HELP!!!. Weird rampent crashes on several 2.0.x machines lead to question on 2.0.x status

[Anthony Barbachan]

Hi,

    I've had a 2.0.x based firewall up and running for 3+ years.  A few
months ago I heard that it had started crashing rarely but occasionally,
usually non-fatally but sometimes fatally.  I figured something had gone
wrong with the hardware and the problem appear to be getting worse so I
switched it out with another machine.  The hard drive, memory, and network
card went into another old but reliable machine that had been laying around.
The problems continued and appeared to start to get worse with crashes
(usually fatal now, also just before the machine switch).  As the problem
followed the system I decided to switch out the old network cards and memory
that had been brought over from the original machine.  The ISA network cards
were swapped for a couple of PCI ones and the memory was completely
replaced.  The problems continued.  I upgraded the kernel from 2.0.36 to
2.0.38.  The swap partition on the hard drive was checked for bad blocks as
well, none found.  The problems continued and got even worse.  The entire
machine except for the hard drive was switch again with another spare old
machine that had been lying around.  Problems continued and actually
appeared to get worse.  By the time of the memory/network cards switch the
machine appears to only be staying up for a couple of hours before freezing
up and many times in only a few minutes.  Machine was completely replaced
again, this time by a very stable and relatively new machine.  Kernel
further upgraded to 2.0.39pre1.  Everything appeared to be fine at last
check but just found out that it froze again.  At this point the kernel has
been recompiled three times and the hard drive has been moved to four
separate machines but the problem continues.  I can't see how anything on
the hard drive could be causing this problem, at least without giving some
sort of error on the drive though at this point its the only thing left to
change.  Since this problem has followed this system across 4 machines I
have a small suspicion that perhaps the crashes are being caused by some
sort of attack.  Are there any known security issues in the 2.0.x kernels?
I'd especially like to know of any that could causes crashes like the ones I
have described.  As for the error messages displayed.  One that appears to
have remained relatively constant is one that says something like "free
skb(something) placed on list".  On one of the Oopses that I got today I
managed to run ksymoops on it.  The oops was a protection fault.  And it
happened in something like sock_free, called from skb(something).  I'll try
making a better log of the error in the future though that may not be
possible as the crashed appear to freeze up the entire system usually.
Thanks for any help.