Netgear Hides Router Backdoor Instead of Fixing It

by
in security on (#3J8)
story imageA very recent firmware analysis from the reverse engineer Eloi Vanderbeken shows that NETGEAR didn't fix the backdoor on port 32764 but instead implemented a knocking feature that is now required to unlock the service.

Summary from the slides: The knocking feature is initiated when a "packet type == 0x201" arrived at "ft_tool" that listens to the Ethernet packets. It only works with EtherType 0x8888 and the payload has to be "45d1bb339b07a6618b2114dbc0d7783e" which is the MD5-hash of the model number DGN1000. If such a packet arrives, the backdoor service /usr/bin/scfgmgr f- is launched.

Ars Technica reports :
The nature of the change, which leverages the same code as was used in the old firmware to provide administrative access over the concealed port, suggests that the backdoor is an intentional feature of the firmware and not just a mistake made in coding. "It's DELIBERATE," Vanderbecken asserted in his presentation.

(Cross posted on Soylentnews)

Re: Okay (Score: 1, Interesting)

by Anonymous Coward on 2014-04-26 11:15 (#16E)

Yeah, Alix boards are nice if you're handling traffic from 2000. Most of those boards cannot route my home internet connection traffic, let alone serious workloads. And serious workloads often need other stuff like proper link aggregation, and multiple default routes (all of these exist in OpenBSD now, but several years after FreeBSD). And for "serious workloads", pf was a limiting factor until recently, because of lack of SMP support.

I have an Alix board I never used for anything. An atom board with 2xGigE has >10x the same routing capacity, and its not much more expensive than an Alix. It will draw more power, sure, and it takes more space. But at least it can handle my home connection.
Post Comment
Subject
Comment
Captcha
31, 6 or eighty three: which of these is the highest?