[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CANn89i+DzusGJEZQYXe+z_zDzj7deYvWrxG7G=9L3Q+r1P0h9g@mail.gmail.com>
Date: Sat, 19 Aug 2023 05:23:28 +0200
From: Eric Dumazet <edumazet@...gle.com>
To: Jakub Kicinski <kuba@...nel.org>
Cc: davem@...emloft.net, netdev@...r.kernel.org, pabeni@...hat.com,
syzbot+5ba06978f34abb058571@...kaller.appspotmail.com, wg@...ndegger.com,
mkl@...gutronix.de, idosch@...dia.com, lucien.xin@...il.com,
xemul@...allels.com, socketcan@...tkopp.net, linux-can@...r.kernel.org
Subject: Re: [PATCH net] net: validate veth and vxcan peer ifindexes
On Sat, Aug 19, 2023 at 3:26 AM Jakub Kicinski <kuba@...nel.org> wrote:
>
> veth and vxcan need to make sure the ifindexes of the peer
> are not negative, core does not validate this.
>
> Using iproute2 with user-space-level checking removed:
>
> Before:
>
> # ./ip link add index 10 type veth peer index -1
> # ip link show
> 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN mode DEFAULT group default qlen 1000
> link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
> 2: enp1s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc fq_codel state UP mode DEFAULT group default qlen 1000
> link/ether 52:54:00:74:b2:03 brd ff:ff:ff:ff:ff:ff
> 10: veth1@...h0: <BROADCAST,MULTICAST,M-DOWN> mtu 1500 qdisc noop state DOWN mode DEFAULT group default qlen 1000
> link/ether 8a:90:ff:57:6d:5d brd ff:ff:ff:ff:ff:ff
> -1: veth0@...h1: <BROADCAST,MULTICAST,M-DOWN> mtu 1500 qdisc noop state DOWN mode DEFAULT group default qlen 1000
> link/ether ae:ed:18:e6:fa:7f brd ff:ff:ff:ff:ff:ff
>
> Now:
>
> $ ./ip link add index 10 type veth peer index -1
> Error: ifindex can't be negative.
>
> This problem surfaced in net-next because an explicit WARN()
> was added, the root cause is older.
>
> Fixes: e6f8f1a739b6 ("veth: Allow to create peer link with given ifindex")
> Fixes: a8f820a380a2 ("can: add Virtual CAN Tunnel driver (vxcan)")
> Reported-by: syzbot+5ba06978f34abb058571@...kaller.appspotmail.com
> Signed-off-by: Jakub Kicinski <kuba@...nel.org>
SGTM, I was not sure how to fix this myself ;)
Reviewed-by: Eric Dumazet <edumazet@...gle.com>
Powered by blists - more mailing lists