Re: I'd like to congratulate you and the team that worked on Judy Arrays

[Doug B. <dou...@ya...>]

Gang:

On modern processors, the problem of branch prediction is more
important than subroutine call/returns, so inline code is not where
your focus should be.  Subroutine call/returns are fully predictable
whereas a simple if <condition> statement is NOT.  I have found
that inlining only helps on the most trivial piece of code (on Intel 386 etc)
and the compiler does a good job on doing it for you.  MACRO's 

are a waste of time, IF used for improving performance.  I now use 

them onlyto abbreviate/document trivial things or to make simple changes
to test code in a 1000 places.

doug

 
Doug Baskins <dou...@ya...>



>________________________________
> From: Jim Lloyd <jl...@si...>
>To: aj...@fr... 
>Cc: dou...@ya...; jud...@li...; sta...@gm... 
>Sent: Tuesday, April 3, 2012 4:05 AM
>Subject: Re: I'd like to congratulate you and the team that worked on Judy Arrays
> 
>
>On Mon, Apr 2, 2012 at 11:35 AM, Alan Silverstein <aj...@fr...> wrote:
>
>Amusing, it reminds me that in the last year I carefully marked some new
>>source functions "inline", but later discovered that gcc was basically
>>ignoring my advice and doing whatever the heck it wanted.  Turns out
>>"inline" is merely a suggestion, and I don't know of a way to force it
>>("I know what I'm doing and I want to waste space for performance")
>>other than reverting to macros.  (Maybe there's some gcc option or
>>#pragma I didn't look for...)
>>
>
>See http://gcc.gnu.org/onlinedocs/gcc/Inline.html
>
>
>Here is a macro that does the trick:
>#define force_inline inline __attribute__((always_inline))
>
>
>------------------------------------------------------------------------------
>Better than sec? Nothing is better than sec when it comes to
>monitoring Big Data applications. Try Boundary one-second 
>resolution app monitoring today. Free.
>http://p.sf.net/sfu/Boundary-dev2dev
>_______________________________________________
>Judy-devel mailing list
>Jud...@li...
>https://lists.sourceforge.net/lists/listinfo/judy-devel
>
>
>