Write Combining for PowerPC?

Write Combining for PowerPC?

[Kendall Bennett]

Hi Guys,

We are working on some PowerPC machines and noticed that the boxes don't 
appear to support the equivalent of Write Combining that we get on x86 
boxes. Copies to Video Memory on our Motorola Sandpoint box run about 
10Mb/s, which is terribly, terribly slow!

Does anyone know if it is possible to do something similar to Write 
Combining for the PowerPC architecture, to speed up CPU access to the 
linear framebuffer? Part of the problem is that for video overlay support 
(not motion compensation) you have to dump the entire YUV frame into 
video memory for the hardware overlay, and even on a 1GHz PPC box playing 
an MPEG2 stream is not possible as X takes up over 80% of the CPU just to 
copy the YUV data to video memory!

Obviously bus mastering will help solve this problem, but it would be 
better if there was a way to enabling faster CPU access to the 
framebuffer as well.

Regards,

---
Kendall Bennett
Chief Executive Officer
SciTech Software, Inc.
Phone: (530) 894 8400
http://www.scitechsoft.com

~ SciTech SNAP - The future of device driver technology! ~

Re: Write Combining for PowerPC?

[Antonio Vargas]

On Fri, 10 Dec 2004 13:23:06 -0800, Kendall Bennett
<KendallB@scitechsoft.com> wrote:
> Hi Guys,
> 
> We are working on some PowerPC machines and noticed that the boxes don't
> appear to support the equivalent of Write Combining that we get on x86
> boxes. Copies to Video Memory on our Motorola Sandpoint box run about
> 10Mb/s, which is terribly, terribly slow!
> 
> Does anyone know if it is possible to do something similar to Write
> Combining for the PowerPC architecture, to speed up CPU access to the
> linear framebuffer? Part of the problem is that for video overlay support

hmmm... G4 processors will do write-combining when you write two
cachelines just one after the other... there is a write-queue on the
cpu and when there is 1 write on the queue and you add another one for
the next cacheline, it will automatically merge the write. Also, if
you manage to complete a sector, then he knows you don't need to
read-then-write the memory and just issues a "kill-this-cacheline" on
the bus instead of "please-i-need-exclusive-access-for-this-cacheline"

> (not motion compensation) you have to dump the entire YUV frame into
> video memory for the hardware overlay, and even on a 1GHz PPC box playing
> an MPEG2 stream is not possible as X takes up over 80% of the CPU just to
> copy the YUV data to video memory!
> 
> Obviously bus mastering will help solve this problem, but it would be
> better if there was a way to enabling faster CPU access to the
> framebuffer as well.

bus mastering is DMA-ing from system memory to gfxcard memory?
 
> Regards,
> 
> ---
> Kendall Bennett
> Chief Executive Officer
> SciTech Software, Inc.
> Phone: (530) 894 8400
> http://www.scitechsoft.com
> 
> ~ SciTech SNAP - The future of device driver technology! ~
> 
> _______________________________________________
> Linuxppc-dev mailing list
> Linuxppc-dev@ozlabs.org
> https://ozlabs.org/mailman/listinfo/linuxppc-dev
> 


-- 
Greetz, Antonio Vargas aka winden of network

http://wind.codepixel.com/

Las cosas no son lo que parecen, excepto cuando parecen lo que si son.