netdev - Re: [PATCH] SCTP: Fix Protocol violation when receiving a error length INIT ACK

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: <47E915E3.2040501@hp.com>
Date:	Tue, 25 Mar 2008 11:10:27 -0400
From:	Vlad Yasevich <vladislav.yasevich@...com>
To:	Gui Jianfeng <guijianfeng@...fujitsu.com>
Cc:	Wei Yongjun <yjwei@...fujitsu.com>,
	netdev <netdev@...r.kernel.org>,
	lksctp-dev <lksctp-developers@...ts.sourceforge.net>,
	David Miller <davem@...emloft.net>
Subject: Re: [PATCH] SCTP: Fix Protocol violation when receiving a error length
 INIT ACK

Gui Jianfeng wrote:
> Wei Yongjun wrote:
>> Hi Gui:
>>
>> I looked the source code and found that only in the state *COOKIE_WAIT*
>> received INIT-ACK need to do this, not all of the states. The other
>> states we should have known the init-tag of peer.
>   yes, it's the only possibility.
> 
>> Gui Jianfeng wrote:
>>> Wei Yongjun wrote:
>>>  
>>>> NACK.
>>>>
>>>> If the INIT-ACK chunk is too short to contain the init-tag, get the
>>>> init-tag of peer may get a unexpected value.
>>>> Such as this:
>>>>  CHUNK_INIT_ACK
>>>>   Type                             = 2
>>>>   Flags                            = 0
>>>>   Length                           = 4
>>>>
>>>> So I think the better way is to set T bit of ABORT chunk and used the
>>>> own's Tag.
>>>>     
>>>   Seems reasonable. Please ignore the previous one, here is a new patch.
>>>
>>> Signed-off-by: Gui Jianfeng <guijianfeng@...fujitsu.com>
>>> ---
>>>  net/sctp/outqueue.c     |    3 +++
>>>  net/sctp/sm_statefuns.c |    5 +++++
>>>  2 files changed, 8 insertions(+), 0 deletions(-)
>>>
>>> diff --git a/net/sctp/outqueue.c b/net/sctp/outqueue.c
>>> index 1bb3c5c..c071446 100644
>>> --- a/net/sctp/outqueue.c
>>> +++ b/net/sctp/outqueue.c
>>> @@ -793,6 +793,9 @@ int sctp_outq_flush(struct sctp_outq *q, int
>>> rtx_timeout)
>>>              break;
>>>  
>>>          case SCTP_CID_ABORT:
>>> +            if (sctp_test_T_bit(chunk)) {
>>> +                packet->vtag = asoc->c.my_vtag;
>>> +            }
>>>          case SCTP_CID_SACK:
>>>          case SCTP_CID_HEARTBEAT:
>>>          case SCTP_CID_HEARTBEAT_ACK:
>>>   
>>> diff --git a/net/sctp/sm_statefuns.c b/net/sctp/sm_statefuns.c
>>> index f2ed647..85e1d63 100644
>>> --- a/net/sctp/sm_statefuns.c
>>> +++ b/net/sctp/sm_statefuns.c
>>> @@ -4144,6 +4144,11 @@ static sctp_disposition_t sctp_sf_abort_violation(
>>>          goto nomem;
>>>  
>>>      if (asoc) {
>>> +        /* Treat INIT-ACK as a special case. */
>>> +        if (chunk->chunk_hdr->type == SCTP_CID_INIT_ACK) {
>>> +            abort->chunk_hdr->flags |= SCTP_CHUNK_FLAG_T;
>>> +        }
>>> +
>>>          sctp_add_cmd_sf(commands, SCTP_CMD_REPLY, SCTP_CHUNK(abort));
>>>          SCTP_INC_STATS(SCTP_MIB_OUTCTRLCHUNKS);
>>>  
>>>   
>> Those code should be move to sctp_make_abort() for
>> common using. And may be need check whether we need the T flags.
>   This rarely happens, so i think it's sufficient to place this block of code here.
>   Moving it into sctp_make_abort() will waste much cpu time when generating each abort chunk .
> 

There is a side-effect to this patch that now we will completely ignore the verification
tag in the INIT-ACK regardless of the violation.

In particular, if the INIT-ACK contains all the fixed parameters but violates structure
in some variable parameters, we'll currently use the Initiate Tag from the INIT-ACK in
the ABORT.   We should not be changing this behavior.

The simple hack is to add some conditional code into the the different violation functions, but
I'd like to see if there is a cleaner way to solve this.

-vlad
--
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