Interrupt handling problem

[Luka Laurent]

Hello!
 
I'm trying to develop support for PCMCIA/CF at our original platform
(Altera EPXA1, ARM922T based, Linux kernel version 2.6, PCMCIA 3.1.22,
Orinoco WiFi driver).
I made a softcore VHDL module and low-level socket service which
controles module and connects it with PCMCIA layer in Linux.
Everything works fine (I got MAC address of local access point and
message 'NET: Registered protocol family 17') until network has to be
established. Then I got a message:
 
'cardmgr[133]: start cmd exited with status 1'. 

I made a research and noticed that I have active IRQ line active from
card during entire period between these two messages. That means some
interrupt (or interrupts) is not handled. If I enable interrupts in my
softcore module I get many interrupts but they are wrongly 'handled'
by socket service. I can see this in /proc/interrupts. It seams like
handling of interrupts between these two drivers (socket service and
Orinoco) in Card service doesn't work well. I think that I made a
mestake in my socket service somewhere.
What is possible cause of this, and does anybody has an idea what
should I do. What are usualy the 'events' that card is sending during
this 'handshaking'?
 
Thanks in advance!
 
Luka