Accton Cheetah 10/100 problems

[Bryan Scaringe]

Damn! That looks just like the situation/error I've been getting
for over a month with my PCI NIC.  I finally solved it on
friday.  How, you ask?  I opened up my case, and move the
card from one PCI slot to the other.  /proc/pci then listed
the card correctly (a generic RealTek).  I compiled the driver
directly into the kernel, and handed the parameters to LILO.

I have no idea if it was the PCI slot that it was in, or if
I simply didn't have the card in all the way.

Mr. Becker suggested that I check that the card was in all the way.
Ths seemed silly at the time, so I put it off for a week.
I wish I hadn't wated...

Bryan Scaringe 


> 
> I have big trouble getting my Accton 10/100 Cheetah PCI NIC detected by
> RedHat 5.1 with kernel 2.0.34.
> Here are some data :
> 
> My /etc/conf.modules looks like this:
> 
> alias eth0 tulip
> options tulip debug=6 options=0
> 
> Running 'insmod tulip' i get the following error :
> 
> tulip.o: init_module: Device or resource busy
> 
> 
> My /proc/pci file looks like this :
> 
> PCI devices found:
>  Bus  0, device  12, function  0:
>   Ethernet controller: Unknown vendor Unknown device (rev 16).
>   Vendor id=1113. Device id=1211.
>   Medium devsel.  IRQ 11.  Master Capable.  Latency=32.  Min
> Gnt=32.Max Lat=64.
>      I/O at 0xd800.
>      Non-prefetchable 32 bit memory at 0xe6000000.
> 
> I have checked the /proc/ioports and /proc/interrupts that there is no other
> device using the I/O or interrupt.
> I notice that the vendor and device is unknown. Something missing in pci.h
> ???
> Probably because the vendor and device is unknown, i must use the -p d800
> option with the diag utility so that it can find the card :
> 
> # tulip-diag -a -e -e
> Unable to find a Tulip card in /proc/pci
> 
> # tulip-diag -p d800 -a -e -e
> 
> tulip-diag.c:v1.06 9/18/98 Donald Becker (becker@cesdis.gsfc.nasa.gov)
> Digital DC21040 Tulip Tulip chip registers at 0xd800:
>   4fe80000 00000000 00002000 00002000 00000000 00000000 00000000 0000fff0
>   60000000 1594a161 000c1000 0000000c 1000000f 00000000 00000024 00000000
>  The Rx process state is 'Stopped'.
>  The Tx process state is 'Stopped'.
> Transmit stopped, Receive stopped, half-duplex.
>  The transmit threshold is 72.
>  Port selection is 10mpbs-serial, half-duplex.
> EEPROM contents:
>   1f0d 382c 4e43 6459 7a6f 9085 a69b bcb1
>   d2c7 e8dd fef3 1409 2a1f 4035 564b 6c61
>   8277 988d aea3 c4b9 dacf f0e5 06fb 1c11
>   3227 483d 5e53 7469 8a7f a095 b6ab ccc1
>   e2d7 f8ed 0e03 2419 3a2f 5045 665b 7c71
>   9287 a89d beb3 d4c9 eadf 00f5 160b 2c21
>   4237 584d 6e63 8479 9a8f b0a5 c6bb dcd1
>   f2e7 08fd 1e13 3429 4a3f 6055 766b 8c81
>  ID CRC 0x91 (vs. 0xc7), complete CRC 8310ce98.
> 
> Is there a way to tell modprobe (in /etc/conf.modules) what I/O address the
> card is at ?
> I have tried  options tulip options=0 debug=6 io=0xd800
> but it will not buy it.
> 
> Steinar
> 
>