Re: RENDER ext. & hardware acceleration, libart integration

Keith Packard <[email protected]> Sat, 24 Aug 2002 15:28:05 -0700

Newsgroups	gmane.comp.xfree86.render
Message-ID	<E17ijO1-0000oQ-00@localhost>

Around 21 o'clock on Aug 24, Juliusz Chroboczek wrote:

> KP> It has been tried with no effect on performance.
> 
> Curious.  Do you understand why?

Yes.  The CPU does no prefetching or pipelining when reading memory mapped 
uncached or write-combining.  Synchronous reads over the AGP bus are 
amazingly slow.

Keith Packard        XFree86 Core Team        HP Cambridge Research Lab