On Sun, Jun 18, 2023 at 11:48 AM Han Zhou wrote:
>
> This patch achieves zero recompute for VIF updates and deletions in
> common scenarios. The performance gain for these scenarios is similar to
> the patch "northd: Incremental processing of VIF additions in 'lflow'
> node."
>
> Signed-off-by:
On Fri, Jun 30, 2023 at 7:00 AM Han Zhou wrote:
>
> On Thu, Jun 29, 2023 at 9:19 AM Dumitru Ceara wrote:
> >
> > On 6/27/23 10:23, Numan Siddique wrote:
> > > On Mon, Jun 26, 2023 at 10:34 PM Han Zhou wrote:
> > >>
> > >> On Mon, Jun 26, 2023 at 7:25 AM Numan Siddique wrote:
> > >>>
> > >>> On
On Fri, Jun 30, 2023 at 7:01 AM Han Zhou wrote:
>
> On Thu, Jun 29, 2023 at 4:57 AM Dumitru Ceara wrote:
> >
> > On 6/18/23 08:17, Han Zhou wrote:
> > > For incremental processing, we need to maintain SB lflow uuids in
> > > northd. For this reason, we generate the row uuid when creating the
> >
Hi Ilya,
Thanks again for reviewing this. I attempted to fix review notes in
https://patchwork.ozlabs.org/project/openvswitch/list/?series=361784 .
Thank you.
On Thu, 29 Jun 2023, Ilya Maximets wrote:
On 6/29/23 15:58, Simon Horman wrote:
On Tue, Jun 06, 2023 at 03:35:32PM +0400, Ivan Malov
Hi Ilya,
Thanks for reviewing this. Please see below.
On Thu, 29 Jun 2023, Ilya Maximets wrote:
On 6/6/23 13:35, Ivan Malov wrote:
Action PORT_ID has been deprecated. Use REPRESENTED_PORT instead.
AFAICT, not all drivers moved to a REPRESENTED_PORT action.
I don't see support in NFP
This may be required by some PMDs in offload scenarios.
Signed-off-by: Ivan Malov
---
lib/netdev-dpdk.c | 51 +++
1 file changed, 51 insertions(+)
diff --git a/lib/netdev-dpdk.c b/lib/netdev-dpdk.c
index 63dac689e..d9d1b43f6 100644
---
Action PORT_ID has been deprecated. Use REPRESENTED_PORT instead.
Signed-off-by: Ivan Malov
---
lib/netdev-offload-dpdk.c | 30 +-
1 file changed, 17 insertions(+), 13 deletions(-)
diff --git a/lib/netdev-offload-dpdk.c b/lib/netdev-offload-dpdk.c
index
Address two problems using two corresponding features in DPDK,
which have been around for a year and are now stable:
1) The need to make sure that metadata generated by flow
rule execution be delivered from NIC to application;
2) Replacing PORT_ID action with REPRESENTED_PORT;
---
v2:
On Thu, 29 Jun 2023, Simon Horman wrote:
On Tue, Jun 06, 2023 at 03:35:32PM +0400, Ivan Malov via dev wrote:
This may be required by some PMDs in offload scenarios.
Signed-off-by: Ivan Malov
---
lib/netdev-dpdk.c | 50 +++
1 file changed, 50
On Thu, Jun 29, 2023 at 4:57 AM Dumitru Ceara wrote:
>
> On 6/18/23 08:17, Han Zhou wrote:
> > For incremental processing, we need to maintain SB lflow uuids in
> > northd. For this reason, we generate the row uuid when creating the
> > Logical_Flow record in SB DB, rather than waiting for SB DB
On Thu, Jun 29, 2023 at 9:19 AM Dumitru Ceara wrote:
>
> On 6/27/23 10:23, Numan Siddique wrote:
> > On Mon, Jun 26, 2023 at 10:34 PM Han Zhou wrote:
> >>
> >> On Mon, Jun 26, 2023 at 7:25 AM Numan Siddique wrote:
> >>>
> >>> On Sun, Jun 18, 2023 at 11:48 AM Han Zhou wrote:
>
> For
On Thu, Jun 29, 2023 at 4:23 AM Dumitru Ceara wrote:
>
> On 6/29/23 04:18, Han Zhou wrote:
> > When multiple LSP deletions are handled in incremental processing, if it
> > hits a LSP that can't be incrementally processed after incrementally
> > processing some LSP deletions, it falls back to
/20230629203005.2137107-3-eric%40garver.life
patch subject: [PATCH net-next 2/2] net: openvswitch: add drop action
config: hexagon-randconfig-r045-20230629
(https://download.01.org/0day-ci/archive/20230630/202306300609.tdrdzscy-...@intel.com/config)
compiler: clang version 17.0.0 (https://github.com/llvm/llvm
/20230629203005.2137107-3-eric%40garver.life
patch subject: [PATCH net-next 2/2] net: openvswitch: add drop action
config: mips-randconfig-r036-20230629
(https://download.01.org/0day-ci/archive/20230630/202306300645.towwst4x-...@intel.com/config)
compiler: clang version 14.0.6 (https://github.com/llvm/llvm
On Thu, Jun 29, 2023 at 3:10 PM Ilya Maximets wrote:
>
> On 6/9/23 17:05, Mike Pattrick wrote:
> > Several xlate actions used in recursive translation currently store a
> > large amount of information on the stack. This can result in handler
> > threads quickly running out of stack space despite
On 6/25/23 19:05, Han Zhou wrote:
> Signed-off-by: Han Zhou
> ---
> ovsdb/raft.h | 3 ++-
> 1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/ovsdb/raft.h b/ovsdb/raft.h
> index 403ed3dd7321..a5b55d9bf031 100644
> --- a/ovsdb/raft.h
> +++ b/ovsdb/raft.h
> @@ -26,7 +26,8 @@
> *
On 6/23/23 12:33, Paolo Valerio wrote:
> The patch introduces a new commands ovs-appctl dpctl/dump-conntrack-exp
> that allows to dump the existing expectations for the userspace ct.
>
> Signed-off-by: Paolo Valerio
> ---
> NEWS |2 +
> lib/conntrack.c
On 6/29/23 16:27, Simon Horman wrote:
> On Tue, Jun 20, 2023 at 06:20:39PM +0100, Kevin Traynor wrote:
>> get_log_next_line_num() was defined in alb.at.
>>
>> As it may be useful in other test files, move to
>> ofproto-macros.at.
>>
>> Suggested-by: David Marchand
>> Signed-off-by: Kevin Traynor
On 6/21/23 21:24, Robin Jarry wrote:
> Some control protocols are used to maintain link status between
> forwarding engines (e.g. LACP). When the system is not sized properly,
> the PMD threads may not be able to process all incoming traffic from the
> configured Rx queues. When a signaling packet
On 6/27/23 16:13, Timothy Redaelli wrote:
> Since on CentOS/RHEL the builds are based on stable branches and not on
> tags for debugging purpose it's better to have the downstream version as
> version so it's easier to know which commits are included in a build.
>
> This commit adds
Signed-off-by: Eric Garver
---
tests/system-traffic.at | 29 +
1 file changed, 29 insertions(+)
diff --git a/tests/system-traffic.at b/tests/system-traffic.at
index 4c378e1d02b0..3268d319381c 100644
--- a/tests/system-traffic.at
+++ b/tests/system-traffic.at
@@
Signed-off-by: Eric Garver
---
include/linux/openvswitch.h | 3 ++-
lib/dpif.c | 6 --
lib/dpif.h | 1 -
ofproto/ofproto-dpif.c | 34 --
4 files changed, 34 insertions(+), 10 deletions(-)
diff --git
This adds an explicit drop action. This is used by OVS to drop packets
for which it cannot determine what to do. An explicit action in the
kernel allows passing the reason _why_ the packet is being dropped. We
can then use perf tracing to match on the drop reason.
e.g. trace all OVS dropped skbs
Prior to this series the "drop" action was implicit by an empty set of
actions. This series adds support for an explicit drop action. The
primary motivation is to allow passing xlate_error from userspace such
that xlater_error can be passed to kfree_skb_reason() and therefore
traced.
Probe the datapath implementation for support of OVS_ACTION_ATTR_DROP.
Also add a new test case.
Eric Garver (3):
dpif: probe support for OVS_ACTION_ATTR_DROP
system-common-macros: check for drop action in datapath
tests: system-traffic: add coverage for drop action
These are counterparts to userspace's xlate_error values.
Signed-off-by: Eric Garver
---
include/net/dropreason.h | 6 ++
net/openvswitch/datapath.c | 17 +
net/openvswitch/drop.h | 34 ++
3 files changed, 57 insertions(+)
create mode
Signed-off-by: Eric Garver
---
tests/system-common-macros.at | 4
1 file changed, 4 insertions(+)
diff --git a/tests/system-common-macros.at b/tests/system-common-macros.at
index 0077a8609c02..afc0df00a329 100644
--- a/tests/system-common-macros.at
+++ b/tests/system-common-macros.at
@@
On 6/20/23 13:26, David Marchand wrote:
> At some point in OVS history, some virtio features were announced as
> supported (ECN and UFO virtio features).
>
> The userspace TSO code, which has been added later, does not support
> those features and tries to disable them.
>
> This breaks OVS
On 6/9/23 17:05, Mike Pattrick wrote:
> Several xlate actions used in recursive translation currently store a
> large amount of information on the stack. This can result in handler
> threads quickly running out of stack space despite before
> xlate_resubmit_resource_check() is able to terminate
On 6/29/23 18:47, Ilya Maximets wrote:
> On 6/29/23 18:17, miter wrote:
>> Hi Ilya,
>>
>> Thank you for your advice.
>>
>> Can I take 2 minutes off of your time?
>>
>> Does this command work correctly? I usually use this command to send
>> patches.
>> "git send-email --to d...@openvswitch.org
On 6/29/23 18:17, miter wrote:
> Hi Ilya,
>
> Thank you for your advice.
>
> Can I take 2 minutes off of your time?
>
> Does this command work correctly? I usually use this command to send
> patches.
> "git send-email --to d...@openvswitch.org --cover-letter --annotate -v6 -4"
Hmm. Yes,
On 6/27/23 10:23, Numan Siddique wrote:
> On Mon, Jun 26, 2023 at 10:34 PM Han Zhou wrote:
>>
>> On Mon, Jun 26, 2023 at 7:25 AM Numan Siddique wrote:
>>>
>>> On Sun, Jun 18, 2023 at 11:48 AM Han Zhou wrote:
For incremental processing, it is important to maintain relationship
Hi Ilya,
Thank you for your advice.
Can I take 2 minutes off of your time?
Does this command work correctly? I usually use this command to send
patches.
"git send-email --to d...@openvswitch.org --cover-letter --annotate -v6 -4"
Best regards, Lin Huang.
On 6/29/2023 10:51 PM, Ilya
On 6/29/23 17:43, Eelco Chaudron wrote:
>
>
> On 22 Jun 2023, at 0:32, Ilya Maximets wrote:
>
>> Current implementation of meters in the userspace datapath takes
>> the meter lock for every packet batch. If more than one thread
>> hits the flow with the same meter, they will lock each other.
On 22 Jun 2023, at 0:32, Ilya Maximets wrote:
> Current implementation of meters in the userspace datapath takes
> the meter lock for every packet batch. If more than one thread
> hits the flow with the same meter, they will lock each other.
>
> Replace the critical section with atomic
On 6/29/23 15:58, Simon Horman wrote:
> On Tue, Jun 06, 2023 at 03:35:32PM +0400, Ivan Malov via dev wrote:
>> This may be required by some PMDs in offload scenarios.
>>
>> Signed-off-by: Ivan Malov
>> ---
>> lib/netdev-dpdk.c | 50 +++
>> 1 file
On Thu, Jun 29, 2023 at 08:25:35AM -0400, Aaron Conole wrote:
> Simon Horman writes:
>
> > On Wed, Jun 28, 2023 at 12:27:10PM -0400, Aaron Conole wrote:
> >> The openvswitch selftests currently contain a few cases for managing the
> >> datapath, which includes creating datapath instances, adding
On 6/28/23 03:12, Simon Jones wrote:
> This patch is same as MR: https://github.com/openvswitch/ovs/pull/415
>
> This is Patch detail
>
> From ddff9ac664083e6cbed97c214935b77f25dd1228 Mon Sep 17 00:00:00 2001
> From: simon
> Date: Tue, 27 Jun 2023 11:20:51 +0800
> Subject: [PATCH V3]
On Thu, Jun 29, 2023 at 04:15:09PM +0200, Ilya Maximets wrote:
> On 6/6/23 13:35, Ivan Malov wrote:
> > Action PORT_ID has been deprecated. Use REPRESENTED_PORT instead.
>
> AFAICT, not all drivers moved to a REPRESENTED_PORT action.
> I don't see support in NFP driver, for example.
>
> Are we
On 6/18/23 15:58, mit...@outlook.com wrote:
> From: Lin Huang
>
> v5->v4: police pkts bps and pps at the same time.
> Get the maximum from the number of dropped packets, and drop that.
>
> v4->v3: rewrite egress_policer_details_to_param func.
> add a new
On 6/18/23 15:58, mit...@outlook.com wrote:
> From: Lin Huang
>
> OvS has supported packet-per-second policer which can be set at ingress
> and egress side in kernel datapath. But the userspace datapath dosen't
> support for ingress and egress packet-per-second policing now.
>
> So, this patch
On 6/18/23 15:58, mit...@outlook.com wrote:
> From: Lin Huang
>
> OvS has supported packet-per-second policer which can be set at ingress
> and egress side in kernel datapath. But the userspace datapath doesn't
> support for ingress and egress packet-per-second policing now.
>
> So, this patch
On Thu, Jun 29, 2023 at 10:06 AM Eelco Chaudron wrote:
>
>
>
> On 26 Jun 2023, at 15:43, Mike Pattrick wrote:
>
> > Currently OVS keeps track of which mirrors that each packet has been
> > sent to for the purpose of deduplication. However, this doesn't consider
> > that openflow rules can make
On Wed, Jun 14, 2023 at 02:36:39PM +0100, Kevin Traynor wrote:
> other_config:pmd-maxsleep is a config option to allow
> PMD thread cores to sleep under low or no load conditions.
>
> Rename it to 'pmd-sleep-max' to allow a more structured
> name and so that additional options or command can
On Tue, Jun 20, 2023 at 06:20:39PM +0100, Kevin Traynor wrote:
> get_log_next_line_num() was defined in alb.at.
>
> As it may be useful in other test files, move to
> ofproto-macros.at.
>
> Suggested-by: David Marchand
> Signed-off-by: Kevin Traynor
Acked-by: Simon Horman
On 6/6/23 13:35, Ivan Malov wrote:
> Action PORT_ID has been deprecated. Use REPRESENTED_PORT instead.
AFAICT, not all drivers moved to a REPRESENTED_PORT action.
I don't see support in NFP driver, for example.
Are we OK with dropping hardware offload support for NFP ?
Or are there plans to
On Sun, Jun 11, 2023 at 06:58:27PM +0300, Eli Britstein via dev wrote:
> When using a userspace vport ("vxlan0"), dpif-netdev adds an additional
> netdev ("vxlan_sys_4789"). The dpif netdev ("vxlan0") is added to the
> netdev-offload ports map, thus flows are associated on this netdev.
>
>
On Sun, Jun 11, 2023 at 06:58:26PM +0300, Eli Britstein via dev 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
> not the ones on other netdevs' maps.
On 26 Jun 2023, at 15:43, Mike Pattrick wrote:
> Currently OVS keeps track of which mirrors that each packet has been
> sent to for the purpose of deduplication. However, this doesn't consider
> that openflow rules can make significant changes to packets after
> ingress.
>
> For example, OVN
On Tue, Jun 06, 2023 at 03:35:32PM +0400, Ivan Malov via dev wrote:
> This may be required by some PMDs in offload scenarios.
>
> Signed-off-by: Ivan Malov
> ---
> lib/netdev-dpdk.c | 50 +++
> 1 file changed, 50 insertions(+)
>
> diff --git
On Tue, Jun 20, 2023 at 01:26:38PM +0200, David Marchand wrote:
> At some point in OVS history, some virtio features were announced as
> supported (ECN and UFO virtio features).
>
> The userspace TSO code, which has been added later, does not support
> those features and tries to disable them.
>
This wil lbe useful for correcting possible spelling
mistakes with ease. Suggestions limited to 3 at first,
but configurable in the future
Signed-off-by: Chandan Somani
---
utilities/checkpatch.py | 3 +++
1 file changed, 3 insertions(+)
diff --git a/utilities/checkpatch.py
This will help narrow down spelling errors that are in the
commit subject
Signed-off-by: Chandan Somani
---
utilities/checkpatch.py | 4
1 file changed, 4 insertions(+)
diff --git a/utilities/checkpatch.py b/utilities/checkpatch.py
index 57f7baf46..c5575ba00 100755
---
A list approach lets us single out each flagged word and
provide more useful details, like spelling suggestions.
This will be used in an upcoming patch.
Signed-off-by: Chandan Somani
---
utilities/checkpatch.py | 12 +---
1 file changed, 9 insertions(+), 3 deletions(-)
diff --git
Simon Horman writes:
> On Wed, Jun 28, 2023 at 12:27:10PM -0400, Aaron Conole wrote:
>> The openvswitch selftests currently contain a few cases for managing the
>> datapath, which includes creating datapath instances, adding interfaces,
>> and doing some basic feature / upcall tests. This is
On 6/18/23 08:17, Han Zhou wrote:
> For incremental processing, we need to maintain SB lflow uuids in
> northd. For this reason, we generate the row uuid when creating the
> Logical_Flow record in SB DB, rather than waiting for SB DB to populate
> back.
>
> Signed-off-by: Han Zhou
> ---
>
On 6/29/23 04:18, Han Zhou wrote:
> When multiple LSP deletions are handled in incremental processing, if it
> hits a LSP that can't be incrementally processed after incrementally
> processing some LSP deletions, it falls back to recompute without
> destroying the ovn_port objects that are already
On 27 Jun 2023, at 12:48, Roi Dayan wrote:
> From: Gavin Li
>
> Add a vxlan gbp offload test case:
>
> vxlan offloads with gbp extention - ping between two ports - offloads
> enabled ok
>
> Signed-off-by: Gavin Li
> Reviewed-by: Roi Dayan
> Reviewed-by: Simon Horman
Some small comments
On 27 Jun 2023, at 12:48, Roi Dayan wrote:
> From: Gavin Li
>
> Add TC offload support for vxlan encap with gbp option
>
> Signed-off-by: Gavin Li
> Reviewed-by: Gavi Teitz
> Reviewed-by: Roi Dayan
> Reviewed-by: Simon Horman
> Acked-by: Eelco Chaudron
> ---
I already ACKed the v5, but
On Thu, Jun 29, 2023 at 10:06:33AM +0800, .贺鹏 wrote:
> [You don't often get email from hepeng.0...@bytedance.com. Learn why this is
> important at https://aka.ms/LearnAboutSenderIdentification ]
>
> On Mon, Jun 26, 2023 at 10:57 PM Simon Horman
> wrote:
> >
> > On Fri, Oct 09, 2020 at
If the ovs port used as output_port in mirror is deleted (either by
mistake or intentionally), ovn-controller would crash in the
check_and_update_interface_table() when trying to sync the mirror port
configuration. e.g.:
0 0x00449d6e in check_and_update_interface_table
61 matches
Mail list logo