On Sunday 09 December 2001 10:28 pm, you wrote:
> And this is true about the kernel-level API as well. I have just figured
> out that agp_memory_t.memory field is an array of addresses to allocated
> pages - but I have not figured out whether these are physical or not.
The memory[] table currentl
On Sun, 9 Dec 2001, Philip Brown wrote:
> Well, unfortunately, the driver isnt THAT far along :-)
> I'm also using that test program you gave me, as a bit more documentation
> as to how agpgart is used. I havent come across a document that describes,
> "This is how you use /dev/agpgart at the use
On Sun, 9 Dec 2001, Philip Brown wrote:
> On Sun, Dec 09, 2001 at 11:29:53PM +0100, Dieter Nützel wrote:
> > ...
> > Have you checked that your AGP module is correctly loaded ?
> > If so, I suggest that you compile and run (as root, and before running X) the
> > C program located at :
> > http:
On Sun, Dec 09, 2001 at 10:00:07PM -0700, Derrik Pates wrote:
> ...
> Well, I don't think that anyone developing AGP GART support thought a lot
> about that - usually, apps aren't going to be talking directly to the AGP
> GART, but via some sort of abstraction layer, like DRI. Which, from what I
>
On Sun, Dec 09, 2001 at 11:29:53PM +0100, Dieter Nützel wrote:
> ...
> Have you checked that your AGP module is correctly loaded ?
> If so, I suggest that you compile and run (as root, and before running X) the
> C program located at :
> http://ltswww.epfl.ch/~aspert/patches/testgart.c
>
> Send t
Am Sonntag, 9. Dezember 2001 22:01 schrieben Sie:
> Sorry if this is a duplicate: It's been 8 hours or more, and I havent seen
> a copy of my email to the list.
>
> I'm looking to write a driver that implements /dev/agpgart, as a first step
> towards doing a full DRI port.
>
> Trouble is, I cant s
