Re: [PATCH net v2] net/smc: fix general protection fault in __smc_diag_dump

[Kuniyuki Iwashima <kuniyu@google.com>]

From: Wang Liang <wangliang74@huawei.com>
Date: Fri, 17 Oct 2025 10:48:27 +0800
> The syzbot report a crash:
> 
>   Oops: general protection fault, probably for non-canonical address 0xfbd5a5d5a0000003: 0000 [#1] SMP KASAN NOPTI
>   KASAN: maybe wild-memory-access in range [0xdead4ead00000018-0xdead4ead0000001f]
>   CPU: 1 UID: 0 PID: 6949 Comm: syz.0.335 Not tainted syzkaller #0 PREEMPT(full)
>   Hardware name: Google Compute Engine/Google Compute Engine, BIOS Google 08/18/2025
>   RIP: 0010:smc_diag_msg_common_fill net/smc/smc_diag.c:44 [inline]
>   RIP: 0010:__smc_diag_dump.constprop.0+0x3ca/0x2550 net/smc/smc_diag.c:89
>   Call Trace:
>    <TASK>
>    smc_diag_dump_proto+0x26d/0x420 net/smc/smc_diag.c:217
>    smc_diag_dump+0x27/0x90 net/smc/smc_diag.c:234
>    netlink_dump+0x539/0xd30 net/netlink/af_netlink.c:2327
>    __netlink_dump_start+0x6d6/0x990 net/netlink/af_netlink.c:2442
>    netlink_dump_start include/linux/netlink.h:341 [inline]
>    smc_diag_handler_dump+0x1f9/0x240 net/smc/smc_diag.c:251
>    __sock_diag_cmd net/core/sock_diag.c:249 [inline]
>    sock_diag_rcv_msg+0x438/0x790 net/core/sock_diag.c:285
>    netlink_rcv_skb+0x158/0x420 net/netlink/af_netlink.c:2552
>    netlink_unicast_kernel net/netlink/af_netlink.c:1320 [inline]
>    netlink_unicast+0x5a7/0x870 net/netlink/af_netlink.c:1346
>    netlink_sendmsg+0x8d1/0xdd0 net/netlink/af_netlink.c:1896
>    sock_sendmsg_nosec net/socket.c:714 [inline]
>    __sock_sendmsg net/socket.c:729 [inline]
>    ____sys_sendmsg+0xa95/0xc70 net/socket.c:2614
>    ___sys_sendmsg+0x134/0x1d0 net/socket.c:2668
>    __sys_sendmsg+0x16d/0x220 net/socket.c:2700
>    do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline]
>    do_syscall_64+0xcd/0x4e0 arch/x86/entry/syscall_64.c:94
>    entry_SYSCALL_64_after_hwframe+0x77/0x7f
>    </TASK>
> 
> The process like this:
> 
>                (CPU1)              |             (CPU2)
>   ---------------------------------|-------------------------------
>   inet_create()                    |
>     // init clcsock to NULL        |
>     sk = sk_alloc()                |
>                                    |
>     // unexpectedly change clcsock |
>     inet_init_csk_locks()          |
>                                    |
>     // add sk to hash table        |
>     smc_inet_init_sock()           |
>       smc_sk_init()                |
>         smc_hash_sk()              |
>                                    | // traverse the hash table
>                                    | smc_diag_dump_proto
>                                    |   __smc_diag_dump()
>                                    |     // visit wrong clcsock
>                                    |     smc_diag_msg_common_fill()
>     // alloc clcsock               |
>     smc_create_clcsk               |
>       sock_create_kern             |
> 
> With CONFIG_DEBUG_LOCK_ALLOC=y, the smc->clcsock is unexpectedly changed
> in inet_init_csk_locks(). The INET_PROTOSW_ICSK flag is no need by smc,
> just remove it.
> 
> After removing the INET_PROTOSW_ICSK flag, this patch alse revert
> commit 6fd27ea183c2 ("net/smc: fix lacks of icsk_syn_mss with IPPROTO_SMC")
> to avoid casting smc_sock to inet_connection_sock.
> 
> Reported-by: syzbot+f775be4458668f7d220e@syzkaller.appspotmail.com
> Closes: https://syzkaller.appspot.com/bug?extid=f775be4458668f7d220e
> Tested-by: syzbot+f775be4458668f7d220e@syzkaller.appspotmail.com

nit: looks like this diff is not tested by syzbot, you may
want to send diff to syzbot.


> Fixes: d25a92ccae6b ("net/smc: Introduce IPPROTO_SMC")
> Signed-off-by: Wang Liang <wangliang74@huawei.com>

Change itself looks good.

Reviewed-by: Kuniyuki Iwashima <kuniyu@google.com>

Thanks!