#define SUM(x,y) ((x)+(y))
void print(){
for(int i=0;i<10;i++)
printf("sum=%d\n", SUM(i,i+1));
}
Is there any benefit in using the macro SUM like above? I read that macros are useful when there is a loop with a function call inside.
-
1No, it's just obfuscation.Fred Foo– Fred Foo2014-04-24 14:08:30 +00:00Commented Apr 24, 2014 at 14:08
-
2"I read" Where? Citation please.n. m. could be an AI– n. m. could be an AI2014-04-24 14:14:03 +00:00Commented Apr 24, 2014 at 14:14
-
a benefit of macros can be not having to duplicate code just to deal with different typesbph– bph2014-04-24 14:15:23 +00:00Commented Apr 24, 2014 at 14:15
Add a comment
|
1 Answer
For very simple computation, call function may have more overhead than the actual computation itself; in a loop, the situation even worse.
However, you need to define your function replacement macro carefully to only evaluate its arguments once. For example, if you have a macro like this
#define DOUBLE(i) ((i) + (i))
and you call it like this DOUBLE(i++), it will be expanded to (i++)+(i++), and this will cause undefined behavior in C. That is why inline function, which will evaluate its arguments only once, is preferable than macro.
5 Comments
Danvil
You example would also be undefined behavior if
SUM would be a function.
Potatoswatter
Would be better to mention inline functions in the first paragraph and mention that they avoid call overhead.
Lee Duhem
@Potatoswatter My intention is to give a direct answer in the first paragraph to the question in the original post. And point out possible pitfalls, alternatives and other additional information in the following paragraph.
Potatoswatter
It's your prerogative. Personally, if an alternative is superior almost all the time due to such a serious pitfall of question topic, I'll give it top billing.
Lee Duhem
@Potatoswatter Good point. We are here not only to answer questions but also, more importantly, to share our experience. I will adopt this organization of answer when I try to answer a question next time. Thank you.