Soyo P4-IS2 motherboard problems with LAN card
 

Hi,
Is anyone out there with LAN card working on SOYO p4-iS2 (i845 chipset)
for RedHat 71./9 or Mandrake 9?

I have tried 3 different LAN cards from Netgear, D-LInk, Intel.. In windows
all work, for Linux ( RH 7.1/9, Mandrake 9) all of these cards are recognised and a suitable driver seems to be loaded for them. But they can't connect to network.

I am trying to connect to internet using my cable modem connection.
I have looked at it from networking perspective and it seems it should be
straight forward to connect using DHCP and I have tried different things.

Now I was wondering whether there could be any problems with the motherboard support. Now since all the LAN cards are recognized and drivers seem to be loaded properly, so these signs are that motherboard is working fine.
But I remember having to load Drivers in Windows case for this motherboard, but there are no drivers provided for Linux and there is no
support from SOYO for Linux.

Did any of you have LAN cards working on a computer with SOYO P4-IS2 motherboard?


Thanks

The problem probably isn't the cards. Intel, Dlink, etc all use relatively generic chipsets, offhand your drivers were probably tulip, 8139too and eepro100s. I would hazard to guess that there's an oddity in how the dhcp request needs to be sent to the modem. Who's your cable provider?

Cheers,

Finegan

Thanks for your response.
The driver loaded are natsemi , 8139too, eepro100 for different cards. The
LAN card is probably fine. I was wondering if MotherBoard p4-IS2 from SOYO could have any problem for Linux.

My cable provider is ComCast. I called those folks and they don't provide
tech support for Linux. But I know many of my friends who have Linux working fine with the same cable connection from ComCast.

Yep, its all Comcast, they do this obnoxious thing where you have to request the dhcp lease AS the hostname that they're going to assign you. Here's a site that gives the lowdown, more or less... some of the tech info is a little dated, for instance RedHat has switched to dhclient since 7.3 instead of dhcpcd, but there's the problem. Post back if you get stuck.

http://www.astro.umd.edu/~teuben/linux/comcast.html

Cheers,

Finegan

I tried after changing ifcfg-eth0 file with recommended settings.. I set
DHCP_HOSTNAME to the hostname I see on my windows machine( when
connected to Comcast).

But unfortunately the same thing, on startup I get the same message
"Determining IP information for eth0 : failed"..
And if I try "ifup eth0" after system Boot up, RedHat 7.1 hangs.

I see ethernet card lights blinking normally.. the driver is loaded properly.

I see the following messages coming from driver code in /var/log/messages, these are some comments for some fix it seems.

***********************
Nov 4 11:00:48 localhost kernel: ip_conntrack (4095 buckets, 32760 max)
Nov 4 11:00:48 localhost kernel: eepro100.c:v1.09j-t 9/29/99 Donald Becker http://cesdis.gsfc.nasa.gov/linux/drivers/eepro100.html
Nov 4 11:00:48 localhost kernel: eepro100.c: $Revision: 1.36 $ 2000/11/17 Modified by Andrey V. Savochkin <saw@saw.sw.com.sg> and others
Nov 4 11:00:48 localhost kernel: PCI: Found IRQ 11 for device 02:01.0
Nov 4 11:00:48 localhost kernel: eth0: Intel Corporation 82557 [Ethernet Pro 100], 00:A0:C9:8A:E0:22, I/O at 0xa000, IRQ 11.
Nov 4 11:00:48 localhost kernel: Receiver lock-up bug exists -- enabling work-around.
Nov 4 11:00:48 localhost kernel: Board assembly 678400-001, Physical connectors present: RJ45
Nov 4 11:00:48 localhost kernel: Primary interface chip i82555 PHY #1.
Nov 4 11:00:48 localhost kernel: General self-test: passed.
Nov 4 11:00:48 localhost kernel: Serial sub-system self-test: passed.
Nov 4 11:00:48 localhost kernel: Internal registers self-test: passed.
Nov 4 11:00:48 localhost kernel: ROM checksum self-test: passed (0x49caa8d6).
Nov 4 11:00:48 localhost kernel: Receiver lock-up workaround activated.
Nov 4 11:00:48 localhost kernel: NETDEV WATCHDOG: eth0: transmit timed out
Nov 4 11:00:48 localhost kernel: eth0: Transmit timed out: status f048 0000 at 0/28 command 0001a000.
*************************************
I get the following messages regarding eth0 interface.

***********************************
Nov 4 10:22:26 localhost kudzu: succeeded
Nov 4 10:22:26 localhost sysctl: net.ipv4.ip_forward = 0
Nov 4 10:22:26 localhost sysctl: net.ipv4.conf.all.rp_filter = 1
Nov 4 10:22:26 localhost sysctl: kernel.sysrq = 0
Nov 4 10:22:26 localhost network: Setting network parameters: succeeded
Nov 4 10:22:26 localhost network: Bringing up interface lo: succeeded
Nov 4 10:22:26 localhost ifup: Determining IP information for eth0...
Nov 4 10:22:26 localhost pumpd[394]: starting at (uptime 0 days, 0:00:27) Tue Nov 4 10:22:26 2003
Nov 4 10:22:37 localhost /etc/hotplug/usb.agent: ... no drivers for USB product 0/0/0
Nov 4 10:22:37 localhost /etc/hotplug/usb.agent: ... no drivers for USB product 0/0/0
Nov 4 10:23:26 localhost ifup: Operation failed.
Nov 4 10:24:26 localhost dhcpcd[413]: timed out waiting for a valid DHCP server response
Nov 4 10:24:26 localhost ifup: failed.
Nov 4 10:24:26 localhost network: Bringing up interface eth0: failed
Nov 4 10:24:29 localhost gpm: gpm startup succeeded
*******************************************


I have no clue what is going on, whether it is in hardware, driver, dhcp config or Comcast?

Any help is greatly appreciated.


Thanks

Nah, there's still a dhcp timeout, so when the machine is up, give it the command from the command line, as root:

dhcpcd -h hostname eth0

Case sensitivity may matter also...

Cheers,

Finegan

When I issue this command, my RedHat 7.1 hangs. And I don't see any log
of it in /var/log/messages.

Next time, I selected network configuration option from GUI Menu, this time also the system hangs.. somthing is definitely wrong when I touch the networking part related to eth0.

Hangs? As in stops dead or that command just takes a long time to time out? Okay, that's just bloody weird that a dhcp request is causing an unlogged kernel hang. If this command causes it to hang, then it is in fact probably something wrong with the card:

ifconfig eth0 192.168.0.1 up
ping -b 192.168.0.0

That'll at least bring up the device with an albeit fake IP and then broadcast ping accross the line. If the driver for the eepro100 has gone kazoo, it should crash the machine just the same as a dhcp request...

Cheers,

Finegan

I have the same problem
 

Hi. I also have a SOYO motherboard with the same chipset. Different model though.
I run the ofconfig eth0 192.168.0.1 up and ping 192.168.0.0 -b ok. If I try to ping anything else, it fails.
DHCPD times out too.
Any advice to get my network up?

Thanks.