Yenta Socket work with 32-bit card but not 16-bit

[David Clark]

We solved our issue; it ended up being a hardware problem.  A mistake
was made when designing the PCB.  Thank you for your assistance.

David Clark
Senior Software Engineer
C&H Technologies, Inc
Web: http:\\www.chtech.com
Phone: 512-733-2621
Fax: 512-733-2629
Email: dlclark at chtech.com
 

-----Original Message-----
From: linux-pcmcia-bounces at lists.infradead.org
[mailto:linux-pcmcia-bounces at lists.infradead.org] On Behalf Of Dominik
Brodowski
Sent: Thursday, February 15, 2007 12:29 PM
To: David Clark
Cc: linux-pcmcia at lists.infradead.org
Subject: Re: Yenta Socket work with 32-bit card but not 16-bit

Hi,

On Thu, Feb 15, 2007 at 12:18:20PM -0600, David Clark wrote:
> I may be way off track here and pleasae tell me if so.  From looking
at
> the source code in "rsrc_nonstatic.c" I gather that the first time the
> socket comes up, the kernel probes (do_mem_probe()) the PCI memory
space
> to find a suitable memory region which can be allocated to the PCMCIA
> cards.  

Right. Such mem resources are needed to access the card "CIS", too,
which in
turn is required to correctly determine which card this is.

> In my case 0x80000000-0xfcffffff is the entire memory range available
to
> the host PCI bridge.  Not all of this memory range is allocated to the
> various PCI devices, so why then does the memory probe not find any
> available resources?

Good question -- to which I don't really have an answer, unfortunately.
Could you try out what happens if you disable "CONFIG_ISA" in your
kernel
config (which in turn should switch of CONFIG_PCMCIA_PROBE)?

> Is this potentially related to my problem?  Again I'm trying my best
to
> understand how this all works.

Most likely this is the root of the problem.

	Dominik

_______________________________________________
Linux PCMCIA reimplementation list
http://lists.infradead.org/mailman/listinfo/linux-pcmcia