engine, public: fix float precision issues in mathlib and monster navigation code

engine/server/sv_move.c

@@ -243,7 +243,7 @@ float SV_VecToYaw( const vec3_t src )
 	}
 	else
 	{
-		yaw = (int)( atan2( src[1], src[0] ) * 180.0f / M_PI_F );
+		yaw = (int)( atan2( src[1], src[0] ) * 180.0 / M_PI );
 		if( yaw < 0 ) yaw += 360.0f;
 	}
 	return yaw;

public/xash3d_mathlib.h

@@ -34,18 +34,18 @@ GNU General Public License for more details.
 #define ROLL		2
 
 #ifndef M_PI
-#define M_PI		(float)3.14159265358979323846
+#define M_PI		(double)3.14159265358979323846
 #endif
 
 #ifndef M_PI2
-#define M_PI2		((float)(M_PI * 2))
+#define M_PI2		((double)(M_PI * 2))
 #endif
 
 #define M_PI_F		((float)(M_PI))
 #define M_PI2_F		((float)(M_PI2))
 
-#define RAD2DEG( x )	((float)(x) * (float)(180.f / M_PI_F))
-#define DEG2RAD( x )	((float)(x) * (float)(M_PI_F / 180.f))
+#define RAD2DEG( x )	((double)(x) * (double)(180.0 / M_PI))
+#define DEG2RAD( x )	((double)(x) * (double)(M_PI / 180.0))
 
 #define NUMVERTEXNORMALS	162