Latest working kernel version: 2.6.25
Earliest failing kernel version: 2.6.18 (didn't test earlier)
Distribution: Debian/Testing (pre-lenny)
Microsoft Windows Server's Hyper-V virtual machine
emulates a Digital Equipment Corporation DECchip 21140 [FasterNet] [1011:0009]
which is handled by the tulip driver.
Software Environment: Debian Testing (pre-lenny).
The driver doesn't reports the device's link state, therefore Gnome's
NetworkManager doesn't bring up the interface.
Steps to reproduce:
Install Debian Etch or Testing inside WindowsServer (see ).
Then login and notice that the interface is down.
Created attachment 16826 [details]
Created attachment 16827 [details]
I don't know If this problem can be solved on the driver only, but it would be great if we could find a solution so Windows admins don't get mad when they try Linux ;)
I've opened a similar bug again NetworkManager :
Your help would be appreciated.
1) I'm now the tulip driver maintainer. Can you reassign this bug to me please?
2) It sounds like the bug is the fact that the interface is "ifconfig up" from the RC scripts but isn't getting a link. "ifdown eth0" and "ifup eth0" makes the link work. Is this correct?
Hello from Austria!
Same error here.
Microsoft Virtual Server 2005 R2, Debian 4.0-Image.
"ifconfig -a" delivers no eth0 (but dmesg does), but there is a eth1 which can not be configured, too.
lsmod | grep tulip
>>> tulip 46560 0
modprobe -v tulip
dmesg | grep eth
>>> eth0: Digital DS21140 Tulip rev 32 at 0001ec00, ...
>>> eth1: Using EEPROM-set media 100baseTc-FDX
Your Tip 2 lead to:
>>> interface eth0 not configured
>>> interface eth1 not configured
Maybe this error refers to the date of the image creation, it is 1.5 years old and was created from a former colleague. In the meantime there were some Windows- and VirtualServer-Updates, maybe the image is only to old and it has to be deleted und build a new Debian-Image.
But this is the job of another colleague which isn't in the office today, I'm dealing only exceptionally with this problem.