[tulip] tulip-diag eeprom writing - some questions

Donald Becker becker@scyld.com
Wed, 19 Jul 2000 10:54:57 -0400 (EDT)

On Wed, 19 Jul 2000 jlhagen@collins.rockwell.com wrote:

> > This is just the result of the 21143 chip not existing in I/O space.
> dmesg reports 'tulip.c:v0.91g-ppc 7/16/99 becker@cesdis.gsfc.nasa.gov' . I
> should mention I'm running a Mandrake 6.1distribution with a clean 2.2.14
> source tree installed. I'm not at the machine right now, does this version
> (0.91g) have the option of using memory space instead of I/O space??? I
> think I should try and grap the tulip.c:v0.92 4/17/2000 version. I see 0.92
> by default does not use 'USE_IO_OPS'  which is what we want, I believe.

I suspect a power-down issue, or an invalid assignment by the BIOS.

> I ran pci-config once with ' -a ' and got this. Really I haven't played
> with it that much. I need to look at the code to
> understand how to use it correctly.
> pci-config.c:v2.00 4/19/2000 Donald Becker (becker@scyld.com)
> http://www.scyld.com/diag/index.html
> Device #8 at bus 0 device/function 17/0, 00241011.
> Device #9 at bus 0 device/function 19/0, 00241011.

Ohh, complication: These are the bus bridges on your multiport card.  You'll
have to find their bus number in /proc/pci and do
  pci-config -B <bus-number>
  pci-config -B <bus-number>
and look at the output for the individual chips.
The 'lspci', if you have it, might do this without as much mucking about.

If the device has ACPI power management (not all 21143 chips do), it must
be in ACPI D0 state, not ACPI D3 state.

Donald Becker				becker@scyld.com
Scyld Computing Corporation		http://www.scyld.com
410 Severn Ave. Suite 210		Beowulf Clusters / Linux Installations
Annapolis MD 21403