]> git.ipfire.org Git - thirdparty/kernel/stable.git/commit
Revert "net/sctp: fix race condition in sctp_destroy_sock"
authorXin Long <lucien.xin@gmail.com>
Sun, 2 May 2021 21:11:41 +0000 (05:11 +0800)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Sat, 22 May 2021 08:38:26 +0000 (10:38 +0200)
commit1b6863dd6a24f84643fe138b46ae9046e1ed52f0
tree976e8f647aaf502b5832a9cf77cb2bdccdc582af
parentbfd21f05118c8e9d9ecaa97b9672eb68c248083f
Revert "net/sctp: fix race condition in sctp_destroy_sock"

commit 01bfe5e8e428b475982a98a46cca5755726f3f7f upstream.

This reverts commit b166a20b07382b8bc1dcee2a448715c9c2c81b5b.

This one has to be reverted as it introduced a dead lock, as
syzbot reported:

       CPU0                    CPU1
       ----                    ----
  lock(&net->sctp.addr_wq_lock);
                               lock(slock-AF_INET6);
                               lock(&net->sctp.addr_wq_lock);
  lock(slock-AF_INET6);

CPU0 is the thread of sctp_addr_wq_timeout_handler(), and CPU1
is that of sctp_close().

The original issue this commit fixed will be fixed in the next
patch.

Reported-by: syzbot+959223586843e69a2674@syzkaller.appspotmail.com
Signed-off-by: Xin Long <lucien.xin@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
net/sctp/socket.c