According to Goody: While burning my CPU.
>
> On Thu, 20 Aug 1998, you wrote:
> >
> >
> >Someone said check your memory, a possablity, more information from you as to
> >just how much memory you had and now have would help more, which kernel
> >version, what hardware??.
> >
> >You could use "gdb" The GNU debugger to try to find out what is causing the
> >segfaults, gdb is not so difficult to use, it might point you the right
> >direction.
> >
> >As to memory, i have a patch here which checks memory twice at boottime, it
> >was written for kernel 2.0.34 (if i remember correctly), but i have patched
> >it to 35 and now 36pre6 succesfullly. If you require it let me know.
> >
>
> The motherboard is a no-name computer show special -- I have no identification
> other than "model 5IW" and "made in Taiwan" :-) It has an Intel 133 MHz Pentium
> and a Triton chipset. I had 16 MB of EDO in two SIMM slots. I added another 32
> MB of EDO in the other two slots. Kernel is 2.0.34.
>
[humor on]
Ah!, "Made in Taiwan" some would say, thats the answer ;-)
[Humor off]
To the point,
There is one kernel option which comes to mind, altho' it may not have any
influence on memory, it will (when configured and compiled into a kernel)
elininate any triton releated problems.
CONFIG_BLK_DEV_TRITON read /usr/src/linux/Documentation/Configure.help
(If you have not already done so);
> Here's what gdb told me :
>
> Core was generated by `/usr/lib/netscape/netscape-communicator
>/usr/doc/HTML/index.html'.
> Program terminated with signal 4, Illegal Instruction.
> #0 0x40028c1f in ?? ()
You have just installed Redhat 5.1 which uses the new libc.so.6 commonly
known as glibc, 'ls -l /lib/libc.so.?' will show you, now if you are using
programs which are compiled against libc.so.5 then this could cause grief, i
am not saying this is the problem, just saying it "can" be a problem.
I have experianced many lib releated problems with netscape in the past.
Swap space comes to mind here when starting large programs, if i am not
being to childish in asking, you do have a swap space and Know its activated
dont you.
Realy the only thing i can think of at the minute which could cause all
these problems is the lib's and memory, or java scripts in the index.html
which is being opened by netscape at starting time. However you say other
programs also segfault, which would contradicht the java idea.!
>
> I'd be interested in the patch, or if you could lead me in the right direction,
> I'll download it myself.
Its quite small, i will gzip it and mail it to you in a seperate mail.
>
> Thanks for your help
>
> Also - Norbert - thanks for your reply - that may also be the case. I've
> switched the SIMMs around to see if the problem goes away..
>
Norbert has a good point, however in my experiance that problem is caused by
not cleaning the slots before adding the simms, dust and grime is our enemy
here, just simply blowing the dust of the slots is not enough, contact
cleaner needs to be used with "CAUTION".
Keep fingers OFF the pins, our fingers may seem dry and clean, i can assure
you the surfice of the pins will become corroded in no time at all. (The
shine we all know they have will dissapier within days).
Remember we all have static charges in our bodys/hands no need to tell
anyone what happens to simms with an electric discharge right!.
Another thing with simms is the size order in which they are placed in the
slots,
bank0 = 8m
bank1 = 16m
The size may stay the same, but normaly speaking CANT be smaller.
If you have a book for your home-brew board check the mem size table page,
some motherboards have "funny" idea's about memory.
Another factor is double sided simms and singel sided simms normaly they
wont work togehter, they wont in my P200, but do in my old 486 (no
comparision realy).
I hope this all helps more that confuses.
> Goody
>
> --
> g o o d y @ v o i c e n e t . c o m
> k a 3 n g h @ q s l . n e t
>
> Amateur Radio : KA3NGH
>
> Running Redhat Linux 5.1 -- a FREE UNIX-based operating system !!!
> http://www.voicenet.com/~goody
> http://www.qsl.net/ka3ngh
>
--
Regards Richard.
[EMAIL PROTECTED]
