[utils] adding non-branching sign functions

src/float_utils.c

@@ -14,6 +14,10 @@ inline bool float_equals(float a, float b) {
     return float_equals_epsilon(a, b, FLT_EPSILON);
 }
 
+inline float fsign(float x) {
+    return (float)((x > 0.0f) - (x < 0.0f));
+}
+
 inline bool double_equals_epsilon(double a, double b, double epsilon) {
     double diff = fabs(a - b);
     a = fabs(a);
@@ -27,3 +31,8 @@ inline bool double_equals_epsilon(double a, double b, double epsilon) {
 inline bool double_equals(double a, double b) {
     return double_equals_epsilon(a, b, DBL_EPSILON);
 }
+
+
+inline double sign(double x) {
+    return (double)((x > 0.0) - (x < 0.0));
+}