The commit in the Fixes tag below made so flow->mask free is scheduled via RCU right after it is removed from the flow table. The pointer stays in the flow structure and it can be accessible while in the same RCU critical section. This is done to avoid requiring ovs_mutex for the ovs_flow_free(). However, while removing the flow during processing of CMD_DEL, we do not take RCU read lock before the removal, and ovs_flow_cmd_fill_info() uses the flow->mask pointer afterwards. The RCU read lock is taken, but it's already late at that point. The comment on that line acknowledges that the lock is cosmetic and doesn't serve a real purpose. This leads to use-after-free if the RCU grace period passes between removal and the filling. It is a short race window, but it is there and can lead to a real crash in case memory allocation for the info takes a bit longer: BUG: KASAN: slab-use-after-free in __ovs_nla_put_key net/openvswitch/flow_netlink.c:1996 BUG: KASAN: slab-use-after-free in ovs_nla_put_key+0x2463/0x2e30 net/openvswitch/flow_netlink.c:2250 Read of size 4 at addr ffff88801ee89970 by task ovs_flow_del_ec/9487 Call Trace: __ovs_nla_put_key net/openvswitch/flow_netlink.c:1996 ovs_nla_put_key+0x2463/0x2e30 net/openvswitch/flow_netlink.c:2250 ovs_flow_cmd_fill_info+0x420/0x9c0 net/openvswitch/datapath.c:930 ovs_flow_cmd_del+0x53a/0x970 net/openvswitch/datapath.c:1467 ... netlink_rcv_skb+0x156/0x420 net/netlink/af_netlink.c:2556 Allocated by task 9487: mask_alloc net/openvswitch/flow_table.c:967 flow_mask_insert net/openvswitch/flow_table.c:1012 ovs_flow_tbl_insert+0xea2/0x1a90 net/openvswitch/flow_table.c:1084 ovs_flow_cmd_new+0x7e3/0xd90 net/openvswitch/datapath.c:1086 ... netlink_rcv_skb+0x156/0x420 net/netlink/af_netlink.c:2556 Freed by task 9485: rcu_free_sheaf+0x1e/0x100 mm/slub.c:5978 rcu_do_batch kernel/rcu/tree.c:2645 rcu_core+0x59c/0x10c0 kernel/rcu/tree.c:2897 handle_softirqs+0x1e4/0x9a0 kernel/softirq.c:622 ... instr_sysvec_apic_timer_interrupt arch/x86/kernel/apic/apic.c:1062 ovs_flow_tbl_remove() must be called after the ovs_flow_cmd_fill_info() to avoid this race. This also helps with cleaning up the forced cast and the cosmetic RCU read lock. Before the commit in the Fixes tag the order did not matter as long as the flow object itself was not freed. A wider RCU critical section could be another option, but we have a GFP_KERNEL allocation in the way. Reported by Trend Micro's Zero Day Initiative as ZDI-CAN-32042. Fixes: 56c19868e115 ("openvswitch: Make flow mask removal symmetric.") Cc: stable@vger.kernel.org Signed-off-by: Ilya Maximets --- net/openvswitch/datapath.c | 45 +++++++++++++++++++------------------- 1 file changed, 23 insertions(+), 22 deletions(-) diff --git a/net/openvswitch/datapath.c b/net/openvswitch/datapath.c index ae69b2cabab9..ded46d993a4e 100644 --- a/net/openvswitch/datapath.c +++ b/net/openvswitch/datapath.c @@ -1473,33 +1473,34 @@ static int ovs_flow_cmd_del(struct sk_buff *skb, struct genl_info *info) goto unlock; } - ovs_flow_tbl_remove(&dp->table, flow); - ovs_unlock(); - - reply = ovs_flow_cmd_alloc_info((const struct sw_flow_actions __force *) flow->sf_acts, + reply = ovs_flow_cmd_alloc_info(ovsl_dereference(flow->sf_acts), &flow->id, info, false, ufid_flags); - if (likely(reply)) { - if (!IS_ERR(reply)) { - rcu_read_lock(); /*To keep RCU checker happy. */ - err = ovs_flow_cmd_fill_info(flow, ovs_header->dp_ifindex, - reply, info->snd_portid, - info->snd_seq, 0, - OVS_FLOW_CMD_DEL, - ufid_flags); - rcu_read_unlock(); - if (WARN_ON_ONCE(err < 0)) { - kfree_skb(reply); - goto out_free; - } + if (IS_ERR(reply)) { + netlink_set_err(sock_net(skb->sk)->genl_sock, 0, 0, + PTR_ERR(reply)); + reply = NULL; + } - ovs_notify(&dp_flow_genl_family, reply, info); - } else { - netlink_set_err(sock_net(skb->sk)->genl_sock, 0, 0, - PTR_ERR(reply)); + if (likely(reply)) { + err = ovs_flow_cmd_fill_info(flow, ovs_header->dp_ifindex, + reply, info->snd_portid, + info->snd_seq, 0, + OVS_FLOW_CMD_DEL, ufid_flags); + if (WARN_ON_ONCE(err < 0)) { + kfree_skb(reply); + reply = NULL; } } + /* Removal has to happen after ovs_flow_cmd_fill_info(), as it uses + * the flow->mask that can be scheduled to be freed by the + * ovs_flow_tbl_remove() and we're not holding the RCU read lock. + */ + ovs_flow_tbl_remove(&dp->table, flow); + ovs_unlock(); + + if (likely(reply)) + ovs_notify(&dp_flow_genl_family, reply, info); -out_free: ovs_flow_free(flow, true); return 0; unlock: -- 2.55.0