I've just upgraded my PC from Suse 9.1 to Suse 10. Complete new install. No hardware changes. The PC is networked by ethernet cable from a Netgear WG311v2 router, is configured for DHCP. Post installation I have no internet or network access.

Find I cannot ping the router, 192.168.0.1 - although I get a dynamic IP address from the router at boot, 192.168.0.105, and I can ping that. Router is firewalled. Makes no difference whether Linux firewall is enabled or not.

PC also has Windows XP installed using exactly the same network hardware, which continues to function as usual.

The story is the same if I take the PC off the router and connect it straight to the cable modem. DHCP gets an IP address from the ISP and updates my nameservers, but I can't ping those nameservers after normal boot.

If I boot in failsafe mode with no GUI then I can __ping__ my router. There is no difference between the output of `ifstatus eth0' in failsafe mode and normal mode.

Any ideas much appreciated.

What says ifconfig?

Hi,

After normal boot, when the router is unreachable, ifconfig says this:

eth0 Link encap:Ethernet HWaddr 00:01:2E:02:97:38
inet addr:192.168.0.105 Bcast:192.168.0.255 Mask:255.255.255.0
inet6 addr: fe80::201:2eff:fe02:9738/64 Scope:Link
UP BROADCAST NOTRAILERS RUNNING MULTICAST MTU:1500 Metric:1
RX packets:123 errors:0 dropped:0 overruns:0 frame:0
TX packets:20 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:38486 (37.5 Kb) TX bytes:2332 (2.2 Kb)
Interrupt:153 Base address:0x4000

After failsafe boot, when the router is reachable, ifconfig says this:

eth0 Link encap:Ethernet HWaddr 00:01:2E:02:97:38
inet addr:192.168.0.105 Bcast:192.168.0.255 Mask:255.255.255.0
inet6 addr: fe80::201:2eff:fe02:9738/64 Scope:Link
UP BROADCAST NOTRAILERS RUNNING MULTICAST MTU:1500 Metric:1
RX packets:78 errors:0 dropped:0 overruns:0 frame:0
TX packets:25 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:22278 (21.7 Kb) TX bytes:2784 (2.7 Kb)
Interrupt:5 Base address:0x2000

I tried going into Yast Network Devices and giving the eth0 module parameters irq=5 io=0x2000, as per failsafe boot. However when Yast saved this and retarted the interface, ifconfig still showed interrupt 153, but with base address 0x2000. I then had no network at all, i.e. ping <anything> returned "No network available".

There are 2 network cards in the PC, one wireless and one ethernet. Both of them worked under Suse 9.1 as long as they weren't both up at the same time. The wireless card is not configured in this new Suse 10 installation. However Yast Hardware Information shows that both cards were configured with interrupt 153, base address 0x4000. Don't know if that is significant or not.

Cheers!

I have the same problem .. any found solution

Are you sure, that the router has the IP-adress 192.168.0.1 ?

Your computer gets a IP-adress: the router is working.

Make that three with the exact same problem..I have tried everything I know(which could fit in a thimble) but cannot get online with 10.0...I jhust don't know enough about networks to get it going..
Just for information, I use a liksys 4port router/sw to connect to the internet via cable modem..I use a hard drive removable rack so when I plug in suse 9.3 or 9.2 I can connect to the internet fine..Tried both DHCP and static IP's but
cannot answer all the configuration questions..(noob)
I'll watch this thread to see if anyone can get theirs going so if yours works please
tell us how you got it going..
jolphil

Just thinking out loud -- I don't have a solution but have you tried removing the unused network card to see if that simplification of your hardware improves things. I may be mistaken but two cards with the same interrupt and base address sounds strange to me.

Hi all,

>Are you sure, that the router has the IP-adress 192.168.0.1 ?

Yes.

>have you tried removing the unused network card to see if that simplification of your hardware improves things.
>I may be mistaken but two cards with the same interrupt and base address sounds strange to me.

I suspected that, but I've now deleted the wireless card from the configured network devices and it makes no difference. And when I boot in failsafe mode I can reach my router, whether or not the wireless card is configured.

I posted a support request to Novell last Sunday and an engineer eventually responded yesterday. He asked me to run `hwinfo' and `siga no -net' and send him the output, which I did last night. No reply yet, but I studied the output myself (never heard of siga before). It has a section where it captures the output of `grep -v SFW2- /var/log/messages | tail -n 100' and in among this I found:

Nov 17 21:02:00 mike kernel: bootsplash: status on console 0 changed to on
Nov 17 21:02:00 mike kernel: eth0: no IPv6 routers present
Nov 17 21:02:00 mike kernel: irq 153: nobody cared (try booting with the "irqpoll" option)
Nov 17 21:02:00 mike kernel: [] __report_bad_irq+0x24/0x90
Nov 17 21:02:00 mike kernel: [] note_interrupt+0x72/0xc0
Nov 17 21:02:00 mike kernel: [] __do_IRQ+0x132/0x150
Nov 17 21:02:00 mike kernel: [] do_IRQ+0x36/0x70
Nov 17 21:02:00 mike kernel: [] common_interrupt+0x1a/0x20
Nov 17 21:02:01 mike kernel: handlers:
Nov 17 21:02:01 mike kernel: [] (rtl8139_interrupt+0x0/0x220 [8139too])
Nov 17 21:02:01 mike kernel: Disabling IRQ #153

This is after the bootsplash as you see, so eth0 has already been brought up some time before and DHCP has got me an IP address. But now it seems that eth0 is trying to talk IPv6 to my router, and then the kernel is disabling my card's IRQ because it can't. I think this is the bug. I haven't done anything to ask for IPv6 and I don't know how to say I don't want it.

Anybody know?

Other victims of this problem can run `siga no -net' and it it will create a report in /tmp/siga/siga.html. Browse that file for section 2.3, messages filtered, and see if you find something like the above.

Cheers

Hi,

Problem is solved.

My theory about IPv6 was nonsense. I found out how to disable IPv6 and it made no difference. Still seeing IRQ 153 disabled in the boot messages.

Then I took the advice in the boot message:

irq 153: nobody cared (try booting with the "irqpoll" option)

Boot option irqpoll fixes it. Googling irqpoll doesn't tell me what this option does. Does anybody know? My eth0 is __still__ on IRQ 153, but now it works.

Thanks for all advice.

Cheers

Hi

This is taken straight out of the "kernel-parameters.txt" file which should be in your /usr/src/linux/Documentation folder (assuming you have the sources installed)

+ irqpoll [HW]
+ When an interrupt is not handled search all handlers
+ for it. Also check all handlers each timer
+ interrupt. Intended to get systems with badly broken
+ firmware running.


Ian

Just picked up on this thread and noticed the irqpoll option and incorrect firmware. In order to get my WG311 to respond correctly under Debian, I first had to update the firmware, as outlined here - http://www.houseofcraig.net/acx100_howto.php . I then had to compile the drivers, but if SuSE does that directly for you, simply follow the instructions for firmware update and then try without passing irqpoll at the boot prompt. Just a suggestion