ColorTiling breaks output on G5

Benjamin Herrenschmidt benh at kernel.crashing.org
Tue Feb 8 20:43:54 PST 2005


On Tue, 2005-02-08 at 19:20 +0100, Roland Scheidegger wrote:

> Hmm ok. Color tiling just is broken on r300 then. Well it wouldn't 
> really do anything useful currently anyway, so this patch should fix 
> that (just always disables color tiling on new chips). I'll check that 
> in if it works.
> Someone with docs (or too much time for reverse-engeneering...) would 
> need to fix that later if the experimental r300 dri driver is going to 
> support tiling for a performance boost.
> 
> Roland
> plain text document attachment (r300_notiling.diff)
> Index: radeon_driver.c
> ===================================================================
> RCS file: /cvs/xorg/xc/programs/Xserver/hw/xfree86/drivers/ati/radeon_driver.c,v
> retrieving revision 1.40
> diff -u -r1.40 radeon_driver.c
> --- radeon_driver.c	2 Feb 2005 21:35:30 -0000	1.40
> +++ radeon_driver.c	8 Feb 2005 18:14:53 -0000
> @@ -4572,7 +4572,16 @@
>  		       "Couldn't load fbdevhw module, not using framebuffer device\n");
>  	}
>      }
> -    
> +
> +    if (!info->FBDev)
> +	if (!RADEONPreInitInt10(pScrn, &pInt10))
> +	    goto fail;
> +
> +    RADEONPostInt10Check(pScrn, int10_save);
> +
> +    if (!RADEONPreInitConfig(pScrn))
> +	goto fail;
> +
>  

Why are you moving that around ? I don't like when a patch is doing more
than advertised :) I also think I had a reason to move that down, though
I'm not 100% what was the reason back then. I'll have a look tomorrow.

Ben.





More information about the xorg mailing list