Fix for bug #8534 (Compile errors in Precision Math code on Windows)

sql/filesort.cc

@@ -719,7 +719,7 @@ static void make_sortkey(register SORTPARAM *param,
             }
             *to++=1;
           }
-          my_decimal2binary(E_DEC_FATAL_ERROR, dec_val, (byte*)to,
+          my_decimal2binary(E_DEC_FATAL_ERROR, dec_val, (char*)to,
                             item->max_length - (item->decimals ? 1:0),
                             item->decimals);
          break;

sql/my_decimal.cc

@@ -116,7 +116,7 @@ int my_decimal2string(uint mask, const my_decimal *d,
     E_DEC_OVERFLOW
 */
 
-int my_decimal2binary(uint mask, const my_decimal *d, byte *bin, int prec,
+int my_decimal2binary(uint mask, const my_decimal *d, char *bin, int prec,
 		      int scale)
 {
   int err1= E_DEC_OK, err2;

sql/my_decimal.h

@@ -150,12 +150,12 @@ void my_decimal2decimal(const my_decimal *from, my_decimal *to)
 }
 
 
-int my_decimal2binary(uint mask, const my_decimal *d, byte *bin, int prec,
+int my_decimal2binary(uint mask, const my_decimal *d, char *bin, int prec,
 		      int scale);
 
 
 inline
-int binary2my_decimal(uint mask, const byte *bin, my_decimal *d, int prec,
+int binary2my_decimal(uint mask, const char *bin, my_decimal *d, int prec,
 		      int scale)
 {
   return check_result(mask, bin2decimal((char *)bin, (decimal *)d, prec,