>Did you use it from the start in your AGP implementation? I didn't
>notice any difference between a couple weeks back and lately.
It was there since I had working AGP. The only difference between
what I had in my tree previously and what I submited here are some
cleanups to remove warnings and n
On Thu, 2001-11-22 at 11:55, [EMAIL PROTECTED] wrote:
> >I suspect it's less stable for us than for the x86 folks though, or
> >there would be more noise about it on the various lists.
> >
> >
> >What's this agp_special_page about?
>
> The agp_special_page is a page of RAM reserved by some code I
On Thu, 2001-11-22 at 12:03, [EMAIL PROTECTED] wrote:
> >> Well, if we are supposed to flush the CPU write buffer, then yes.
> >
> >mb() doesn't do that?
>
> mb() is a lot more expensive, and I'm not completely sure how it
> behaves regarding main memory accessed by UniNorth via AGP. It works
> f
>> Well, if we are supposed to flush the CPU write buffer, then yes.
>
>mb() doesn't do that?
mb() is a lot more expensive, and I'm not completely sure how it
behaves regarding main memory accessed by UniNorth via AGP. It works
for SMP barriers, but knowing how buggy UniNorth can be with AGP,
I p
>I suspect it's less stable for us than for the x86 folks though, or
>there would be more noise about it on the various lists.
>
>
>What's this agp_special_page about?
The agp_special_page is a page of RAM reserved by some code I have
in the kernel's MM init stuff which is as high as possible, th
On Thu, 2001-11-22 at 11:53, [EMAIL PROTECTED] wrote:
> >Is this really useful?
>
> Well, if we are supposed to flush the CPU write buffer, then yes.
mb() doesn't do that?
--
Earthling Michel Dänzer (MrCooper)/ Debian GNU/Linux (powerpc) developer
XFree86 and DRI project member / CS studen
>Is this really useful?
Well, if we are supposed to flush the CPU write buffer, then yes.
Ben.
___
Dri-devel mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/dri-devel
On Sat, 2001-11-10 at 20:58, Benjamin Herrenschmidt wrote:
> Note that it doesn't work correctly with all chipsets yet (the Rage M3
> loves to lockup) with XFree 4.1. I beleive the problem is specific to
> that chip and possibly not related to those AGP fixes as other r128
> cards seem to work pr
>Hey Ben, I don't suppose you would consider helping a Darwin port of DRI
Sorry, I really don't have time for that. Also, my knowledge of
Darwin's mach kernel internals isn't good enough I beleive ;)
Ben.
___
Dri-devel mailing list
[EMAIL PROTECTED]
Hey Ben, I don't suppose you would consider helping a Darwin port of DRI
___
Dri-devel mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/dri-devel
It has been some time...
So, here's is my latest cut at latest AGP support for UniNorth.
It consists of 2 patches:
- uninorth_agp.diff
This one can probably be merged as-is, it adds support for the UniNorth
chipset to the AGP driver backend
- mac_r128_drm.diff
That one cannot be merged as-
11 matches
Mail list logo
