Re: Advice on iterator vs. index efficiency

[Tim Prince <tprince at computer dot org>]

Neal Becker wrote:
> Based on advice from some time back, I write algorithms to use iterators
> rather than index-based loops.  Often, this requires jumping through
> considerable hoops.
>
> I'm just wondering, with current gcc loop optimizations, it this advice
> still relevant?  For example:
>
> < iterator style loop >
> for (; i != end (container); ++i)
>   do something with (*i);
>
> < index style loop >
> for (int i = 0; i < size (container); ++i)
>   do something with container[i];
>
> Is there (in general) a difference in speed for these 2 styles?  What about
> if the containers are multi-dimensional and loops are nested?
>
>   
You would have to give a lot more context for anyone to make a 
judgement.  If it were not that the STL uses the first style a lot, and 
has made it mandatory for compilers to find ways to optimize it, the 
second version would be much more conducive to optimizations like 
unrolling and vectorization. I suspect the second style is better for 
nested loops with current technology. gcc hasn't got very far with 
optimization of nested loops, so it's up to you to optimize the nesting 
in source.  It's pretty difficult to optimize nested C++ loops anyway.