Merge pull request #1379 from IMRCLab/feature_print_nan

printf: add support to print NaNs

Author: ataffanel

src/utils/src/eprintf.c

@@ -29,6 +29,7 @@
 #include <stdint.h>
 #include <stdbool.h>
 #include <ctype.h>
+#include <math.h>
 
 static const char digit[] = {'0', '1', '2', '3', '4', '5', '6', '7', '8', '9', 
                              'A', 'B', 'C', 'D', 'E', 'F'};
@@ -275,15 +276,21 @@ int evprintf(putc_t putcf, const char * fmt, va_list ap)
           break;
         case 'f':
           num = va_arg(ap, double);
-          if(num<0)
-          {
-            putcf('-');
-            num = -num;
-            len++;
+          if (isnan(num)) {
+            putcf('n');len++;
+            putcf('a');len++;
+            putcf('n');len++;
+          } else {
+            if(num<0)
+            {
+              putcf('-');
+              num = -num;
+              len++;
+            }
+            len += itoa10(putcf, (int)num, 0);
+            putcf('.'); len++;
+            len += itoa10(putcf, (num - (int)num) * power(10,precision), precision);
           }
-          len += itoa10(putcf, (int)num, 0);
-          putcf('.'); len++;
-          len += itoa10(putcf, (num - (int)num) * power(10,precision), precision);
           break;
         case 's':
           str = va_arg(ap, char* );