WARNING: proc registration bug in hashlimit_mt_check_common

WARNING: proc registration bug in hashlimit_mt_check_common

[syzbot]

Hello,

syzbot found the following crash on:

HEAD commit:    2981de74 Add linux-next specific files for 20200210
git tree:       linux-next
console output: https://syzkaller.appspot.com/x/log.txt?x=104b16b5e00000
kernel config:  https://syzkaller.appspot.com/x/.config?x=53fc3c3fcb36274f
dashboard link: https://syzkaller.appspot.com/bug?extid=d195fd3b9a364ddd6731
compiler:       gcc (GCC) 9.0.0 20181231 (experimental)
syz repro:      https://syzkaller.appspot.com/x/repro.syz?x=136321d9e00000
C reproducer:   https://syzkaller.appspot.com/x/repro.c?x=159f7431e00000

The bug was bisected to:

commit 8d0015a7ab76b8b1e89a3e5f5710a6e5103f2dd5
Author: Cong Wang <xiyou.wangcong@gmail.com>
Date:   Mon Feb 3 04:30:53 2020 +0000

    netfilter: xt_hashlimit: limit the max size of hashtable

bisection log:  https://syzkaller.appspot.com/x/bisect.txt?x=12a7f25ee00000
final crash:    https://syzkaller.appspot.com/x/report.txt?x=11a7f25ee00000
console output: https://syzkaller.appspot.com/x/log.txt?x=16a7f25ee00000

IMPORTANT: if you fix the bug, please add the following tag to the commit:
Reported-by: syzbot+d195fd3b9a364ddd6731@syzkaller.appspotmail.com
Fixes: 8d0015a7ab76 ("netfilter: xt_hashlimit: limit the max size of hashtable")

xt_hashlimit: size too large, truncated to 1048576
xt_hashlimit: max too large, truncated to 1048576
------------[ cut here ]------------
proc_dir_entry 'ip6t_hashlimit/syzkaller1' already registered
WARNING: CPU: 1 PID: 9858 at fs/proc/generic.c:362 proc_register+0x41e/0x590 fs/proc/generic.c:362
Kernel panic - not syncing: panic_on_warn set ...
CPU: 1 PID: 9858 Comm: syz-executor532 Not tainted 5.5.0-next-20200210-syzkaller #0
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+0x197/0x210 lib/dump_stack.c:118
 panic+0x2e3/0x75c kernel/panic.c:221
 __warn.cold+0x2f/0x3e kernel/panic.c:582
 report_bug+0x289/0x300 lib/bug.c:195
 fixup_bug arch/x86/kernel/traps.c:174 [inline]
 fixup_bug arch/x86/kernel/traps.c:169 [inline]
 do_error_trap+0x11b/0x200 arch/x86/kernel/traps.c:267
 do_invalid_op+0x37/0x50 arch/x86/kernel/traps.c:286
 invalid_op+0x23/0x30 arch/x86/entry/entry_64.S:1027
RIP: 0010:proc_register+0x41e/0x590 fs/proc/generic.c:362
Code: ff df 48 89 f9 48 c1 e9 03 80 3c 01 00 0f 85 5a 01 00 00 48 8b 45 d0 48 c7 c7 e0 1e 59 88 48 8b b0 d0 00 00 00 e8 11 af 5e ff <0f> 0b 48 c7 c7 e0 44 cb 89 e8 24 77 08 06 48 8b 4d a0 48 b8 00 00
RSP: 0018:ffffc90002117550 EFLAGS: 00010286
RAX: 0000000000000000 RBX: ffff8880a88a0330 RCX: 0000000000000000
RDX: 0000000000000000 RSI: ffffffff815ec996 RDI: fffff52000422e9c
RBP: ffffc900021175b8 R08: ffff8880a45343c0 R09: fffffbfff16a376e
R10: fffffbfff16a376d R11: ffffffff8b51bb6f R12: ffff88809ce38b80
R13: 0000000000000000 R14: ffff88809de6f9f8 R15: dffffc0000000000
 proc_create_seq_private+0x12b/0x190 fs/proc/generic.c:593
 htable_create net/netfilter/xt_hashlimit.c:341 [inline]
 hashlimit_mt_check_common.isra.0+0xb30/0x1680 net/netfilter/xt_hashlimit.c:902
 hashlimit_mt_check_v1+0x325/0x3ab net/netfilter/xt_hashlimit.c:928
 xt_check_match+0x280/0x690 net/netfilter/x_tables.c:501
 check_match net/ipv6/netfilter/ip6_tables.c:489 [inline]
 find_check_match net/ipv6/netfilter/ip6_tables.c:506 [inline]
 find_check_entry.isra.0+0x389/0x9d0 net/ipv6/netfilter/ip6_tables.c:557
 translate_table+0xd15/0x1860 net/ipv6/netfilter/ip6_tables.c:734
 do_replace net/ipv6/netfilter/ip6_tables.c:1153 [inline]
 do_ip6t_set_ctl+0x2fe/0x4c8 net/ipv6/netfilter/ip6_tables.c:1681
 nf_sockopt net/netfilter/nf_sockopt.c:106 [inline]
 nf_setsockopt+0x77/0xd0 net/netfilter/nf_sockopt.c:115
 ipv6_setsockopt net/ipv6/ipv6_sockglue.c:949 [inline]
 ipv6_setsockopt+0x147/0x180 net/ipv6/ipv6_sockglue.c:933
 rawv6_setsockopt+0x5e/0x150 net/ipv6/raw.c:1081
 sock_common_setsockopt+0x94/0xd0 net/core/sock.c:3149
 __sys_setsockopt+0x261/0x4c0 net/socket.c:2130
 __do_sys_setsockopt net/socket.c:2146 [inline]
 __se_sys_setsockopt net/socket.c:2143 [inline]
 __x64_sys_setsockopt+0xbe/0x150 net/socket.c:2143
 do_syscall_64+0xfa/0x790 arch/x86/entry/common.c:294
 entry_SYSCALL_64_after_hwframe+0x49/0xbe
RIP: 0033:0x4482b9
Code: e8 ec 14 03 00 48 83 c4 18 c3 0f 1f 80 00 00 00 00 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 0f 83 9b 0c fc ff c3 66 2e 0f 1f 84 00 00 00 00
RSP: 002b:00007f5cbd09cda8 EFLAGS: 00000246 ORIG_RAX: 0000000000000036
RAX: ffffffffffffffda RBX: 00000000006dec48 RCX: 00000000004482b9
RDX: 0000000000000040 RSI: 0000000000000029 RDI: 0000000000000004
RBP: 00000000006dec40 R08: 0000000000000001 R09: 0000000000000000
R10: 0000000020000000 R11: 0000000000000246 R12: 00000000006dec4c
R13: 0000000020000000 R14: 00000000004b09c0 R15: 000000000000002d
Kernel Offset: disabled
Rebooting in 86400 seconds..


---
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 syzkaller@googlegroups.com.

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

Re: WARNING: proc registration bug in hashlimit_mt_check_common

[Cong Wang]

On Mon, Feb 10, 2020 at 10:35 AM syzbot
<syzbot+d195fd3b9a364ddd6731@syzkaller.appspotmail.com> wrote:
>
> Hello,
>
> syzbot found the following crash on:
>
> HEAD commit:    2981de74 Add linux-next specific files for 20200210
> git tree:       linux-next
> console output: https://syzkaller.appspot.com/x/log.txt?x=104b16b5e00000
> kernel config:  https://syzkaller.appspot.com/x/.config?x=53fc3c3fcb36274f
> dashboard link: https://syzkaller.appspot.com/bug?extid=d195fd3b9a364ddd6731
> compiler:       gcc (GCC) 9.0.0 20181231 (experimental)
> syz repro:      https://syzkaller.appspot.com/x/repro.syz?x=136321d9e00000
> C reproducer:   https://syzkaller.appspot.com/x/repro.c?x=159f7431e00000
>
> The bug was bisected to:
>
> commit 8d0015a7ab76b8b1e89a3e5f5710a6e5103f2dd5
> Author: Cong Wang <xiyou.wangcong@gmail.com>
> Date:   Mon Feb 3 04:30:53 2020 +0000
>
>     netfilter: xt_hashlimit: limit the max size of hashtable
>
> bisection log:  https://syzkaller.appspot.com/x/bisect.txt?x=12a7f25ee00000
> final crash:    https://syzkaller.appspot.com/x/report.txt?x=11a7f25ee00000
> console output: https://syzkaller.appspot.com/x/log.txt?x=16a7f25ee00000
>
> IMPORTANT: if you fix the bug, please add the following tag to the commit:
> Reported-by: syzbot+d195fd3b9a364ddd6731@syzkaller.appspotmail.com
> Fixes: 8d0015a7ab76 ("netfilter: xt_hashlimit: limit the max size of hashtable")
>
> xt_hashlimit: size too large, truncated to 1048576
> xt_hashlimit: max too large, truncated to 1048576
> ------------[ cut here ]------------
> proc_dir_entry 'ip6t_hashlimit/syzkaller1' already registered

I think we probably have to remove the procfs file too before releasing
the global mutex. Or, we can mark the table as deleted before actually
delete it, but this requires to change the search logic as well.

I will work on a patch.

Thanks.