linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* Looking for tunable hardware parameters in 2.4.22
@ 2003-08-26 23:13 Chris Rankin
  2003-08-28 15:46 ` David Zaffiro
  0 siblings, 1 reply; 2+ messages in thread
From: Chris Rankin @ 2003-08-26 23:13 UTC (permalink / raw)
  To: linux-kernel

Hi,

This could well be a motherboard problem, but the motherboard support
person has given up and all BIOS development has long since
stopped. So I'm wondering if just *maybe* there's a magic parameter I
could try tweaking on the Linux command line, please? I've already
tried "noapic", "acpi=off" and "nosmp".

Background:
- a dual PIII Coppermine board, i840 chipset, 133 MHz FSB
- 1GB of P100 memory
- CPUs are identical, and rated at 933 MHz on a 133 MHz FSB

This board has suddenly become unstable when there is a CPU in what
the motherboard calls "slot 2" and when the FSB is 133 MHz. The
motherboard is fine when either:
a) both CPUs are installed, and the FSB is set to 100 MHz, or
b) the CPU is removed from slot 2 and the FSB is set to 133 MHz.

I've swapped the CPUs around and tried replacing them entirely to no
avail, and so I know these CPUs are OK. I've also run memtest 3.0 over
the memory for almost 6 hours (with FSB=133 MHz; 3 times through the
entire RAM) with no problems, so I'm sure the RAM is OK too. However,
the support person thought that the problem was a mis-timing between
the FSB and the memory bus. (This board's memory bus runs at 100 MHz,
of course.)

The fact that the board runs OK with the underclocked FSB gives me
hope that the board hasn't burnt out or anything, because surely such
a catastrophic failure would have rendered it completely useless?

Unfortunately, the only timing the BIOS lets me change is the PCI
latency, which is currently set to 64. Would it be wise to adjust this
to (say) 96? I really don't want to risk damaging anything via a
"randomly tweaking monkey" approach.

Any advice appreciated, even if it's just "Give up, you have tried
everything."

Thanks in advance,
Chris



^ permalink raw reply	[flat|nested] 2+ messages in thread

* Re: Looking for tunable hardware parameters in 2.4.22
  2003-08-26 23:13 Looking for tunable hardware parameters in 2.4.22 Chris Rankin
@ 2003-08-28 15:46 ` David Zaffiro
  0 siblings, 0 replies; 2+ messages in thread
From: David Zaffiro @ 2003-08-28 15:46 UTC (permalink / raw)
  To: linux-kernel

> This board has suddenly become unstable when there is a CPU in what
> the motherboard calls "slot 2" and when the FSB is 133 MHz. The
> motherboard is fine when either:
> a) both CPUs are installed, and the FSB is set to 100 MHz, or
> b) the CPU is removed from slot 2 and the FSB is set to 133 MHz.
> 
> I've swapped the CPUs around and tried replacing them entirely to no
> avail, and so I know these CPUs are OK. I've also run memtest 3.0 over
> the memory for almost 6 hours (with FSB=133 MHz; 3 times through the
> entire RAM) with no problems, so I'm sure the RAM is OK too.

I'm probably of no use to you at all, but there's a slight chance that you're suffering from the same problem as I did a while ago... Since two CPU's at 100MHz consume less power than at 133MHz, it just might be that your PSU doesn't deliver enough power, you could messure the voltages and check this... I've suffered from that problem, buying a more powerfull 350W PSU *really* helped me...

Of course there's also a chance that it's somewhat software related... ;-)



^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2003-08-28 15:48 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2003-08-26 23:13 Looking for tunable hardware parameters in 2.4.22 Chris Rankin
2003-08-28 15:46 ` David Zaffiro

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).