Fix for Security Hole in Android G1
In one fell swoop and with an automatically distributed patch, Google and T-Mobile fixed a problem with the G1 mobile phone whereby users could access root privileges and possibly raise all kinds of havoc.
In its originally delivered state, the G1 interpreted keyboard input as a remote shell request. This could be pretty annoying if you happen to type in "reboot" with applications running.
T-Mobile has since plugged the security hole with firmware update RC30, hoping thereby to raise the bar for any future hacks. The problem was discovered in early November, but got special notice after an experience by user jdhorvat. While talking on the G1 with his girlfriend, he restarted it.
When she asked why he wasn't responding, he IM'd her with the natural response "Reboot." He was surprised to see the device do just that.
Source of the security hole was boiled down to two lines of code in the init.rc file, according to the bug report. The file is a script that drives the boot process. A number of websites declared the problem one of the most embarrassing in recent history.
The G1 is manufactured by HTC and based on Google's Android platform, which is itself a knockoff of Linux and other Open Source components. Since its introduction, the G1 proved to be a favorite sport for hackers, who even managed to install and start Debian Lenny on it. The device is available in the U.S., but when other parts of the world start seeing it early 2009, all security holes will likely be plugged.