sctp: use call_rcu to free endpoint

This patch is to delay the endpoint free by calling call_rcu() to fix
another use-after-free issue in sctp_sock_dump():

  BUG: KASAN: use-after-free in __lock_acquire+0x36d9/0x4c20
  Call Trace:
    __lock_acquire+0x36d9/0x4c20 kernel/locking/lockdep.c:3218
    lock_acquire+0x1ed/0x520 kernel/locking/lockdep.c:3844
    __raw_spin_lock_bh include/linux/spinlock_api_smp.h:135 [inline]
    _raw_spin_lock_bh+0x31/0x40 kernel/locking/spinlock.c:168
    spin_lock_bh include/linux/spinlock.h:334 [inline]
    __lock_sock+0x203/0x350 net/core/sock.c:2253
    lock_sock_nested+0xfe/0x120 net/core/sock.c:2774
    lock_sock include/net/sock.h:1492 [inline]
    sctp_sock_dump+0x122/0xb20 net/sctp/diag.c:324
    sctp_for_each_transport+0x2b5/0x370 net/sctp/socket.c:5091
    sctp_diag_dump+0x3ac/0x660 net/sctp/diag.c:527
    __inet_diag_dump+0xa8/0x140 net/ipv4/inet_diag.c:1049
    inet_diag_dump+0x9b/0x110 net/ipv4/inet_diag.c:1065
    netlink_dump+0x606/0x1080 net/netlink/af_netlink.c:2244
    __netlink_dump_start+0x59a/0x7c0 net/netlink/af_netlink.c:2352
    netlink_dump_start include/linux/netlink.h:216 [inline]
    inet_diag_handler_cmd+0x2ce/0x3f0 net/ipv4/inet_diag.c:1170
    __sock_diag_cmd net/core/sock_diag.c:232 [inline]
    sock_diag_rcv_msg+0x31d/0x410 net/core/sock_diag.c:263
    netlink_rcv_skb+0x172/0x440 net/netlink/af_netlink.c:2477
    sock_diag_rcv+0x2a/0x40 net/core/sock_diag.c:274

This issue occurs when asoc is peeled off and the old sk is freed after
getting it by asoc->base.sk and before calling lock_sock(sk).

To prevent the sk free, as a holder of the sk, ep should be alive when
calling lock_sock(). This patch uses call_rcu() and moves sock_put and
ep free into sctp_endpoint_destroy_rcu(), so that it's safe to try to
hold the ep under rcu_read_lock in sctp_transport_traverse_process().

If sctp_endpoint_hold() returns true, it means this ep is still alive
and we have held it and can continue to dump it; If it returns false,
it means this ep is dead and can be freed after rcu_read_unlock, and
we should skip it.

In sctp_sock_dump(), after locking the sk, if this ep is different from
tsp->asoc->ep, it means during this dumping, this asoc was peeled off
before calling lock_sock(), and the sk should be skipped; If this ep is
the same with tsp->asoc->ep, it means no peeloff happens on this asoc,
and due to lock_sock, no peeloff will happen either until release_sock.

Note that delaying endpoint free won't delay the port release, as the
port release happens in sctp_endpoint_destroy() before calling call_rcu().
Also, freeing endpoint by call_rcu() makes it safe to access the sk by
asoc->base.sk in sctp_assocs_seq_show() and sctp_rcv().

Thanks Jones to bring this issue up.

v1->v2:
  - improve the changelog.
  - add kfree(ep) into sctp_endpoint_destroy_rcu(), as Jakub noticed.

Reported-by:  <syzbot+9276d76e83e3bcde6c99@syzkaller.appspotmail.com>
Reported-by: Lee Jones <lee.jones@linaro.org>
Fixes: d25adbeb ("sctp: fix an use-after-free issue in sctp_sock_dump")
Signed-off-by: Xin Long <lucien.xin@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>