[RFC] setting resources available to pcmcia

Russell King rmk at arm.linux.org.uk
Fri Jun 13 18:07:58 BST 2003


On Fri, Jun 13, 2003 at 05:53:46PM +0200, Dominik Brodowski wrote:
> On Fri, Jun 13, 2003 at 07:12:27AM +0100, David Woodhouse wrote:
> > On Thu, 2003-06-12 at 20:02, Dominik Brodowski wrote:
> > > OK, let's get to more serious stuff now: the way PCMCIA reserves
> > > resources. Here are my suggestions - at first without code, but this
> > > may change soon.
> > > 
> > > _Requirements_
> > > 
> > > 1.) We need to deal with both static and window sockets -
> > 
> > I have a window socket on a PCI bridge which is far from the root and
> > doesn't have an ISA bridge anywhere near it -- hence doesn't see cycles
> > for the whole address space.
> 
> I assume that a PCI-to-PCI bridge needs to be "above" the PCMCIA socket for
> this to become true?
> 
> > It _must_ use memory in the range 0xf5000000-0xf50fffff and I/O in the
> > range 0x2000-0x2fff, or it doesn't even get to see the PCI transactions.
> > Currently this is done by an evil hack -- I configure cardmgr
> > accordingly, and also see the 'membase' and 'isa_setup' module
> > parameters in Marcelo's drivers/pcmcia/i82092.c.
> 
> Can the socket driver know this mem and I/O range?
> 
> I'm brewing up a new plan...

Yes it can, and this case is already covered by my cardbus stuff.  If we
have a PCI to PCMCIA bridge behind a PCI to PCI bridge, or even a host
bridge which only forwards a limited set of resources to the PCMCIA
controller, we must allocate the resource from the bus which the
controller is sitting on.

Unfortunately, GregKH broke all that in his PCI muckup patches, so it
needs to be reworked.

-- 
Russell King (rmk at arm.linux.org.uk)                The developer of ARM Linux
             http://www.arm.linux.org.uk/personal/aboutme.html




More information about the linux-pcmcia mailing list