Kernel Crash Exploit detected...
Hello all,
This article on slashdot is quite serious, much more so if you have users on your server with shell accounts:
http://slashdot.org/article.pl?sid=04/06/14/118209
In fact anything that triggers a SIGFPE will cause the kernel to spiral into a lockstate.
If you implement the i387 (gentoo) patch on this site, I've found that the kernel seems to only partly recover.
The signal.c patch appears to work well on my system and throws a message to syslog whenever the SIGFPE is caught by the kernel.
This article on slashdot is quite serious, much more so if you have users on your server with shell accounts:
http://slashdot.org/article.pl?sid=04/06/14/118209
In fact anything that triggers a SIGFPE will cause the kernel to spiral into a lockstate.
If you implement the i387 (gentoo) patch on this site, I've found that the kernel seems to only partly recover.
The signal.c patch appears to work well on my system and throws a message to syslog whenever the SIGFPE is caught by the kernel.
