segmentation faults running rampant

[Alan E.Davis <adavis@saipan.com>]

A have been having much trouble with a good machine, and I have to finally ask some advice.  

Maching specs:

      ASUS A7V133 
      AMD Tbird 800
      512 MB SDRAM (2 @ 256 MB DIMMS)
      Adaptec 19160 LVD-capable SCSI adaptor
      LVD cable
      Seagate Cheetah 18 GB 10000 RPM SCSI Drive
      etc.

This machine has been running Debian Woody pretty nicely for a few months.  Recently, 
segfaults have been plaguing me.  I suspect Hardware, but must ask for advice.  Because the only evidence I have is the error messages (Debian GNU/Linux), I am asking this list.  Please forgive me if this is off-topic.  

I get error messages, beginning with a segfault, when fscking, running any e2fs utility, or sometimes when doing an fdisk.  Perhaps at other times also.

Here are the main guts of one of the long error message, while booting, and before I reinstalled Debian potato:  

/dev/sda2 contains a filesystem w/ errors, check forced.
/dev/sda2:
   Inode 407458 has illegal blocks(s).
/dev/sda2: UNEXPECTED INCONSISTENCY; RUN fsck MANUALLY (ie., without -a or -p optoins).
Unable to handle kernel paging request
  address 00800030
printing eip:
C012ddf1
*pdl = 00000000   [I'm not sure about the "l" in pdl]
Oops: 0002
CUP: 0
EIP: 0010:[<c012ddf1>]
EFLAGS: 00010206
[here are the contents of a bunch of registers]
eax:        ebx:        ecx:            edx:     
esi:        edi:        edp:            esp:
ds:         es:          ss:
Process fsck.ext2 (pid: 54, stackpage dfa3f000)
Stack: c012e101 [etc]
.
.
.
Call Trace: [etc]
Code 8b 7b 20 0f [etc]
Warning... fsck.ext2 for /dev/sda6 exited with signal 11
fsck failed.  Please repair manually.
[end of error messages I jotted down]



This was after a number of file system errors.   I had the system running ok for a morning, was planning to compile a new kernel.  Then someone unplugged the machine.  Eventually I gave up, and started trying to reinstall.

After a few tries, still getting segmentation faults, and these kinds of long messages, I pulled out one dimm.  The system worked better.  But eventually I did get these messages.  

Can someone offer any suggestions?  

I did notice this: When the system crashes, I have little trouble getting an e2fsck the first time down.  If problems persist, and th esystem is forced down again, especially if more than one more time, the fs problems seem quite intractible.  I thought of running ext3?

Sorry for the bandwidth.  

[Those images on the TV when I woke up...  Too improbable.  Impossible.  I'm dreaming, shake the cobwebs out of my head...... Oh, my God...   Our airport is closed on Saipan, too.] 

Alan Davis
Marianas High School
Saipan NMI