Re: [OpenBIOS] QEMU PPC support for OpenBIOS

[Greg Watson]

Take a look at the briq code. I hand crafted the device tree, but I  
was able to get to the point of booting Linux from an IDE drive.  
Things failed at that point because I didn't have the interrupts set  
up correctly. There should not be any serious problems getting PPC to  
work.

Greg

On Jun 15, 2006, at 4:08 PM, Wessel, Jason wrote:
> I started porting the OpenBIOS to the PPC qemu, and I was curious  
> of the state of the PearPC portion of the OpenBIOS.  I thought I  
> check and ask if there is another branch or if someone was already  
> working in this area or had it working.
>
> I would be happy to collaborate and share some patches.
>
> Nothing really works at this point, but it is further along then  
> having nothing:
> > > =============================================================
> > > OpenBIOS 1.0RC1 [Jun 15 2006 16:42]
> > > current nvram:
> > > ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
> > > ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
> > > ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
> > > ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
> > > nvram error detected, zapping pram
> > > Initializing PCI devices...
> > > ide0: [io ports 0x1f0-0x1f7,0x3f6]
> > >     drive0 [ATA disk]: QEMU HARDDISK
> > >     drive1 [ATAPI floppy]: QEMU CD-ROM
> > > Initializing ADB driver
> > > Autobooting disabled - dropping into OpenFirmware
>
> No cdrom found
> Output device /pci/pci6666,6666 not found.
> /pci/via-cuda/adb has no read method.
> Output device screen not found.
> Input device keyboard not found.
>
> 0 >
>
> I get that on the serial port for PPC qemu, and now I have to work  
> backwards to fix it up correctly.  Serial input doesn't work quite  
> right and that will probably be the first hurdle.
>
> Any pointers to active or recent projects in this area would be  
> appreciated.
>
> Cheers,
> Jason.

QEMU PPC support for OpenBIOS

[Stefan Reinauer]

* Greg Watson  [060616 05:03]:
> Take a look at the briq code. I hand crafted the device tree, but I  
> was able to get to the point of booting Linux from an IDE drive.  
> Things failed at that point because I didn't have the interrupts set  
> up correctly. There should not be any serious problems getting PPC to  
> work.

Maybe with another target (qemu) we can get a more dynamic creation of
the device tree going.

Stefan