On Sat, 3 Aug 2002, Jake Page wrote:
>
> Looking into it a bit more, it looks like DRI is only supported well for
> AGP cards. There is a "pcigart" implementation for the Radeon DRI kernel
> module that is disabled by default (actually enabled for the alpha and
> ppc, but not x86) - going to try rebuilding it with it enabled to see what
> happens... (for anyone interested, look for PCIGART_ENABLE in radeon_cp.c
> AND radeon_dri.c)
>
> Anyway, I wouldn't be surprised if an AGP Radeon works with DRI & HW
> accelerated OpenGL with little or no changes (unfortunately I don't have a
> %$@! AGP slot on this board...)
Bit of a follow up on this -
I built a DRI server + kernel module from the latest DRI-CVS tree. Now I
get a couple different errors...
This one is the same as before, but now it looks like it may be an
issue after all...
(II) Loading /usr/X11R6/lib/modules/linux/libint10.a
(II) Module int10: vendor="The XFree86 Project"
compiled for 4.2.0, module version = 1.0.0
ABI class: XFree86 Video Driver, version 0.5
(II) RADEON(0): initializing int10
(EE) RADEON(0): No V_BIOS found
(--) RADEON(0): Chipset: "ATI Radeon QD (AGP)" (ChipID = 0x5144)
(--) RADEON(0): Linear framebuffer at 0x98000000
(--) RADEON(0): MMIO registers at 0x88080000
(--) RADEON(0): VideoRAM: 32768 kByte (64-bit SDR SDRAM)
(II) Truncating PCI BIOS Length to 45056
This one shows that DRI is still being disabled...
drmOpenByBusid: busid is PCI:0:9:0
drmOpenDevice: minor is 0
drmOpenDevice: node name is /dev/dri/card0
drmOpenDevice: open result is 6, (OK)
drmOpenByBusid: drmOpenMinor returns 6
drmOpenByBusid: drmGetBusid reports PCI:0:9:0
(II) RADEON(0): [drm] drmSetBusid failed (6, PCI:0:9:0), Device or
resource busy
(EE) RADEON(0): [dri] DRIScreenInit failed. Disabling DRI.
Best hint on this I found searching the dri-devel mailing list...
=====
> - about 2D (plain X): will Radeon PCI (the 2001 T&L model, not the 7xxx)
> Xserver run on ppc?
Basically yes, but I think the drivers still rely on the firmware setting
up the card sanely, which would mean getting a Mac card in this case.
=====
ie. it looks like the card's BIOS may be necessary to get it working...
(unless/until someone figures out what it's doing and puts it into the
driver, at least :) Heh, it feels so CLOSE I'm tempted to go pick up a
Mac card to see if it works ;)
-Jake
