[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <8f815f67-b0e7-4425-8c0e-6a6449ee35ca@blackwall.org>
Date: Sun, 6 Oct 2024 20:42:17 +0300
From: Nikolay Aleksandrov <razor@...ckwall.org>
To: Amedeo Baragiola <ingamedeo@...il.com>
Cc: Roopa Prabhu <roopa@...dia.com>, "David S. Miller" <davem@...emloft.net>,
Eric Dumazet <edumazet@...gle.com>, Jakub Kicinski <kuba@...nel.org>,
Paolo Abeni <pabeni@...hat.com>, bridge@...ts.linux.dev,
netdev@...r.kernel.org, linux-kernel@...r.kernel.org,
Pablo Neira Ayuso <pablo@...filter.org>
Subject: Re: [PATCH] bridge: use promisc arg instead of skb flags
On 06/10/2024 20:24, Amedeo Baragiola wrote:
> I agree, just patch actually changes the behaviour when a BR_FDB_LOCAL
> dst is found and drops the traffic because promisc is *always* set to
> false when a BR_FDB_LOCAL dst is found in br_handle_frame_finish().
> I guess the problem I was trying to solve was that since the
> introduction of the promisc flag we still use brdev->flags &
> IFF_PROMISC in br_pass_frame_up() which is essentially the value of
> promisc (except in the BR_FDB_LOCAL case above) instead of promisc
> itself.
>
> Amedeo
>
>
[snip]
Please don't top post on netdev@.
The current code works correctly, my question to Pablo was more about if the warn
can still be triggered by adding a BR_FDB_LOCAL fdb and setting bridge
promisc on, then we'll hit that codepath with promisc == false and it's
kind of correct because traffic would've been passed up anyway, but the
promisc flag can be actually set on the device..
Cheers,
Nik
Powered by blists - more mailing lists