Re: [ovs-dev] [PATCH net-next 2/4] net: move add ct helper function to nf_conntrack_helper for ovs and tc

/5957a293eadac6403657d079ddc6e360cd3f0895.1666034595.git.lucien.xin%40gmail.com patch subject: [PATCH net-next 2/4] net: move add ct helper function to nf_conntrack_helper for ovs and tc config: x86_64-randconfig-c002-20221017 compiler: gcc-11 (Debian 11.3.0-8) 11.3.0 reproduce (this is a W=1 build

Re: [ovs-dev] [PATCH] tests: fix reference output for meter offload stats

W Get Outlook for Android From: Ashluke Wells Sent: Saturday, October 8, 2022 7:06:55 AM To: j...@mojatatu.com Cc: dcara...@redhat.com ; d...@openvswitch.org ; echau...@redhat.com ; i.maxim...@ovn.org ; mleit...@redhat.com ;

Re: [ovs-dev] [RFC ovn] Multi-chassis port + MTU behavior

Initial (incomplete) approach to the idea can be seen here: https://patchwork.ozlabs.org/project/ovn/patch/20221017210546.120517-1-ihrac...@redhat.com/ There are lots of missing parts still (support for one of two directions only; there's some issue with local port size check to investigate;

[ovs-dev] [PATCH RFC ovn] Generate ICMP Fragmentation Needed for multichassis ports

When multichassis ports attached to localnet switches are forced to use tunneling to communicate with other ports, MTU of the path between ports may effectively change (to accommodate tunnel headers or otherwise). In this case, it may be helpful to inform TCP sessions about the change, so that

[ovs-dev] [PATCH net-next 4/4] net: sched: add helper support in act_ct

This patch is to add helper support in act_ct for OVS actions=ct(alg=xxx) offloading, which is corresponding to Commit cae3a2627520 ("openvswitch: Allow attaching helpers to ct action") in OVS kernel part. The difference is when adding TC actions family and proto cannot be got from the

[ovs-dev] [PATCH net-next 3/4] net: sched: call tcf_ct_params_free to free params in tcf_ct_init

This patch is to make the err path simple by calling tcf_ct_params_free(), so that it won't cause problems when more members are added into param and need freeing on the err path. Signed-off-by: Xin Long --- net/sched/act_ct.c | 35 ++- 1 file changed, 18

[ovs-dev] [PATCH net-next 2/4] net: move add ct helper function to nf_conntrack_helper for ovs and tc

Move ovs_ct_add_helper from openvswitch to nf_conntrack_helper and rename as nf_ct_add_helper, so that it can be used in TC act_ct in the next patch. Signed-off-by: Xin Long --- include/net/netfilter/nf_conntrack_helper.h | 2 + net/netfilter/nf_conntrack_helper.c | 31 +++

[ovs-dev] [PATCH net-next 1/4] net: move the ct helper function to nf_conntrack_helper for ovs and tc

Move ovs_ct_helper from openvswitch to nf_conntrack_helper and rename as nf_ct_helper so that it can be used in TC act_ct in the next patch. Note that it also adds the checks for the family and proto, as in TC act_ct, the packets with correct family and proto are not guaranteed. Signed-off-by:

[ovs-dev] [PATCH net-next 0/4] net: add helper support in tc act_ct for ovs offloading

Ilya reported an issue that FTP traffic would be broken when the OVS flow with ct(commit,alg=ftp) installed in the OVS kernel module, and it was caused by that TC didn't support the ftp helper offloaded from OVS. This patchset is to add the helper support in act_ct for OVS offloading in kernel

Re: [ovs-dev] OVS building with afxdp fails

Ilya Maximets: Thank you very much for the help. I installed libbpf-dev and able to compile ovs Thanks Manickavel -Original Message- From: Ilya Maximets Sent: Monday, October 17, 2022 1:50 PM To: Manickavel Subramani ; d...@openvswitch.org Cc: ovs-discuss ; i.maxim...@ovn.org Subject:

Re: [ovs-dev] OVS building with afxdp fails

On 10/17/22 17:11, Manickavel Subramani wrote: > I been have issue with building ovs with afxdp enabled on ubuntu. > Distributor ID: Ubuntu > Description:Ubuntu 22.04.1 LTS > Release:22.04 > Codename: jammy > > I just followed the steps listed in the Open vSwitch with AF_XDP >

Re: [ovs-dev] [syzbot] WARNING in ovs_dp_reset_user_features

On Mon, 2022-10-17 at 09:26 -0400, Aaron Conole wrote: > Paolo Abeni writes: > > On Wed, 2022-10-12 at 10:43 -0700, syzbot wrote: > > > Hello, > > > > > > syzbot found the following issue on: > > > > > > HEAD commit:e8bc52cb8df8 Merge tag 'driver-core-6.1-rc1' of > > > git://git... > > >

[ovs-dev] OVS building with afxdp fails

I been have issue with building ovs with afxdp enabled on ubuntu. Distributor ID: Ubuntu Description:Ubuntu 22.04.1 LTS Release:22.04 Codename: jammy I just followed the steps listed in the Open vSwitch with AF_XDP https://docs.openvswitch.org/en/latest/intro/install/afxdp/

Re: [ovs-dev] [RFC PATCH 1/2] dpif: Add support for CT flush with partial tuple

Hello Ales, overall the approach is ok, the only concern is that, unless I'm missing something, in case of many connections, the exact match deletion could potentially take a while, whereas in the previous case the cost was basically a lookup (constant time) and of course the remaining deletion

Re: [ovs-dev] [syzbot] WARNING in ovs_dp_reset_user_features

Paolo Abeni writes: > On Wed, 2022-10-12 at 10:43 -0700, syzbot wrote: >> Hello, >> >> syzbot found the following issue on: >> >> HEAD commit:e8bc52cb8df8 Merge tag 'driver-core-6.1-rc1' of git://git... >> git tree: upstream >> console+strace:

Re: [ovs-dev] [PATCH ovn v4 3/3] northd: add drop sampling

Bleep bloop. Greetings Adrian Moreno, 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 spacing around cast #75 FILE: controller/ovn-controller.c:3204:

[ovs-dev] [PATCH 8/8] python: don't exit OFPFlow constructor

returning None in a constructor does not make sense and is just error prone. Removing what was a lefover from an attempt to handle a common error case of trying to parse what is commonly outputted by ovs-ofctl. This should be done by the caller anyway. Signed-off-by: Adrian Moreno ---

[ovs-dev] [PATCH 4/8] python: make key-value matching strict by default

Currently, if a key is not found in the decoder information, we use the default decoder which typically returns a string. This not only means we can go out of sync with the C code without noticing but it's also error prone as malformed flows could be parsed without warning. Make KeyValue parsing

[ovs-dev] [PATCH 7/8] python: return list of actions for odp action clone

Sometimes we don't want to return the result of a nested key-value decoding as a dictionary but as a list of dictionaries. This happens when we parse actions where keys can be repeated. Refactor code that already takes that into account from ofp_act.py to kv.py and use it for datapath action

[ovs-dev] [PATCH 6/8] tests: verify flows in odp.at are parseable

Create a small helper script and check that flows tested in odp.at are parseable. Signed-off-by: Adrian Moreno --- tests/automake.mk | 2 ++ tests/odp.at | 12 +++- tests/test-dpparse.py | 45 +++ 3 files changed, 58 insertions(+), 1

[ovs-dev] [PATCH 5/8] tests: verify flows in ofp-actions are parseable

Create a small helper script and check that flows used in ofp-actions.at are parseable. Signed-off-by: Adrian Moreno --- tests/automake.mk | 2 ++ tests/ofp-actions.at | 18 + tests/test-ofparse.py | 45 +++ 3 files changed, 65

[ovs-dev] [PATCH 3/8] python: add explicit decoders for all ofp actions

We were silently relying on some ofp actions to be decoded by the default decoder which would yield decent string values. In order to be more safe and robust, add an explicit decoder for all missing actions. This patch also reworks the learn action decoding to make it more explicit and verify

[ovs-dev] [PATCH 2/8] python: include aliases in ofp_fields.py

We currently auto-generate a dictionary of field names and decoders. However, sometimes fields can be specified by their cannonical NXM or OXM names. Modify gen_ofp_field_decoders to also generate a dictionary of aliases so it's easy to map OXM/NXM names to their fields and decoding information.

[ovs-dev] [PATCH 1/8] python: fix datapath flow decoders

Fix the following erros in odp decoding: - Missing push_mpls action - Typos in collector_set_id, tp_src/tp_dst and csum - Missing two fields in vxlan match Signed-off-by: Adrian Moreno --- python/ovs/flow/odp.py | 27 ++- 1 file changed, 22 insertions(+), 5 deletions(-)

[ovs-dev] [PATCH ovn v4 3/3] northd: add drop sampling

Two new options are added to NB_Global table that enable drop sampling by specifying the collector_set_id and the obs_domain_id of the sample actions added to all drop flows. For drops coming from an lflow, the sample has the following fields: - obs_domain_id (32-bit): obs_domain_id << 8 |

[ovs-dev] [PATCH ovn v4 1/3] actions: add sample action

sample ovn action encodes into the OFPACT_SAMPLE ovs action. OVN action allows the following parameters: - obs_domain_id: 8-bit integer that identifies the sampling application. This value will be combined with the datapath's tunnel_id to form the final observation_domain_id that will be

[ovs-dev] [PATCH ovn v4 2/3] northd: make default drops explicit

By default, traffic that doesn't match any configured flow will be dropped. But having that behavior implicit makes those drops more difficult to visualize. Make default drops explicit both as default logical flows and as default openflow flows (e.g: for physical tables). Signed-off-by: Adrian

[ovs-dev] [PATCH ovn v4 0/3] Add ovn drop debugging

Very often when troubleshooting networking issues in an OVN cluster one would like to know if any packet (or a specific one) is being dropped by OVN. Currently, this cannot be known because of two main reasons: 1 - Implicit drops: Some tables do not have a default action (priority=0, match=1).

[ovs-dev] [PATCH ovn] ovn-controller: Fixed missing flows after interface deletion

In the following scenario: - interface "old" is created and external_ids:iface-id is set (to lp) - interface "new" is created and external_ids:iface-id is set (to same lp) - interface "old" is deleted flows related to lp were deleted. Note that after "new" interface is created, flows use "new"

Re: [ovs-dev] [PATCH v6 3/4] userspace: Enable IP checksum offloading by default.

On Wed, Jul 13, 2022 at 8:45 PM Mike Pattrick wrote: > > From: Flavio Leitner > > The netdev receiving packets is supposed to provide the flags > indicating if the IP csum was verified and it is GOOD or BAD, > otherwise the stack will check when appropriate by software. > > If the packet comes

Re: [ovs-dev] [PATCH v6 1/4] Documentation: Document netdev offload.

On Wed, Jul 13, 2022 at 8:45 PM Mike Pattrick wrote: > > From: Flavio Leitner > > Document the implementation of netdev hardware offloading > in userspace datapath. > > Signed-off-by: Flavio Leitner > Co-authored-by: Mike Pattrick > Signed-off-by: Mike Pattrick > --- >