• Well, I am not against hardfp. If it is faster or otherwise better then so be it. I was thinking that since JavaScript uses doubles, it is working with software floats anyway. Not sure how the calling convention works for double precision numbers with hardfp and single precision fpu, are they put into two vfp registers before calling and unloaded back in software emulation? Need to try (EDIT: here https://gist.github.com/fanoush/0da3e47a­ee9e20fb11b010cc3aa4e16e - updated my older test, now using -mcpu=cortex-m4 -mfpu=fpv4-sp-d16). Also I was thinking that cortex m4 has some lazy optimization when fpu registers are not used between interrupts they are not saved on stack so it is faster. this may use them just to pass it in arguments between functions without actually using them to do the math.


Avatar for fanoush @fanoush started