(All output is from the latest PCMCIA snapshot as of 1/14/99 with
kernel 2.2.0-pre7) Hi there, I'm trying to get a Linksys 10/100
cardbus card to work on a Toshiba Tecra 8000 which is apparently
powered by the infamous ToPic95 chipset. The card exhibits identical
behaviour in a Dell Inspiron 3500 which seems to have a TI cardbus
controller. The card works in both machines on NT, so it's not a
problem with the card itself...in fact it looks like the card *almost*
works under Linux which is very frustrating. I've moved its base
address and irq around to no avail, which means it shouldn't be a
conflict of some sort. The cardmgr and modules seem to load up fine:
Linux PCMCIA Card Services 3.0.8
kernel build: 2.2.0-pre7 #1 Thu Jan 14 18:52:34 EST 1999
options: [pci] [cardbus] [apm]
Intel PCIC probe:
Toshiba ToPIC95-B PCI-to-CardBus at bus 0 slot 11, mem 0x68000000, 2 sockets
host opts [0]: [slot 0xf0] [ccr 0x10] [cdr 0x86] [rcr 0x02] [no pci irq] [lat
168/176] [bus 20/20]
host opts [1]: [slot 0xf0] [ccr 0x10] [cdr 0x86] [rcr 0x02] [no pci irq] [lat
168/176] [bus 21/21]
ISA irqs (default) = 3,4,5,7,9,10 polling interval = 1000 ms
Previous incarnations of the pcmcia drivers generally pick up a pci
irq of 11, which is set in the BIOS. I can get a normal PCMCIA
Linksys NE2K compat PC-Card to work inboth machines fine, however the
Cardbus card using the Tulip driver isn't very happy. Inserting the
card doesn't evoke a response from the cardmgr, but you can get it to
recognise the card with a 'cardctl insert' at which point:
cs: IO port probe 0x0a00-0x0aff: clean.
cs: cb_config(bus 20): vendor 0x1011, device 0x0019
fn 0 bar 1: io 0xa00-0xa7f
fn 0 bar 2: mem 0xa00c0000-0xa00c03ff
fn 0 rom: mem 0xa0080000-0xa00bffff
cs: cb_enable(bus 20)
bridge io map 0 (flags 0x21): 0xa00-0xa7f
bridge mem map 0 (flags 0x1): 0xa0080000-0xa00c0fff
tulip_attach(bus 20, function 0)
tulip.c:v0.89L 10/9/98 [EMAIL PROTECTED]
eth0: Digital DS21143 Tulip at 0xa00, 00 80 00 80 00 80, IRQ 10.
eth0: EEPROM default media type 10baseT.
The big tip off that something is wrong is of course the hardware
address...00:80:00:80:00:80 isn't it. At this point on the Toshiba,
2.0 kernels generally get extremely unstable and lock hard within a
minute or so, while 2.2 kernels seem fine. The Dell and the 2.2 kernel on the Toshiba
both eventually give you:
eth0: The transmitter stopped! CSR5 is f0678006, CSR6 b3862002.
I've made a bunch of files available at
http://www.snurgle.org/~griffon/topic/ which contains as much
debugging info as I thought to gather off the top of my head, mostly
from /proc, various incantations of lspci and dmesg, any help/advice
anyone could provide would be greatly appreciated.
Thanks in advance.
--
+------- --- -- -- -
| [EMAIL PROTECTED] / My opinions represent snurgle.org :
! Unspoilt by progress / http://www.snurgle.org/~griffon |
- -- -- --- -------+
-
To unsubscribe from this list: send the line "unsubscribe linux-net" in
the body of a message to [EMAIL PROTECTED]
