[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <e8a7cdf3-3b8c-ed50-92fa-c9a4ed5a2f85@gmail.com>
Date: Thu, 14 Mar 2019 05:58:03 -0700
From: Eric Dumazet <eric.dumazet@...il.com>
To: Sabrina Dubroca <sd@...asysnail.net>, netdev@...r.kernel.org
Cc: Eric Dumazet <eric.dumazet@...il.com>,
Jianlin Shi <jishi@...hat.com>,
Stefano Brivio <sbrivio@...hat.com>
Subject: Re: [PATCH net] net: enforce xmit_recursion for devices with a queue
On 03/14/2019 03:15 AM, Sabrina Dubroca wrote:
> Commit 745e20f1b626 ("net: add a recursion limit in xmit path")
> introduced a recursion limit, but it only applies to devices without a
> queue. Virtual devices with a queue (either because they don't have
> the IFF_NO_QUEUE flag, or because the administrator added one) can
> still cause an unbounded recursion, via __dev_queue_xmit ->
> __dev_xmit_skb -> qdisc_run -> __qdisc_run -> qdisc_restart ->
> sch_direct_xmit -> dev_hard_start_xmit . Jianlin reported this in a
> setup with 16 gretap devices stacked on top of one another.
>
> This patch prevents the stack overflow by incrementing xmit_recursion in
> code paths that can call dev_hard_start_xmit() (like commit 745e20f1b626
> did). If the recursion limit is exceeded, the packet is enqueued and the
> qdisc is scheduled.
>
> Reported-by: Jianlin Shi <jishi@...hat.com>
> Signed-off-by: Sabrina Dubroca <sd@...asysnail.net>
> Reviewed-by: Stefano Brivio <sbrivio@...hat.com>
Hi Sabrina, thanks for the patch.
Can't we detect this in the control path instead ?
Surely this makes no sense in real world to stack 16 devices like that.
Powered by blists - more mailing lists