Carsten Haitzler (The Rasterman) wrote: > please please please PLEASE! accelerate everythig you can :) > > http://www.rasterman.com/files/render_bench.tar.gz > http://www.rasterman.com/files/imlib2-1.1.0.tar.gz > this might be handy for you as a way of 1. measuring correctness (well so it > "looks right" - imlib2's algorithms are intended to do the right thing and look > right - not follow a strict formula - so ymmv here, but its better than > nothing). and 2. - especially useful, measuring speed vs. what can be done in > software on the client-side.
I already use this for testing as it was the only benchmark for Xrender that I found. The non-scaled onscreen XRender went from 9.7 to 0.7 seconds... :) (imlib is at 2.3 seconds). The scaled versions are not accelerated, and they all are much slower than Xrender... > if xrender can't beat client side rendering with the cpu either 1. the gfx card > simply is too slow and the driver probably should test its own xrender software > routines vs. hardware ones on init and chose the faster for the situation and > also if using software routines try and force source and destination into system > memory, not keep them in video ram. also accounting for coimbinations of ops, > sizes of source and dest etc. Since the hardware I' dealing with supports stretched bitblits, I'll look if that could be made of any use. Thomas -- Thomas Winischhofer Vienna/Austria thomas AT winischhofer DOT net *** http://www.winischhofer.net/ twini AT xfree86 DOT org _______________________________________________ Devel mailing list [EMAIL PROTECTED] http://XFree86.Org/mailman/listinfo/devel
