[lug] Various Arch/Compiler Binaries living together
Timothy C. Klein
teece at silverklein.net
Tue Jun 25 14:15:10 MDT 2002
* Nate Duehr (nate at natetech.com) wrote:
> inline below...
>
> > If you really want to speed up some code, buy a faster machine or pick a
> > project that you think is "slow" and work with the developers to
> > implement better algorithms and/or data structures. Yes, thats either
> > expensive or a lot of work. But both of those approaches will actually
> > accomplish something whereas re-compiling stuff with optimization tweaks
> > is (again, for most general codes) an exercise in futility...
>
> Those are interesting approaches. Very few people REALLY understand that's
> the purpose of having the source open and available, though... heh.
>
> Another approach would be to fire up Gentoo Linux, and set it to do the
> optimizations to a "reasonable" level, since you build the entire distro
> from source anyway... (GRIN).
>
Seems the general consensus is that it is not worth the trouble to
compile with arch-specific optimizations. I, however, am a skeptic. I
want to see the numbers myself. So I will try it one of these days.
The thing I like most about Linux is that it gave me my computer back.
If were to gain a 1% increase from a recompile, great. It is *my*
choice. As far as optimizing code, I agree that should be done.
However, I can spend a few months trying to figure out why KDE starts up
so damn slow, or I can spend a few mintues (plus a few hours of compile
time, but I am thinking in terms of *my* time) recompiling it and see
what happens. KDE is the one I was really thinking of, too. I remember
reading somewhere once that someone found significant speed increase
from a specific compile time tweak with KDE. Unfortunately, KDE is
giving me a headache right from the outset. While the difference
between an Athlon and i686 architectures in gcc output is small, I figured
I might as well try to use the Athlon.
The idea that a recompile with opts if futile is not quite right. You
ever look at the bicycle of a real cycle enthusiast? They will spend
hundreds (or thousands) of dollars on super light weight parts. Is the
speed increase big? No. Is is worth it to them? Yes. This is Linux.
It belongs to me, it belongs to you, it belongs to everyone. If it is
worth it to me to milk an extra 1% from existing code, then I can do it.
And it aint futile. I do this stuff for fun.
Tim
--
==============================================
== Timothy Klein || teece at silverklein.net ==
== ---------------------------------------- ==
== "Hello, World" 17 Errors, 31 Warnings... ==
==============================================
More information about the LUG
mailing list