I'm compiling for a very small embedded platform where every kilobyte matters. Sometimes the compilation fails because a certain function is inlined excessively causing code bloat.
Is there a way to get a mapping between function name and the number of times it was inlined so that I can tell gcc explicitly not to inline a function where the size vs speed trade-off is not worth it? I otherwise want to optimise for speed, so just compiling with -Os is not an option.