difference bet. macro and inline

[sachin_mzn]

Hi,

It may be a silly question but I want to know
the difference between #define macro and inline functions
Is there any performance issue related to it.

-Sachin

 

[Gernot Frisch]

Hi,

It may be a silly question but I want to know
the difference between #define macro and inline functions
Is there any performance issue related to it.

-Sachin

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

inline template<class T> T max(const T& t1, const T& t2)
{
return t1>t2 ? t1 : t2;
}

will produce the same results on good compilers.


--
-Gernot
int main(int argc, char** argv) {printf
("%silto%c%cf%cgl%ssic%ccom%c", "ma", 58, 'g', 64, "ba", 46, 10);}

________________________________________
Looking for a good game? Do it yourself!
GLBasic - you can do
www.GLBasic.com

 

[Mole Wang]

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

inline template<class T> T max(const T& t1, const T& t2)
{
return t1>t2 ? t1 : t2;
}

will produce the same results on good compilers.

Really? Consider the following code:
int i = 3, j = 2;
int res = max(++i, j);
The inline function and macro result in different return value "res".

 

[BB]

Hi,

It may be a silly question but I want to know
the difference between #define macro and inline functions
Is there any performance issue related to it.

-Sachin

This is addressed in the FAQ.

 

[Ron Natalie]

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

inline template<class T> T max(const T& t1, const T& t2)
{
return t1>t2 ? t1 : t2;
}

will produce the same results on good compilers.

No it will NOT.

max(++a, ++b)

will yield different results.

 

[Niels Dekker - no reply address]

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

Two more parentheses, please:
#define max(a,b) ((a)>(b) ? (a) : (b))

(Try calling max(1, 1 & 2) to spot the difference!)

inline template<class T> T max(const T& t1, const T& t2)
{
return t1>t2 ? t1 : t2;
}

will produce the same results on good compilers.

Ron Natalie replied:

No it will NOT.

max(++a, ++b)

will yield different results.

max(a, b) = 0;

will yield different results as well. Apparently, the macro approach is
the better one, in this case!

Scott Meyers wrote about the implementation of max(a, b), in his article
"min, max, and more": <quote> I increasingly find myself telling people
that the macro approach may well be best, and I hate macros. </quote>

Source: http://www.aristeia.com/Papers/C++ReportColumns/jan95.pdf


Kind regards,

Niels Dekker
http://www.xs4all.nl/~nd/dekkerware

 

[Ron Natalie]

max(a, b) = 0;

add:
inline template<class T> T& max(T& t1, T& t2)
{
return t1 > t2 ? t1 : t2;
}

 

[Niels Dekker - no reply address]

inline template<class T> T max(const T& t1, const T& t2)
{
return t1>t2 ? t1 : t2;
}

add:
inline template<class T> T& max(T& t1, T& t2)
{
return t1 > t2 ? t1 : t2;
}

The Scott Meyers article "min, max, and more" (1995) said that this
would still lead to troubles when mixing const and non-const arguments:

void g(const BigNumber& n1)
{
BigNumber n2 = 22;
BigNumber n3 = max(n1, n2); // call which max?...
}

(From http://www.aristeia.com/Papers/C++ReportColumns/jan95.pdf again)
But all of the compilers I just tried accept the code. So may I assume
that this issue has been solved by a revision of the C++ language?


Kind regards,

Niels Dekker
http://www.xs4all.nl/~nd/dekkerware