Re: unpredictable gcc behavior (not signal 11) with recent slink system
On Wed, Aug 04, 1999 at 12:56:30PM -0700, George Bonser wrote:
>
> Heat.
>
> Get to a nearby Radio Shack and look back in the electronics parts section
> where the power transistor mounting kits and heatsinks are and find a
> small tube of silicon heatsink compound. Make sure you have something hady
> to wipe your hands with and remove the CPU heatsink. Apply a VERY TINY dab
> to the heatsink and spread it out over the area where it contacts the CPU
> case.
This was a very good suggestion-- I tried the heatsink compound, made
sure CPU and case fans were running, and I even tried leaving the case
partially open. Also, I tried to avoid having lots of background processes
by bringing the system into single user mode. Still, the very same errors
pop up during a big make. cc1 catches signal 6 and reports something like:
namei.c: In function `minix_rename':
namei.c:836: internal error--unrecognizable insn:
(jump_insn 93 83 94 (return) -1 (nil)
(nil))
gcc: Internal compiler error: program cc1 got fatal signal 6
If I keep retrying make, it will go progressively deeper into the
compilation, and sometimes it actually finishes making everything.
I've also tested the memory, both by using the sdrams individually
and by disabling the mboard cache. Still no change.
Could it be some other motherboard problem? Or even the cpu?
Thanks again for the help and advice!
--tim
__________________________________________________________________
Timothy Burt
Duke University, Dept. of Physics
Box 90305, Durham N.C. 27708-0305, USA tburt@phy.duke.edu
Reply to: