Re: Booting Red Hat Linux 4 boot.img on AJP 6200 MMX LapTop ? [Anybody]
Wed, 18 Jun 1997 14:53:36 -0600

[This is a courtesy copy of an article posted to Usenet via Deja News]

In article <>, wrote:
> wrote:
> > >
> > > I am a systems developer normally using Solaris 2.4 etc
> > >
> > > I just bought a Win95 laptop computer [British, London
> > > made AJP 62-200M]. Does anyone have success installing Red Hat
> > > Linux 4.1 on an AJP machine or similar. Maybe there is another
> > > person in the UK who has done this already.
> > > The Red Hat Linux came on the May 1997 edition of
> > > Personal Computer World Magazine.
> > >
> > > I made the boot disk with RAWRITE.EXE with \images\boot.img
> > > and restarted by computer and type return at the `boot:'
> > > prompt and let the LILO boot up. The final message I got
> > > happens to be :-
> > >
> > > LILO ...
> > >
> > > uncompressing linux ... done
> > > now booting kernel
> > > console: 16 point foint font, 400 scans
> > > console: colour VGA + 80x25 1 x virtual console (max 63)
> > > pcibios_init: BIOS 32 Service Directory Structure at 0x000ec050
> > > pcibios_init: BIOS 32 Service Directory entry at 0xee600
> > > pcibios_init: PCI BIOS Revision 2.10 at entry 0xee63e
> > > probing PCI hardware.
> > > warning: Unknown PCI device (1039:4107) Please read
> > include/linux/pci.h
> > >
> > > ciao
> > >
> > > Peter Pilgrim
> > > Systems Development,
> > > News International, London, England
> I'm using the same disk and going through similar problems. Please note
> that I
> too am fairly new to this, so if anyone knows any more, please
> contribute.
> Linux, like Windows 95, tries to search your system for known devices,
> so it can configure its device drivers. As part of this search it
> interrogates all devices on the PCI bus for their unique identifiers. In
> this case it has
> found an identifier that the standard kernel does not know about. If you
> want
> a comparison, the situation is similar to Windows'95 'Add new hardware'
> feature coming up with 'unknown device' and prompting you to install the
> correct
> driver.
> However, this may not be fatal, since it may be able to run a device
> driver later which does understand the relevant device. To cut a long
> story short,
> this message is probably not the reason for your system failing to boot.
> Once you have Linux installed, you can make this message go away by
> rebuilding the kernel after modifying (carefully):
> /usr/src/linux/include/linux/pci.h (include device IDs here)
> /usr/src/linux/drivers/pci/pci.c (put device into list here)
> Note that devices are in numerical order in the above list.
> From a quick look in pci.h, linux is having problems with a device from
> SI, which I think is Silicon Integrated Systems, who amongst other
> things
> make chipsets for portables......your AJP machine may have a PCI/ hot
> docking
> controller from them which is not recognisable/ controllable.


#define PCI_VENDOR_ID_SI 0x1039 /* pci.h */

I found the Linux Source Navigator at sunsite to grok the files
myself and this evidence does seem to concurr. I would say that
the Silicon Integrated Systems chips are well supported
according to the include file. So why is my machine playing
me up?

But what does the 4107 means ?
Why is the number presented as 1039:4107 ?

My machine shows in the Control Panel /
System Properties dialog (Win95)

# View devices by connection

o PCI bus
o OMEGA Micro PCI to PCMCIA Bridge
o OMEGA Micro PCI to PCMCIA Bridge
o Silicon Integrated Systems PCI to ISA bridge
o IO read data port for ISA Plug and Play enumerator
o Silicon Integrated Systems Pentium(r) to PCI bridge [original model]

If I read the dialog resource part properly, then Windows has
not or can not allocate any resources for SI PCI ISA bridge?!

[Question]: How can I block this device from being detected
at boot time by Linux? I understand from the that these Unknown PCI device
are supposedly harmless. Hopefully I can give an argument at
the boot prompt. I hope it is not really a kernel driver
problem, otherwise curtains for time being ...

Your analysis seems to be more concrete than my own hypothesis
about Calibrating BogoMIPS ... I have not got anywhere further
with it in the last six days or so.

Meanwhile there was an AJP 5400 owner who thought it might be a
MPEG card, but I dont think I have one in my machine.

Any help appreciated for all AJP 62xx series owners in future.



Peter Pilgrim
System Development, News International

-------------------==== Posted via Deja News ====----------------------- Search, Read, Post to Usenet