🚀 go-pugleaf

In article <48rn0t$51l@adam.telalink.net>, jreader@telalink.net (jreader) writes:
>...
>Basically, I have a Pentium 75, EIDE controller, PCI BIOS(of course)
>when using gcc (gcc, g++, make, and related) I sometimes get random 
>errors that halt the compilation process...they usually happen on 
>larger files, and are quite intermittant.  It screws up the compiling, 
>if the files are small enough i can keep typing and get it to run 
>through, but heaven forbid i try to compile the kernel...
>Suggestions so far ( Parity Memory (don't have) wait states (checked 
>and rechecked) power supply failure (checked)) anny suggestions 
>comments appreciated.

Two possibilities come to mind:

    - one or more of your SIMMS could be marginal; gcc is famous for detecting
      bad memory chips; to check this, you could try:
        - swapping SIMMs one at a time (and compile a kernel as a test :-);
        - using the SIMMs in someone else's PC;
        - configuring the kernel only to use only a certain amount of memory
          (I'm sure you can do this at boot time); start off with small
          amount (4MB, or 8MB) and increase it;
       
    - your motherboard could have unreliable timing (ie the worst-case timing
      for that design may not be guaranteed), and is just on the edge of
      failing; to check this, you could try:
        - swapping the motherboard;
        - varying the motherboard's ambient temperature (see below).

If your system has marginal timing, then the rate of failure will be sensitive
to temperature. As a general rule, MOS-based logic (DRAMs etc) get slower as 
they get hotter.

Do the errors occur more often soon after you first power up the PC, or on a
cold day? Do the errors occur more often after the PC has been running for
hours, or on a hot day?

If you keep the motherboard and power supply unusually cool (eg with fans), or
unusually warm (eg *modest* application of a hair dryer or something similar),
what happens?

Carl.