>>Has anyone had the experience with ls yielding a "segmentation fault"
>>error while the rest of the machine operates normally (vi, nslookup, etc)?
[ variant on suggestion: ]
Quote:>1) Corrupted binary for ls
I had a dying motherboard - maybe memory, maybe disk controller, who cares? -
which exhibited a _very_ confusing syndrome:
Ordinarily harmless programs like ls, df, ... would run once or twice, then
crash the system after some other activity. I correlated many of these to
swapping in and out of memory, and felt at first my swap usage wasn't working.
(In DOS, equally mysterious: MS Word would start, then lock up after a few or
many minutes of use; WP on the other hand showed no problems.)
After swapping out the guts of the machine and having the problems vanish,
I poked through /var/adm/messages and found some error messages supporting
the hardware-failure explanation. Fortunately I didn't do much which wrote
to my active disk partitions, as the potential for severe destruction is
fairly obvious!!! (I did lose a text file which got written onto two
We are used to near-bulletproof hardware, but _sometimes_ the electrons
get lost. Actually I'm amazed the OS was robust enough to ride through
a flakey board's tricks, and still generate cogent error messages.
Georgia Tech/ GTRI/ SDL, 7220 Richardson Rd., Smyrna, GA 30080
"Software grows to fill or exceed hardware." -Parkinson's Computing Law