|Re: [eigen] a few things|
[ Thread Index | Date Index | More lists.tuxfamily.org/eigen Archives ]
Hi Christian, Ben, Ramon, thanks a lot for your answers On Thursday 10 January 2008 21:13:13 ollupaC De La Pradera wrote: > However, I'm not sure if "static -const- int" is not enough for the > compiler to optimize it over, have you checked it? Yes, that's what we currently do in Eigen (for instance in alpha2). > (Sorry, my "artistic" name confused you, my real name is Ramon) > Ramon Ah OK, I was not too sure. So your full name is... Ramon De La Pradera ? Please edit the TODO to put your full name; I think that anyone can register at techbase.kde.org, AFAIK you don't need a KDE SVN account. Ben wrote: > I think that the static const ints are unnecessarily stored in data segment > of the compiled program. It would be good to check the size difference of > the executables before you revert back to static const ints. Good idea. I checked the executables created for "benchmark" with g++-4.2 -O3 -DNDEBUG. They have the same size byte for byte. Christian wrote: > A static variable has to be somewhere as other objects might access it. > But it's only created once. It might stop compilers from optimizing > things away (that GCC and ICC do work doesn't say that it works for all > compilers...) Yes, I completely agree. So I conclude that I should go ahead with enums, even if it doesn't bring any improvement for the particular compilers we are using. > We are using C++ here, so please avoid all #define (as far as possible...) I agree. I only used #defines here to make absolutely sure that the absence of performance improvement was not caused by const ints being unoptimized. Now that I see it makes no difference, i'll go back to const ints. OK, thanks a lot for your advice, you all. Cheers, Benoit
Description: This is a digitally signed message part.
|Mail converted by MHonArc 2.6.19+||http://listengine.tuxfamily.org/|