[PyOpenGL-Devel] Error checking induce performance problems
Brought to you by:
mcfletch
From: <ir...@gm...> - 2007-06-20 23:23:35
|
Hi: Profiling a pyopengl application I have been developing I obtained this results: ncalls tottime percall cumtime percall filename:lineno(function) 213189 11.436 0.000 26.370 0.000 globjects.py:27(render) 2145235 6.653 0.000 12.704 0.000 error.py:162(glCheckError) 2144179 6.050 0.000 6.050 0.000 win32.py:72(safeGetError) Where render is the function that does the main work of the app. As you can see, error checking takes almost one half of the real work time. I think it could be a good idea to be able to disable error checking once you know your code works ok. After Reading http://pyopengl.sourceforge.net/ctypes/using.html, I tried this: from OpenGL import error def foobar(): pass error.ErrorChecker.registerChecker(foobar) But there is no gain in performance. Regards, Jorge |