Cisco TelePresence Multiple Vulnerabilities – SOS-11-010

Standard


Sense of Security - Security Advisory - SOS-11-010

Release Date. 19-Sep-2011
Last Update. -
Vendor Notification Date. 21-Feb-2011
Product. Cisco TelePresence Series
Platform. Cisco
Affected versions. C < = TC4.1.2, MXP <= F9.1 Severity Rating. Low - Medium Impact. Cookie/credential theft, impersonation, loss of confidentiality, client-side code execution, denial of service. Solution Status. Vendor patch References. 1. CVE-2011-2544 (CSCtq46488) 2. CVE-2011-2543 (CSCtq46496) 3. CVE-2011-2577 (CSCtq46500) Details. Cisco TelePresence is an umbrella term for Video Conferencing Hardware and Software, Infrastructure and Endpoints. The C & MXP Series are the Endpoints used on desks or in boardrooms to provide users with a termination point for Video Conferencing. 1. Post-authentication HTML Injection - CVE-2011-2544 (CSCtq46488): Cisco TelePresence Endpoints have a web interface (HTTP or HTTPS) for managing, configuring and reporting. It is possible to set the Call ID (with H.323 or SIP) to a HTML value. If a call is made to another endpoint and an authenticated user browses to the web interface on the endpoint receiving the call (e.g. to view call statistics), the HTML will render locally within the context of the logged in user. From this point it is possible to make changes to the system as the authenticated user. The flaw is due to the flexibility of the H.323 ID or SIP Display Name fields and failure to correctly validate user input. Examples (MXP): Rebooting the system:
The attacker may also choose to change passwords in the system, disable
encryption or enable telnet:

2. Post-authentication Memory Corruption - CVE-2011-2543 (CSCtq46496):
Cisco TelePresence systems (Endpoints and Infrastructure) use XPath for
setting and getting configuration.

Example syntax is:
http://ip/getxml?location=/Configuration/Video
The request is sent to a locally listening shell (tshell). This is the
case for all requests relating to performing an action on the system (e.g.
config get or set). The shell then sends the input to the "main"
application (/app/main, id=0), and the data is passed as a parameter.

It was discovered that the getXML handle does not properly perform
length checking on the user supplied input before passing it to the
tshell. Furthermore, there is no length checking performed in the tshell
and no bounds checking performed in the main application where the
parameter is consumed. As such, it is possible to send input that
exceeds the size of the receiving buffer, subsequently causing an
invalid address to be read. This causes a reboot on the Endpoints. The
VCS will not reboot, the process will crash by SIGSEGV (or sigabrt) but
it will restart the process itself which drops all calls.

Proof of Concept: GET
/wsgi/getxml?location="+("A"*5200)+("x60"*4)+("X"*4)+"HTTP/1.1rn
Host: 192.168.6.99rnrn"

Received signal SIGSEGV (11) in thread 0x129e8480, TID 2670
Illegal memory access at: 0x5858585c
Registers:
GPR00: 00f2c908 129e5960 129ef920 00000005 00000040 0000000c 00000037
0f315580
GPR08: 00000005 129e5a70 129e5a80 58585858 0f3272d4 11589858 129e6896
0000000b
GPR16: 129e6084 11164a1c 00000000 129e6894 00000037 1299ca18 00000005
00000002
GPR24: 129e59a8 00000002 0f3ea3a4 129e5a64 00000037 00000005 0f410bac
129e5960
GPR24: 129e59a8 00000002 0f3ea3a4 129e5a64 00000037 00000005 0f410bac
129e5960
NIP: 0f39abc8 MSR: 0000d032 OGPR3: 00000002

As you can see, the crash string is passed as a parameter in GPR 8.
The severity of this issue is compounded by the fact that the main
application runs as root, this could potentially lead to arbitrary code
execution.

3. Pre-authentication SIP Denial of Service - CVE-2011-2577 (CSCtq46500):
Cisco TelePresence Endpoints utilise SIP for the call setup protocol.
Sending a SIP INVITE with a 4x8 a"s in the MAC Address field and the
receive field causes the system to reboot.

Proof of Concept: MXP:
Exception 0x1100 : Data TLB load miss Active task
FsmMain FSM process : SipTrnsp(0) FSM message : SipTrnsp_Send_Msg_Req
from SipTrnsp(0) Data TLB miss (DMISS) : 0x00000000 (illegal addr.
accessed)

Solution.
Upgrade to TC4.2 for the C series to fix validation issues.

Discovered by.
David Klein, Sense of Security Labs.

About us.
Sense of Security is a leading provider of information
security and risk management solutions. Our team has expert
skills in assessment and assurance, strategy and architecture,
and deployment through to ongoing management. We are
Australia's premier application penetration testing firm and
trusted IT security advisor to many of the countries largest
organisations.

Sense of Security Pty Ltd
Level 8, 66 King St
Sydney NSW 2000
AUSTRALIA

T: +61 (0)2 9290 4444
F: +61 (0)2 9290 4455
W: http://www.senseofsecurity.com.au
E: info@senseofsecurity.com.au
Twitter: @ITsecurityAU

The latest version of this advisory can be found at:
http://www.senseofsecurity.com.au/advisories/SOS-11-010.pdf

Other Sense of Security advisories can be found at:
http://www.senseofsecurity.com.au/research/it-security-advisories.php

Fonte: http://www.exploit-db.com/exploits/17871/

Como ativar alt2 atheros attansic 100mb no Centos 5.x facilmente

Standard

 

Após terminar a instalação de uma CentOs 5.5 i386 pude perceber que a placa de rede não havia sido atrelada a um módulo dentro do meu Gnu/Linux Centos. Achei estranho, principalmente por que kernels da linha 2.6.x costumam ter muitos módulos prontos para uso e a maior parte deles só precisa de um simples modprobe. Como não havia obtido NENHUM êxito com muitas coisas que vi acabei tendo de preparar um how to simples para vocês.

A placa-mãe que usei é uma P5GC-MX 1333 da ASUS, uma placa simples mas muito bacana pois é bem barata (e lembrem-se, essa é velhinha 😉 ).
Para instalar de maneira fácil baixe os arquivos abaixo com os respectivos comandos:

rpm -Uhv http://www.libphp.net/module-init-tools-3.3-0.pre3.1.60.el5.elrepo.i386.rpm

e

rpm -Uhv http://www.libphp.net/kmod-atl2-2.2.3-4.el5.elrepo.i686.rpm

Após instalados os arquivos com os comandos citados reinicie seu Centos e após o boot já prepare aquela caixa de BIS pretinha para agradar a equipe nerd do AppUnix.
Abraços e breve voltaremos.
Att: L.O.

 

how to configure network on red hat Configurando interfaces de rede manualmente no Centos Fedora ou Red Hat

Standard

Boa noite,

Estou usando o CentOs 5.3 (na realidade era o 5.1 e o maravilhoso yum fez a upgrade tranquilamente) e necessitei configurar a interface de rede manualmente.
Supondo que sua interface seja a eth0 (a primeira) você deve acessar seu arquivo assim:

vim /etc/sysconfig/network-scripts/ifcfg-eth0

O vim não é obrigação, entretanto gosto dele.

Observe a sintaxe abaixo:

DEVICE=eth0
BOOTPROTO=static
DHCPCLASS=
HWADDR=00:30:48:56:A6:NE
IPADDR=192.168.1.101
NETMASK=255.255.255.0
ONBOOT=yes

Fazendo isto (tornando sua sintaxe semelhante) você estará:

colocando o ip da interface estaticamente como sendo 192.168.1.101 e dizendo que isto será feito no momento do boot do sistema.

Em seguida entre aqui:

vim /etc/sysconfig/network

NETWORKING=yes
HOSTNAME=libphp1
GATEWAY=192.168.1.1

Aqui definimos a rede como sendo o hostname libphp1 e o gateway 192.168.1.1

Após isto somente faça isto:

/etc/init.d/network restart

Depois podemos até setar um dns para acesso a web:

vim /etc/resolv.conf

Dentro dele eu coloquei assim:

nameserver 200.165.132.148
nameserver 200.165.132.155

Este é o modo simples de definir o dns da velox.

Abração e tenham uma semana excelente e abençoada por Deus.