suddenly, strange segfaults, kernel logs

[Peter Molnar <peter.molnar@wm.sk>]

Hi,

i have a debian machine that has been stabily working for 1 month. 
On Friday users have reported, the services were not working anymore and have 
done a reboot.

At that time approximately (when it stoped working), i had the following in /
var/log/messages:

May  6 16:44:50 *** kernel: KERNEL: assertion (sk->forward_alloc == 0) failed 
at tcp.c(1863)
May  6 16:44:50 *** kernel: KERNEL: assertion (atomic_read(&sk->rmem_alloc) == 
0) failed at af_inet.c(173)
May  6 16:44:50 *** kernel: KERNEL: assertion (sk->forward_alloc == 0) failed 
at af_inet.c(176)

When I logged into the machine, i've got the following errors:

$ ls -l 
ls: error while loading shared libraries: /lib/libc.so.6: unexpected PLT reloc 
type 0x1a@$???E
$ lastlog
Segmentation fault
$ which lastlog
Segmentation fault
$ which ls
/bin/ls
$ ls
ls: error while loading shared libraries: /lib/libc.so.6: unexpected PLT reloc 
type 0x1a@T???E
$ passwd
Segmentation fault

named was also down (crashed ?)

I have checked the md5 sums of the binaries and the libc against the ones 
generated from fresh debs downloaded from debian.org - they are all ok.

Does anybody have an idea, what the problem could be? Suddenly everything is 
broken, without having touched the machine. 

I suppose a hardware issue. The kernel messages and segfaults could indicate 
corrupted memory or a corrupted bus.

Thanks, Peter