Guest User

Untitled

a guest
Dec 18th, 2014
496
0
Never
Not a member of Pastebin yet? Sign Up, it unlocks many cool features!
text 14.10 KB | None | 0 0
  1. Misfortune Cookie is a critical vulnerability that allows an intruder to
  2. remotely take over an Internet router and use it to attack home and business
  3. networks.
  4.  
  5.  
  6.  
  7.  
  8. What is the Misfortune Cookie vulnerability?
  9. --------------------------------------------
  10.  
  11. Researchers from Check Point’s Malware and Vulnerability Research Group
  12. recently uncovered this critical vulnerability present on millions of
  13. residential gateway (SOHO router) devices from different models and makers. It
  14. has been assigned the CVE-2014-9222 identifier. This severe vulnerability
  15. allows an attacker to remotely take over the device with administrative
  16. privileges.
  17.  
  18.  
  19.  
  20. How many devices are affected?
  21. ------------------------------
  22.  
  23. To date, researchers have distinctly detected approximately 12 million readily
  24. exploitable unique devices connected to the Internet present in 189 countries
  25. across the globe, making this one of the most widespread vulnerabilities
  26. revealed in recent years. Research suggests the true number of affected devices
  27. may be even greater.
  28.  
  29.  
  30.  
  31. How does it affect me?
  32. ----------------------
  33.  
  34. If your gateway device is vulnerable, then any device connected to your network
  35. - including computers, phones, tablets, printers, security cameras,
  36. refrigerators, toasters or any other networked device in your home or office
  37. network - may have increased risk of compromise. An attacker exploiting the
  38. Misfortune Cookie vulnerability can easily monitor your Internet connection,
  39. steal your credentials and personal or business data, attempt to infect your
  40. machines with malware, and over-crisp your toast.
  41.  
  42.  
  43.  
  44. Is it that bad?
  45. ---------------
  46.  
  47. Yes.
  48.  
  49.  
  50.  
  51. What can I do to protect against the vulnerability?
  52. ---------------------------------------------------
  53.  
  54. First of all, be smart about your privacy. Make sure your devices and any
  55. documents or folders containing sensitive information are password protected.
  56. Consider adding more privacy to your browsing by using HTTPS connections to
  57. encrypt all your browser activity.
  58.  
  59. Watch for firmware updates from your device vendor addressing Misfortune
  60. Cookie, apply the update as it is released.
  61.  
  62. More technical users may flash alternative firmware to their device, replacing
  63. the vulnerable service (note this may void the warranty by your vendor).
  64. Another option would be configuring your current gateway as a bridge and using
  65. a second secure device as your Internet dialer/gateway.
  66.  
  67. Check Point IPS does block any attempt to exploit Misfortune Cookie if deployed
  68. over live relevant traffic. If you are a service provider in control of device
  69. fleets, please read our ‘Protecting against Misfortune Cookie and TR-069 ACS
  70. Vulnerabilities' whitepaper. If you have a vulnerable device owned and managed
  71. by your service provider, you can contact your customer support with the
  72. request to fix.
  73.  
  74. Remember that your gateway’s security is another layer in your network security
  75. defenses – you should have endpoint protections in place, including firewalls,
  76. anti-virus software, and a freshly updated operating system.
  77.  
  78. What makes Misfortune Cookie more dangerous than many other embedded device
  79. vulnerabilities?
  80.  
  81. Misfortune Cookie is unique due to a combination of multiple factors, including
  82. its severity, ease of exploitability, lacking of almost any preconditions, and
  83. the sheer volume of vulnerable networks. In certain countries, we have measured
  84. eye-opening vulnerability rates of anywhere between 10% to an estimated 50% of
  85. used IP addresses in that country (this is not a typo, 1 in 2 Internet users in
  86. these countries is likely vulnerable).
  87.  
  88. All an attacker needs in order to exploit Misfortune Cookie is to send a single
  89. packet to your public IP address. No hacking tools required, just a simple
  90. modern browser.
  91.  
  92. This should be considered an alarming wake-up call for the embedded device
  93. industry and consumers alike, highlighting the importance of increased security
  94. and privacy for consumer and enterprise networks.
  95.  
  96.  
  97.  
  98. Why have you named it Misfortune Cookie?
  99. ----------------------------------------
  100.  
  101. The Misfortune Cookie vulnerability is due to an error within the HTTP cookie
  102. management mechanism present in the affected software, allowing an attacker to
  103. determine the ‘fortune’ of a request by manipulating cookies. Attackers can
  104. send specially crafted HTTP cookies that exploit the vulnerability to corrupt
  105. memory and alter the application and system state. This, in effect, can trick
  106. the attacked device to treat the current session with administrative privileges
  107. - to the misfortune of the device owner.
  108.  
  109.  
  110.  
  111. What software component is vulnerable?
  112. --------------------------------------
  113.  
  114. The affected software is the embedded web server RomPager from AllegroSoft.
  115. Internet-wide scans suggest RomPager is likely the most popular web server
  116. software in the world with respect to number of available endpoints. RomPager
  117. is typically embedded in the firmware released with the device. This specific
  118. vulnerability was introduced to the code base in 2002.
  119.  
  120.  
  121.  
  122. Which devices are affected?
  123. ---------------------------
  124.  
  125. Our research has detected at least 200 different models of devices of various
  126. manufacturers and brands currently exposing a vulnerable service on the public
  127. Internet address space. The majority of these devices are residential gateways.
  128. The list includes models by D-Link, Edimax, Huawei, TP-Link, ZTE, and ZyXEL,
  129. among others. We suspect that the source for inclusion of the vulnerable piece
  130. of software is a common chipset SDK (distributed to the different
  131. manufacturers), however this cannot be confirmed at this point.
  132.  
  133. Prior to this publication and the expected firmware patches, we believe that
  134. devices exposing RomPager services with versions before 4.34 (and specifically
  135. 4.07) are vulnerable. Note that some vendor firmware updates may patch RomPager
  136. to fix Misfortune Cookie without changing the displayed version number,
  137. invalidating this as an indicator of vulnerability.
  138.  
  139.  
  140.  
  141. Can I detect if I was compromised using Misfortune Cookie?
  142. ----------------------------------------------------------
  143.  
  144. Typically you would not have logs or other traces of Misfortune Cookie
  145. exploitation. General warning signs may be the inability to log in to the web
  146. interface or the discovery of changed settings in your device.
  147.  
  148.  
  149.  
  150. Are any Check Point devices affected by Misfortune Cookie?
  151. ----------------------------------------------------------
  152.  
  153. No.
  154.  
  155.  
  156.  
  157. Can you further explain the technical risk?
  158. -------------------------------------------
  159.  
  160. An attacker with administrative access to your gateway holds an alarming
  161. control over your wired and/or wireless network (local area network, a.k.a.
  162. LAN) infrastructure. Such control puts devices at risk of Man-in-The-Middle
  163. attacks, greatly increases the attack surface for LAN-side vulnerabilities, and
  164. gives attackers the ability to directly monitor connections and identifiers
  165. belonging to your devices.
  166.  
  167. The implications of these risks mean more than just a privacy violation – they
  168. also set the stage for further attacks, such as installing malware on devices
  169. and making permanent configuration changes. This WAN-to-LAN free-crossing is
  170. also bypassing any firewall or isolation functionality previously provided by
  171. your gateway and breaks common threat models. For example, an attacker can try
  172. to access your home webcam (potentially using default credentials) or extract
  173. data from your business NAS backup drive.
  174.  
  175.  
  176.  
  177. Are you aware of attackers exploiting Misfortune Cookie?
  178. --------------------------------------------------------
  179.  
  180. Not yet, although we can assume certain attackers have already discovered and
  181. exploited the vulnerability. It can possibly explain some SOHO router attacks
  182. we have seen in recent years, though we have no direct evidence yet.
  183.  
  184.  
  185.  
  186. Has it been fixed?
  187. ------------------
  188.  
  189. Technically, yes; but it’s complicated. AllegroSoft issued a fixed version to
  190. address the Misfortune Cookie vulnerability in 2005, which was provided to
  191. licensed manufacturers. The patch propagation cycle, however, is incredibly
  192. slow (sometimes non-existent) with these types of devices. We can confirm many
  193. devices today still ship with the vulnerable version in place. We believe this
  194. is a serious problem that the industry needs to solve; secure automatic
  195. software updates should be offered for all modern devices, if not as a default
  196. setting.
  197.  
  198.  
  199.  
  200. What needs to happen for a patch to arrive at my device?
  201. --------------------------------------------------------
  202.  
  203. Generally, all vulnerable device makers need to obtain an updated version of
  204. RomPager or patch it manually, integrate the fixed version into their current
  205. firmware for all vulnerable lines and models, test that nothing was broken
  206. during the process, release the firmware version, which would then have to be
  207. installed on every vulnerable device in the world.
  208.  
  209. If your service provider uses TR-069, it may be much easier for them to install
  210. the firmware update in bulk.
  211.  
  212.  
  213.  
  214. That patching process sounds unlikely to happen any time soon.
  215. --------------------------------------------------------------
  216.  
  217. We know. That’s why we consider this a serious problem in need of attention.
  218.  
  219.  
  220.  
  221. Can’t you just use the vulnerability to patch it everywhere?
  222. ------------------------------------------------------------
  223.  
  224. While theoretically that might be possible, performing such pervasive action on
  225. devices that are not in your possession would constitute a criminal charge in
  226. many countries, regardless of the well intention of the originator.
  227.  
  228.  
  229.  
  230. Why is Check Point performing this type of research?
  231. ----------------------------------------------------
  232.  
  233. Since ancient history to modern day, it was proven that in order to understand
  234. the adversary, you must become one. These efforts are a part of what makes
  235. Check Point a 21-year leader in an ever-changing security landscape.
  236.  
  237. Check Point actively contributes to the security community by making
  238. independent research progress and working towards better public security
  239. awareness and education. We’ve been doing this for a while now, you may have
  240. noticed some of our recent work.
  241.  
  242.  
  243.  
  244. How did you find it?
  245. --------------------
  246.  
  247. We reversed the firmware binary and looked for vulnerabilities. Check Point
  248. researcher Lior Oppenheim independently discovered this vulnerability among
  249. several other ones. Later it was clarified that the vulnerability had already
  250. been known and fixed during an internal code review by AllegroSoft – but is
  251. still a serious hole in many devices. The details of our research and methods
  252. will be presented at the 31st CCC conference in Germany later this month.
  253.  
  254.  
  255.  
  256. What other vulnerabilities did you find?
  257. ----------------------------------------
  258.  
  259. We identified a couple of buffer overflow vulnerabilities that have been
  260. assigned the CVE-2014-9223 identifier. While these can also lead to complete
  261. compromise of the device given knowledge of the target firmware, we decided to
  262. focus our research on the more exploitable Misfortune Cookie vulnerability.
  263.  
  264.  
  265.  
  266. How can I exploit Misfortune Cookie?
  267. ------------------------------------
  268.  
  269. Check Point does not share exploitation tools or exploit code as a policy, nor
  270. will we give you the detailed instructions for creating one. One can assume,
  271. however, that other researchers will independently develop such code
  272. eventually. We can only encourage you to use it professionally and responsibly.
  273.  
  274.  
  275.  
  276. Which vendors did you disclose these findings to? What was their response?
  277. --------------------------------------------------------------------------
  278.  
  279. We contacted AllegroSoft, as well as several major vendors we could identify as
  280. having significant numbers of vulnerable model instances. Responses vary from
  281. immediate recognition and following to correctly patch firmware binaries to
  282. complete lack of response.
  283.  
  284.  
  285.  
  286. Why disclose at all? Aren’t you helping the bad guys?
  287. -----------------------------------------------------
  288.  
  289. We gave this a lot of thought. As always with vulnerability disclosure there is
  290. a consideration including balances between the values of public awareness/
  291. knowledge and the associated risks. In this case, we felt public education was
  292. the most critical deciding factor, specifically due to the expected lengthy
  293. patching process caused by the popularity of the vulnerable software and the
  294. diversity of makers and models. This public awareness may serve as a better
  295. incentive for the makers to release updated firmware faster.
  296.  
  297.  
  298.  
  299. What does Misfortune Cookie have to do with TR-069?
  300. ---------------------------------------------------
  301.  
  302. We began this research by surveying client-side implementations of TR-069
  303. (CWMP), after noticing the extreme prevalence of endpoints listening on the
  304. default CWMP Connection-Request port (7547), second only to HTTP (port 80)
  305. listening endpoints. Misfortune Cookie was uncovered during the examination of
  306. RomPager - the most popular recognized service on this port.
  307.  
  308.  
  309.  
  310. Is this a problem with the TR-069 protocol specification?
  311. ---------------------------------------------------------
  312.  
  313. While the proliferation of devices managed by TR-069 is responsible for
  314. creating a very large vulnerable client population, Misfortune Cookie is not a
  315. vulnerability related to the TR-069/CWMP per se. Misfortune Cookie affects any
  316. implementation of a service using the old version of RomPager’s HTTP parsing
  317. code, on port 80, 8080, 443, 7547, and others.
  318.  
  319.  
  320.  
  321. Is this a problem with the security design of RomPager?
  322. -------------------------------------------------------
  323.  
  324. Not at all; AllegroSoft were very responsive and security-aware. We just
  325. happened to research an old version of their software due to its popularity. We
  326. have no reason to believe it is any better or worse than comparable software in
  327. 2002.
  328.  
  329.  
  330.  
  331. Is this an intentionally placed back door?
  332. ------------------------------------------
  333.  
  334. It doesn’t look like one.
  335.  
  336.  
  337.  
  338. I need someone to blame.
  339. ------------------------
  340.  
  341. That’s not even a question.
  342.  
  343. Page last updated at
  344. Media Contact
  345. Lindsay Bubbico
  346. Check Point Software Technologies
  347. +1-917-754-3013
  348. [misfortune-cookie]
  349. [CP_horz_wh]
Advertisement
Add Comment
Please, Sign In to add comment