Fix float number not working properly for option min/max (#11211)

* ConfigOptionDef: min/max values type are changed from INT to FLOAT.

(cherry picked from commit f277bc80c22e0c9a067481a4301922e2c96aed47)

* Fix infinite loop and crash when `fuzzy_skin_point_distance` = 0 (SoftFever/OrcaSlicer#11069)

* Fix Linux build issue

* Fix float comparison due to precision loss

src/libslic3r/Config.hpp

@@ -4,6 +4,7 @@
 #include <assert.h>
 #include <map>
 #include <climits>
+#include <cfloat>
 #include <cstdio>
 #include <cstdlib>
 #include <functional>
@@ -2455,10 +2456,11 @@ public:
     // Optional width of an input field.
     int                                 width           = -1;
     // <min, max> limit of a numeric input.
-    // If not set, the <min, max> is set to <INT_MIN, INT_MAX>
+    // If not set, the <min, max> is set to <-FLT_MAX, FLT_MAX>
     // By setting min=0, only nonnegative input is allowed.
-    int                                 min = INT_MIN;
-    int                                 max = INT_MAX;
+    float                               min = -FLT_MAX;
+    float                               max =  FLT_MAX;
+    bool                                is_value_valid(double value, int max_precision = 4) const;
     // To check if it's not a typo and a % is missing
     double                              max_literal = 1;
     ConfigOptionMode                    mode = comSimple;