possible deadlock in start_sync_thread

[syzbot]

Hello,

syzbot found the following crash on:

HEAD commit: d7e64f80 ANDROID: x86_64_cuttlefish_defconfig: Enable F2FS
git tree: android-4.9
console output: https://syzkaller.appspot.com/x/log.txt?x=1726c59f800000
kernel config: https://syzkaller.appspot.com/x/.config?x=f99aced9dd6a7628
dashboard link: https://syzkaller.appspot.com/bug?extid=da6fbb7d5548c83a9a19
compiler: gcc (GCC) 8.0.1 20180413 (experimental)
syz repro: https://syzkaller.appspot.com/x/repro.syz?x=15a6d48f800000
C reproducer: https://syzkaller.appspot.com/x/repro.c?x=14fc81d7800000

IMPORTANT: if you fix the bug, please add the following tag to the commit:
Reported-by: syzbot+da6fbb...@syzkaller.appspotmail.com

IPv6: ADDRCONF(NETDEV_UP): veth1: link is not ready
IPv6: ADDRCONF(NETDEV_CHANGE): veth1: link becomes ready
IPv6: ADDRCONF(NETDEV_CHANGE): veth0: link becomes ready

======================================================
[ INFO: possible circular locking dependency detected ]
4.9.105-gd7e64f8 #40 Not tainted
-------------------------------------------------------
syz-executor124/3795 is trying to acquire lock:
(sk_lock-AF_INET){+.+.+.}, at: [<ffffffff832f6a9b>] lock_sock
include/net/sock.h:1404 [inline]
(sk_lock-AF_INET){+.+.+.}, at: [<ffffffff832f6a9b>] set_mcast_if
net/netfilter/ipvs/ip_vs_sync.c:1370 [inline]
(sk_lock-AF_INET){+.+.+.}, at: [<ffffffff832f6a9b>] make_send_sock
net/netfilter/ipvs/ip_vs_sync.c:1490 [inline]
(sk_lock-AF_INET){+.+.+.}, at: [<ffffffff832f6a9b>]
start_sync_thread+0x88b/0x1e80 net/netfilter/ipvs/ip_vs_sync.c:1885
but task is already holding lock:
(rtnl_mutex){+.+.+.}, at: [<ffffffff830b5937>] rtnl_lock+0x17/0x20
net/core/rtnetlink.c:70
which lock already depends on the new lock.


the existing dependency chain (in reverse order) is:

lock_acquire+0x130/0x3e0 kernel/locking/lockdep.c:3756
__mutex_lock_common kernel/locking/mutex.c:521 [inline]
mutex_lock_nested+0xc0/0x870 kernel/locking/mutex.c:621
rtnl_lock+0x17/0x20 net/core/rtnetlink.c:70
mrtsock_destruct+0x3b/0x1e0 net/ipv4/ipmr.c:1231
ip_ra_control+0x2c2/0x420 net/ipv4/ip_sockglue.c:360
do_ip_setsockopt.isra.13+0x15ff/0x2b10 net/ipv4/ip_sockglue.c:1139
ip_setsockopt+0x3a/0xb0 net/ipv4/ip_sockglue.c:1242
raw_setsockopt+0xb7/0xd0 net/ipv4/raw.c:833
sock_common_setsockopt+0x9a/0xe0 net/core/sock.c:2706
SYSC_setsockopt net/socket.c:1772 [inline]
SyS_setsockopt+0x166/0x260 net/socket.c:1751
do_syscall_64+0x1a6/0x490 arch/x86/entry/common.c:282
entry_SYSCALL_64_after_swapgs+0x5d/0xdb

check_prev_add kernel/locking/lockdep.c:1828 [inline]
check_prevs_add kernel/locking/lockdep.c:1938 [inline]
validate_chain kernel/locking/lockdep.c:2265 [inline]
__lock_acquire+0x3019/0x4070 kernel/locking/lockdep.c:3345
lock_acquire+0x130/0x3e0 kernel/locking/lockdep.c:3756
lock_sock_nested+0xc6/0x120 net/core/sock.c:2511
lock_sock include/net/sock.h:1404 [inline]
set_mcast_if net/netfilter/ipvs/ip_vs_sync.c:1370 [inline]
make_send_sock net/netfilter/ipvs/ip_vs_sync.c:1490 [inline]
start_sync_thread+0x88b/0x1e80 net/netfilter/ipvs/ip_vs_sync.c:1885
do_ip_vs_set_ctl+0x90f/0xbb0 net/netfilter/ipvs/ip_vs_ctl.c:2396
nf_sockopt net/netfilter/nf_sockopt.c:105 [inline]
nf_setsockopt+0x6d/0xc0 net/netfilter/nf_sockopt.c:114
ip_setsockopt+0x9a/0xb0 net/ipv4/ip_sockglue.c:1249
udp_setsockopt+0x4a/0x90 net/ipv4/udp.c:2091
ipv6_setsockopt+0x10a/0x130 net/ipv6/ipv6_sockglue.c:901
tcp_setsockopt+0x88/0xe0 net/ipv4/tcp.c:2758
sock_common_setsockopt+0x9a/0xe0 net/core/sock.c:2706
SYSC_setsockopt net/socket.c:1772 [inline]
SyS_setsockopt+0x166/0x260 net/socket.c:1751
do_syscall_64+0x1a6/0x490 arch/x86/entry/common.c:282
entry_SYSCALL_64_after_swapgs+0x5d/0xdb

other info that might help us debug this:

Possible unsafe locking scenario:

CPU0 CPU1
---- ----
lock(rtnl_mutex);
lock(sk_lock-AF_INET);
lock(rtnl_mutex);
lock(sk_lock-AF_INET);

*** DEADLOCK ***

2 locks held by syz-executor124/3795:
#0: (rtnl_mutex){+.+.+.}, at: [<ffffffff830b5937>] rtnl_lock+0x17/0x20
net/core/rtnetlink.c:70
#1: (ipvs->sync_mutex){+.+.+.}, at: [<ffffffff832f6328>]
start_sync_thread+0x118/0x1e80 net/netfilter/ipvs/ip_vs_sync.c:1779

stack backtrace:
CPU: 0 PID: 3795 Comm: syz-executor124 Not tainted 4.9.105-gd7e64f8 #40
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS
Google 01/01/2011
ffff8801d9747538 ffffffff81eb41a9 ffffffff855e7690 ffffffff8559cf30
ffffffff855e7690 ffff8801b752a0e8 ffff8801b7529800 ffff8801d9747580
ffffffff8142679d 0000000000000001 00000000b7529800 0000000000000001
Call Trace:
[<ffffffff81eb41a9>] __dump_stack lib/dump_stack.c:15 [inline]
[<ffffffff81eb41a9>] dump_stack+0xc1/0x128 lib/dump_stack.c:51
[<ffffffff8142679d>] print_circular_bug.cold.51+0x1bd/0x27d
kernel/locking/lockdep.c:1202
[<ffffffff812396b9>] check_prev_add kernel/locking/lockdep.c:1828 [inline]
[<ffffffff812396b9>] check_prevs_add kernel/locking/lockdep.c:1938 [inline]
[<ffffffff812396b9>] validate_chain kernel/locking/lockdep.c:2265 [inline]
[<ffffffff812396b9>] __lock_acquire+0x3019/0x4070
kernel/locking/lockdep.c:3345
[<ffffffff8123b180>] lock_acquire+0x130/0x3e0 kernel/locking/lockdep.c:3756
[<ffffffff830246a6>] lock_sock_nested+0xc6/0x120 net/core/sock.c:2511
[<ffffffff832f6a9b>] lock_sock include/net/sock.h:1404 [inline]
[<ffffffff832f6a9b>] set_mcast_if net/netfilter/ipvs/ip_vs_sync.c:1370
[inline]
[<ffffffff832f6a9b>] make_send_sock net/netfilter/ipvs/ip_vs_sync.c:1490
[inline]
[<ffffffff832f6a9b>] start_sync_thread+0x88b/0x1e80
net/netfilter/ipvs/ip_vs_sync.c:1885
[<ffffffff832e399f>] do_ip_vs_set_ctl+0x90f/0xbb0
net/netfilter/ipvs/ip_vs_ctl.c:2396
[<ffffffff831df12d>] nf_sockopt net/netfilter/nf_sockopt.c:105 [inline]
[<ffffffff831df12d>] nf_setsockopt+0x6d/0xc0 net/netfilter/nf_sockopt.c:114
[<ffffffff8334931a>] ip_setsockopt+0x9a/0xb0 net/ipv4/ip_sockglue.c:1249
[<ffffffff833f7d7a>] udp_setsockopt+0x4a/0x90 net/ipv4/udp.c:2091
[<ffffffff835b4cea>] ipv6_setsockopt+0x10a/0x130
net/ipv6/ipv6_sockglue.c:901
[<ffffffff833691f8>] tcp_setsockopt+0x88/0xe0 net/ipv4/tcp.c:2758
[<ffffffff8301d75a>] sock_common_setsockopt+0x9a/0xe0 net/core/sock.c:2706
[<ffffffff8301a516>] SYSC_setsockopt net/socket.c:1772 [inline]
[<ffffffff8301a516>] SyS_setsockopt+0x166/0x260 net/socket.c:1751
[<ffffffff81006316>] do_syscall_64+0x1a6/0x490 arch/x86/entry/common.c:282
[<ffffffff839f9b13>] entry_SYSCALL_64_after_swapgs+0x5d/0xdb
IPVS: sync thread started: state = MASTER, mcast_ifn = veth0_to_bridge,
syncid = 3, id = 0


---
This bug is generated by a bot. It may contain errors.
See https://goo.gl/tpsmEJ for more information about syzbot.
syzbot engineers can be reached at syzk...@googlegroups.com.

syzbot will keep track of this bug report. See:
https://goo.gl/tpsmEJ#status for how to communicate with syzbot.
syzbot can test patches for this bug, for details see:
https://goo.gl/tpsmEJ#testing-patches