New security flaw discovered in Debian 64-bit Linux OS
November 21, 2012
A new Linux malware and rootkit has been discovered late yesterday that security researchers say is designed to inject iFrames and viruses into specific websites and then push traffic to malicious sites that then propagate the malware even further.
News of the rootkit has circulated for the past few days after an anonymous user of the Full Disclosure mailing list posted about it online. Since then, researchers at Kaspersky Lab and CrowdStrike have looked into the malware and shared their findings.
Originally designed for 64-bit Linux systems, it specifically targets kernel version 2.6.32-5-amd64, which is the latest kernel used in the 64-bit Debian Squeezy Linux flavor.
"The rootkit at hand seems to be the next step in iFrame injecting cyber crime operations, driving traffic to exploit rootkits," says George Wicherski, senior security researcher at CrowdStrike. "It could also be used in a Waterhole attack to conduct a targeted attack against a specific target audience without leaving much forensic trail," he added.
According to CrowdStrike, the malware doesn't appear to be a modified version of any known malware, and appears to be the work of an intermediate-level programmer. It is believed that it could be the work of a Russian software contractor.
"The malware ensures its startup by adding an entry to the /etc/rc.local script: insmod /lib/modules/2.6.32-5-amd64/kernel/sound/module_init.ko," says Martha Janus of Kaspersky Lab. "After loading it into memory, the rootkit uses one of two methods to retrieve kernel symbols and write them to the /.kallsyms_tmp file:
"Then it extracts the memory addresses of several kernel functions and variables and stores them in the memory for later use," she says.
In order to hide files and the startup entry, the rootkit hooks a number of kernel functions, including: vfs_readdir, vfs_read, filldir64 and filldir.
"The TCP code will then later retrieve data from that buffer and encapsulate it in a TCP packet for transmission," he adds.
"Based on the Tools, Techniques, and Procedures employed and some background information we cannot publicly disclose, a Russia-based attacker is likely," he adds. "It remains an open question regarding how the attackers have gained the root privileges to install the rootkit. However, considering the code quality, a custom privilege escalation exploit seems very unlikely."
Janus further speculates in her analysis that the malware is still in the initial development stage due to the presence of debugging information and some of the functions do not seem to be fully working or implemented as of now.
"So far, in most of the drive-by download scenarios we've seen, an automated injection mechanism is first implemented as a simple PHP script," Janus states. "In the case described above, we are dealing with something far more sophisticated-- a kernel-mode binary component that uses advanced hooking techniques to ensure that the injection process is more transparent and low-level than ever before," he said.
"This rootkit, though it's still in the initial development stage, reveals a new approach to the drive-by download scheme and we can certainly expect more such malware in the future," he added.
In other internet security news
BSD software developers say that hackers broke into two of its FreeBSD project servers using a stolen SSH authentication key, with admin login credentials that appear to have belonged to one of the developers.
The lead project developer behind the open-source operating system has launched a full-fledged investigation into the security breach and has taken a few of the servers offline during his probe. However, early indications are that the damage might have been far worse than was initially thought.
None of the so-called base repositories - stores of core components such as the kernel, system libraries, compiler and daemons were hit, however. And only servers hosting source code for third-party packages were exposed by the attack, which was detected on November 11 and announced on Saturday, November 17, following a preliminary investigation.
The intrusion itself may have happened as far back as September 19, according to the lead developer. On November 11, an intrusion was detected on two servers within the FreeBSD.org cluster. The affected machines were taken offline for analysis, and probably won't be reconnected until sometime next week.
Additionally, a large portion of the remaining infrastructure machines were also taken offline as a precautionary gesture. "We have found no evidence of any modifications that would put any end user at risk. However, we do urge all BSD users to read the report available on our site and decide on any required actions themselves. We will continue to update you as further information becomes known. We do not currently believe users have been affected given current forensic analysis," read a FreeBSD statement on their site.
"And no Trojanized packages have been uncovered, at least as yet. But FreeBSD users have been urged to carefully check third-party packages installed or updated between September 19 and November 11 nonetheless, as a precaution," it continued.
The FreeBSD.org team has promised to tighten up security, in particular by phasing out legacy services such as the distribution of FreeBSD source via CV Sup, in favor of the more robust Subversion, freebsd-update, and portsnap distribution methods. The hack was "not due to any vulnerability or code exploit within FreeBSD", according to the BSD developers.
The whole incident raises some embarassing and troubling questions since it seems that the unknown attackers behind the hacking attempt managed to steal both SSH (remote administration) key file and passwords from a developer.
Analysis of the attack can be found in an informative blog post by Paul Ducklin of Sophos. Attacks on open-source repositories are far from unprecedented. Kernel.org was suspended for a month in July 2011 following a much more serious malware attack and a server compromise.
Then in August 2011 another breach on the MySQL.com website left visitors exposed to malware that could infiltrate said MySQL databases.
But perhaps the most similar attack to the FreeBSD hacking attempt occurred in 2009, with a breach against the Apache Software Foundation, also facilitated by the misuse of SSH keys.
In other internet security news
The U.S. Transportation Security Administration (TSA) has taken yet another bad doze of publicity with the recent discovery that its questionable security system allows passengers in its PreCheck system to choose their own security status, and thus compromising other security features.
The TSA's PreCheck system allows some frequent fliers willing to pay $100 for a background check to skip some of the onerous security checks, like taking off shoes and unpacking laptops or toiletries. PreCheck customers are still subject to more intensive searches on a randomized basis, however.
Aviation blogger John Butler discovered that the barcode information used for the boarding passes of Precheck fliers wasn't encoded, and could be read by a simple smartphone app. It contained the flier's name, flight details, and a number, either a 1 or a 3, with the latter confirming the passenger was cleared for lesser screening.
Ordinarily, it would be a relatively simple task to just scan the issued boarding pass, decode it, and then change the security setting if you are planning to bring something suspicious aboard, or even change the name on the ticket to match a fake ID.
But after placing the new information into a barcode, and a couple of minutes of cut and paste, the new boarding pass would work as normal, Butler explained, and that's where all the issue lies.
"The really scary part in all of that is both the TSA document checker, because the scanners the TSA use are just barcode decoders, they don't check against the real time information," he said. "So the TSA document checker will not pick up on the alterations."
This means that, as long as their boarding pass has a 3 on it, they can always use the Pre-Check line. But the agency that appears to devote so much time to irradiating fliers, fondling vibrators, promoting the homosexual agenda, or just plain stealing fliers' belongings doesn't seem to have thought of that.
The TSA only deems it necessary to have barcode readers for checking the data itself against the presented ID, not the accuracy of the boarding pass itself. And simply encrypting the data would also work as well, so how come they didn't think of that?
According to the TSA's vision statement, the agency strives to "continuously set the standard for excellence in transportation security through its people, processes, and technology." Really? Wow!
In other security news
According to a new study recently released, on average, hackers exploit security vulnerabilities in software for about ten to eleven months before the full details of the security issues surface to the public.
Researchers from Symantec say that these zero-day attacks, so called because they are launched well before security firms and industry vendors are even aware of the vulnerabilities per se, are more prevalent and more potent than previously believed.
Overall, zero-day exploits are often closely guarded secrets and the simple reason is that they can be very valuable to potential hackers. However, once the details of the exploited security flaws emerge in public, application developers and system admins alike can rapidly get to work to mitigate or halt the attacks dead in their tracks.
But in today's imperfect cyber world, this comes at a huge price-- it also tips off the world that these security vulnerabilities also exist in systems.
Case in point-- Leyla Bilge and Tudor Dumitras, both of Symantec Research Labs, identified no less than eighteen zero-day attacks between January 2008 and December 2011, and eleven of them were previously undetected.
"A typical zero-day attack lasts an average of about 312 days and, after vulnerabilities are disclosed publicly, the volume of attacks exploiting them increases by up to five orders of magnitude," the security researchers note.
The study is based on data from customers who had opted into Symantec's anti-virus telemetry service.
A paper on the research-- "Before We Knew It: An Empirical Study of Zero-Day Attacks In The Real World" was presented at the ACM Conference on Computer and Communications Security in Raleigh, North Carolina last week.
In other internet security news
U.S. federal police and the Department of Justice (DoJ) are increasingly gaining real-time access to Americans' social network accounts, such as Twitter, Facebook and Google+, but prior to obtaining search warrants, newly released documents reveal.
And the numbers are really dramatic-- live interception requests made by the U.S. Department of Justice to social-networking sites and email providers jumped over 80 percent from 2010 to 2011 alone, and the trend is rapidly increasing.
Documents the ACLU released yesterday reveal that U.S. federal police are using a 1986 law originally intended to tell police what phone numbers were dialed for far more invasive surveillance-- monitoring of whom specific social-network users communicate with, what IP addresses they're connecting from, and perhaps even likes and +1s.
The DoJ conducted 1,662 live intercepts on social networks and email providers last year, up from only 922 a year earlier, the reports demonstrate.
The ACLU hopes that the disclosure of the documents it sued to obtain under the Freedom of Information Act will persuade Congress to tighten up the requirements for police to intercept "noncontent" data -- a broad category that excludes e-mail messages and direct messages.
The current legal standard "allows the government to use these powerful surveillance tools with very little oversight in place to safeguard Americans' privacy," says Catherine Crump, an ACLU staff attorney.
And it could work. On September 25, Rep. Zoe Lofgren, D-Calif., introduced a new bill that would require police to get warrants to access Americans' email and track their mobile phones. But last week, senators delayed a vote on a similar bill after law enforcement groups vehemently objected to it.
The U.S. DoJ didn't immediately respond to questions about social-network surveillance. We'll update this story if we receive a response.
It still isn't clear on just how many of those 1,662 real-time intercepts last year -- which do require a judge's approval -- targeted social networks, and how many were aimed at email providers themselves.
Traditional phone intercepts remain far more frequent-- for example, the U.S. Marshals Service says that 409 of its noncontent intercepts were for internet service providers, while 14,568 were for telephone call data.
The largest number of them fell into the fugitive-finding category, including parole or probation violations. To perform noncontent intercepts on social networks, police must generally seek court authorization for a pen register or trap and trace order, both of which are terms borrowed from decades-old surveillance law.
They were originally designed to allow law enforcement to easily collect the phone numbers associated with incoming and outgoing calls, and were extended to the Internet by the Patriot Act eleven years ago.
However, the Patriot Act didn't make it any more difficult for law enforcement to ask for such an order. Police must merely claim their request is "relevant" to an ongoing investigation. A search warrant, by contrast, requires probable cause, and a live wiretap order is even more privacy-protective.
What's also unclear is what kind of real-time data police are seeking from social networks through these orders. It's clear that they can obtain the current IP address of a Facebook user, for instance, and the port number, which is increasingly important.
Source: Kapersky Labs.
You can link to the Internet Security web site as much as you like.