FLT_EVAL_METHOD
From cppreference.com
                    
                                        
                    
                    
                                                            
                    |   Defined in header  <cfloat>
  | 
||
|   #define FLT_EVAL_METHOD /* implementation defined */  | 
(since C++11) | |
Specifies the precision in which all floating-point arithmetic operations other than assignment and cast are done.
| Value | Explanation | 
  negative values except -1
 | 
implementation-defined behavior | 
  -1
 | 
the default precision is not known | 
  0
 | 
all operations and constants evaluate in the range and precision of the type used. Additionally, float_t and double_t are equivalent to float and double respectively | 
  1
 | 
all operations and constants evaluate in the range and precision of double. Additionally, both float_t and double_t are equivalent to double | 
  2
 | 
all operations and constants evaluate in the range and precision of long double. Additionally, both float_t and double_t are equivalent to long double | 
Notes
Regardless of the value of FLT_EVAL_METHOD, any floating-point expression may be contracted, that is, calculated as if all intermediate results have infinite range and precision (unless #pragma STDC FP_CONTRACT is off)
Cast and assignment strip away any extraneous range and precision: this models the action of storing a value from an extended-precision FPU register into a standard-sized memory location.