The hypot function is very useful since the naive sqrt(x*x + y*y) has overflow and hence mathematical accuracy issues, see e.g http://www.johndcook.com/blog/2010/06/02/whats-so-hard-about-finding-a-hypotenuse/.
1/2 adds configure+libm support. The fallback is currently a naive one, and may be improved in a separate patch. 2/2 gives a simple illustration in avutil/eval. It may be useful elsewhere as well. Ganesh Ajjanagadde (2): configure+libm.h: add hypot emulation avutil/eval: change sqrt to hypot configure | 2 ++ libavutil/eval.c | 2 +- libavutil/libm.h | 8 ++++++++ 3 files changed, 11 insertions(+), 1 deletion(-) -- 2.6.2 _______________________________________________ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
