[Orinoco-users] Re: [PATCH] Disable (false?) IRQ check on 2.6 in orinoco_cs (was: yenta_socket.c:yenta_get_status returns bad value compared to 2.4)

[Stefan Jones]

Joseph Gruber said:
> Are you talking about hanging at "Starting pcmcia" during the boot
> process?
>

No, when I insert a pcmcia Netgear MA401 wireless card. "Starting pcmcia"
works fine, with no pcmcia card in the slot.

When I insert the pcmcia card, cardmgr starts it's ioctl things apon
recieving a 0x04 from the pcmcia charater device.

It first calls ds_ioctrl(DS_GET_STATUS) and then
ds_ioctrl(DS_ADJUST_RESOURCE_INFO). This works fine

Then it does ds_ioctrl(DS_VALIDATE_CIS)

Which goes to cistpl.c which then goes to the "duff" funtion
validate_mem() inrsrc_mgr.c .
This is with ISA on, so PCMCIA_PROBE is also on.

The machine is now hung with a blank screen with all the hardware going
screwy!

Now if I unset ISA then it uses a different validate_mem() and all works
fine. ( well the wireless card appears on ifconfig eth1 after the patch to
orinoco_cs, I haven't got a base station though ... the dhcpd packets it
sends out are logged in ifconfig )

If I zap/hide all pcmcia device drivers then the hang still occours so it
is in the core pcmcia system, not a faulty device driver.

I have also tried the cardmgr patch from the linux-pcmcia list.

For full system config see:
http://www.ussg.iu.edu/hypermail/linux/kernel/0307.3/0166.html
The machine is an Acer Aspire 1300.

Thanks for reading,

Stefan