linux_dsm_epyc7002/net/bridge
Zhang Zhengming 82646b1844 bridge: Fix possible races between assigning rx_handler_data and setting IFF_BRIDGE_PORT bit
[ Upstream commit 59259ff7a81b9eb6213891c6451221e567f8f22f ]

There is a crash in the function br_get_link_af_size_filtered,
as the port_exists(dev) is true and the rx_handler_data of dev is NULL.
But the rx_handler_data of dev is correct saved in vmcore.

The oops looks something like:
 ...
 pc : br_get_link_af_size_filtered+0x28/0x1c8 [bridge]
 ...
 Call trace:
  br_get_link_af_size_filtered+0x28/0x1c8 [bridge]
  if_nlmsg_size+0x180/0x1b0
  rtnl_calcit.isra.12+0xf8/0x148
  rtnetlink_rcv_msg+0x334/0x370
  netlink_rcv_skb+0x64/0x130
  rtnetlink_rcv+0x28/0x38
  netlink_unicast+0x1f0/0x250
  netlink_sendmsg+0x310/0x378
  sock_sendmsg+0x4c/0x70
  __sys_sendto+0x120/0x150
  __arm64_sys_sendto+0x30/0x40
  el0_svc_common+0x78/0x130
  el0_svc_handler+0x38/0x78
  el0_svc+0x8/0xc

In br_add_if(), we found there is no guarantee that
assigning rx_handler_data to dev->rx_handler_data
will before setting the IFF_BRIDGE_PORT bit of priv_flags.
So there is a possible data competition:

CPU 0:                                                        CPU 1:
(RCU read lock)                                               (RTNL lock)
rtnl_calcit()                                                 br_add_slave()
  if_nlmsg_size()                                               br_add_if()
    br_get_link_af_size_filtered()                              -> netdev_rx_handler_register
                                                                    ...
                                                                    // The order is not guaranteed
      ...                                                           -> dev->priv_flags |= IFF_BRIDGE_PORT;
      // The IFF_BRIDGE_PORT bit of priv_flags has been set
      -> if (br_port_exists(dev)) {
        // The dev->rx_handler_data has NOT been assigned
        -> p = br_port_get_rcu(dev);
        ....
                                                                    -> rcu_assign_pointer(dev->rx_handler_data, rx_handler_data);
                                                                     ...

Fix it in br_get_link_af_size_filtered, using br_port_get_check_rcu() and checking the return value.

Signed-off-by: Zhang Zhengming <zhangzhengming@huawei.com>
Reviewed-by: Zhao Lei <zhaolei69@huawei.com>
Reviewed-by: Wang Xiaogang <wangxiaogang3@huawei.com>
Suggested-by: Nikolay Aleksandrov <nikolay@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
2021-05-22 11:40:54 +02:00
..
netfilter netfilter: bridge: add pre_exit hooks for ebtable unregistration 2021-04-21 13:00:55 +02:00
br_arp_nd_proxy.c net: bridge: when suppression is enabled exclude RARP packets 2021-05-19 10:12:53 +02:00
br_device.c bridge: Fix a deadlock when enabling multicast snooping 2020-12-07 17:14:43 -08:00
br_fdb.c net: bridge: fdb: don't flush ext_learn entries 2020-09-28 12:47:43 -07:00
br_forward.c net: bridge: mcast: when forwarding handle filter mode and blocked flag 2020-09-23 13:24:35 -07:00
br_if.c net: bridge: allow enslaving some DSA master network devices 2020-05-10 19:52:33 -07:00
br_input.c net: bridge: allow enslaving some DSA master network devices 2020-05-10 19:52:33 -07:00
br_ioctl.c net: bridge: delete duplicated words 2020-09-18 14:12:43 -07:00
br_mdb.c net: bridge: mcast: add support for blocked port groups 2020-09-23 13:24:34 -07:00
br_mrp_netlink.c bridge: mrp: Extend br_mrp_fill_info 2020-07-14 13:46:43 -07:00
br_mrp_switchdev.c bridge: mrp: Fix the usage of br_mrp_port_switchdev_set_state 2021-02-17 11:02:29 +01:00
br_mrp.c bridge: mrp: Fix the usage of br_mrp_port_switchdev_set_state 2021-02-17 11:02:29 +01:00
br_multicast.c net: bridge: mcast: fix broken length + header check for MRDv6 Adv. 2021-05-14 09:50:44 +02:00
br_netfilter_hooks.c netfilter: bridge: reset skb->pkt_type after NF_INET_POST_ROUTING traversal 2020-11-28 11:46:51 -08:00
br_netfilter_ipv6.c Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next 2019-06-25 01:32:59 +02:00
br_netlink_tunnel.c net: bridge: notify on vlan tunnel changes done via the old api 2020-07-12 15:18:24 -07:00
br_netlink.c bridge: Fix possible races between assigning rx_handler_data and setting IFF_BRIDGE_PORT bit 2021-05-22 11:40:54 +02:00
br_nf_core.c net: add bool confirm_neigh parameter for dst_ops.update_pmtu 2019-12-24 22:28:54 -08:00
br_private_mrp.h bridge: mrp: Fix the usage of br_mrp_port_switchdev_set_state 2021-02-17 11:02:29 +01:00
br_private_stp.h treewide: Replace GPLv2 boilerplate/reference with SPDX - rule 152 2019-05-30 11:26:32 -07:00
br_private_tunnel.h net: bridge: vlan options: add support for tunnel mapping set/del 2020-03-17 22:47:12 -07:00
br_private.h bridge: Fix a deadlock when enabling multicast snooping 2020-12-07 17:14:43 -08:00
br_stp_bpdu.c net: bridge: add STP xstats 2019-12-14 20:02:36 -08:00
br_stp_if.c net: remove newlines in NL_SET_ERR_MSG_MOD 2020-05-07 17:56:14 -07:00
br_stp_timer.c treewide: Replace GPLv2 boilerplate/reference with SPDX - rule 152 2019-05-30 11:26:32 -07:00
br_stp.c net: bridge: Add checks for enabling the STP. 2020-04-27 11:40:25 -07:00
br_switchdev.c net: bridge: don't notify switchdev for local FDB addresses 2021-03-30 14:32:04 +02:00
br_sysfs_br.c net: bridge: Add checks for enabling the STP. 2020-04-27 11:40:25 -07:00
br_sysfs_if.c net: bridge: use switchdev for port flags set through sysfs too 2021-03-07 12:34:07 +01:00
br_vlan_options.c net: bridge: vlan options: move the tunnel command to the nested attribute 2020-03-20 08:52:20 -07:00
br_vlan_tunnel.c net: bridge: vlan tunnel: constify bridge and port arguments 2020-03-17 22:47:12 -07:00
br_vlan.c net: bridge: vlan: fix error return code in __vlan_add() 2020-12-04 15:41:06 -08:00
br.c net: bridge: Fix a warning when del bridge sysfs 2021-02-23 15:53:23 +01:00
Kconfig treewide: replace '---help---' in Kconfig files with 'help' 2020-06-14 01:57:21 +09:00
Makefile bridge: mrp: Connect MRP API with the switchdev API 2020-04-27 11:40:25 -07:00