header files. Each header file is responsible for the datastructures of that component. This logical split allows better reuse and modularity of the code, and reduces the very large fi

[ovs-dev] [v14 00/11] DPIF Framework + Optimizations

v14 Summary: - Prefetch 2 packets ahead when processing in AVX512 DPIF. This was found to perform best when testing. - Make stores to a PMDs DPIF function pointer atomic. - Update AVX512 DPIF to use the latest PHWOL implementation. This changed in the scalar DPIF from v13. - Change CMD names to

Re: [ovs-dev] [PATCH v2] checkpatch: Ignore macro definitions of FOR_EACH

On 6/25/21 7:18 PM, Aaron Conole wrote: > When defining a FOR_EACH macro, checkpatch freaks out and generates a > control block whitespace error. Create an exception so that it doesn't > generate errors for this case. > > Reported-at: > https://mail.openvswitch.org/pipermail/ovs-dev/2020-August/

Re: [ovs-dev] [PATCH v2] dpif-netdev: apply subtable-lookup-prio-set on any datapath

On 6/24/21 11:18 AM, Van Haaren, Harry wrote: >> -Original Message- >> From: Timothy Redaelli >> Sent: Wednesday, June 23, 2021 11:10 PM >> To: d...@openvswitch.org >> Cc: Van Haaren, Harry >> Subject: [PATCH v2] dpif-netdev: apply subtable-lookup-prio-set on any >> datapath >> >> Curren

Re: [ovs-dev] [PATCH] ovs-actions.xml: add missing bracket

On 5/26/21 12:26 AM, Paolo Valerio wrote: > a bracket is apparently missing in ovs-actions(7). > > The patch changes the relevant row from: > > ct(argument]...) > > to: > > ct([argument]...) > > Signed-off-by: Paolo Valerio > --- Thanks! Applied. Backported down to 2.13. Best regards, Ily

Re: [ovs-dev] [PATCH] netdev-offload-tc: use nl_msg_put_flag for OVS_TUNNEL_KEY_ATTR_CSUM

On 4/7/21 2:32 PM, Eelco Chaudron wrote: > > > On 7 Apr 2021, at 13:43, Paolo Valerio wrote: > >> when a tunnel port gets added to the bridge setting the checksum option >> to true: >> >> ovs-vsctl add-port br0 geneve0 -- set interface geneve0 type=geneve >> options:remote_ip= options:key= opti

Re: [ovs-dev] [PATCH] conntrack: increment coverage counter for all bad checksum cases

On 6/23/21 3:44 PM, Tonghao Zhang wrote: > On Wed, Jun 23, 2021 at 1:01 AM Paolo Valerio wrote: >> >> Paolo Valerio writes: >> >>> conntrack_l4csum_err gets incremented only when corrupted icmp >>> pass through conntrack. >>> Increase it for the remaining bad checksum cases including when >>> che

Re: [ovs-dev] [PATCH ovn v4] ovn-controller: Fix incremental processing for logical port references.

On Tue, Jun 29, 2021 at 3:23 PM Han Zhou wrote: > > If a lflow has an lport name in the match, but when the lflow is > processed the port-binding is not seen by ovn-controller, the > corresponding openflow will not be created. Later if the port-binding is > created/monitored by ovn-controller, the

Re: [ovs-dev] [PATCH] dpif/dpcls: limit count subtable search info logs

> -Original Message- > From: Flavio Leitner > Sent: Thursday, July 1, 2021 2:55 PM > To: Van Haaren, Harry > Cc: Amber, Kumar ; d...@openvswitch.org; > i.maxim...@ovn.org > Subject: Re: [ovs-dev] [PATCH] dpif/dpcls: limit count subtable search info > logs > > On Thu, Jul 01, 2021 at 12:

Re: [ovs-dev] [PATCH] dpif/dpcls: limit count subtable search info logs

On Thu, Jul 01, 2021 at 12:12:18PM +, Van Haaren, Harry wrote: > > -Original Message- > > From: Flavio Leitner > > Sent: Thursday, July 1, 2021 12:15 PM > > To: Van Haaren, Harry > > Cc: Amber, Kumar ; d...@openvswitch.org; > > i.maxim...@ovn.org > > Subject: Re: [ovs-dev] [PATCH] dpi

Re: [ovs-dev] [v4 10/12] dpif-netdev/mfex: Add AVX512 based optimized miniflow extract

On Wed, Jun 30, 2021 at 04:54:22PM +, Van Haaren, Harry wrote: > > -Original Message- > > From: Eelco Chaudron > > Sent: Wednesday, June 30, 2021 3:35 PM > > To: Van Haaren, Harry > > Cc: Amber, Kumar ; d...@openvswitch.org; > > i.maxim...@ovn.org; Flavio Leitner ; Stokes, Ian > > >

Re: [ovs-dev] [PATCH ovn v2] ovn-controller: Fix port group I-P when they contain non-vif ports.

On 7/1/21 2:53 PM, Numan Siddique wrote: >> Thanks Dumitru! >> Acked-by: Han Zhou >> >> Not sure if Numan would like to take a second look as well, so let's wait >> for one or two days before merging. > Thanks Dumitru and Han. > > I looked at it and LGTM. I applied the patch to the main branch.

Re: [ovs-dev] [PATCH ovn v2] northd-ddlog: Add proxy arp flows for configured addresses in lsp router port.

Bleep bloop. Greetings Numan Siddique, I am a robot and I have tried out your patch. Thanks for your contribution. I encountered some error that I wasn't expecting. See the details below. checkpatch: WARNING: Line is 85 characters long (recommended limit is 79) #37 FILE: northd/ovn.rs:187: pu

Re: [ovs-dev] [PATCH ovn v2] ovn-controller: Fix port group I-P when they contain non-vif ports.

On Wed, Jun 30, 2021 at 7:04 PM Han Zhou wrote: > > On Wed, Jun 30, 2021 at 7:01 AM Dumitru Ceara wrote: > > > > It's valid that port_groups contain non-vif ports, they can actually > > contain any type of logical_switch_port. > > > > Also, there's no need to allocate a new sset containing the lo

Re: [ovs-dev] ovn-northd-ddlog scale issues

On Thu, Jul 1, 2021 at 4:31 AM Dumitru Ceara wrote: > > On 7/1/21 9:18 AM, Ben Pfaff wrote: > > On Tue, Jun 29, 2021 at 03:32:26PM -0700, Ben Pfaff wrote: > >> On Mon, Jun 28, 2021 at 05:40:53PM +0200, Dumitru Ceara wrote: > >>> On 5/20/21 5:50 PM, Ben Pfaff wrote: > On Thu, May 20, 2021 at 0

[ovs-dev] [PATCH ovn v2] northd-ddlog: Add proxy arp flows for configured addresses in lsp router port.

From: Numan Siddique The commit [1] didn't add the ddlog part. [1] - 8087cbc7462("ovn-northd.c: Add proxy ARP support to OVN") Signed-off-by: Numan Siddique --- northd/ovn.dl| 1 + northd/ovn.rs| 13 + northd/ovn_northd.dl | 38

[ovs-dev] [PATCH] ovs-save : fix "ovs-ofctl dump-tlv-map" command with right ofp_version

"ovs-ofctl dump-tlv-map ${bridge}" command need right ofp_version. Signed-off-by: linhuang --- utilities/ovs-save | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/utilities/ovs-save b/utilities/ovs-save index 7d810cb40..069c5932e 100755 --- a/utilities/ovs-save +++ b/utilities

Re: [ovs-dev] [PATCH] dpif/dpcls: limit count subtable search info logs

> -Original Message- > From: Flavio Leitner > Sent: Thursday, July 1, 2021 12:15 PM > To: Van Haaren, Harry > Cc: Amber, Kumar ; d...@openvswitch.org; > i.maxim...@ovn.org > Subject: Re: [ovs-dev] [PATCH] dpif/dpcls: limit count subtable search info > logs > > On Thu, Jul 01, 2021 at 10

Re: [ovs-dev] [PATCH] dpif/dpcls: limit count subtable search info logs

On Thu, Jul 01, 2021 at 10:55:55AM +, Van Haaren, Harry wrote: > > -Original Message- > > From: Amber, Kumar > > Sent: Thursday, July 1, 2021 10:50 AM > > To: Van Haaren, Harry > > Cc: d...@openvswitch.org; i.maxim...@ovn.org; Flavio Leitner > > > > Subject: RE: [ovs-dev] [PATCH] dpi

Re: [ovs-dev] [PATCH] dpif/dpcls: limit count subtable search info logs

> -Original Message- > From: Amber, Kumar > Sent: Thursday, July 1, 2021 10:50 AM > To: Van Haaren, Harry > Cc: d...@openvswitch.org; i.maxim...@ovn.org; Flavio Leitner > > Subject: RE: [ovs-dev] [PATCH] dpif/dpcls: limit count subtable search info > logs > > Hi Harry, > > Any Insight

Re: [ovs-dev] [PATCH V7 00/13] Netdev vxlan-decap offload

On 6/29/21 1:53 PM, Van Haaren, Harry wrote: >> -Original Message- >> From: Ilya Maximets >> Sent: Monday, June 28, 2021 3:33 PM >> To: Van Haaren, Harry ; Ilya Maximets >> ; Sriharsha Basavapatna >> >> Cc: Eli Britstein ; ovs dev ; Ivan >> Malov >> ; Majd Dibbiny ; Stokes, Ian >> ; Ferr

Re: [ovs-dev] [PATCH] dpif/dpcls: limit count subtable search info logs

Hi Harry, Any Insights on the following 😊 > -Original Message- > From: Flavio Leitner > Sent: Wednesday, June 30, 2021 12:28 AM > To: Amber, Kumar > Cc: d...@openvswitch.org; i.maxim...@ovn.org > Subject: Re: [ovs-dev] [PATCH] dpif/dpcls: limit count subtable search info > logs > > On

Re: [ovs-dev] [v4 11/12] dpif-netdev/mfex: add more AVX512 traffic profiles

> -Original Message- > From: dev On Behalf Of Amber, Kumar > Sent: Thursday, July 1, 2021 9:43 AM > To: Flavio Leitner > Cc: d...@openvswitch.org; i.maxim...@ovn.org > Subject: Re: [ovs-dev] [v4 11/12] dpif-netdev/mfex: add more AVX512 traffic > profiles > > Hi Flavio, > > > +static v

Re: [ovs-dev] [ovn] howto? L3 support in HW VTEP

I looked at the patch and couldn’t understand how to use it in my case. Can you clarify your idea please? Regards, Vladislav Odintsov > On 28 Jun 2021, at 23:34, Numan Siddique wrote: > > On Fri, Jun 25, 2021 at 4:13 AM Vladislav Odintsov > wrote: >> >> Hi Numan, >>

Re: [ovs-dev] [v4 11/12] dpif-netdev/mfex: add more AVX512 traffic profiles

Hi Flavio, Pls find the replies inline, > > +mfex_vlan_pcp(const uint8_t vlan_pcp, uint64_t *block) { > > +/* Bitwise-OR in the CFI flag, keeping other data the same. */ > > +uint8_t *cfi_byte = (uint8_t *) block; > > +cfi_byte[2] = 0x10 | vlan_pcp; > > Trying to reduce the magic n

Re: [ovs-dev] ovn-northd-ddlog scale issues

On 7/1/21 9:18 AM, Ben Pfaff wrote: > On Tue, Jun 29, 2021 at 03:32:26PM -0700, Ben Pfaff wrote: >> On Mon, Jun 28, 2021 at 05:40:53PM +0200, Dumitru Ceara wrote: >>> On 5/20/21 5:50 PM, Ben Pfaff wrote: On Thu, May 20, 2021 at 05:06:26PM +0200, Dumitru Ceara wrote: > On 4/7/21 6:49 PM, Be

  1   2   >