[haiku-appserver] Re: user needs to know about the clipping region?

  • From: Adi Oanca <adioanca@xxxxxxxxx>
  • To: haiku-appserver@xxxxxxxxxxxxx
  • Date: Sat, 26 Nov 2005 23:40:27 +0200

Hi Axel,

        Sorry to respond so late, been quite busy. :(

Axel Dörfler wrote:
Adi Oanca <adioanca@xxxxxxxxxxxxxx> wrote:

I'm actually not sure if that assumption is true - it depends on how you draw, and that can be optimized for this case. Video memory should be mapped in "write-combined" mode which makes sure that the writes leave the cache as 32 byte bursts.

It's not an assumption. Telling the graphics card to take a piece of data from main memory is faster (uses AGP) than writing 32 bytes times and times over.

I'm pretty sure you're wrong here.

I may be, but every time I read about AGP related stuff, the text said something about the card accessing the memory bypassing the CPU or fetching big quantities of data. Also, I don't remember from where I know (or what left me that impression), the graphics card uses the power of AGP when performing actions to the main memory. The other way around, CPU writing in graphics memory, I member it simply uses the PCI bus and sometimes and extension named FastWrites (the one Rudolf enabled with its agp driver).
I may be wrong, but AFAIK the above is right.
Anyway, excuse the text: "It's not an assumption.". :-)

The bus memory bandwidth is precious for graphics, and you can easily fill it up by having a wrong design in the app_server during video playback or something like this, and this will hurt overall performance.

Sure, more documentation is needed. We all aim at improving the performance not make it worse. :-)
I just said an idea based on my knowledges so far.

bye, Adi.

Other related posts: