rmk+pcmcia at arm.linux.org.uk
Fri Mar 26 22:22:41 GMT 2004
On Fri, Mar 26, 2004 at 11:12:18PM +0100, Dominik Brodowski wrote:
> On Fri, Mar 26, 2004 at 08:40:05PM +0000, Russell King wrote:
> > > b) use the iomem and ioport "behind" the parent PCI bridge, if such a bridge
> > > exists,
> > We do this already; we have to.
> Not necessarily:
Sorry, misread. We should not use iomem and ioport ranges outside the
range of the parent bridge. If we did, we may allocate addresses which
are not forwarded to the bus with the cardbus bridge on.
> a) 11-pcmcia allowed me to use ioports well outside the "ioport behind
> bridge" area, without even looking at the rsrc_mgr.c database [see other
11-pcmcia didn't change this behaviour, or it wasn't supposed to.
> b) there's a config bit in the bridge which enables and/or disables the
> forwarding of other ioport/iomem access.
I suspect clearing the NoISA bit will be extremely dodgy. By clearing it,
you forward all ISA accesses within the bridge windows to the subordinate
bus instead of being forwarded to the ISA bridge.
There may be some ports in the ISA bridge which need to be accessed at
that address, or maybe the VGA card (I know some VGA cards have resources
scattered around at, eg, 0x43f8, which VGA drivers and XFree may need
Really, it is only safe to allocate within the first 256 bytes of every
1K of memory on PCI buses in x86 PCs, and for PCMCIA only the lower 1K,
provided you have full usage information for the motherboard IO ports.
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
More information about the linux-pcmcia