netdev - Re: general protection fault in tcp_v6

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Sun, 2 Jun 2019 21:29:16 -0600
From:   David Ahern <dsahern@...il.com>
To:     syzbot <syzbot+5ee26b4e30c45930bd3c@...kaller.appspotmail.com>,
        davem@...emloft.net, edumazet@...gle.com, netdev@...r.kernel.org,
        syzkaller-bugs@...glegroups.com
Subject: Re: general protection fault in tcp_v6_connect

On 6/1/19 12:05 AM, syzbot wrote:
> Hello,
> 
> syzbot found the following crash on:
> 
> HEAD commit:    f4aa8012 cxgb4: Make t4_get_tp_e2c_map static
> git tree:       net-next
> console output: https://syzkaller.appspot.com/x/log.txt?x=1662cb12a00000
> kernel config:  https://syzkaller.appspot.com/x/.config?x=d137eb988ffd93c3
> dashboard link:
> https://syzkaller.appspot.com/bug?extid=5ee26b4e30c45930bd3c
> compiler:       gcc (GCC) 9.0.0 20181231 (experimental)
> 
> Unfortunately, I don't have any reproducer for this crash yet.
> 
> IMPORTANT: if you fix the bug, please add the following tag to the commit:
> Reported-by: syzbot+5ee26b4e30c45930bd3c@...kaller.appspotmail.com
> 
> 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: 1 PID: 17324 Comm: syz-executor.5 Not tainted 5.2.0-rc1+ #2
> Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS
> Google 01/01/2011
> RIP: 0010:__read_once_size include/linux/compiler.h:194 [inline]
> RIP: 0010:rt6_get_cookie include/net/ip6_fib.h:264 [inline]
> RIP: 0010:ip6_dst_store include/net/ip6_route.h:213 [inline]
> RIP: 0010:tcp_v6_connect+0xfd0/0x20a0 net/ipv6/tcp_ipv6.c:298
> Code: 89 e6 e8 83 a2 48 fb 45 84 e4 0f 84 90 09 00 00 e8 35 a1 48 fb 49
> 8d 7e 70 48 b8 00 00 00 00 00 fc ff df 48 89 fa 48 c1 ea 03 <80> 3c 02
> 00 0f 85 57 0e 00 00 4d 8b 66 70 e8 4d 88 35 fb 31 ff 89
> RSP: 0018:ffff888066547800 EFLAGS: 00010207
> RAX: dffffc0000000000 RBX: ffff888064e839f0 RCX: ffffc90010e49000
> RDX: 000000000000002b RSI: ffffffff8628033b RDI: 000000000000015f
> RBP: ffff888066547980 R08: ffff8880a9412080 R09: ffffed1015d26be0

This one is not so obvious.

The error has to be a bad dst from ip6_dst_lookup_flow called by
tcp_v6_connect which then is attempted to be stored in the socket via
ip6_dst_store. ip6_dst_store calls rt6_get_cookie with dst as the
argument. RDI (first arg for x86) shows 0x15f which is not a valid and
would cause a fault.

None of the ip6_dst_* functions in net/ipv6/ip6_output.c have changed
recently (5.2-next definitely but I believe this true for many releases
prior). Further, all of the FIB lookup functions (called by
ip6_dst_lookup_flow) always return a non-NULL dst.

If my hunch about the other splat is correct (pcpu corruption) that
could explain this one: FIB lookup is fine and finds an entry, the entry
has a pcpu cache entry so it is returned. If the pcpu entry was stomped
on then it would be invalid and the above would result.