Re: possible deadlock in bpf_tcp_close

[syzbot]

syzbot has found a reproducer for the following crash on:

HEAD commit:    7a1a98c171ea Merge branch 'bpf-sendmsg-hook'
git tree:       bpf-next
console output: https://syzkaller.appspot.com/x/log.txt?x=149ae2b7800000
kernel config:  https://syzkaller.appspot.com/x/.config?x=e4078980b886800c
dashboard link: https://syzkaller.appspot.com/bug?extid=47ed903f50684f046b15
compiler:       gcc (GCC) 8.0.1 20180413 (experimental)
syzkaller repro:https://syzkaller.appspot.com/x/repro.syz?x=1553b17b800000
C reproducer:   https://syzkaller.appspot.com/x/repro.c?x=1460be2f800000

IMPORTANT: if you fix the bug, please add the following tag to the commit:
Reported-by: syzbot+47ed903f50684f046b15@xxxxxxxxxxxxxxxxxxxxxxxxx

random: sshd: uninitialized urandom read (32 bytes read)
random: sshd: uninitialized urandom read (32 bytes read)
random: sshd: uninitialized urandom read (32 bytes read)

======================================================
WARNING: possible circular locking dependency detected
4.17.0-rc6+ #25 Not tainted
------------------------------------------------------
syz-executor800/4527 is trying to acquire lock:
        (ptrval) (&htab->buckets[i].lock){+...}, at:  
bpf_tcp_close+0x822/0x10b0 kernel/bpf/sockmap.c:285

but task is already holding lock:
        (ptrval) (clock-AF_INET6){++..}, at: bpf_tcp_close+0x241/0x10b0  
kernel/bpf/sockmap.c:260

which lock already depends on the new lock.


the existing dependency chain (in reverse order) is:

-> #1 (clock-AF_INET6){++..}:
       __raw_write_lock_bh include/linux/rwlock_api_smp.h:203 [inline]
       _raw_write_lock_bh+0x31/0x40 kernel/locking/spinlock.c:312
       sock_hash_delete_elem+0x7c6/0xaf0 kernel/bpf/sockmap.c:2338
       map_delete_elem+0x32e/0x4e0 kernel/bpf/syscall.c:815
       __do_sys_bpf kernel/bpf/syscall.c:2349 [inline]
       __se_sys_bpf kernel/bpf/syscall.c:2317 [inline]
       __x64_sys_bpf+0x342/0x510 kernel/bpf/syscall.c:2317
       do_syscall_64+0x1b1/0x800 arch/x86/entry/common.c:287
       entry_SYSCALL_64_after_hwframe+0x49/0xbe

-> #0 (&htab->buckets[i].lock){+...}:
       lock_acquire+0x1dc/0x520 kernel/locking/lockdep.c:3920
       __raw_spin_lock_bh include/linux/spinlock_api_smp.h:135 [inline]
       _raw_spin_lock_bh+0x31/0x40 kernel/locking/spinlock.c:168
       bpf_tcp_close+0x822/0x10b0 kernel/bpf/sockmap.c:285
       inet_release+0x104/0x1f0 net/ipv4/af_inet.c:427
       inet6_release+0x50/0x70 net/ipv6/af_inet6.c:459
       sock_release+0x96/0x1b0 net/socket.c:594
       sock_close+0x16/0x20 net/socket.c:1149
       __fput+0x34d/0x890 fs/file_table.c:209
       ____fput+0x15/0x20 fs/file_table.c:243
       task_work_run+0x1e4/0x290 kernel/task_work.c:113
       exit_task_work include/linux/task_work.h:22 [inline]
       do_exit+0x1aee/0x2730 kernel/exit.c:865
       do_group_exit+0x16f/0x430 kernel/exit.c:968
       get_signal+0x886/0x1960 kernel/signal.c:2482
       do_signal+0x98/0x2040 arch/x86/kernel/signal.c:810
       exit_to_usermode_loop+0x28a/0x310 arch/x86/entry/common.c:162
       prepare_exit_to_usermode arch/x86/entry/common.c:196 [inline]
       syscall_return_slowpath arch/x86/entry/common.c:265 [inline]
       do_syscall_64+0x6ac/0x800 arch/x86/entry/common.c:290
       entry_SYSCALL_64_after_hwframe+0x49/0xbe

other info that might help us debug this:

 Possible unsafe locking scenario:

       CPU0                    CPU1
       ----                    ----
  lock(clock-AF_INET6);
                               lock(&htab->buckets[i].lock);
                               lock(clock-AF_INET6);
  lock(&htab->buckets[i].lock);

 *** DEADLOCK ***

2 locks held by syz-executor800/4527:
 #0:         (ptrval) (rcu_read_lock){....}, at: bpf_tcp_close+0x0/0x10b0  
kernel/bpf/sockmap.c:2106
 #1:         (ptrval) (clock-AF_INET6){++..}, at:  
bpf_tcp_close+0x241/0x10b0 kernel/bpf/sockmap.c:260

stack backtrace:
CPU: 0 PID: 4527 Comm: syz-executor800 Not tainted 4.17.0-rc6+ #25
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS  
Google 01/01/2011
Call Trace:
 __dump_stack lib/dump_stack.c:77 [inline]
 dump_stack+0x1b9/0x294 lib/dump_stack.c:113
 print_circular_bug.isra.36.cold.54+0x1bd/0x27d  
kernel/locking/lockdep.c:1223
 check_prev_add kernel/locking/lockdep.c:1863 [inline]
 check_prevs_add kernel/locking/lockdep.c:1976 [inline]
 validate_chain kernel/locking/lockdep.c:2417 [inline]
 __lock_acquire+0x343e/0x5140 kernel/locking/lockdep.c:3431
 lock_acquire+0x1dc/0x520 kernel/locking/lockdep.c:3920
 __raw_spin_lock_bh include/linux/spinlock_api_smp.h:135 [inline]
 _raw_spin_lock_bh+0x31/0x40 kernel/locking/spinlock.c:168
 bpf_tcp_close+0x822/0x10b0 kernel/bpf/sockmap.c:285
 inet_release+0x104/0x1f0 net/ipv4/af_inet.c:427
 inet6_release+0x50/0x70 net/ipv6/af_inet6.c:459
 sock_release+0x96/0x1b0 net/socket.c:594
 sock_close+0x16/0x20 net/socket.c:1149
 __fput+0x34d/0x890 fs/file_table.c:209
 ____fput+0x15/0x20 fs/file_table.c:243
 task_work_run+0x1e4/0x290 kernel/task_work.c:113
 exit_task_work include/linux/task_work.h:22 [inline]
 do_exit+0x1aee/0x2730 kernel/exit.c:865
 do_group_exit+0x16f/0x430 kernel/exit.c:968
 get_signal+0x886/0x1960 kernel/signal.c:2482
 do_signal+0x98/0x2040 arch/x86/kernel/signal.c:810
 exit_to_usermode_loop+0x28a/0x310 arch/x86/entry/common.c:162
 prepare_exit_to_usermode arch/x86/entry/common.c:196 [inline]
 syscall_return_slowpath arch/x86/entry/common.c:265 [inline]
 do_syscall_64+0x6ac/0x800 arch/x86/entry/common.c:290
 entry_SYSCALL_64_after_hwframe+0x49/0xbe
RIP: 0033:0x445709
RSP: 002b:00007f36c605ddb8 EFLAGS: 00000246 ORIG_RAX: 00000000000000ca
RAX: fffffffffffffe00 RBX: 00000000006dac3c RCX: 0000000000445709
RDX: 0000000000000000 RSI: 0000000000000000 RDI: 00000000006dac3c
RBP: 00000000006dac38 R08: 0000000000000000 R09: 000