RE: AJP 62-200 LapTop

Pilgrim, Peter (
Mon, 16 Jun 1997 12:42:44 +0100

>From: Albert D. Cahalan[]
>> After searching dejanews with <pcibios_init> I found there are
>> many articles with the error Unknown PCI Device error,
>This is considered a harmless wart. It means you have some new
>hardware that can not be listed by name in /proc/pci. If you can
>identify the part, send in the information so that future Linux
>versions will be able to identify the part.

Question how can I identify the part, assuming I can install Linux in
future :-).

>> but the boot continues and is regardlessly successful with
>> the next message from the kernel:-
>> `Calibrating Delay Loop .. BogoMips'.
>> It is possible that I am barking up the wrong tree !
>> Nothing to do with PCI or BIOS32, perhaps.
>> Where could the kernel be failing ?
>You have the answer, but refuse to believe it. The kernel is
>not able to calibrate the delay loop (BogoMIPS number).

Yes I thought so in my logical deduction.

>This means you have a problem with your IRQ controller,
>or maybe the clock chips. This problem occurs on machines
>with the Microchannel (MCA) bus, usually genuine IBM 386s.
>You may want to try a Microchannel boot disk. Find them on
>the web if you need to. You may get best results if you
>can compile an unusual kernel for both Microchannel and PCI.

This is very interesting because my laptop is 200 MHz MMX edition
using the desktop Pentium socket 7 type, not the Intel new Tape
Carrier Package (assuming Intel will start production this year!).
I thought that this could push the BogoMips off the Richter scale so
to speak. Any Ideas ? What is the upper limit ?

Why would I need to MCA boot on my machine ? I have seen an
Omega Micro PCI to ISA bridge in the Windows 95 Systems Dialog.
I have not seen anything that says MCA. I am assuming that
I have PCI and ISA. I will look again thoroughly tonight. Cheers.

Anyway my BIOS is from my notepad on the startup screen:-

SystemSoft BIOS PCI BIOS for SiS-510X Ver1.01.26 (6X00-05-11)
Copyright 1983-1996 System Soft Corp. A. R. R.
System Processor: 200MHz Intel Pentium P55C with MMX CPU
External Cache: 256kb Enabled,
Video Chipset: Trident Cyber 9385 with 2MB Video RAM
SystemSoft Plug-n-Play BIOS Ver1.0A

Free Memory 640kb
Extend Memory 31744kb
Shadow Memory 256kb
Reserved Memory 128kb
Total Memory 32678MB

CD ROM Type: CD-211E 5.0D

Found 78Mb Suspension to Disk Partition okay.

(boot os ....)

>Check linux-kernel mailing list archives for a patch that
>changes IRQ behavior. It was posted a week or two ago. The
>patch was intended for broken old 386 hardware. It affects
>EOI operation (End-Of-Interrupt). (the subject may have
>been "386 12.5" or "obsolete")

I do not have a linux machine yet. So how could I apply a patch?

>If all that fails, report that you think the problem is with
>the IRQ controller. Someone may know how to fix it.

So you think it is the IRQ controller. You have not set why this the
behaviour should be fixed. What are the problems with IRQ controllers
in any case. I looked on the Windows 95 Systems on the Device Manager
Dialog at the IRQ section , and the dialog reports no conflicts.

>In any case, report your success or failure so that the kernel
>can be adjusted to work on such hardware.
>#include <standard/disclaimer> // "Objects, dontcha just love em!"
>Peter A. Pilgrim Email: `'
> ACCU member [ enquiries `'
> rather recent ` ' ]
>News International Newspapers Limited, London, England.
>Tel: +44 (0)171-782-3602 Fax: +44 (0)171-782-3624