packet: fix race condition in packet_set_ring 72/103572/3 accepted/tizen/mobile/20161209.060720 submit/tizen/20161209.020350
authorPhilip Pettersson <philip.pettersson@gmail.com>
Thu, 8 Dec 2016 08:18:53 +0000 (17:18 +0900)
committerSeung-Woo Kim <sw0312.kim@samsung.com>
Fri, 9 Dec 2016 00:51:31 +0000 (16:51 -0800)
commit842e874249aeb1c5a274e1bb3006e9d4d40590b6
tree098fefa370c8130db6219ee29845598f44ad4c84
parent916b094f76ae8b8b0da443094a21c5b0e116d92e
packet: fix race condition in packet_set_ring

When packet_set_ring creates a ring buffer it will initialize a
struct timer_list if the packet version is TPACKET_V3. This value
can then be raced by a different thread calling setsockopt to
set the version to TPACKET_V1 before packet_set_ring has finished.

This leads to a use-after-free on a function pointer in the
struct timer_list when the socket is closed as the previously
initialized timer will not be deleted.

The bug is fixed by taking lock_sock(sk) in packet_setsockopt when
changing the packet version while also taking the lock at the start
of packet_set_ring.

Ps. This is CVE-2016-8655 patch,
http://seclists.org/oss-sec/2016/q4/607

Change-Id: I3396f1bfe60b03082a981ae9d8a787b41cb5a529
Fixes: f6fb8f100b80 ("af-packet: TPACKET_V3 flexible buffer implementation.")
Signed-off-by: Philip Pettersson <philip.pettersson@gmail.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
net/packet/af_packet.c