Re: libssh-4.0.2 client code problem

[Eugene Starozhilov <estarozhilov@xxxxxxxxx>]

Aris,

I didn't mention a compilation issue but to be able to compile libssh I changed WITH_VISIBILITY_HIDDEN:INTERNAL=1
to 
WITH_VISIBILITY_HIDDEN:INTERNAL=
in CMakeCache.txt

Thanks,
Eugene

--- On Tue, 4/13/10, Aris Adamantiadis <aris@xxxxxxxxxxxx> wrote:

From: Aris Adamantiadis <aris@xxxxxxxxxxxx>
Subject: Re: libssh-4.0.2 client code problem
To: libssh@xxxxxxxxxx
Date: Tuesday, April 13, 2010, 4:22 PM

Eugene,

Replying to myself, but I managed to reproduce the problem (+ a
compilation issue with the latest stable git). We will keep you
informed of the status.

Aris

Aris Adamantiadis a écrit :
> Hi Eugene,
> 
> I was not able to reproduce the problem on a CentOS 4.8 with
> openssh-server-3.9p1-11.el4_7.x86-64.rpm, at least not when libssh is
> running on my ubuntu workstation.
> What's the configuration of the client in your test ? Are you able to
> connect to rhel using libssh 0.4.2 from something else than Rhel/centos ?
> I'm trying now to compile libssh on Centos.
> 
> Regards,
> 
> Aris
> 
> Eugene Starozhilov a écrit :
>> Aris,
>>
>> Just checked:
>>
>> samplessh -l oracle -p 1778 -c aes128-ctr alpha-dev30.edf.gxs.com
>> result: sshd[22710]: Disconnecting: Corrupted MAC on input.
>>
>> samplessh -l oracle -p 1778  alpha-dev30.edf.gxs.com
>> result: sshd[22778]: Disconnecting: Corrupted MAC on input.
>>
>> samplessh -l oracle -p 1778 -c aes256-cbc  alpha-dev30.edf.gxs.com
>> result: OK
>>
>> samplessh -l oracle -p 1778 -c blowfish-cbc alpha-dev30.edf.gxs.com
>> result: OK
>>
>> The box has openssh-server-3.9p1-11.el4_7.x86_64.rpm. It is RedHat
>> Enterprise Server 4 distribution.
>>
>> Thanks,
>> Eugene
>>
>> --- On *Mon, 4/12/10, Aris Adamantiadis /<aris@xxxxxxxxxxxx>/* wrote:
>>
>>
>>     From: Aris Adamantiadis <aris@xxxxxxxxxxxx>
>>     Subject: Re: libssh-4.0.2 client code problem
>>     To: libssh@xxxxxxxxxx
>>     Date: Monday, April 12, 2010, 12:14 PM
>>
>>     Hi Eugene,
>>
>>     That's very strange. I will check out what has happened.
>>
>>     Have you got any information on the distribution/ssh server installed
>>     on the server ? I suspect it's a problem with aes256-ctr.
>>
>>     In order to verify:
>>     samplessh -l oracle -p 1778 -c aes256-cbc alpha-dev30.edf.gxs.com
>>     samplessh -l oracle -p 1778 -c blowfish-cbc alpha-dev30.edf.gxs.com
>>
>>     It would also help if you tried aes128-ctr.
>>
>>     Thanks,
>>
>>     Aris
>>
>>     Eugene Starozhilov a écrit :
>>     >
>>     > Hi Aris,
>>     >
>>     > Thanks for fixing server api bug. But I ran into another problem with
>>     > client api.
>>     > These are traces from sshd and  libssh client code:
>>     >
>>     > ---------------------------  sshd
>>     > --------------------------------------------------
>>     > sshd[22309]: Connection from ::ffff:10.160.123.184 port 44351
>>     > sshd[22309]: debug1: Client protocol version 2.0; client software
>>     > version libssh-0.4.2
>>     > sshd[22309]: debug1: no match: libssh-0.4.2
>>     > sshd[22309]: debug1: Enabling compatibility mode for protocol 2.0
>>     > sshd[22309]: debug1: Local version string SSH-1.99-OpenSSH_3.9p1
>>     > sshd[22341]: debug1: permanently_set_uid: 74/74
>>     > sshd[22341]: debug1: list_hostkey_types: ssh-rsa,ssh-dss
>>     > sshd[22341]: debug1: SSH2_MSG_KEXINIT sent
>>     > sshd[22341]: debug1: SSH2_MSG_KEXINIT received
>>     > sshd[22341]: debug1: kex: client->server aes256-ctr hmac-sha1 none
>>     > sshd[22341]: debug1: kex: server->client aes256-ctr hmac-sha1 none
>>     > sshd[22341]: debug1: expecting SSH2_MSG_KEXDH_INIT
>>     > sshd[22341]: debug1: SSH2_MSG_NEWKEYS sent
>>     > sshd[22341]: debug1: expecting SSH2_MSG_NEWKEYS
>>     > sshd[22341]: debug1: SSH2_MSG_NEWKEYS received
>>     > sshd[22341]: debug1: KEX done
>>     > sshd[22341]: Disconnecting: Corrupted MAC on input.
>>     >
>>     ------------------------------------------------------------------------------
>>     >
>>     > -------------------- libssh client code -----------------------------
>>     > [3] host 10.160.31.50 matches an IP address
>>     > [1] Trying to connect to host: 10.160.31.50:1778 with timeout 1800.0
>>     > [3] Socket connected with timeout
>>     >
>>     > [1] SSH server banner: SSH-1.99-OpenSSH_3.9p1
>>     > [1] Analyzing banner: SSH-1.99-OpenSSH_3.9p1
>>     > [1] We are talking to an OpenSSH server version: 3.9 (30900)
>>     > [3] Packet size decrypted: 636 (0x27c)
>>     > [3] Read a 636 bytes packet
>>     > [3] 11 bytes padding, 635 bytes left in buffer
>>     > [3] After padding, 624 bytes left in buffer
>>     > [3] Final size 624
>>     > [3] Type 20
>>     > [3] Writing on the wire a packet having 141 bytes before
>>     > [3] 141 bytes after comp + 6 padding bytes = 148 bytes packet
>>     > [3] Writing on the wire a packet having 133 bytes before
>>     > [3] 133 bytes after comp + 6 padding bytes = 140 bytes packet
>>     > [3] Packet size decrypted: 444 (0x1bc)
>>     > [3] Read a 444 bytes packet
>>     > [3] 10 bytes padding, 443 bytes left in buffer
>>     > [3] After padding, 433 bytes left in buffer
>>     > [3] Final size 433
>>     > [3] Type 31
>>     > [3] Writing on the wire a packet having 1 bytes before
>>     > [3] 1 bytes after comp + 10 padding bytes = 12 bytes packet
>>     > [1] SSH_MSG_NEWKEYS sent
>>     >
>>     > [3] Packet size decrypted: 12 (0xc)
>>     > [3] Read a 12 bytes packet
>>     > [3] 10 bytes padding, 11 bytes left in buffer
>>     > [3] After padding, 1 bytes left in buffer
>>     > [3] Final size 1
>>     > [3] Type 21
>>     > [1] Got SSH_MSG_NEWKEYS
>>     >
>>     > [3] Set output algorithm to aes256-ctr
>>     > [3] Set input algorithm to aes256-ctr
>>     >
>>     > [3] Writing on the wire a packet having 17 bytes before
>>     > [3] 17 bytes after comp + 10 padding bytes = 28 bytes packet
>>     > [3] Encrypting packet with seq num: 3, len: 32
>>     > [3] Sent SSH_MSG_SERVICE_REQUEST (service ssh-userauth)
>>     > [3] Decrypting 16 bytes
>>     > [3] Packet size decrypted: 44 (0x2c)
>>     > [3] Read a 44 bytes packet
>>     > [3] Decrypting 32 bytes
>>     > 2010-04-12 13:14:54,211557; 1126189408 procSrvAuth;  Did not receive
>>     > SERVICE_ACCEPT
>>     >
>>     --------------------------------------------------------------------------------------------
>>     >
>>     > sshd receives corrupted MAC and drops connection. It happens when
>>     client
>>     > calls ssh_service_request(session, "ssh-userauth");
>>     >
>>     > I built libssh-4.0.2 from git repository:
>>     >
>>     > git clone git://git.libssh.org/projects/libssh/libssh.git libssh
>>     > git checkout -b v0-4 origin/v0-4
>>     >
>>     > with libcrypto library (openssl 0.9.7a).
>>     >
>>     >
>>     > The same error can be reproduced using samplessh example with sshd.
>>     >
>>     > /usr/sbin/sshd -d  -p 1778
>>     > samplessh -l oracle -p 1778 alpha-dev30.edf.gxs.com
>>     >
>>     >
>>     > Thank you,
>>     > Eugene
>>     >
>>     >
>>     > --- On *Sun, 3/28/10, Aris Adamantiadis /<aris@xxxxxxxxxxxx
>>     </mc/compose?to=aris@xxxxxxxxxxxx>>/* wrote:
>>     >
>>     >
>>     >     From: Aris Adamantiadis <aris@xxxxxxxxxxxx
>>     </mc/compose?to=aris@xxxxxxxxxxxx>>
>>     >     Subject: Re: libssh-0.4.2 server api bug
>>     >     To: libssh@xxxxxxxxxx </mc/compose?to=libssh@xxxxxxxxxx>
>>     >     Date: Sunday, March 28, 2010, 3:51 PM
>>     >
>>     >     Hi,
>>     >
>>     >     This was resolved in d2bb97c1c6f32c167e1a6093201e01a52bfe0e0d.
>>     Thanks
>>     >     for your feedback on this bug.
>>     >
>>     >     Regards,
>>     >
>>     >     Aris
>>     >
>>     >     Aris Adamantiadis a écrit :
>>     >     > Hi
>>     >     > Oops, we missed that bug for the release. I will find a
>>     solution.
>>     >     >
>>     >     > Aris
>>     >     >
>>     >     > Eugene Starozhilov a écrit :
>>     >     >> Hi Aris,
>>     >     >>
>>     >     >> The new release libssh-0.4.2 has the same problem as
>>     libssh-0.4.1
>>     >     >> (described below). samplesshd doesn't work with standard
>>     LINUX ssh
>>     >     >> client. Is any chance to get it fixed soon?
>>     >     >>
>>     >     >>
>>     >     >> Thank you,
>>     >     >> Eugene
>>     >     >>
>>     >
>>     >
>>
>>
>