[Gegl-developer] Is there a reason why -ffast-math is not included in the CFLAGS?

[Nicolas Robidoux]

At least on my machine (recent Intel Ubuntu laptop), it makes a small,
but noticeable, difference.

Why not include it? I doubt very much that any of us is doing
something which depends too closely on exactly how floating point is
done. I certainly trust that the samplers will not be affected (maybe
nearest neighbour, but nitpicking about a discontinuous method...).

In the gegl/gegl/buffer Makefile, for example:

CFLAGS = -g -DGEGL_ENABLE_DEBUG -g -O2 -Wall -Wdeclaration-after-statement -Wmissing-prototypes -Wmissing-declarations -Winit-self -Wpointer-arith -Wold-style-definition -mmmx -msse

Nicolas Robidoux
Laurentian University/Universite Laurentienne