Yenta and IRQ0

[Russell King]

On Tue, Aug 10, 2004 at 09:42:40AM -0700, David Hinds wrote:
> On Tue, Aug 10, 2004 at 05:30:30PM +0100, Russell King wrote:
> > 
> > Is there any reason that we're not just out-right refusing to allow
> > yenta to initialise if we find that we don't have the PCI interrupt
> > assigned by the PCI subsystem/BIOS ?
> > 
> > FWICS, we try to emulate the status change interrupt handling via
> > a timer in yenta.c, but this only half-solves the issue - PCI drivers
> > for cardbus cards themselves won't work with this arrangement.
> > 
> > Could we instead refuse to initialise and print a (hopefully) useful
> > message to prompt users to check their BIOS settings/ACPI settings?
> 
> The message is certainly a good idea.  I'm not sure that bombing out
> at that point is a good idea... because in some cases the problem is
> not correctable (either due to BIOS issues, or hardware design) and
> you would be breaking systems that work fine for 16 bit cards.

Ok, in which case we should at least drop SS_CAP_CARDBUS from the
socket capabilities, since the socket is not cardbus capable if it
doesn't have an IRQ assigned.  We should also test that capability
in cs.c before even powering up a cardbus card.  Do you agree?

-- 
Russell King
 Linux kernel    2.6 ARM Linux   - http://www.arm.linux.org.uk/
 maintainer of:  2.6 PCMCIA      - http://pcmcia.arm.linux.org.uk/
                 2.6 Serial core