[ovs-dev] [PATCH] dpif-netdev:The ukey is deleted during del-flows

From: Shaohua Wu description: Using the command line to delete flow tables is a risky behavior. When the netdev flow table was deleted, only the dpflow was deleted. The ukey associated with the dpflow is not deleted. When the packet sending persists, the packet fails to match the flow table after

Re: [ovs-dev] [PATCH v10 01/11] tests: Include working system-traffic tests into the system-offloads-testsuite.

On 2/3/23 17:48, Simon Horman wrote: > On Thu, Feb 02, 2023 at 12:09:47PM +0100, Eelco Chaudron wrote: >> Include and run the system-traffic.at tests as part of the system offload >> testsuite. Exclude all the tests that will not run without any special >> modifications. >> >> Signed-off-by: Eelco

[ovs-dev] OVN 23.03 soft freeze today

Hi everyone, Today is the soft freeze for OVN 23.03. As a reminder, this is the deadline for submitting feature patches to be considered for OVN 23.03.0. In two weeks (17 February) we will branch 23.03 and at that point, all new features must be merged. Two weeks from then (3 March), we will

Re: [ovs-dev] [PATCH v4] netdev-offload-tc: Preserve tc statistics when flow gets modified.

On 2/1/23 12:12, Eelco Chaudron wrote: > When a flow gets modified, i.e. the actions are changes, the tc layer will > remove, and re-add the flow. This is causing all the counters to be reset. > > This patch will remember the previous tc counters and adjust any requests > for statistics. This is d

Re: [ovs-dev] [PATCH] sparse: Fix numa.h for libnuma >= 2.0.13.

On 2/2/23 09:31, David Marchand wrote: > On Tue, Jan 17, 2023 at 6:13 PM Ilya Maximets wrote: >> >> Current numa.h header for sparse re-defines functions in a way >> that breaks the header from libnuma 2.0.13+, because the original >> issue was fixed in that version: >> >> https://github.com/nu

Re: [ovs-dev] [PATCH v10 01/11] tests: Include working system-traffic tests into the system-offloads-testsuite.

On Thu, Feb 02, 2023 at 12:09:47PM +0100, Eelco Chaudron wrote: > Include and run the system-traffic.at tests as part of the system offload > testsuite. Exclude all the tests that will not run without any special > modifications. > > Signed-off-by: Eelco Chaudron > Acked-by: Roi Dayan ... > di

Re: [ovs-dev] [PATCH ovn v6] northd: Refactor build_lrouter_nat_flows_for_lb function

Hi Ales, Thanks for providing the diff between v5 and v6, that made it a lot easier to go through the changes. Acked-by: Mark Michelson On 1/24/23 06:29, Ales Musil wrote: On Tue, Jan 24, 2023 at 12:28 PM Ales Musil > wrote: To make it easier to add flows t

Re: [ovs-dev] [PATCH ovn] northd, controller: Commit flows dropped by ACLs to conntrack

On 1/25/23 05:36, Abhiram Sangana wrote: Hi Mark, I have replied to your comments. Can you please have a look when you get a chance? I had a look at the code itself, and from a purely mechanical perspective, I can't see anything wrong with it. I have some high level comments down below, th

Re: [ovs-dev] 回复: [PATCH net-next v7 1/1] net:openvswitch:reduce cpu_used_mask memory

Fri, Feb 03, 2023 at 04:50:31PM CET, taoyuan_e...@hotmail.com wrote: >Change between V7 and V6: >move initialization of cpu_used_mask up to follow stats_last_writer Okay, please stop sending stuff and begin to read. > >thanks >eddy ___ dev mailing list

[ovs-dev] 回复: [PATCH net-next v7 1/1] net:openvswitch:reduce cpu_used_mask memory

Change between V7 and V6: move initialization of cpu_used_mask up to follow stats_last_writer thanks eddy ___ dev mailing list d...@openvswitch.org https://mail.openvswitch.org/mailman/listinfo/ovs-dev

[ovs-dev] 回复: [PATCH net-next v7 1/1] net:openvswitch:reduce cpu_used_mask memory

Change between V7 and V6: move initialization of cpu_used_mask up to follow stats_last_writer Thanks ___ dev mailing list d...@openvswitch.org https://mail.openvswitch.org/mailman/listinfo/ovs-dev

[ovs-dev] [PATCH net-next v7 1/1] net:openvswitch:reduce cpu_used_mask memory

Use actual CPU number instead of hardcoded value to decide the size of 'cpu_used_mask' in 'struct sw_flow'. Below is the reason. 'struct cpumask cpu_used_mask' is embedded in struct sw_flow. Its size is hardcoded to CONFIG_NR_CPUS bits, which can be 8192 by default, it costs memory and slows down

Re: [ovs-dev] [PATCH] sparse: Fix build with DPDK and GCC 12.

On 2/2/23 10:21, David Marchand wrote: > On Wed, Feb 1, 2023 at 5:54 PM David Marchand > wrote: >> >> rte_vect.h pulls some AVX512 instrinsics headers added in GCC 12 [1] >> trigger a lot of warnings: >> >> libtool: compile: env "REAL_CC=ccache gcc" "CHECK=sparse -Wsparse-error >> -I ../

Re: [ovs-dev] [PATCH v5] conntrack: Properly unNAT inner header of related traffic

On 2/3/23 13:46, Ales Musil wrote: > The inner header was not handled properly. > Simplify the code which allows proper handling > of the inner headers. > > Reported-at: https://bugzilla.redhat.com/2137754 > Signed-off-by: Ales Musil > --- [...] > +nat_packet(struct dp_packet *pkt, struct conn

[ovs-dev] [PATCH v1 1/3] string_helpers: Move string_is_valid() to the header

Move string_is_valid() to the header for wider use. Signed-off-by: Andy Shevchenko --- include/linux/string_helpers.h | 5 + net/tipc/netlink_compat.c | 6 +- 2 files changed, 6 insertions(+), 5 deletions(-) diff --git a/include/linux/string_helpers.h b/include/linux/string_helpers

[ovs-dev] [PATCH v1 3/3] openvswitch: Use string_is_valid() helper

Use string_is_valid() helper instead of cpecific memchr() call. This shows better the intention of the call. Signed-off-by: Andy Shevchenko --- net/openvswitch/conntrack.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/net/openvswitch/conntrack.c b/net/openvswitch/connt

[ovs-dev] [PATCH v1 2/3] genetlink: Use string_is_valid() helper

Use string_is_valid() helper instead of cpecific memchr() call. This shows better the intention of the call. Signed-off-by: Andy Shevchenko --- net/netlink/genetlink.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/net/netlink/genetlink.c b/net/netlink/genetlink.c index

[ovs-dev] 回复: 回复: [PATCH net-next v5 1/1] net:openvswitch:reduce cpu_used_mask memory

Hi, Jiri: Difference between v5 and v4 are a wrap to fix max-line-length warning - cpumask_set_cpu(cpu, &flow->cpu_used_mask); + cpumask_set_cpu(cpu, + flow->cpu_used_mas

[ovs-dev] 回复: [PATCH net-next v6 1/1] net:openvswitch:reduce cpu_used_mask memory

Hi, Eelco:   I almost missed this mail, i see your point - aggregating the code for the initialization of sw_flow I am good with this revision, will incorporate in v7. Jiri, though i hoped that V6 is the last one, till I saw this mail, i will wait a couple of hours before making the next la

[ovs-dev] 回复: [PATCH net-next v6 1/1] net:openvswitch:reduce cpu_used_mask mmory

Yes, Jiri, it is the last V, i do not see other revision suggestions. Sorry for the toppost storm. Again, thank you all for the time, I am grateful for the kindness i saw. have a good day eddy 发件人: Jiri Pirko 发送时间: 2023年2月3日 12:00 收件人: Eddy Tao 抄送: net...@vger.

[ovs-dev] [PATCH v5] conntrack: Properly unNAT inner header of related traffic

The inner header was not handled properly. Simplify the code which allows proper handling of the inner headers. Reported-at: https://bugzilla.redhat.com/2137754 Signed-off-by: Ales Musil --- v5: Rebase on top of current master. Address comments from Dumitru: - Use explicit struct sizes fo

Re: [ovs-dev] [PATCH v4] conntrack: Properly unNAT inner header of related traffic

On Fri, Feb 3, 2023 at 11:37 AM Dumitru Ceara wrote: > On 1/27/23 14:52, Ales Musil wrote: > > The inner header was not handled properly. > > Simplify the code which allows proper handling > > of the inner headers. > > > > Reported-at: https://bugzilla.redhat.com/2137754 > > Signed-off-by: Ales M

Re: [ovs-dev] [PATCH v3 3/3] dpif-netlink: add revalidator for offload of meters

On 30 Jan 2023, at 17:42, Simon Horman wrote: > From: Tianyu Yuan > > Allow revalidator to continuously delete police in kernel > tc datapath until it is deleted. > > In current implementation, polices in tc datapath will not > deleted when they are being used and these remaining polices > wil

Re: [ovs-dev] [PATCH net-next v6 1/1] net:openvswitch:reduce cpu_used_mask memory

Fri, Feb 03, 2023 at 10:51:18AM CET, taoyuan_e...@hotmail.com wrote: >Use actual CPU number instead of hardcoded value to decide the size >of 'cpu_used_mask' in 'struct sw_flow'. Below is the reason. > >'struct cpumask cpu_used_mask' is embedded in struct sw_flow. >Its size is hardcoded to CONFIG_N

Re: [ovs-dev] 回复: [PATCH net-next v5 1/1] net:openvswitch:reduce cpu_used_mask memory

Fri, Feb 03, 2023 at 10:03:28AM CET, taoyuan_e...@hotmail.com wrote: >sorry, Jiri, my bad, will hold on before the next submission/hope it is the >final one Don't top-post please. So what is the change comparing to v4? > >eddy > >发件人: Jiri Pirko >发送时间: 2023年2月3

Re: [ovs-dev] [PATCH net-next v6 1/1] net:openvswitch:reduce cpu_used_mask memory

On 3 Feb 2023, at 10:51, Eddy Tao wrote: > Use actual CPU number instead of hardcoded value to decide the size > of 'cpu_used_mask' in 'struct sw_flow'. Below is the reason. > > 'struct cpumask cpu_used_mask' is embedded in struct sw_flow. > Its size is hardcoded to CONFIG_NR_CPUS bits, which c

[ovs-dev] [PATCH v3 2/2] ofproto-dpif-upcall: Include hardware offloaded flows in total flows.

The revalidator process uses the internal call udpif_get_n_flows() to get the total number of flows installed in the system. It uses this value for various decisions on flow installation and removal. With the tc offload this values is incorrect, as the hardware offloaded are not included. With rte_

[ovs-dev] [PATCH v3 1/2] ofproto-dpif-upcall: Reset ukey's last stats value if the datapath changed.

When the ukey's action set changes, it could caus the flow to use a different datapath, for example, when it moves from tc to kernel. This will cause the the cached previous datapath statistics to be used. This change will reset the cached statistics when a change in datapath is discovered. Signe

Re: [ovs-dev] [PATCH v4] conntrack: Properly unNAT inner header of related traffic

On 1/27/23 14:52, Ales Musil wrote: > The inner header was not handled properly. > Simplify the code which allows proper handling > of the inner headers. > > Reported-at: https://bugzilla.redhat.com/2137754 > Signed-off-by: Ales Musil > --- Hi, Ales, The changes look mostly good to me. I only

[ovs-dev] [PATCH net-next v6 1/1] net:openvswitch:reduce cpu_used_mask memory

Use actual CPU number instead of hardcoded value to decide the size of 'cpu_used_mask' in 'struct sw_flow'. Below is the reason. 'struct cpumask cpu_used_mask' is embedded in struct sw_flow. Its size is hardcoded to CONFIG_NR_CPUS bits, which can be 8192 by default, it costs memory and slows down

[ovs-dev] 回复: [PATCH net-next v5 1/1] net:openvswitch:reduce cpu_used_mask memory

sorry, Jiri, my bad, will hold on before the next submission/hope it is the final one eddy 发件人: Jiri Pirko 发送时间: 2023年2月3日 9:00 收件人: Eddy Tao 抄送: net...@vger.kernel.org ; Pravin B Shelar ; David S. Miller ; Eric Dumazet ; Jakub Kicinski ; Paolo Abeni ; d...@o

Re: [ovs-dev] [PATCH net-next v5 1/1] net:openvswitch:reduce cpu_used_mask memory

Fri, Feb 03, 2023 at 09:52:56AM CET, taoyuan_e...@hotmail.com wrote: >Use actual CPU number instead of hardcoded value to decide the size >of 'cpu_used_mask' in 'struct sw_flow'. Below is the reason. > >'struct cpumask cpu_used_mask' is embedded in struct sw_flow. >Its size is hardcoded to CONFIG_N

Re: [ovs-dev] [PATCH net-next v4 1/1] net:openvswitch:reduce cpu_used_mask memory

Fri, Feb 03, 2023 at 09:40:36AM CET, taoyuan_e...@hotmail.com wrote: >Use actual CPU number instead of hardcoded value to decide the size >of 'cpu_used_mask' in 'struct sw_flow'. Below is the reason. > >'struct cpumask cpu_used_mask' is embedded in struct sw_flow. >Its size is hardcoded to CONFIG_N

[ovs-dev] [PATCH net-next v5 1/1] net:openvswitch:reduce cpu_used_mask memory

Use actual CPU number instead of hardcoded value to decide the size of 'cpu_used_mask' in 'struct sw_flow'. Below is the reason. 'struct cpumask cpu_used_mask' is embedded in struct sw_flow. Its size is hardcoded to CONFIG_NR_CPUS bits, which can be 8192 by default, it costs memory and slows down

Re: [ovs-dev] [PATCH ovn] northd.c: Validate port type to avoid unexpected behavior.

On 2/3/23 01:43, Han Zhou wrote: > On Thu, Feb 2, 2023 at 11:51 AM Dumitru Ceara wrote: >> >> On 2/2/23 01:30, Han Zhou wrote: >>> In ovn_igmp_group_get_ports(), it accesses a union member that should >>> exist only if the port is a LSP: port->peer->od->mcast_info.rtr.relay. >>> But in theory it i

Re: [ovs-dev] [PATCH v3 ovn] northd: do not create flows for reserved multicast IPv6 groups

On 2/2/23 23:33, Lorenzo Bianconi wrote: > Avoid creating logical flows for Link-Local reserved multicast addresses if > advertised in a MLD reports since this interferes with Slaac IPv6 address > resolution implemented in OVN. > > Acked-by: Acked-by: Dumitru Ceara One "Acked-by: " too many but

[ovs-dev] [PATCH net-next v4 1/1] net:openvswitch:reduce cpu_used_mask memory

Use actual CPU number instead of hardcoded value to decide the size of 'cpu_used_mask' in 'struct sw_flow'. Below is the reason. 'struct cpumask cpu_used_mask' is embedded in struct sw_flow. Its size is hardcoded to CONFIG_NR_CPUS bits, which can be 8192 by default, it costs memory and slows down