Build with profiling (GCC)
Brought to you by:
marcus256
It should be possible to optimize the speed of the OpenCTM library by using code profiling in GCC (MinGW, Linux, Mac).
Create a simple profiling program, profiler.c, which cycles all compression methods and compression levels etc, using a synthetic 3D mesh. Use this in conjunction with -fprofile-generate / -fprofile-use to make a two-pass compilation.
Two-pass compilation should perhaps be optional?