Microsoft Corp. on Tuesday released 10 software security patches for its products, including seven it deemed critical and that could allow remote attackers to take control of systems running the company’s software. The Redmond, Wash., software maker advised customers to download and install critical patches for a wide range of products as soon as possible, including its Windows operating system, Exchange e-mail server and Microsoft Office productivity software.
The company published the software updates, labeled MS04-029 through MS04-038, on its Web site.
The security alerts detail holes in a number of critical components, including Windows components for handling SMTP (Simple Mail Transfer Protocol), which is used for sending and receiving e-mail, and NNTP (Network News Transfer Protocol) traffic, as well as a Windows feature for processing compressed ZIP files.
The slew of vulnerabilities — more than 21 spread across the 10 security bulletins — are sure to cause headaches for network managers, who will be rushing to distribute the patches before software code to exploit the vulnerabilities is released on the Internet. Among the most critical for enterprises are MS04-035, which patches the SMTP hole, and MS04-036, which plugs the hole in Windows handling of NNTP, a protocol used to manage traffic to and from Internet news groups, said Brian Mann, outbreak manager for McAfee Inc.’s Antivirus Emergency Response Team.
Both the vulnerabilities described inMS04-035 and MS04-036 affect servers running at the enterprise gateway and will need to be patched as soon as possible, especially with the threat of remote exploit and code execution, Mann said.
SANS Institute Internet Storm Center director, Marcus Sachs, said MS04-035 is “one to watch” since it affects the way Windows Server 2003 handles Domain Name Server (DNS) lookups. “DNS is so critical to the operation of the Internet…(that) there is a lot more behind this, (it is) one that needs to be kept an eye on.”
The disclosed vulnerability in Windows handling of ZIP folders (MS04-34) is also dangerous, because it affects machines running Windows Server and Windows XP, Mann said.
For that vulnerability, a buffer overflow can be created on Windows by ZIP files specially crafted to trigger the vulnerability. Windows users would have to download and open the files from a Web site or double click on a malicious ZIP file in an e-mail to trigger the buffer overflow, Microsoft said. However, ZIPs are a common form of e-mail attachment, and virus writers are already fond of using the compressed files to deliver malicious payloads, Mann said.
Administrators should also hurry to apply cumulative software patches for Windows (MS04-032) and the Internet Explorer Web browser (MS04-038), said Thor Larholm, senior security researcher at PivX Solutions LLC. Sachs said MS04-032 is of particular concern since it is the only one with no known workaround, meaning patching systems is the only viable solution.
Additionally, since XP machines with SP2 installed are protected against all the announced vulnerabilities, save MS04-038, “it seems that they have been known about for quite a while,” Sachs said. As for MS04-038, Sachs said it is “a big one” since it affects even SP2 patched XP machines and is IE specific. Since IE is deeply tied into the OS the vulnerability (there are actually eight vulnerabilities in 038, according to SANS) can allow for everything from privilege escalation to remote code execution.
Malicious code that exploits vulnerabilities covered by those patches, including flaws in a Windows component called Windows Shell and a vulnerability in the way Internet Explorer handles drag and drop events, is already circulating on the Internet, Larholm said.
While there were no major surprises in the October batch of patches, the sheer number of vulnerabilities disclosed will keep administrators busy, especially with the short window of time between the publication of a software patch the development of exploit code that takes advantage of it, he said.
PivX estimates that new exploit code for vulnerabilities typically appears within 10 days of a patch being released or publicized, Larholm said.
– with files from Chris Conrath, ComputerWorld Canada
