Radeon 9000 Pro corruption problem on PPC when DRI is enabled

[Dave Airlie]

> On 8/23/06, Roland Scheidegger <rscheidegger_lists at hispeed.ch> wrote:
> > Ari Entlich wrote:
> > >>>> I could /try/ to look at the code, but I know nothing about Radeon
> > >>>> hardware and my C skills aren't quite professional-grade just yet, so
> > >>>>  I can't promise anything :-).
> > >>> The attached patch might fix things. Completely untested, I think it
> > >>> should apply to pretty much any version of radeon_bios.c you can find...
> > >>>
> > >> Yeah, ok, I'll try that. God, I love Gentoo :-). Is that the test
> > >> radeonfb uses?
> > >>
> > >
> > > Awesome!!! This fixed both the DRI corruption and UseFBDev problems!
> > > Well, that patch as you sent it didn't fix it (keep reading), but the
> > > problem has definitely been identified! The check you used to test for
> > > an x86 BIOS doesn't work
> > The test is bogus (the 0x14 value is relative to the adress at 0x18, not
> > relative to the vbios start). So it wouldn't work.
> > That said, I don't understand why it would get identified as a x86 bios
> > if it isn't, as xf86ReadPciBIOS checks that itself, so in fact there
> > shouldn't be any need to put in the same verification again in the
> > radeon driver.
> > There is some debug code in xserver/xorg/hw/xfree86/os-support/bus/Pci.c
> > (PRINT_PCI) maybe this would provide some insight what's going on. There
> > seems to be absolutely no code difference between 6.8.2 and 7.1 there,
> > just as there doesn't seem to be a difference in the radeon driver
> > itself wrt to bios reading...
>
> perhaps this is related to airlied's change to use sysfs access to the
> rom in 7.1 on linux.
>

That it is, my code just left everything up to the kernel, which must
return OF ROMS as well as PCI ones, so feel free to add code to the X
server to do the checks as opposed to the drivers..

Dave.