Re: [PATCH net] tcp: process the 3rd ACK with sk_socket for for TFO/MPTCP

[Paolo Abeni <pabeni@redhat.com>]

On 7/17/24 17:09, Matthieu Baerts wrote:
> On 17/07/2024 16:57, Eric Dumazet wrote:
>> I had no time yet to run all our packetdrill tests with Kuniyuki patch
>> because of the ongoing netdev conference.
>>
>> Is it ok for you if we hold your patch for about 5 days ?
> 
> Sure, no problem, take your time!
> 
>> I would like to make sure we did not miss anything else.
> 
> I understand!
> 
>> I am CCing Neal, perhaps he can help to expedite the testing part
>> while I am busy.
> 
> Thank you, no urgency here.
> 
> If it's OK with you, I can send a v2 using Kuniyuki's suggestion --
> simply limiting the bypass to SYN+ACK only -- because it is simpler and
> ready to be sent, but also to please the CI because my v1 was rejected
> by the CI because I sent it just before the sync with Linus tree. We can
> choose later to pick the v2, the previous one, or a future one.

I think it would be better to have this patch going through the netdev 
CI, so a repost would be appreciated. I also thing Kuniyuki's suggestion 
should be preferred, so I would say go for it :)

@Neal could you please run the pktdrill tests on the new, upcoming 
version, instead?

Thanks,

Paolo