Comments on: C++ inline abuse http://www.losingfight.com/blog/2006/08/21/c-inline-abuse/ software engineering on the mac Fri, 05 Dec 2008 09:19:59 +0000 http://wordpress.org/?v=2.5 By: Greg http://www.losingfight.com/blog/2006/08/21/c-inline-abuse/#comment-20100 Greg Wed, 20 Jun 2007 01:30:07 +0000 http://www.losingfight.com/blog/2006/08/21/c-inline-abuse/#comment-20100 Excellent article, Andy, and I am in agreement. I'm in the process of writing code standards for our program, and although I have cited many of the same reasons you have for not inlining, your voice helps lend additional perspective. Thanks. Excellent article, Andy, and I am in agreement.

I’m in the process of writing code standards for our program, and although I have cited many of the same reasons you have for not inlining, your voice helps lend additional perspective. Thanks.

]]> By: Edd http://www.losingfight.com/blog/2006/08/21/c-inline-abuse/#comment-17330 Edd Fri, 18 May 2007 09:05:59 +0000 http://www.losingfight.com/blog/2006/08/21/c-inline-abuse/#comment-17330 There is another benefit of inline. If a function is inline => it is visible to the compiler at the point of calling so besides prologue and epilogue the compiler can extend the optimization inside the function for that inlining. But I agree with you that you should do inline only when after a benchmarking you see that you need to improve the performance. There is another benefit of inline.
If a function is inline => it is visible to the compiler at the point of calling so besides prologue and epilogue the compiler can extend the optimization inside the function for that inlining.
But I agree with you that you should do inline only when after a benchmarking you see that you need to improve the performance.

]]> By: Mike http://www.losingfight.com/blog/2006/08/21/c-inline-abuse/#comment-11817 Mike Thu, 05 Apr 2007 21:19:40 +0000 http://www.losingfight.com/blog/2006/08/21/c-inline-abuse/#comment-11817 Good Article: Another reason for inline... code size optimization?! When using templates, I sometimes inherit a non-templated class from a templated class. The non-templated class does most of the work, while the templated class uses inlined calls to its base class. That way, if the template code is instantiated many times for different types (as with smart pointers), the method code shouldn't be instantiated for each type. Good Article:

Another reason for inline… code size optimization?!

When using templates, I sometimes inherit a non-templated class from a templated class. The non-templated class does most of the work, while the templated class uses inlined calls to its base class.

That way, if the template code is instantiated many times for different types (as with smart pointers), the method code shouldn’t be instantiated for each type.

]]> By: DMINATOR http://www.losingfight.com/blog/2006/08/21/c-inline-abuse/#comment-4178 DMINATOR Fri, 02 Feb 2007 15:28:04 +0000 http://www.losingfight.com/blog/2006/08/21/c-inline-abuse/#comment-4178 Good article and well written. I agree with you about how and when the inlines should be used. I think the biggest limitation for me is that they should be declared in header file, I generally don't like putting the code there as it makes the header file less readable. But I still consider them as an option. Good article and well written. I agree with you about how and when the inlines should be used.

I think the biggest limitation for me is that they should be declared in header file, I generally don’t like putting the code there as it makes the header file less readable.

But I still consider them as an option.

]]>