Clock speeds and system freeze-ups...

Clock speeds and system freeze-ups...

Post by Tracy Lawrence Fulghu » Sat, 01 Jun 1996 04:00:00



I'd looked at some items in the newsgroups about segment violations and
"signal 6" and "signal 11" internal compiler errors having to do with
overheating Pentiums. I'd  had similar things happen, in addition to
the system freezing up.

The guy who put my PC together had the clock jumpered up to 120 MHz and
the Pentium is rated to 75 MHz.  Of course, he did this with my
knowledge,
and he wasn't trying to bilk me.  Anyway, it works fine with Windoze,
and Linux, until I really push it with some real multitasking, or try
to build a kernel. Basically, I was in denial: "CAN'T be the hardware!
It would've showed up a long time ago! Yeah, I KNOW it's pushing the
envelope already but..."

I'm stepped back to 100 MHz, and things are purring along.
So, lesson learned for me. It's not a black or white case whether
things will work when you crank up the clock speed...some spooky
things can happen.

Just though I'd pass this on.

Kudos to Thomas Lostaunau for hitting the nail right on the head.

--

========================================================================
|      Tracy L. Fulghum, P.E.(Fla.)        N.C. State University       |


|                                          Box 7911                    |
|                                          Raleigh, NC 27695-7911      |
========================================================================

 
 
 

Clock speeds and system freeze-ups...

Post by David E. F » Sun, 09 Jun 1996 04:00:00



>I'd looked at some items in the newsgroups about segment violations and >"signal 6" and "signal 11" internal compiler errors having to do with
>overheating Pentiums. I'd  had similar things happen, in addition to
>the system freezing up.

>The guy who put my PC together had the clock jumpered up to 120 MHz and
>the Pentium is rated to 75 MHz.  Of course, he did this with my

Well, I suppose I'll put in my two cents, because I observed the
same behavior when I tried jumping my new P-100 to 133mhz. Making
the kernel resulted in several sig11s. The system runs absolutely
fine at 100 mhz.

My motherboard (some model MP058) features the Intel Triton chipset
and is configured with 256K of pipeline burst cache (on a separate
plugin card = seems proprietary) and there's 16 megs of 60ns
EDO RAM (8x32) in there. And, of course, I have a CPU
fan, so overheating might not be an issue. (But how hot can
a (overclocked) P100, or a P133 get under Linux?) The motherboard
supports speeds up to 180 mhz.

In view of that last sentence, something doesn't seem right. If
the motherboard can't support a 100 mhz Pentium overclocked to
133 mhz (and "can't support" can be defined as getting sig11s
at that speed, imho) then why claim it can go to 180mhz, with
*existing* cache module and wait state configurations? (I'm
assuming, based on what I've read about the sig11 problem, that
the main cause is cache and/or RAM timing problems.)

Actually, it's not that big of a problem, in view of the fact
I just upgraded from a 386sx/16. :) A P100 is plenty fast
for me.

Quote:>and he wasn't trying to bilk me.  Anyway, it works fine with Windoze,
>and Linux, until I really push it with some real multitasking, or try
>to build a kernel. Basically, I was in denial: "CAN'T be the hardware!

Then something's wrong, to put it bluntly. In surfing usenet, I
came across a quote from Terry Lambert (who is mainly developing
FreeBSD) who says if the system you bring home can't do a 'make
world' then there's something wrong with it, and that most systems
are marginally set up to make Windows work on them, but as we
all know, doing 'make world', compiling the kernel, etc., really
pushes the hardware.

>========================================================================
>|      Tracy L. Fulghum, P.E.(Fla.)        N.C. State University       |


>|                                          Box 7911                    |
>|                                          Raleigh, NC 27695-7911      |
>========================================================================

--
------------------------------------------------------------------------
David E. Fox                 Tax              Thanks for lettimg me


-----------------------------------------------------------------------

 
 
 

1. Strange Freeze-ups

Normally I would not put a post about hardware freeze-ups here, but I cannot
think of anywhere else to put it.

Basically, I have a machine that runs a floppy-based firewall (Dachstein to
be specific), using two PCI net cards.  I've been using it for over a year
with never a single freeze-up ever.  The machine is a beautiful old DEC
P75.

However, when I put SuSE 8.0 minimal install onto a HDD and drop it in
there, and run my own script, the machine will freeze up with a blinking
CAPS LOCK and SCROLL LOCK.  The lockup occurs anywhere from 1 minute to 60
minutes after power up.  

Before the lockup, it works fine, SNAT'ing all outbound traffic, and
DNAT'ing a few selected ports to machines behind the firewall.

Now here I'd say bad HDD or something, except that:

-> A completely different hard drive did it also
-> A completely different *machine* did this to me, this time a DEC
   P166 that I was using to first learn about firewall scripts.  This
   machine also has been running for a year with no problems until
   I started playing with firewall scripts.

The only thing these machines had in common after the HDD was an ISA NIC,
except that I was using that very NIC with SuSE 8.0 for an X-terminal box
for about 8 months w/o ever one single problem, so I don't think it was the
NIC.

Unfortunately, eliminating the HDD and the NIC leaves me with no hardware
culprit.  The SuSE bootup is starting virtually no services except for
CRON, atd, sshd (for local traffic only), and that's about it.  Things like
SuSE's personal firewall, SMTP and other stuff in a default install are all
removed.

I have tried a lot of variations, but before I list everthing, I am
wondering if there is some known combination of kernel parameters used in
firewalling that can cause this, or should I go back to looking at
hardware?  Am I way out in left field here, or has this happened to anyone
else?

If anybody thinks this could actually by my firewall script, I will be happy
to post the script.

Another detail, the lockups always occur when actually connected to the
internet, so I'm wondering if I'm leaving out protection against something
that could be causing this.

TIA,

--
The Student
Programming.  Think about it.

2. How to avoid user's .kshrc

3. SunOS freeze-ups...

4. Upgrading X, latex, or kernel

5. Incorrect speed of system clock

6. Oracle client 817 for RedHat7.1

7. System clock speed whacko with XView 3l5

8. Anonymous FTP

9. system clock speed too high?

10. System Clock Speed

11. System clock running 1/2 speed

12. system clock/bios clock setting

13. how to sync system clock to CMOS clock