[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <CANn89iLtXW-MFJ74UhP4WyC3a60LrevAxddBjJ1nGu78eSG1DQ@mail.gmail.com>
Date: Tue, 15 Feb 2022 09:06:39 -0800
From: Eric Dumazet <edumazet@...gle.com>
To: Willem de Bruijn <willemdebruijn.kernel@...il.com>
Cc: Tao Liu <thomas.liu@...oud.cn>, David Miller <davem@...emloft.net>,
Hideaki YOSHIFUJI <yoshfuji@...ux-ipv6.org>,
David Ahern <dsahern@...nel.org>,
Jakub Kicinski <kuba@...nel.org>,
"Samudrala, Sridhar" <sridhar.samudrala@...el.com>,
Network Development <netdev@...r.kernel.org>,
LKML <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] gso: do not skip outer ip header in case of ipip and net_failover
On Tue, Feb 15, 2022 at 7:46 AM Willem de Bruijn
<willemdebruijn.kernel@...il.com> wrote:
> > Stuffing state in the skb has been a mistake I think.
>
> If we could unwind those skb inner_* fields (and reclaim the skbuff
> space!) that would be fantastic.
Not sure we can easily remove the space, many networking drivers need them,
we probably do not want to dissect packets in their ndo_start_xmit()
>
> Immediately for this bug: perhaps it can be fixed by resetting the
> network_header on the gso skb if segs == NULL. As the offset is stored
> on the stack.
It seems correct. Any other fields we need to take care of ?
Powered by blists - more mailing lists