Re:pci_fixup_video change blows up on sparc64

[<eiichiro.oiwa.nm@hitachi.com>]

>
>This change in 2.6.19-GIT:
>
>commit b5e4efe7e061ff52ac97b9fa45acca529d8daeea
>Author: eiichiro.oiwa.nm@hitachi.com <eiichiro.oiwa.nm@hitachi.com>
>Date:   Thu Sep 28 13:55:47 2006 +0900
>
>    PCI: Turn pci_fixup_video into generic for embedded VGA
>
>breaks sparc64 with ATI Radeon and ATY128 cards.
>
>The problem is that there is no system rom at 0xc0000 on sparc64, and
>therefore nothing copies the VGA bios of the graphics card there on
>bootup.  Therefore all of this code is bogus and will just result in
>bus errors when the Radeon or ATY128 driver tries to pci_map_rom() and
>read the graphics card ROM.  Nothing will respond to accesses at the
>0xc0000 region on sparc64.
>
>The existence of a primary video ROM at 0xc0000 is quite platform
>specific.  If some non-x86 systems have this too, that's great.
>However, assuming all systems do is not correct.
>

Does ATI Radeon card have an expansion ROM (video ROM)?
Could you show me "lspci -vv" on sparc64?

If an expansion ROM exists on ATI Radeon or ATY128 card, pci_map_rom returns
the expansion ROM base address instead of 0xC0000 because fixup_video checks
the VGA Enable bit in the Bridge Control register.
The Bridge Control register describes in "PCI-to-PCI Bridge Architecture
Specification Revision 1.2".
This specification is the standard specification in PCI.