Wired network works for a few seconds and then stops
 

I have SuSE 9.2 installed on a Dell Inspirion 9300 laptop. The wireless network works just fine. The wired network does not. The symptoms are as follows:

Immediately after enabling the wired network, I get the following console message:

Disabling IRQ #9

The network appears to initialize correctly - it successfully negotiates an IP address via DHCP and for a few seconds, it works exactly as expected. After a few seconds, however, the network fails. Pinging the gateway results in "network unreachable". After this point, nothing I've tried has any effect. ifdown followed by ifup does not successfully connect at all (not even for a few seconds).

The network card is a "Broadcom Corporation BCM4401-B0 100Base-TX (rev 02)". YaST has (correctly, I believe) configured it to use the "b44" driver.

The system is fully updated with all the patches provided by SuSE.

Everything works just fine if I boot into Windows.

Any assistance or suggestions will be very gratefully received!

Here is a trace showing what I've just described above:

--------------------------------------------------------------------------
paul-laptop-linux:~ # ifup eth0
eth0 device: Broadcom Corporation BCM4401-B0 100Base-TX (rev 02)
eth0 configuration: eth-id-00:11:43:7a:fa:67

Message from syslogd@paul-laptop-linux at Mon Jul 4 16:01:07 2005 ...
paul-laptop-linux kernel: Disabling IRQ #9
Starting DHCP Client Daemon on eth0... . IP/Netmask: 192.168.1.70 / 255.255.255.0 ('linux')

paul-laptop-linux:~ # ifconfig eth0
eth0 Link encap:Ethernet HWaddr 00:11:43:7A:FA:67
inet addr:192.168.1.70 Bcast:192.168.1.255 Mask:255.255.255.0
inet6 addr: fe80::211:43ff:fe7a:fa67/64 Scope:Link
UP BROADCAST NOTRAILERS RUNNING MULTICAST MTU:1500 Metric:1
RX packets:38 errors:0 dropped:0 overruns:0 frame:0
TX packets:26 errors:0 dropped:0 overruns:0 carrier:7
collisions:0 txqueuelen:1000
RX bytes:12835 (12.5 Kb) TX bytes:4123 (4.0 Kb)
Interrupt:11

paul-laptop-linux:~ # ping 192.168.1.254
PING 192.168.1.254 (192.168.1.254) 56(84) bytes of data.
From 192.168.1.70: icmp_seq=10 Destination Host Unreachable
From 192.168.1.70 icmp_seq=10 Destination Host Unreachable
From 192.168.1.70 icmp_seq=11 Destination Host Unreachable
From 192.168.1.70 icmp_seq=12 Destination Host Unreachable

--- 192.168.1.254 ping statistics ---
15 packets transmitted, 0 received, +4 errors, 100% packet loss, time 13998ms
, pipe 3

could you please post the output of "route -n" once the card "crashed" ?

your "IRQ" might be caused because you are missing PnP (plug and play) in your kernel. Some laptop need Pnp to be activated in order to use ACPI and PCMCIA... is your card a PCMCIA device?

Here it is:

Nope, not a PCMCIA device - it's a built-in network adapter (Mini-PCI, I believe).

Thanks a lot for your response - any light you can cast would be very welcome!

Paul.

hummm okay, so except the "169.254.0.0 " line (but this should be a minor error) your routing seems fine... so it may be a bug with the drivers or something. My guess would be to make sure you do have PnP included in your kernel, even if you don't use PCMCIA, it may be related. You might want to recompile your kernel to make sure you have it (eeh, good time to learn recompiling kernel, isn't it? :) )

Out of that... try searching google about this network card... maybe someone got the same bug and fixed it already. Trying an older/newer kernel may be a good gues too. Sorry I'm running out of idea :(

I'm pretty sure that the kernel does have PnP compiled in (can't imagine that SuSE wouldn't?). Although I'm not sure how I would go about finding out for sure?

And no, you're right, learning how to recompile the kernel isn't top of my things to do right now :-)

I've done the Google thing - and everything that I can find suggests that things should work (assuming that I've got the latest patches from SuSE). For example http://linux-laptop.net/hosted/dell-inspiron-9300.html.

I'm kinda suspicious that the problem is power management related - the fact that everything works to start off with and then stops suggests to me that something's being powered down which shouldn't be, and isn't IRQ #9 the ACPI interrupt? I'm not a Linux config expert though, so I'm kinda at a loss to work out how to pursue this hunch. Is there some kind of global "switch of power management" that I could do to test this, for example?

Thanks for your help so far!

yup I'm pretty sure it has something to do about power management too, or ACPI, as it is called on laptop :D
SUSE probably have PnP compiled as module... you might try to just load the module, I'm not on a linux computer right now but I'll take a look on my ystem later to give you the module name. But I'm pretty sure that kind of thing should be loaded at boot so it has to be in initrd (assuming Suse use initrd... initrd is a way to load module "driver" just "before" the system bootup, so you can have essential drivers compiled as optionnal module). I can't tell you for sure right now... post me the output of "dmesg" just after the boot (it's the kernel logging... lot of stuff, I warn you) and I could maybe tell you, by comparing the output with my own.

Yup - PnP is definitely there - lots of PnP-related messages in the log.

I've made some more progress - the issue is definitely ACPI-related. If I boot with "pci=noacpi", then everything works just fine. What I'm not sure about is what bad things may occur as a result of booting this way. I will post a question to that effect on the laptop forum, I think!

Very many thanks for all your assistance.

Hmmm maybe I have it wrong since the beginning... maybe you should disable PnP? Back in old time, a lot of PCI device were having troubles with PnP, you often had to disable it. But these days, it should be pretty stable.

Giving "pci=noacpi" isn't a real problem, it shouldn't cause problem except maybe about hardwares able to suspend or to use power management, that you won't be able to use (might cause some bugs too I guess but I can't really tell you). If it works fine for you, it could be a possible fix I guess.