[OpenBIOS] [Qemu-devel] Re: [PATCHv8 00/16] boot order specification

Benjamin Herrenschmidt benh at kernel.crashing.org
Tue Dec 14 23:02:20 CET 2010


On Tue, 2010-12-14 at 22:34 +0100, Andreas Färber wrote:
> > That looks like the reg property for ide at 3 - it's a pci child, so  
> > the addressing is in the pci bus. It looks like it has five BARs
> in  
> > IO space, each between 4 and 10 bytes long.
> 
> It was an excerpt from `dev /pci/pci-ata .properties` on
> `qemu-system- 
> ppc -nographic -prom-env 'auto-boot?=false'`. 

Ah ok, that's fine then. I though you were looking at the children of
pci-ata.

On a real B&W G3 it looks like this:

compatible       "pci1095,646"
		 "pci1095,646"
		 "pciclass,01018f"
assigned-addresses 81010810 00000000 00001440 00000000 00000010
		 81010814 00000000 00001430 00000000 00000010
		 81010818 00000000 00001420 00000000 00000010
		 8101081c 00000000 00001410 00000000 00000010
		 81010820 00000000 00001400 00000000 00000010
#address-cells   00000001
fast-back-to-back
#size-cells      00000000
subsystem-id     00000646 (1606)
vendor-id        00001095 (4245)
device_type      "pci-ide"
min-grant        00000002
revision-id      00000005
subsystem-vendor-id 00001095 (4245)
name             "pci-ata"
max-latency      00000004
devsel-speed     00000001
device-id        00000646 (1606)
class-code       0001018f (65935)
reg              00010800 00000000 00000000 00000000 00000000 01010810
		 00000000 00000000 00000000 00000010 01010814 00000000
		 00000000 00000000 00000010 01010818 00000000 00000000
		 00000000 00000010 0101081c 00000000 00000000 00000000
		 00000010 01010820 00000000 00000000 00000000 00000010
interrupts       00000001
linux,phandle    ff873960

ata-4 at 0:
compatible       "cmd646-ata"
#address-cells   00000001
#size-cells      00000000
device_type      "ata"
name             "ata-4"
reg              00000000
linux,phandle    ff874828


Cheers,
Ben.




More information about the OpenBIOS mailing list