Re: Installing Debian on a Qube 1
Peter Horton wrote:
On Sat, Sep 18, 2004 at 11:07:43AM +0200, Christoph Pfisterer wrote:
* Christoph Pfisterer <cp@chrisp.de> [2004-09-17 08:47]:
What stumbles me is that CoLo doesn't seem to load the file
default.colo from the NFS server.
Did you tell your DHCP server that it should serve default.colo by
default (filename "default.colo";)?
Yes, I have that setting on my DHCP server. I've now peered at the
CoLo
source code and realized that CoLo does its own DHCP query to get that
name. My server's log show only the BOOTP query done by the built-in
firmware, which happens to be consistent with the LCD displaying only
"Booting..." and no IP address. My prime suspect is now the tulip
driver code in CoLo. I'm not even sure the Cube 1 has the same
networking hardware as the Cube 2...
I'm pretty sure the NIC is still a Tulip. I wonder if it's still in the
same PCI position and with the same interrupt ?
Anyone got a dmesg from one of these beasts ?
I used the Cobalt CD to restore the original OS. (Very nice to have the
box actually running for a change. ;-) ). Here's the relevant part from
dmesg:
>>>
Linux version 2.0.33 (root@ninka.cobaltmicro.com) (gcc version 2.7.2)
#1 Tue Mar 10 18:00:37 PST 1998
<<<
>>>
tulip.c:v0.76 4/22/97 becker@cesdis.gsfc.nasa.gov
eth0: DEC 21041 Tulip at 0x10100000, 21041 mode, 00 10 e0 00 37 68, IRQ
13
eth0: Old style EEPROM -- no media selection information.
<<<
And from /proc/pci:
>>>
Bus 0, device 7, function 0:
Ethernet controller: DEC DC21041 (rev 33).
Medium devsel. Fast back-to-back capable. IRQ 13. Master
Capable. Latency=128.
I/O at 0x10100000.
Non-prefetchable 32 bit memory at 0xab224800.
<<<
(0,0,0) is the host bridge, (0,9,[0-2]) is the VIA chipset.
Hope this helps,
chrisp
--
chrisp a.k.a. Christoph Pfisterer "Any sufficiently advanced
cp@chrisp.de - http://chrisp.de bug is indistinguishable
PGP key & geek code available from a feature."
Reply to: