Re: general protection fault in sctp_assoc_rwnd_increase

[Dmitry Vyukov]

On Tue, Mar 12, 2019 at 5:30 PM Xin Long <lucien.xin@xxxxxxxxx> wrote:
>
> On Tue, Mar 12, 2019 at 9:52 PM syzbot
> <syzbot+85e0b422ff140b03672a@xxxxxxxxxxxxxxxxxxxxxxxxx> wrote:
> >
> > Hello,
> >
> > syzbot found the following crash on:
> >
> > HEAD commit:    a089e4fe Merge tag 'linux-watchdog-5.1-rc1' of git://www.l..
> > git tree:       upstream
> > console output: https://syzkaller.appspot.com/x/log.txt?x=144e9ee7200000
> > kernel config:  https://syzkaller.appspot.com/x/.config?x=b613f0327d980b6b
> > dashboard link: https://syzkaller.appspot.com/bug?extid=85e0b422ff140b03672a
> > compiler:       gcc (GCC) 9.0.0 20181231 (experimental)
> > userspace arch: amd64
> > syz repro:      https://syzkaller.appspot.com/x/repro.syz?x=12613b13200000
> > C reproducer:   https://syzkaller.appspot.com/x/repro.c?x=10f1884d200000
> Not sure why I got this Call Trace with the reproducer, and it's
> different, is that expected?

Hi Xin,

This is no a crash. This is fault injection stack trace.
Fault injection subsystem prints stack whenever it injects a fault so
that it's possible o figure out what happened and where.

You can see a fault injection stack trace in the "console output' for
the crash too.
But if seems that your stack trace is slightly different, and that
most likely explains why you can't reproduce the bug.
This seems to be a bug that it triggered by a very particular failing
kmalloc. If on your kernel fault is injected into a different kmalloc,
you won't reproduce the bug.
For fault injection-triggered bugs you need precisely the same kernel.
You can actually see this in action: for smack and apparmor-based
kernels fault needs to be injected into 7-th and 8-th kmalloc in the
syscall respectively:

https://syzkaller.appspot.com/text?tag=ReproSyz&x=12613b13200000
https://syzkaller.appspot.com/text?tag=ReproSyz&x=14425577200000

Or you can play with the fault_nth parameter to target fault injection
into the right place... Or of course just change the kernel code to
fail that kmalloc explicitly.


> [256577.101170] FAULT_INJECTION: forcing a failure.
> [256577.101170] name failslab, interval 1, probability 0, space 0, times 0
> [256577.103628] CPU: 0 PID: 2453 Comm: syz-executor Not tainted
> 5.0.0.test.syz #235
> [256577.105201] Hardware name: Red Hat KVM, BIOS seabios-1.7.5-8.el7 04/01/2014
> [256577.106680] Call Trace:
> [256577.107248]  dump_stack+0x7c/0xc0
> [256577.107981]  should_fail.cold.4+0x5/0x13
> [256577.108853]  ? fault_create_debugfs_attr+0x190/0x190
> [256577.109926]  ? lock_downgrade+0x5d0/0x5d0
> [256577.110816]  ? selinux_sk_alloc_security+0x72/0x190
> [256577.111878]  ? selinux_sk_alloc_security+0x72/0x190
> [256577.112931]  should_failslab+0xa/0x20
> [256577.113738]  kmem_cache_alloc_trace+0x27a/0x350
> [256577.114733]  selinux_sk_alloc_security+0x72/0x190
> [256577.115757]  ? kmem_cache_alloc+0x2dc/0x310
> [256577.116674]  security_sk_alloc+0x4f/0x90
> [256577.117543]  sk_prot_alloc+0x82/0x250
> [256577.118362]  sk_alloc+0x35/0xc80
> [256577.119100]  inet6_create+0x26e/0xec0
> [256577.119924]  __sock_create+0x277/0x550
> [256577.120804]  sctp_do_peeloff+0x162/0x3b0 [sctp]
> [256577.121795]  ? sched_clock+0x5/0x10
> [256577.122596]  ? sctp_copy_sock+0xfa0/0xfa0 [sctp]
> [256577.123646]  sctp_getsockopt_peeloff_common.isra.31+0x78/0x2c0 [sctp]
> [256577.125064]  ? sctp_do_peeloff+0x3b0/0x3b0 [sctp]
> [256577.126129]  sctp_getsockopt.part.32+0x2d76/0x4a90 [sctp]
>
> >
> > IMPORTANT: if you fix the bug, please add the following tag to the commit:
> > Reported-by: syzbot+85e0b422ff140b03672a@xxxxxxxxxxxxxxxxxxxxxxxxx
> >
> > RAX: ffffffffffffffda RBX: 00000000006dbc38 RCX: 0000000000446679
> > RDX: 0000000000000066 RSI: 0000000000000084 RDI: 0000000000000003
> > RBP: 00000000006dbc30 R08: 0000000020000140 R09: 0000000000000038
> > R10: 0000000020000040 R11: 0000000000000246 R12: 00000000006dbc3c
> > R13: 00007f47f49ded90 R14: 0000000000000004 R15: 20c49ba5e353f7cf
> > kasan: CONFIG_KASAN_INLINE enabled
> > kasan: GPF could be caused by NULL-ptr deref or user memory access
> > general protection fault: 0000 [#1] PREEMPT SMP KASAN
> > CPU: 0 PID: 7680 Comm: syz-executor161 Not tainted 5.0.0+ #18
> > Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS
> > Google 01/01/2011
> > RIP: 0010:sctp_assoc_rwnd_increase+0x34/0x520 net/sctp/associola.c:1498
> > Code: 41 54 49 89 fc 53 89 f3 48 83 ec 10 e8 95 90 f1 fa 49 8d bc 24 60 06
> > 00 00 48 b8 00 00 00 00 00 fc ff df 48 89 fa 48 c1 ea 03 <0f> b6 04 02 84
> > c0 74 08 3c 03 0f 8e 2f 04 00 00 45 8b ac 24 60 06
> > RSP: 0018:ffff88808da476f8 EFLAGS: 00010203
> > RAX: dffffc0000000000 RBX: 0000000000000000 RCX: 0000000000000001
> > RDX: 00000000000000cb RSI: ffffffff867ecd5b RDI: 000000000000065f
> > RBP: ffff88808da47730 R08: ffff88808c5d64c0 R09: ffff88808c5d6d88
> > R10: 0000000000000000 R11: 0000000000000000 R12: ffffffffffffffff
> > R13: 0000000000000000 R14: ffff88808c4d6f00 R15: 0000000000000000
> > FS:  00007f47f49df700(0000) GS:ffff8880ae800000(0000) knlGS:0000000000000000
> > CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> > CR2: 0000000020000044 CR3: 0000000097508000 CR4: 00000000001406f0
> > Call Trace:
> >   sctp_ulpevent_release_data net/sctp/ulpevent.c:1092 [inline]
> >   sctp_ulpevent_free+0x21f/0x4e0 net/sctp/ulpevent.c:1129
> >   sctp_queue_purge_ulpevents+0xc4/0x110 net/sctp/ulpevent.c:1146
> >   sctp_close+0x148/0x860 net/sctp/socket.c:1515
> >   inet_release+0x105/0x1f0 net/ipv4/af_inet.c:428
> >   inet6_release+0x53/0x80 net/ipv6/af_inet6.c:473
> >   __sock_release+0x1fe/0x2b0 net/socket.c:579
> >   sock_release+0x18/0x20 net/socket.c:599
> >   sctp_do_peeloff+0x38a/0x470 net/sctp/socket.c:5649
> >   sctp_getsockopt_peeloff_common.isra.0+0x8e/0x270 net/sctp/socket.c:5665
> >   sctp_getsockopt_peeloff net/sctp/socket.c:5707 [inline]
> >   sctp_getsockopt net/sctp/socket.c:7802 [inline]
> >   sctp_getsockopt+0x1ec1/0x6741 net/sctp/socket.c:7758
> >   sock_common_getsockopt+0x9a/0xe0 net/core/sock.c:3079
> >   __sys_getsockopt+0x168/0x250 net/socket.c:1960
> >   __do_sys_getsockopt net/socket.c:1971 [inline]
> >   __se_sys_getsockopt net/socket.c:1968 [inline]
> >   __x64_sys_getsockopt+0xbe/0x150 net/socket.c:1968
> >   do_syscall_64+0x103/0x610 arch/x86/entry/common.c:290
> >   entry_SYSCALL_64_after_hwframe+0x49/0xbe
> > RIP: 0033:0x446679
> > Code: e8 6c b4 02 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 2b 09 fc ff c3 66 2e 0f 1f 84 00 00 00 00
> > RSP: 002b:00007f47f49ded88 EFLAGS: 00000246 ORIG_RAX: 0000000000000037
> > RAX: ffffffffffffffda RBX: 00000000006dbc38 RCX: 0000000000446679
> > RDX: 0000000000000066 RSI: 0000000000000084 RDI: 0000000000000003
> > RBP: 00000000006dbc30 R08: 0000000020000140 R09: 0000000000000038
> > R10: 0000000020000040 R11: 0000000000000246 R12: 00000000006dbc3c
> > R13: 00007f47f49ded90 R14: 0000000000000004 R15: 20c49ba5e353f7cf
> > Modules linked in:
> > ---[ end trace dfa9a15945f164b7 ]---
> > RIP: 0010:sctp_assoc_rwnd_increase+0x34/0x520 net/sctp/associola.c:1498
> > Code: 41 54 49 89 fc 53 89 f3 48 83 ec 10 e8 95 90 f1 fa 49 8d bc 24 60 06
> > 00 00 48 b8 00 00 00 00 00 fc ff df 48 89 fa 48 c1 ea 03 <0f> b6 04 02 84
> > c0 74 08 3c 03 0f 8e 2f 04 00 00 45 8b ac 24 60 06
> > RSP: 0018:ffff88808da476f8 EFLAGS: 00010203
> > RAX: dffffc0000000000 RBX: 0000000000000000 RCX: 0000000000000001
> > RDX: 00000000000000cb RSI: ffffffff867ecd5b RDI: 000000000000065f
> > RBP: ffff88808da47730 R08: ffff88808c5d64c0 R09: ffff88808c5d6d88
> > R10: 0000000000000000 R11: 0000000000000000 R12: ffffffffffffffff
> > R13: 0000000000000000 R14: ffff88808c4d6f00 R15: 0000000000000000
> > FS:  00007f47f49df700(0000) GS:ffff8880ae800000(0000) knlGS:0000000000000000
> > CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> > CR2: 0000000020000044 CR3: 0000000097508000 CR4: 00000000001406f0
> >