FW: TI1510 probelm - can you help direct me?

[Warren Melnick]

Funny that you mention the pcmcia-cs drivers and the irq_mode=0, because
that is exactly right.  I discovered that a while ago.  The problem with
that is that the prism54 drivers are not compatible with the pcmcia-cs
code, so I cannot use it, as the whole point here is to try to get
current PCMCIA wireless adapters (Netgear, DLink and/or Linksys b and/or
g) to work. Also, I cannot go up to kernel 2.6 because the motherboard
has an MPEG decoder chip built onto it that has no 2.6 support.

I will try the patch from your other email and let you know the results.

I am also happy to send one or two of the older versions of this
motherboard out to anyone who can help with this problem - the changes
to the newer versions had nothing to do with the PCMCIA, mostly sound
and networking problems.

Regards,
Warren

-----Original Message-----
From: Pavel Roskin [mailto:proski at gnu.org] 
Sent: Wednesday, May 05, 2004 7:47 PM
To: Warren Melnick
Cc: linux-pcmcia at lists.infradead.org
Subject: Re: FW: TI1510 probelm - can you help direct me?

On Wed, 5 May 2004, Warren Melnick wrote:

> One time it made it through:
> Linux Kernel Card Services 3.1.22
>   options:  [pci] [cardbus] [pm]
> Yenta TI: socket 00:0b.0, mfunc 0x00c01002, devctl 0x66
> Yenta TI: socket 00:0b.0 probing PCI interrupt failed, trying to fix
> Yenta TI: socket 00:0b.0 falling back to parallel PCI interrupts
> hub.c: port 5 over-current change
> hub.c: port 6 over-current change
> Yenta TI: socket 00:0b.0 no PCI interrupts. Fish. Please report.
> Yenta ISA IRQ mask 0x0040, PCI irq 0
> Socket status: 30000007

hub.c is a part of the USB driver.  Something is happening to the USB
bus
while the PCMCIA bridge is being probed.

I've seen similar problems.  The symptoms are different, but the hang is
more likely if a 5V card is used and some USB devices are attached.  It
seems that the PCMCIA bridge starts sucking too much power or does
something wrong with the PCI bus.

This problem doesn't happen with the i82365 driver from pcmcia-cs.
Since
you are using a 2.4 kernel it should be possible for you.  Use the
"irq_mode=0" parameter to the i82365 module to force PCI interrupts.

I'm trying to figure out what is wrong in the kernel yenta_socket
driver,
but it may take some time.

> Any ideas or steps forward would be appreciated.  I am perfectly
willing
> and happy to spend whatever time and effort it takes to get this
> working.

All you need to do is to find out what i82365 does right but
yenta_socket does wrong.

-- 
Regards,
Pavel Roskin