Re: [PATCH net v3] mptcp: fix stale skb->sk reference on subflow close

[Matthieu Baerts <matttbe@kernel.org>]

Hi Paolo,

On 29/05/2026 20:38, Paolo Abeni wrote:
> On 5/27/26 1:31 PM, Matthieu Baerts wrote:
>> Hi Paolo,
>>
>> Thank you for your review!
>>
>> On 27/05/2026 19:40, Paolo Abeni wrote:
>>> On 5/27/26 7:49 AM, Kalpan Jani wrote:
>>>> Gentle ping on this patch sent on 19 May 2026 — I haven't seen any review feedback yet. If anyone has had a chance to look, I'd appreciate comments; otherwise I'm happy to rebase or resend if it slipped through.
>>>
>>> Note that maintainers are under an ever increasing influx of AI
>>> generated contents that make processing incoming patches more
>>> challenging every day.
>>
>> +1
>>
>>> FTR this revision LGTM.
>>
>> Just to be sure, does it mean that Sashiko was wrong?
>>
>>  https://sashiko.dev/#/patchset/20260519092243.1242351-1-kalpan.jani%40mpiricsoftware.com
> 
> AFAICS, yes: the msk socket is set to TCP_CLOSE status before calling
> mptcp_backlog_purge(): __mptcp_add_backlog() can't add anything to the
> backlog.

Thank you for having checked!

The code then looks good, but it seems the commit message is no longer
correct with the last version (and a Fixes tag is still missing). I sent
a dedicated message for that.

>> If yes, can I convert this "LGTM" as an Acked-by? :)
> 
> There is already my Suggested-by: tag, I think the latter would be
> redundant?!?

Personally, I think it is clearer with an extra Acked-by, just to say
that what you had in mind (the suggestion) is what is actually done in
the patch. But if you prefer without, fine by me :)

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