F11 Preview problems

[Adam Williamson]

On Mon, 2009-05-04 at 14:00 -0400, Adam Jackson wrote:
> On Thu, 2009-04-30 at 15:14 -0700, Adam Williamson wrote:
> 
> > There's also 0x10DE 0x0008, 0x0009 and 0x0010. From what I can find, the
> > 0x0008 and 0x0009 were the original NV1 chips - Diamond Edge 3D, as they
> > were released. lshw claims the 0x0010 is the NV2, which was never
> > released to the public.
> 
> nv1 hasn't had a native driver since XFree86 3.3.6.  Also I think I own
> two-thirds of the world's collection of them by this point.  (The third
> card being in a glass case at the NVIDIA office in Santa Clara.)

There's a picture of one on Wikipedia uploaded by some dude last year.
So there's at least one you missed. :) I also found a reference to one
being used in a system by a guy in Malaysia in 2005.

(Why yes, I am a little unhealthily obsessive about graphics card
identification, thank you for asking :>)

> > The oddball 0x10b4 0x1b1d is, according to old kudzu and ldetect-lst,
> > the STB Velocity 128 3D, which Google (and my own slightly unreliable
> > memory) agree is a Riva 128-based card. It seems to be the only time
> > this PCI vendor ID was ever used, so we can just send that vendor ID
> > straight to nv.
> 
> Technically the nv driver doesn't recognize 0x10b4 at all, so I'm happy
> to let it continue to fall back to vesa.

OK. It probably *should*, but of course it's hard for anyone outside
NVIDIA to work on nv :\

> Anyway.  nouveau patch in the server changed to special-case nv1 and nv3
> onto vesa and nv, respectively.  Thanks!

FWIW, I looked into this some more after sending my mail, and it gets
even murkier. Everyone seems to have copied their data off each other,
but if you try and find references to actual hardware, it gets
difficult. Most sources seem to act as if the 0x10DE and 0x12D2 vendor
IDs are sort of concurrent - as if NV1 and NV3 chips existed with the
same device IDs (0x0008, 0x0009, 0x0018, 0x0019) but with both 0x10DE
and 0x12D2 vendor IDs. I suspect this isn't actually true, and really
only the 0x12D2 vendor ID was used for NV1 and NV3, and then handed off
to 0x10DE for NV4 and higher. So I suspect 0x10DE 0x0008, 0x10DE 0x0009,
0x10DE 0x0018 and 0x10DE 0x0019 never existed (and neither does 0x12D2
0x0020, for e.g.) But I suspect it'd be hard to prove that with
reference to real hardware at this point, and there's probably no
drawback to acting as if some cards exist that might not have.

(what are the actual PCI IDs of your nv1 cards, out of interest? maybe
they can deny my theory...)
-- 
Adam Williamson
Fedora QA Community Monkey
IRC: adamw | Fedora Talk: adamwill AT fedoraproject DOT org
http://www.happyassassin.net