Re: RTL8139 Card and forcedeth : Network Card all confused ..Help !



Hello,

Vidyut a écrit :

I am trying to connect to my BroadBand ADSL connection using my
Ethernet card ( Intex technologies RTL 8139 D ). I tried most of the
things that were mentioned in many threads that .I read in the
Forums :

* Configured Static IP Address, Gateway, Mask etc

On which network interface ? The nVidia one or the Intex one ?

* Tried Unloading forcedeth module ( using sudo modprobe -r forcedeth)

Why would you do such a thing ?

and then loading the drivers 8139cp.o and 8139too.o ( using sudo
modprobe 8139cp and sudo modprobe 8139too)

And what were the resulting kernel messages ?

$ ifconfig

eth0 Link encap:Ethernet HWaddr 00:11:2F:35:0A7
[...]
Interrupt:22

This MAC address is allocated to Asustek, so I guess it belongs to the built-in nVidia ethernet interface on the motherboard.

The output of lspci :

$ lspci
[...]
0000:00:04.0 Ethernet controller: nVidia Corporation nForce2 Ethernet
Controller (rev a1)

If you check with lspci -v, you'll see that interrupt 22 has been allocated to this card.

0000:01:08.0 Ethernet controller: Unknown device 1904:2031 (rev 01)

I've already seen these PCI ID numbers once. This is an Intex/Silan card based on a RTL8139 chip but with non standard PCI ID, I/O and memory layout. Unfortunately, it is not compatible with the standard RTL 8139 driver 8139too at all. When I ckeched on their website (not so long ago), the manufacturer provided only a driver source code for 2.4 kernel. I tried to hack the 2.6 kernel 8139too code to adapt it to the non standard PCI ID and PCI I/O and memory regions, but had no luck. The modified driver recognized the card but failed to use it, so differences with a standard RTL8139 card must be deeper.

If you want to use this card with a 2.6 kernel you may use the Windows NDIS driver with ndiswrapper.

The output of dmesg ( cut out the irrelevant stuff ) :

[4295102.495000] forcedeth.c: Reverse Engineered nForce ethernet
driver. Version 0.41.
[4295102.497000] ACPI: PCI Interrupt 0000:00:04.0[A] -> Link [LKLN] ->
GSI 22 (level, high) -> IRQ 22
[4295102.497000] PCI: Setting latency timer of device 0000:00:04.0 to
64
[4295103.010000] eth0: forcedeth.c: subsystem: 01043:80a7 bound to
0000:00:04.0
[4295103.156000] eth0: no link during initialization.
[4295113.903000] eth0: no IPv6 routers present

That's only related to the built-in nVidia ethernet interface.

This is when the system is first booted. After i unload forcedeth and
insert 8139too and 8139cp, dmesg shows :

[4297822.345000] ACPI: PCI interrupt for device 0000:00:04.0 disabled
[4297831.077000] 8139cp: 10/100 PCI Ethernet driver v1.2 (Mar 22,
2004)
[4297832.941000] 8139too Fast Ethernet driver 0.9.27

If a supported RTL8139 card had been found, a message would have been printed out.

Anyway, why don't you just use the built-in nVidia ethernet interface instead ?
.



Relevant Pages

  • Wired LAN Compact Flash on PPC2003
    ... I need help setting a wired 10/100 MB Ethernet card on pocket pc 2003. ... The driver in the Cd ... Out of desperation I even tried to set it up under the Wireless Network ...
    (microsoft.public.pocketpc.developer.networking)
  • PCI ethernet card from ebay doesnt work at all. Possibly fake, but which?
    ... I needed a PCI ethernet card for an old P3 system and so I did some ... but doing ifup eth0 gives ... Silan cards and the correct driver is sc92031. ... does anyone know what card I really have? ...
    (Debian-User)
  • Wired CF LAN on Pocket PC 2003
    ... I need help setting a wired 10/100 MB Ethernet card on pocket pc 2003. ... The driver in the Cd ... Out of desperation I even tried to set it up under the Wireless Network ...
    (microsoft.public.pocketpc)
  • Replacing an ethernet card
    ... I have a box in production that gives some problems with the ethernet ... card it has installed because of its driver is not fully supported... ...
    (Debian-User)
  • RE: [UPDATED PATCH] EFI support for ia32 kernels
    ... >> reuse a single driver image for multiple architectures assuming there ... As one of the people responsible for the EFI Specification and our ... Perhaps the UNDI network card interface that Intel developed ... BIOS can't shadow that much ROM code. ...
    (Linux-Kernel)