EverSpring working shop

# re: The Remark of Usage for ++ and -- 2009-02-21 21:20 everspring79

Another Remark - The evil resides in the combination of below operations:

1) ++/--
2) Macro
3) Function Parms
4) Condition Operation.

Example:
A)
int foo(a, b, c)
{
return (a+b+c);
}

void main()
{
int aa = 5, bb = 4, cc = 3, result;

result = foo(aa++, --bb, cc); // result = (5+3+3), aa = 6 now, bb = 3
result = foo(aa++, ++aa, cc); // result = (7+7+3), aa = 8 now
}

B)
#define MIN(a,b) (((a)<=(b))?(a):(b))

int a = 10, b = 7
d = MIN(++a, b);
//May work well according to gcc.
// d = 7, a still = 10
d = (((a++)<=(b))?(a++):(b)) //as the condition is false, the second a++ will not execut.
// d = 7, a = 11
a = 7;
b = 10;
d = (((a++)<=(b))?(a++):(b)) //as the condition is true, the second a++ is executed, a= 9, but d = 7
d = (((a++)<=(b))?(++a):(b)) // as the condtiion is true, the 2nd ++a is executed, this leads to d to be 9, finally a = 9.
//above is an evil. the first a++ will lead a to be 7+1 before ?.
a = 9; b=10
d= (((a++)<=(b))?(a++):(b)) // antoher evil, the d = 10 because the d is assigned by a = 10 before the 2nd a++. After this a = a+1. so the result is d = 10, a =11.

  回復(fù)  更多評論