• AVR Freaks

Problem enabling –fast-math compiler option

Author
Alex Z
New Users
  • Total Posts : 37
  • Reward points : 0
  • Joined: 2016/03/01 23:27:22
  • Location: New Jersey, USA
  • Status: offline
2016/03/20 18:03:58 (permalink)
0

Problem enabling –fast-math compiler option

I have a MPLab X 3.26 project that works with floating point numbers - using XC16 1.26 everything compiles and works fine; however after I enabled Fast Floating point math in the "XC16 (Global Options)" configuration page (generated command line: -fast-math), building project results in multiple linker errors like the following:

In function `finitenonzero':
: undefined reference to `___divsf3'
In function `finitenonzero':
: undefined reference to `_atanf'
In function `finitenonzero':
: undefined reference to `___subsf3'

 
Please advise what else need to be done to enable this option?
#1

12 Replies Related Threads

    Jump to:
    © 2019 APG vNext Commercial Version 4.5