the attached file implements a simple way to set the AVR Baudrate registers. The method is: Use floating point calculations and let compiler/optimiser remove all of the floatingpoint code. The method iself is applicable for other applications, too. It is also being used in
The question I have: In lines 107-111 an error condition is checked and an error/warning is supposed to be issued, but only if the error condition is really met.
So far, I didn't find any solution for this (tried divide by zero, ...). Any ideas?