On 9 June 2013 12:55, Martin Renold <[email protected]> wrote:

> hi Jon
>
> I measured the performance on my PC and attempted to visualize it:
> http://maxy.homeip.net/misc/openmp_perf/figure_1.png
>
> Summary: visible improvement of "brushengine_paint_hires" benchmark,
> probably a degradation of the standard "paint" benchmark (could be noise).
>

Hi Martin,
thanks for testing!
I see that the paint and paint-rotated benchmarks have had a regression
since 1.1 (due to GTK+3?). I pushed a commit which should help a bit, by
caching the transformation matrix:
http://gitorious.org/mypaint/mypaint/commit/4b0cf83174cfe0ef3c796b497df913a8e06560da

I've pushed a commit making OpenMP opt-out, to give it some more testing.
http://gitorious.org/mypaint/mypaint/commit/321c1f333a28be6cdc2ce205233e528974cd916d

I have also worked on multi-threaded compositing of the layer stack[1], but
we are currently at a point
where most things are limited by how fast we can fetch tiles from the tile
store, ie: _get_tile_numpy() in lib/tiledsurface.py

Time spent in (and below) _get_tile_numpy() on my system:
layerpaint_zoomed_out_5x: 48%
brushengine_paint_hires: 35%
paint_zoomed_out_5x: 52%

I am working on fixing this now, though it may take a while to complete.

1. http://gitorious.org/~jonnor/mypaint/jonnors-clone/commits/compositing-mt

-- 
Jon Nordby - www.jonnor.com
_______________________________________________
Mypaint-discuss mailing list
[email protected]
https://mail.gna.org/listinfo/mypaint-discuss

Reply via email to