Hi All
2.6.12, TitaniumIV, unstable.
This Debian boot system seems to be obsessed to find my ethernet crap
at boot time: I want to load my airport drivers, and only those, at
boot time, and not the ethernet crap, and I set
/etc/network/interfaces
accordingly:
Excerpt:
__________________________________
# /etc/network/interfaces -- configuration file for ifup(8), ifdown(8)
# The loopback interface
auto lo
iface lo inet loopback
# wireless orinoco:
auto eth0
iface eth0 inet dhcp
pre-up /home/shorty/scripts/firewall.sh
pre-up modprobe airport
[ ... ]
# START/STOP WEP DRIVERS:
pre-up modprobe arc4
pre-up modprobe crc32c
post-down rmmod arc4
post-down rmmod crc32c
post-down rmmod libcrc32c
# END WEP DRIVERS
[ ... ]
________________________
But nearly every reboot the system somehow loads these sungem(_phy)
drivers, that, as far as I can see, are responsible for the mess
following when the system is trying to load unsuccessfully eth0
(airport) according to the interfaces file above.
Excerpt from /var/log/kern.log from, IINM, boot time, mentioning
sungem:
-----------------------------------------------------------------
try `ip1394'
Oct 12 21:42:43 debby kernel: [ 48.217962] ohci1394: $Rev: 1250 $ Ben Collins
<[EMAIL PROTECTED]>
Oct 12 21:42:43 debby kernel: [ 48.246573] ohci1394: fw-host0: Unexpected PCI
resource length of 1000!
Oct 12 21:42:43 debby kernel: [ 48.305601] ohci1394: fw-host0: OHCI-1394 1.0 (
PCI): IRQ=[40] MMIO=[f5000000-f50007ff] Max Packet=[2048]
...skipping...
Oct 12 21:42:43 debby kernel: [ 48.979710] sungem.c:v0.98 8/24/03 David S.
Miller (davem@redhat.com)
Oct 12 21:42:43 debby kernel: [ 49.049083] PHY ID: 2060e1, addr: 0
Oct 12 21:42:43 debby kernel: [ 49.066211] eth0: Sun GEM (PCI)
10/100/1000BaseOct 12 21:42:43 debby kernel: [ 49.073393] eth0: Found BCM5421
PHY
023] GUID[000393fffecde4c4]
[EMAIL PROTECTED]>
1394 Ethernet (fw-host0)
e, 2048kB Cache, DMA
0
Oct 12 21:42:43 debby kernel: [ 55.835366] NET: Registered protocol family 10
ce c02cb330(lo)
Oct 12 21:42:43 debby kernel: [ 55.846722] IPv6 over IPv4 tunneling driver
core team
ets, 49152 max) - 248 bytes per conntrack
---------------------------------------------------------------------
So anyone knows which app/config at boot time might be responsible for
loading the sungem stuff?
I already completely removed discover to fix this:
apt-get remove discover1-data discover1 libdiscover1
but to no avail so far ...
Where are these apps/files that load this ethernet crap? Hal, udev,
hotplug?
I already grepped the whole /etc/ area for the "eth" pattern and
such, but nothing so far ...
And sure, I know I can blacklist sungem altogether somewhere in
/etc/hotplug .. But I think it's brain-dead to firstly let some
routine/script try loading a driver I don't want, and then remove it
again via the hotplug stuff. That's why I'm looking for this boot
routine that's obsessed by the idea to load sungem ...
Just in case: No, sungem is not loaded via /etc/modules.
Ah yes, nearly forgot that: Loading my airport card manually after
booting works like charm, after unloading the airport
etc./sungem(_phy) drivers. So it seems it's simply a lousy boot time
set up being responsible for the mess ....
Again: My primary target is to get *sensibly* rid of sungem and
sungem_phy being loaded at boot time.
Any hints?
Thanks in anticipation
Wolfgang
--
Wolfgang Pfeiffer
http://profiles.yahoo.com/wolfgangpfeiffer
Key ID: E3037113
Key fingerprint = A8CA 9D8C 54C4 4CC1 0B26 AA3C 9108 FB42 E303 7113
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
