On 23 Feb 2022, at 12:35, Roi Dayan wrote:

> On 2022-02-23 11:31 AM, Eelco Chaudron wrote:
>>
>>
>> On 23 Feb 2022, at 8:09, Roi Dayan wrote:
>>
>>> On 2022-02-22 5:26 PM, Eelco Chaudron wrote:
>>>> This patch checks for none offloadable ct_state match flag combinations.
>>>> If they exist force the +trk flag down to TC Flower
>>>>
>>>> Signed-off-by: Eelco Chaudron <[email protected]>
>>>> ---
>>>> v3:
>>>>    - Instead of warning about an invalid flag combination fix it.
>>>>
>>>>    lib/netdev-offload-tc.c |    6 ++++++
>>>>    1 file changed, 6 insertions(+)
>>>>
>>>> diff --git a/lib/netdev-offload-tc.c b/lib/netdev-offload-tc.c
>>>> index 0105d883f..3d2c1d844 100644
>>>> --- a/lib/netdev-offload-tc.c
>>>> +++ b/lib/netdev-offload-tc.c
>>>> @@ -1541,6 +1541,12 @@ parse_match_ct_state_to_flower(struct tc_flower 
>>>> *flower, struct match *match)
>>>>                flower->key.ct_state &= ~(TCA_FLOWER_KEY_CT_FLAGS_NEW);
>>>>                flower->mask.ct_state &= ~(TCA_FLOWER_KEY_CT_FLAGS_NEW);
>>>>            }
>>>> +
>>>> +        if (flower->key.ct_state &&
>>>> +            !(flower->key.ct_state & TCA_FLOWER_KEY_CT_FLAGS_TRACKED)) {
>>>> +            flower->key.ct_state |= TCA_FLOWER_KEY_CT_FLAGS_TRACKED;
>>>> +            flower->mask.ct_state |= TCA_FLOWER_KEY_CT_FLAGS_TRACKED;
>>>> +        }
>>>>        }
>>>>         if (mask->ct_zone) {
>>>>
>>>
>>> just to be sure, the check is if we have a ct state flag
>>> that can be offloaded but no +trk then force add +trk ?
>>> so any +state will always be with +trk ?
>>
>> Yes correct. The TC framework in the kernel will only accept any other flags 
>> if +trk is set. This makes sense, as these flags can only be set for a 
>> tracked connection.
>>
>> See also the discussion around this patch for the v2.
>>
>> //Eelco
>>
>
> yes i read the discussion just wanted to be sure. just getting confused
> with the commit msg "checks for none offloadable ct_state match flag 
> combinations".. it's checking for offloadable flags.
> maybe remove the "none" ? maybe i'm reading it wrong.

Maybe it’s my English, I’m trying to say “check for ct_state flag combinations 
that are not offloadable”.

If it really needs a change I can send out a v4, but if it’s the only change, 
maybe Ilya can change it on commit.

I’ll wait a bit for more reviews, or Ilya’s comments if any.

//Eelco

_______________________________________________
dev mailing list
[email protected]
https://mail.openvswitch.org/mailman/listinfo/ovs-dev

Reply via email to