Re: [PATCH mptcp-next v2 1/7] mptcp: prevent excessive coalescing on receive

[Matthieu Baerts <matttbe@kernel.org>]

On 27/12/2024 10:40, Paolo Abeni wrote:
> On 12/21/24 11:00, Matthieu Baerts wrote:
>> On 10/12/2024 13:03, Matthieu Baerts wrote:
>>> On 06/12/2024 13:10, Paolo Abeni wrote:
>>>> Currently the skb size after coalescing is only limited by the skb
>>>> layout (the skb must not carry frag_list). A single coalesced skb
>>>> covering several MSS can potentially fill completely the receive
>>>> buffer. In such a case, the snd win will zero until the receive buffer
>>>> will be empty again, affecting tput badly.
>>>
>>> Good idea! Thank you for having looked at this!
>>>
>>>> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
>>>> ---
>>>> No fixes tag because the problem is not very visible in practice
>>>> currently, but will be apparent after the rx refactor.
>>>> Still I hope this could affect positively the simlut flows self-tests
>>>
>>> The modification looks safe enough to me: it should not make thing worst.
>>
>> Do you still prefer not targetting -net for this patch here? Or is it OK
>> to do so and add a Fixes tag?
> 
> Whatever is easier to do in practice works for me. Perhaps monitoring
> the simult_flows ST flakes could give an idea - if that get more
> stable/less unstable we should possibly target 'net'.

Thank you for your reply!

It looks OK so far. I just sent this patch to netdev, so we can also get
some tests over there. I will monitor if it helps on NIPA as well.

Cheers,
Matt
-- 
Sponsored by the NGI0 Core fund.