CODEWITHSUNDEEP
c++functioninline
Inquirer
Inquirer

Reputation: 7

Inline Function more or less safe

Is "inline" safer, or less safe than using a pre-compiler definition for "max" (see below)?

#define max(a,b)    (a > b) ? a : b

Given this function definition:

inline int max (int a, int b)
{
  return (a > b) ? a : b;
}

Upvotes: 0

Views: 166

Answers (3)

Mark B
Mark B

Reputation: 96281

The macro version will evaluate its arguments multiple times which will result in semantic difference (error) if the argument has side effects (like calling a non-const function).

Further the arguments aren't each enclosed in parens so there may be other ways you can concoct parameters such that precedence completely changes the meaning of what's being attempted. For example max(c, x < y) would fold out as (c > x < y) ? c : x < y which will evaluate c > x and then compare the truth of that against y. That's completely not what was intended.

The inline function on the other hand works only with integral types without truncations.

A better approach is to not reinvent the wheel and use std::max which takes care of all the problems above and is handily provided by your implementation.

Upvotes: 4

Vlad from Moscow
Vlad from Moscow

Reputation: 311048

The macro has many negative side-effects. For example consider

max( ++a, b );

Upvotes: 1

Kerrek SB
Kerrek SB

Reputation: 477348

There's a semantic difference:

max(pay_wife(1000), get_salary());

Upvotes: 8

Related Questions