Re: [ovs-dev] [OVN patch v4 0/9] OVN Release Policy Updates

On Wed, Jun 7, 2023 at 12:47 PM Mark Michelson wrote: > > The current state of development in OVN has resulted in many discussions > happening about how things should change. The main points of the > discussions have been: > > * Recent releases have been much more bare than releases made two or >

Re: [ovs-dev] [OVN patch v4 3/9] release-policy: Change "release" to "branch" in LTS section.

On Wed, Jun 7, 2023 at 12:47 PM Mark Michelson wrote: > > Upcoming changes to the documentation will add documentation for > standard term support branches and will document a policy for when > releases are made for these branches. Our current use of "release" in > the documentation is interchange

Re: [ovs-dev] [PATCH ovn v2 06/14] northd: Track lsp names referenced by LB health-check.

On Wed, Jun 7, 2023 at 2:05 PM Numan Siddique wrote: > > On Fri, Jun 2, 2023 at 12:12 AM Han Zhou wrote: > > > > Maintain a set of lsp names that are referenced in the column > > "ip_port_mappings" of load_balancer records (used for health check > > purpose), which will be used for lsp incrementa

Re: [ovs-dev] [PATCH ovn v2 07/14] northd: Incremental processing of VIF changes in 'northd' node.

On Wed, Jun 7, 2023 at 12:39 PM Numan Siddique wrote: > > On Fri, Jun 2, 2023 at 12:13 AM Han Zhou wrote: > > > > This patch introduces a change handler for NB logical_switch within the > > 'northd' node. It specifically handles cases where logical switch ports > > in the tracked logical switches

Re: [ovs-dev] [PATCH] ovsdb: monitor: Destroy initial change set when new columns added.

On Wed, Jun 7, 2023 at 6:07 AM Ilya Maximets wrote: > > Initial change set is preserved for as long as the monitor itself. > However, if a new client has a condition on a column that is not > one of the monitored columns, this column will be added to the > monitor via ovsdb_monitor_condition_bind(

Re: [ovs-dev] [PATCH v7] tc: fix crash on malformed reply from kernel.

On Wed, Jun 7, 2023 at 11:59 PM Ilya Maximets wrote: > > On 6/6/23 20:33, Frode Nordahl wrote: > > The tc module combines the use of the `tc_transact` helper > > function for communication with the in-kernel tc infrastructure > > with assertions on the reply data by `ofpbuf_at_assert` on the > > r

Re: [ovs-dev] [PATCH] cpu: Fix cpuid check for some AMD processors.

On 6/7/23 10:24, David Marchand wrote: > Some venerable AMD processors do not support querying extended features > (EAX=7) with cpuid. > In this case, it is not a programmatic error and the runtime check should > simply return the isa is unsupported. > > Reported-by: Davide Repetto > Reported-at:

Re: [ovs-dev] [PATCH v7] tc: fix crash on malformed reply from kernel.

On 6/6/23 20:33, Frode Nordahl wrote: > The tc module combines the use of the `tc_transact` helper > function for communication with the in-kernel tc infrastructure > with assertions on the reply data by `ofpbuf_at_assert` on the > received data prior to further processing. > > With the presence o

Re: [ovs-dev] [PATCH ovn v2 06/14] northd: Track lsp names referenced by LB health-check.

On Fri, Jun 2, 2023 at 12:12 AM Han Zhou wrote: > > Maintain a set of lsp names that are referenced in the column > "ip_port_mappings" of load_balancer records (used for health check > purpose), which will be used for lsp incremental processing. > > Signed-off-by: Han Zhou > Reviewed-by: Ales Mus

Re: [ovs-dev] [PATCH ovn 0/5] Add FDB aging mechanism

Thanks Ales, Acked-by: Mark Michelson On 5/18/23 07:32, Ales Musil wrote: Add FDB mechanism that is very similar to already existing MAc binding aging mechanism. There are two new config options to achieve that "fdb_age_threshold" that si configurable per logical switch and "fdb_removal_limit

Re: [ovs-dev] [PATCH ovn v2 07/14] northd: Incremental processing of VIF changes in 'northd' node.

On Fri, Jun 2, 2023 at 12:13 AM Han Zhou wrote: > > This patch introduces a change handler for NB logical_switch within the > 'northd' node. It specifically handles cases where logical switch ports > in the tracked logical switches are changed (added/updated/deleted). > Only regular logical switch

[ovs-dev] [OVN patch v4 1/9] release-process: Use more accurate example releases.

We don't use a four-digit major version, just the final two digits of the year. Signed-off-by: Mark Michelson Acked-by: Dumitru Ceara --- Documentation/internals/release-process.rst | 10 +- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/Documentation/internals/release-pr

[ovs-dev] [OVN patch v4 9/9] backporting-patches: Add concrete policy for where backports go.

This is based on discussions that have happened both on the ovs-dev mailing list as well as public IRC developer meetings. Signed-off-by: Mark Michelson --- .../contributing/backporting-patches.rst | 82 +++ 1 file changed, 82 insertions(+) diff --git a/Documentation/intern

[ovs-dev] [OVN patch v4 8/9] backporting-patches: Fix inaccuracies in the document.

This document appears to be based on the OVS document, but things are done slightly differently in OVN. * OVN is 100% userspace, so there is no need to clarify "userspace" anywhere. * The commit message example is not how we do things in OVN. Backport commit messages are typically identical to

[ovs-dev] [OVN patch v4 3/9] release-policy: Change "release" to "branch" in LTS section.

Upcoming changes to the documentation will add documentation for standard term support branches and will document a policy for when releases are made for these branches. Our current use of "release" in the documentation is interchangeable with "branch" but upcoming changes will differentiate these

[ovs-dev] [OVN patch v4 7/9] Documentation: Change "master" to "main".

OVN hasn't used the name "master" for quite a while, and it looks like we missed some documents when the change to "main" was made. Signed-off-by: Mark Michelson --- .../internals/committer-grant-revocation.rst | 2 +- .../internals/committer-responsibilities.rst | 8

[ovs-dev] [OVN patch v4 6/9] release-policy: Document when versions receive new releases.

The OVN team has had no concrete policy regarding when releases are made for a given version. In the past, new releases have been made only when something critical has been found, meaning that most bug fixes applied to a given OVN version never end up in a release. This change proposes a regular r

[ovs-dev] [OVN patch v4 2/9] release-process: Switch to two releases per year.

Based on discussions that have occurred on the development mailing list and public IRC meetings, we have decided to reduce OVN to having two releases per year instead of four. When OVN initially spun off from OVS, the amount of feature requests and performance improvements were happening at a rapi

[ovs-dev] [OVN patch v4 0/9] OVN Release Policy Updates

The current state of development in OVN has resulted in many discussions happening about how things should change. The main points of the discussions have been: * Recent releases have been much more bare than releases made two or three years ago. * Supporting as many versions of OVN as we have i

[ovs-dev] [OVN patch v4 5/9] release-process: Add section for standard-term support releases.

Signed-off-by: Mark Michelson Acked-by: Dumitru Ceara --- Documentation/internals/release-process.rst | 16 1 file changed, 16 insertions(+) diff --git a/Documentation/internals/release-process.rst b/Documentation/internals/release-process.rst index 37744e3c4..24b5fc060 100644

[ovs-dev] [OVN patch v4 4/9] release-process: State release policy for LTS versions.

The definition of "regular releases" is being kept purposely vague in this commit because a later commit will define what the interval is for these releases. Signed-off-by: Mark Michelson Acked-by: Dumitru Ceara --- Documentation/internals/release-process.rst | 13 +++-- 1 file changed,

Re: [ovs-dev] [PATCH ovn] call ovsrcu_exit() before exit in ovn-northd and ovn-controller to make valgrind happy

Thanks, Igor. Acked-by: Mark Michelson On 6/6/23 20:37, Igor Zhukov wrote: From: Igor Zhukov You can check logs by running: make check-valgrind TESTSUITEFLAGS="246" (Actually almost every test affected but for example we need one test) Valgrind message looks like ==65437== 304 bytes in 1

Re: [ovs-dev] [PATCH ovn 2/2] binding: fixed qos when no iface

> If iface does not exist when qos I+P is run the first time > a qos queue is still created and qos I+P will never apply qos to OVS > qos might still be applied later, if/when (for any reason) runtime_date must > be recomputed > > Fixes: 7d1d111ff213 ("controller: configure qos through ovs qos t

Re: [ovs-dev] [PATCH ovn 1/2] qos: fix potential double deletion of ovs idl row

> If an interface with an qos option is deleted at the same > time as an ofport notification from ovs (causing runtime_data recompute) > is received, the binding module was trying to delete twice the same qos > queue, causing ovs to raise an exception. Hi Xavier, thx for fixing it. Reviewed-by:

[ovs-dev] [PATCH ovn 1/2] qos: fix potential double deletion of ovs idl row

If an interface with an qos option is deleted at the same time as an ofport notification from ovs (causing runtime_data recompute) is received, the binding module was trying to delete twice the same qos queue, causing ovs to raise an exception. Reported-at: https://bugzilla.redhat.com/show_bug.cgi

[ovs-dev] [PATCH ovn 2/2] binding: fixed qos when no iface

If iface does not exist when qos I+P is run the first time a qos queue is still created and qos I+P will never apply qos to OVS qos might still be applied later, if/when (for any reason) runtime_date must be recomputed Fixes: 7d1d111ff213 ("controller: configure qos through ovs qos table and do

[ovs-dev] [PATCH ovn v2] controller: Turn OFTABLE_OUTPUT_INIT into an alias.

Make it alias OFTABLE_OUTPUT_LARGE_PKT_DETECT in order to avoid always doing an extra resubmit in slow path. The OFTABLE_OUTPUT_INIT table didn't have more than the "advance to the next table" default rule. Fixes: 2d6e586e6777 ("Implement MTU Path Discovery for multichassis ports") Fixes: 740f23c

Re: [ovs-dev] [PATCH ovn v3] northd: Add logical flow to skip GARP with LLA

On 6/7/23 16:53, Ales Musil wrote: > > > On Wed, Jun 7, 2023 at 4:48 PM Dumitru Ceara > wrote: > > On 6/7/23 14:41, Ihar Hrachyshka wrote: > > Thank you Ales! > > > > Acked-By: Ihar Hrachyshka > > > > > Hi Ales,

Re: [ovs-dev] [PATCH ovn v3] northd: Add logical flow to skip GARP with LLA

On Wed, Jun 7, 2023 at 4:48 PM Dumitru Ceara wrote: > On 6/7/23 14:41, Ihar Hrachyshka wrote: > > Thank you Ales! > > > > Acked-By: Ihar Hrachyshka > > > > Hi Ales, Ihar, > Hi Dumitru, > > > On Wed, Jun 7, 2023 at 1:58 AM Ales Musil wrote: > > > >> Skip GARP packet with link-local address be

Re: [ovs-dev] [PATCH ovn v3] northd: Add logical flow to skip GARP with LLA

On 6/7/23 14:41, Ihar Hrachyshka wrote: > Thank you Ales! > > Acked-By: Ihar Hrachyshka > Hi Ales, Ihar, > On Wed, Jun 7, 2023 at 1:58 AM Ales Musil wrote: > >> Skip GARP packet with link-local address being advertised >> when "always_learn_from_arp_request=false", this should >> prevent hug

Re: [ovs-dev] [PATCH ovn] controller: Turn OFTABLE_OUTPUT_INIT into an alias.

Thanks for the patch, and sorry for missing before that we could macro it instead. You should also update ovn-architechture.xml that was modified in patches 4/5 and 5/5 of the original series. Thanks! On Wed, Jun 7, 2023 at 7:02 AM Dumitru Ceara wrote: > Make it alias OFTABLE_OUTPUT_LARGE_PKT_D

Re: [ovs-dev] [PATCH 2/2] conntrack: Release nat_conn in case both keys have the same hash.

Hi Paolo, On 4/19/23 2:40 PM, Paolo Valerio wrote: During the creation of a new connection, there's a chance both key and rev_key end up having the same hash. This is more common in the case of all-zero snat with no collisions. In that case, once the connection is expired, but not cleaned up, if

[ovs-dev] [PATCH] ovsdb: monitor: Destroy initial change set when new columns added.

Initial change set is preserved for as long as the monitor itself. However, if a new client has a condition on a column that is not one of the monitored columns, this column will be added to the monitor via ovsdb_monitor_condition_bind(). This new column, however, doesn't exist in the initial chan

Re: [ovs-dev] [PATCH ovn v3] northd: Add logical flow to skip GARP with LLA

Thank you Ales! Acked-By: Ihar Hrachyshka On Wed, Jun 7, 2023 at 1:58 AM Ales Musil wrote: > Skip GARP packet with link-local address being advertised > when "always_learn_from_arp_request=false", this should > prevent huge grow of MAC Binding table. To keep the option > consistent overwrite t

Re: [ovs-dev] [PATCH net v2] net: openvswitch: fix upcall counter access before allocation

Hello: This patch was applied to netdev/net.git (main) by David S. Miller : On Tue, 6 Jun 2023 13:56:35 +0200 you wrote: > Currently, the per cpu upcall counters are allocated after the vport is > created and inserted into the system. This could lead to the datapath > accessing the counters befo

[ovs-dev] [PATCH ovn] controller: Turn OFTABLE_OUTPUT_INIT into an alias.

Make it alias OFTABLE_OUTPUT_LARGE_PKT_DETECT in order to avoid always doing an extra resubmit in slow path. The OFTABLE_OUTPUT_INIT table didn't have more than the "advance to the next table" default rule. Fixes: 740f23c19577 ("Add new egress tables to accommodate for too-big packets handling")

Re: [ovs-dev] [PATCH 6/6] ovsdb: monitor: Keep and maintain the initial change set.

On 6/7/23 00:48, Han Zhou wrote: > > > On Tue, Jun 6, 2023 at 3:08 PM Ilya Maximets > wrote: >> >> On 6/6/23 22:12, Han Zhou wrote: >> > >> > >> > On Tue, Jun 6, 2023 at 12:10 PM Ilya Maximets > > > >

Re: [ovs-dev] [PATCH 1/2] netdev-offload-dpdk: Fix flushing of a physdev

Hello, On Tue, Jun 6, 2023 at 7:10 AM Eli Britstein wrote: > >> > > >> > Vport's offloads are done on the tracked orig-in-port, but the flow > >> > itself is associated in the vport's map. > >> > > >> > Removing the physdev will flush all the ports that are on its map, > >> > but > >> > >> all th

Re: [ovs-dev] [PATCH ovn 2/2] ci: Run the new check-system-dpdk tests as part of the ci.

Bleep bloop. Greetings Eelco Chaudron, 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 89 characters long (recommended limit is 79) #218 FILE: .ci/linux-build.sh:1

[ovs-dev] [PATCH ovn 2/2] ci: Run the new check-system-dpdk tests as part of the ci.

This patch includes changes made earlier by David in the ovs branch to cache the dpdk builds. Co-authored-by: David Marchand Signed-off-by: David Marchand Signed-off-by: Eelco Chaudron --- .ci/ci.sh | 12 ++- .ci/dpdk-build.sh | 54

[ovs-dev] [PATCH ovn 1/2] tests: add make check-system-dpdk to test suite.

Allow the ovn-system tests to run on the OVS-DPDK infrastructure, i.e., DPDK ports and mbuf memory. Co-authored-by: David Marchand Signed-off-by: David Marchand Signed-off-by: Eelco Chaudron --- tests/automake.mk | 23 +++-- tests/ofproto-macros.at|9 ++- tests/s

Re: [ovs-dev] [PATCH 1/2] tests: add make check-system-dpdk to test suite.

Bleep bloop. Greetings Eelco Chaudron, 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. git-am: error: sha1 information is lacking or useless (tests/automake.mk). error: could not build fake an

Re: [ovs-dev] [PATCH 2/2] ci: Run the new check-system-dpdk tests as part of the ci.

Bleep bloop. Greetings Eelco Chaudron, 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. Patch skipped due to previous failure. Please check this out. If you feel there has been an error, plea

Re: [ovs-dev] [PATCH 1/2] tests: add make check-system-dpdk to test suite.

Forgot to put OVN in the title will resend it, so the correct patchwork will pick it up :( //Eelco On 7 Jun 2023, at 10:58, Eelco Chaudron wrote: > Allow the ovn-system tests to run on the OVS-DPDK infrastructure, > i.e., DPDK ports and mbuf memory. > > Co-authored-by: David Marchand > Signed

Re: [ovs-dev] [PATCH net v2] net: openvswitch: fix upcall counter access before allocation

On 7 Jun 2023, at 3:05, wangchuanlei wrote: > Thanks for fix this, in common enviroment, it's a > small probability event. Well, on ARM, they could replicate it a couple of times, but I guess the system was under memory pressure and has a lot of cores. >> Eelco Chaudron writes: > >> Currentl

[ovs-dev] [PATCH 2/2] ci: Run the new check-system-dpdk tests as part of the ci.

This patch includes changes made earlier by David in the ovs branch to cache the dpdk builds. Co-authored-by: David Marchand Signed-off-by: David Marchand Signed-off-by: Eelco Chaudron --- .ci/ci.sh | 12 ++- .ci/dpdk-build.sh | 54

[ovs-dev] [PATCH 1/2] tests: add make check-system-dpdk to test suite.

Allow the ovn-system tests to run on the OVS-DPDK infrastructure, i.e., DPDK ports and mbuf memory. Co-authored-by: David Marchand Signed-off-by: David Marchand Signed-off-by: Eelco Chaudron --- tests/automake.mk | 23 +++-- tests/ofproto-macros.at|9 ++- tests/s

[ovs-dev] [PATCH] cpu: Fix cpuid check for some AMD processors.

Some venerable AMD processors do not support querying extended features (EAX=7) with cpuid. In this case, it is not a programmatic error and the runtime check should simply return the isa is unsupported. Reported-by: Davide Repetto Reported-at: https://bugzilla.redhat.com/show_bug.cgi?id=2211747