[syzbot] [net?] [s390?] possible deadlock in smc_switch_to_fallback (2)

[syzbot]

Hello,

syzbot found the following issue on:

HEAD commit: 568ebdaba637 MAINTAINERS: adjust file entry in FREESCALE Q..
git tree: net-next
console output: https://syzkaller.appspot.com/x/log.txt?x=12f58a61980000
kernel config: https://syzkaller.appspot.com/x/.config?x=e78fc116033e0ab7
dashboard link: https://syzkaller.appspot.com/bug?extid=bef85a6996d1737c1a2f
compiler: Debian clang version 15.0.6, GNU ld (GNU Binutils for Debian) 2.40

Unfortunately, I don't have any reproducer for this issue yet.

Downloadable assets:
disk image: https://storage.googleapis.com/syzbot-assets/1d754ea220a6/disk-568ebdab.raw.xz
vmlinux: https://storage.googleapis.com/syzbot-assets/232f2545fca4/vmlinux-568ebdab.xz
kernel image: https://storage.googleapis.com/syzbot-assets/6398bb41810d/bzImage-568ebdab.xz

IMPORTANT: if you fix the issue, please add the following tag to the commit:
Reported-by: syzbot+bef85a...@syzkaller.appspotmail.com

======================================================
WARNING: possible circular locking dependency detected
6.10.0-rc4-syzkaller-00875-g568ebdaba637 #0 Not tainted
------------------------------------------------------
syz-executor.1/11818 is trying to acquire lock:
ffff888023600a50 (&smc->clcsock_release_lock){+.+.}-{3:3}, at: smc_switch_to_fallback+0x35/0xd00 net/smc/af_smc.c:902

but task is already holding lock:
ffff888023600258 (sk_lock-AF_INET){+.+.}-{0:0}, at: lock_sock include/net/sock.h:1602 [inline]
ffff888023600258 (sk_lock-AF_INET){+.+.}-{0:0}, at: smc_sendmsg+0x55/0x530 net/smc/af_smc.c:2773

which lock already depends on the new lock.


the existing dependency chain (in reverse order) is:

-> #2 (sk_lock-AF_INET){+.+.}-{0:0}:
lock_acquire+0x1ed/0x550 kernel/locking/lockdep.c:5754
lock_sock_nested+0x48/0x100 net/core/sock.c:3543
do_ip_setsockopt+0x1a2d/0x3cd0 net/ipv4/ip_sockglue.c:1078
ip_setsockopt+0x63/0x100 net/ipv4/ip_sockglue.c:1417
do_sock_setsockopt+0x3af/0x720 net/socket.c:2312
__sys_setsockopt+0x1ae/0x250 net/socket.c:2335
__do_sys_setsockopt net/socket.c:2344 [inline]
__se_sys_setsockopt net/socket.c:2341 [inline]
__x64_sys_setsockopt+0xb5/0xd0 net/socket.c:2341
do_syscall_x64 arch/x86/entry/common.c:52 [inline]
do_syscall_64+0xf3/0x230 arch/x86/entry/common.c:83
entry_SYSCALL_64_after_hwframe+0x77/0x7f

-> #1 (rtnl_mutex){+.+.}-{3:3}:
lock_acquire+0x1ed/0x550 kernel/locking/lockdep.c:5754
__mutex_lock_common kernel/locking/mutex.c:608 [inline]
__mutex_lock+0x136/0xd70 kernel/locking/mutex.c:752
do_ip_setsockopt+0x127d/0x3cd0 net/ipv4/ip_sockglue.c:1077
ip_setsockopt+0x63/0x100 net/ipv4/ip_sockglue.c:1417
smc_setsockopt+0x275/0xe50 net/smc/af_smc.c:3072
do_sock_setsockopt+0x3af/0x720 net/socket.c:2312
__sys_setsockopt+0x1ae/0x250 net/socket.c:2335
__do_sys_setsockopt net/socket.c:2344 [inline]
__se_sys_setsockopt net/socket.c:2341 [inline]
__x64_sys_setsockopt+0xb5/0xd0 net/socket.c:2341
do_syscall_x64 arch/x86/entry/common.c:52 [inline]
do_syscall_64+0xf3/0x230 arch/x86/entry/common.c:83
entry_SYSCALL_64_after_hwframe+0x77/0x7f

-> #0 (&smc->clcsock_release_lock){+.+.}-{3:3}:
check_prev_add kernel/locking/lockdep.c:3134 [inline]
check_prevs_add kernel/locking/lockdep.c:3253 [inline]
validate_chain+0x18e0/0x5900 kernel/locking/lockdep.c:3869
__lock_acquire+0x1346/0x1fd0 kernel/locking/lockdep.c:5137
lock_acquire+0x1ed/0x550 kernel/locking/lockdep.c:5754
__mutex_lock_common kernel/locking/mutex.c:608 [inline]
__mutex_lock+0x136/0xd70 kernel/locking/mutex.c:752
smc_switch_to_fallback+0x35/0xd00 net/smc/af_smc.c:902
smc_sendmsg+0x11f/0x530 net/smc/af_smc.c:2779
sock_sendmsg_nosec net/socket.c:730 [inline]
__sock_sendmsg+0x221/0x270 net/socket.c:745
____sys_sendmsg+0x525/0x7d0 net/socket.c:2585
___sys_sendmsg net/socket.c:2639 [inline]
__sys_sendmsg+0x2b0/0x3a0 net/socket.c:2668
do_syscall_x64 arch/x86/entry/common.c:52 [inline]
do_syscall_64+0xf3/0x230 arch/x86/entry/common.c:83
entry_SYSCALL_64_after_hwframe+0x77/0x7f

other info that might help us debug this:

Chain exists of:
&smc->clcsock_release_lock --> rtnl_mutex --> sk_lock-AF_INET

Possible unsafe locking scenario:

CPU0 CPU1
---- ----
lock(sk_lock-AF_INET);
lock(rtnl_mutex);
lock(sk_lock-AF_INET);
lock(&smc->clcsock_release_lock);

*** DEADLOCK ***

1 lock held by syz-executor.1/11818:
#0: ffff888023600258 (sk_lock-AF_INET){+.+.}-{0:0}, at: lock_sock include/net/sock.h:1602 [inline]
#0: ffff888023600258 (sk_lock-AF_INET){+.+.}-{0:0}, at: smc_sendmsg+0x55/0x530 net/smc/af_smc.c:2773

stack backtrace:
CPU: 0 PID: 11818 Comm: syz-executor.1 Not tainted 6.10.0-rc4-syzkaller-00875-g568ebdaba637 #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 06/07/2024
Call Trace:
<TASK>
__dump_stack lib/dump_stack.c:88 [inline]
dump_stack_lvl+0x241/0x360 lib/dump_stack.c:114


---
This report 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 issue. See:
https://goo.gl/tpsmEJ#status for how to communicate with syzbot.

If the report is already addressed, let syzbot know by replying with:
#syz fix: exact-commit-title

If you want to overwrite report's subsystems, reply with:
#syz set subsystems: new-subsystem
(See the list of subsystem names on the web dashboard)

If the report is a duplicate of another one, reply with:
#syz dup: exact-subject-of-another-report

If you want to undo deduplication, reply with:
#syz undup