Stable Linux Kernel 2.6.17.5 Released: Vulnerability Fix

Another 2.6.17 series Linux kernel has been announced by Greg KH. This one has a single patch from Linux to fix a "nasty /proc vulnerability". It looks like users were unintentionally allowed to modify some of the /proc files.

Word from the LKML suggests this kernel breaks on machines using some versions of HAL, so you may want to hold off until the kernel devs have a better fix.

An exploit for another Linux kernel privilege escalation vulnerability has been posted.

The exploit seems to be working on all 2.6.x kernels and is not related to the previous exploit we've written about.

pcburn.com
isc.sans.org

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.

Linux kernels 2.6.x are not secure ?

This stable kernel release 2.6.17.5 has not fixed the fundamental issue of developers are not supervised. More fixes are to be expected for non root priviledged people that can bypass security measures in many modules, because password protection is done in KDE, passwrd.

Open source is a laugh when Linux kernel developers left codes in to breach their own work. Shame on you.

Codes on security issues had to be supervised by directives which have iron clad philosophy for maximum security(trusted source, identity approval and virtual monitor to correct intrusion). Virtualization layer intercepts packets and dump the suspected intruding packet to investigate the hacker's hdd?

Linux world can not be run by a bunch of teenagers. Third party coders have to be brought in to make money on lack of Linux security. IBM quality control people taught me that you have to have policemen(independent inspectors) to make quality control happen.

They're not teenagers...

...but acting like they were!

If you read this interview with Greg KH, you'll notice how anxious are kernel developers to implement new features in the kernel! To support new devices, etc.

This is just crazy. Compromising the security of thousands of GNU/Linux servers just because some developers are so anxious to add something into the kernel...

Horse must be in front of the cart(with engine) in Linux ?

The value of operating system was drivers, to which you can send instructions and data to be processed. Then there are the 18,000 applications that can make money for people.

So, obviously, the kernel can be a minimum of functions; and applications, each different, has to have the versatility to work magic. Some done by parallel processing in virtualization.

If you add all the functions in the kernels, then you are carrying dead weight for small applications. Ian Murdock had the right idea for Debian to be kernel independent. Make applications change for more advanced functions. And most application programmers do just that. Maybe K3B is not a good example. But most burners do cdrw or dvdrw compatibility then burn files with many functions all inclusive. Burners override kernel functions. Browsers are self conyained. KDE qt system override kernel drivers and can and sometimes must change config files.

Linus should learn microkernels and componentize the drivers for applications to use whatever component they please. Vista when finished; will have more versatility in complex functions in applications(maybe not yet in toolbars, but someday hopefully).