netdev - Re: Oops in Unix sockets code

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]

Message-ID: <4B05542D.7060401@gmail.com>
Date:	Thu, 19 Nov 2009 15:20:29 +0100
From:	Eric Dumazet <eric.dumazet@...il.com>
To:	Christian Borntraeger <borntraeger@...ibm.com>
CC:	Blaschka <frank.blaschka@...ibm.com>, netdev@...r.kernel.org,
	linux-s390@...r.kernel.org
Subject: Re: Oops in Unix sockets code

Christian Borntraeger a écrit :
> Am Donnerstag 19 November 2009 14:20:28 schrieb Blaschka:
>>     <1>Unable to handle kernel pointer dereference at virtual kernel address 000000007575e000
>>     <4>Oops: 0011 [#1] PREEMPT SMP DEBUG_PAGEALLOC
> 0011(page translation excepton) and DEBUG_PAGEALLOC might indicate a use after free.
> 
>>     <4>Modules linked in: sunrpc qeth_l3 dm_multipath dm_mod qeth ccwgroup chsc_sch
>>     <4>CPU: 0 Not tainted 2.6.31-39.x.20091102-s390xdefault #1
>>     <4>Process hald (pid: 2117, task: 000000007d200c40, ksp: 000000007ab33880)
>>     <4>Krnl PSW : 0704100180000000 00000000003a15f8 (_raw_read_trylock+0x0/0x28)
>>     <4>           R:0 T:1 IO:1 EX:1 Key:0 M:1 W:0 P:0 AS:0 CC:1 PM:0 EA:3
>>     <4>Krnl GPRS: 16c8a00000000000 000000007d200c40 000000007575ed18 0000000000000003
>>     <4>           00000000005853d2 000000007d201470 0000000000000002 000000007ab33c30
>>     <4>           0000000075746c78 000000007a74da48 000000000051a16a 000000007575ed18
>>     <4>           000000007575ed30 00000000005da190 00000000005853dc 000000007ab338c8
>>     <4>Krnl Code: 00000000003a15e8: c03000185811        larl    %r3,6ac60a
>>     <4>           00000000003a15ee: c0e5fffffdd9        brasl   %r14,3a11a0
>>     <4>           00000000003a15f4: a7f4ffce            brc     15,3a1590
>>     <4>          >00000000003a15f8: 58302000            l       %r3,0(%r2)
>>     <4>           00000000003a15fc: b9170033            llgtr   %r3,%r3
>>     <4>           00000000003a1600: 1853                lr      %r5,%r3
>>     <4>           00000000003a1602: 1813                lr      %r1,%r3
>>     <4>           00000000003a1604: a75a0001            ahi     %r5,1
>>     <4>Call Trace:
>>     <4>([<00000000005853d2>] _read_lock+0x5a/0x98)
>>     <4> [<000000000051a16a>] unix_write_space+0x36/0xb0
> [...]
> 
> So it looks like that struct sock *sk is already gone in unix_write_space.
> Since I have no clue about the socket code, I can only guess that there is a
> locking or refcount issue.

2.6.31 has a known bug

2.6.31.4 should correct it

commit 657453424a3c382035983f9a47306fafea730f6d
Author: Eric Dumazet <eric.dumazet@...il.com>
Date:   Thu Sep 24 10:49:24 2009 +0000

    net: Fix sock_wfree() race
    
    [ Upstream commit d99927f4d93f36553699573b279e0ff98ad7dea6 ]
    
    Commit 2b85a34e911bf483c27cfdd124aeb1605145dc80
    (net: No more expensive sock_hold()/sock_put() on each tx)
    opens a window in sock_wfree() where another cpu
    might free the socket we are working on.
    
    A fix is to call sk->sk_write_space(sk) while still
    holding a reference on sk.
    
    Reported-by: Jike Song <albcamus@...il.com>
    Signed-off-by: Eric Dumazet <eric.dumazet@...il.com>
    Signed-off-by: David S. Miller <davem@...emloft.net>
    Signed-off-by: Greg Kroah-Hartman <gregkh@...e.de>


Please try 2.6.31.6 ;)
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html