[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CANn89iKzm0uf_Sy0mQ=uDWeKnU-LULUnTNS5jNoLQroqbUuBEw@mail.gmail.com>
Date: Fri, 8 Mar 2024 13:40:49 +0100
From: Eric Dumazet <edumazet@...gle.com>
To: Paolo Abeni <pabeni@...hat.com>
Cc: "David S . Miller" <davem@...emloft.net>, Jakub Kicinski <kuba@...nel.org>, netdev@...r.kernel.org,
eric.dumazet@...il.com, Martin KaFai Lau <kafai@...com>, Joe Stringer <joe@...d.net.nz>,
Alexei Starovoitov <ast@...nel.org>, Willem de Bruijn <willemdebruijn.kernel@...il.com>,
Kuniyuki Iwashima <kuniyu@...zon.com>, Florian Westphal <fw@...len.de>
Subject: Re: [PATCH net-next] udp: no longer touch sk->sk_refcnt in early demux
On Fri, Mar 8, 2024 at 12:11 PM Paolo Abeni <pabeni@...hat.com> wrote:
> Right, looks safe.
>
> Acked-by: Paolo Abeni <pabeni@...hat.com>
Thanks !
BTW, I was thinking (for next cycle) to let users (or TCP stack with
ad-hoc heuristics )
opt-in TCP sockets to SOCK_RCU_FREE.
This would avoid the refcnt dance for their incoming packets.
Powered by blists - more mailing lists