SCOUG-HELP Mailing List Archives

<< Previous Message << >> Next Message >>

Date:	Tue, 13 Sep 2005 16:02:35 PDT7
From:	"Michael Rakijas" <mrakijas@adelphia.net >
Reply-To:	scoug-help@scoug.com
To:	scoug-help@scoug.com
Subject:	SCOUG-Help: Ethernet IRQ and SPCIIRQ related questions

** Reply to message from "Steven Levine" on Sun, 11 Sep
2005 20:56:12 PDT7

> In <20050912025927.VWSF12165.mta10.adelphia.net@[192.168.1.99]>, on
> 09/11/05
> at 07:59 PM, "Michael Rakijas" said:
>
> >So far, so good. I found the proper driver for the VIA chipset,
> >FETND.OS2 and the associated .NIF file.
>
> Unless I'm confused, this driver has a less than wonderful reputation.
> VIA chipsets are not all that well regarded in general.

Interesting. Before having loaded any drivers, NICPAK reported up to 6 options
after chipset detection in addition to the FETND. It suggested the Realtek
8139 as well, which I thought was curious (I've had good luck with these) but
since the machine is a VIA based system, I decided to go with the FETND. Is
the Realtek 8139 driver a real option or a lesser/inaccurate detection?
Curiously, after having loaded the FETND, NICPAK reports only the FETND option.

> >During boot up, the data shown at the loading of the
> >driver says IRQ=0xB, which implies to me that it is using IRQ11.
>
> Agreed.
>
> >I go to
> >the device manager and it shows the Ethernet hardware device in the
> >equipment list as using IRQ10.
>
> That sounds like a defect to me.

In the driver with the less-than-stellar reputation, you think or elsewhere?

> >They both for UHCI Compatible USB Host Controllers. LANTRAN.LOG seems to
> >show an IRQ conflict as well. It reports an LT00042 error that FETND_nif
> >was not able to bind because multiple adapters are trying to use the same
> >IRQ level.
>
> Any reasonably well implemented PCI driver can share IRQs.

Apparently not this one, huh?

> >Fine, I say. So, I go to SPCIIRQ to try and get a change in the IRQ ...
> >but which? Report in SPCIIRQ says for the Ethernet device as Bus=0,
> >Device=18, Function=0 Ethernet(11) despite what the device manager says.
>
> Which report? Do you mean the pci.exe output?

When you run the report.cmd that comes with SPCIIRQ (which contains
individually many of the reports you are asking for below) which aggregates the
output into one file called report.log.

> >So, the device line in CONFIG.SYS reads either BASEDEV=SPCIIRQ.SYS 0 18 0
> >1 10 or BASEDEV=SPCIIRQ.SYS 0 18 0 1 11 but both report an SPCIIRQ
> >parameter error when trying to load.
>
> I probably need to see the output of:
>
> pci -s
> pci -bdp
> show_pir
> show_lnk
> rmview /irq

This group of commands was executed in sequence with output collected into a
file called probes.log (although pci does not take composite options as shown -
I think it works as "pci -b -d -p" and that's what was included) and ...

> and your current
>
> config.sys

both are included in the attached DIAGNOSE.ZIP file.

> before I can make any reasonable guesses on what to do next.

> >Is there some way to tell the FETND device driver to
> >change which IRQ to trigger on?
>
> Check the NIF for an IRQ setting.

Strangely, the FETND.NIF file does not have an entry for the IRQ option range.
I've seen these in other NIF files usually specifying a default value and the
possible values but this NIF does not.

> Regards,
>
> Steven

Thanks for your help, as always.

-Rocky

File attachment: DIAGNOSE.ZIP

Return to [ 13 | September | 2005 ]

SCOUG, Warp Expo West, and Warpfest are trademarks of the Southern California OS/2 User Group. OS/2, Workplace Shell, and IBM are registered trademarks of International Business Machines Corporation. All other trademarks remain the property of their respective owners.