https://devzone.nordicsemi.com/f/nordic-q-a/3456/gcc-emitting-wrong-floating-point-instructionsHi there, I&#39;m currently facing the problem, that an example as simple as volatile float sum; extern &quot;C&quot; int _start() { float a = 1.0; double d = 2.2; sum = a + d; for(;;); } is causing problems. The compiler is invoked with -mcpuen-USTelligent Community 13Wed, 13 Aug 2014 15:01:03 GMThttps://devzone.nordicsemi.com/thread/12508?ContentTypeID=1Wed, 13 Aug 2014 15:01:03 GMT137ad170-7792-4731-bb38-c0d22fbe4515:3565c70d-fad5-4e74-8235-902618cd78e4Torsten Robitzki<p>Hi, thanks for commenting on this. I will experiment a little with command line switches.</p><div style="clear:both;"></div>https://devzone.nordicsemi.com/thread/12507?ContentTypeID=1Tue, 12 Aug 2014 22:57:07 GMT137ad170-7792-4731-bb38-c0d22fbe4515:007c6ea9-838a-4894-8539-788ac10219b8RK<p>Well the code is correct, sort of, and I don&#39;t know whether it&#39;s being generated by gcc or linked in from a library, probably generated. That &#39;bx pc&#39; jumps to the PC+4 and switches to ARM mode, which is why you&#39;re ending up in the hard fault handler, in Thumb mode that&#39;s not allowed. There&#39;s a clue in the function name, __aeabi_f2d_from_thumb, that looks like a thunk to the f2d (I guess that&#39;s float division) library function &#39;from thumb&#39;.</p> <p>So it seems to me the compiler thinks it should be using an ARM FP library which it&#39;s linked in, and has generated a correct call into ARM code for it, but you&#39;re on a Thumb device.</p> <p>Possibilities, there&#39;s another flag on the command line confusing it, perhaps try to compile that line by hand with different flags to see what you get. I see you&#39;re using 4.8.4 of gcc, have you tried 4.8.3, may be a regression.</p><div style="clear:both;"></div>