Optimized distro for i686

[Christopher Browne]

On 12/30/05, Steve <bassix-Re5JQEeQqe8AvxtiuMwx3w at public.gmane.org> wrote:
> On 12/30/05, paul sutton <zen14920-1HOZaDBbGgxaa/9Udqfwiw at public.gmane.org> wrote:
> > You can recompile .src.rpm packatges for a specific processor,  I would
> > guess that to optimise FC4 for i686 you could perhaps recompile all the
> > packages,  and burn these on to a cdrom along with modified installation
> > routines,  which would give you a new cd that is optimised for that
> > specific processor.
> >
> > Ok there is more to it than that, but I have managed to recompile one or
> > two packages for optimisation (from i386 to i586 I think) and then
> > instaled these with no problem.
> >
> > it can take a while to do this to lots of packages,  I would be a good
> > idea to perhaps do this at a time when you can leave the computer on for
> > a few hours I would guess you need ALL packages installed so you can
> > avoid problems with dependancies, while compiling.
> >
> > any comments
> >
> > Paul
>
> Is this pretty much what Gentoo does? So probably the Fox distro
> people have done these recompiles and therefore will save people the
> time required to do it themselves.

This is pretty much what Gentoo does.

The trouble is that sometimes bugs emerge when you use unusual
compiler options; figuring out what flags are safe is something of a
package-by-package thing.  Universally applying all the same compiler
options everywhere doesn't work.

It's not at all obvious that the result will actually be meaningfully
faster.  For instance, database apps tend to have disk I/O as their
bottleneck, and changing compiler options can't change that.
--
http://www3.sympatico.ca/cbbrowne/linux.html
"The true  measure of a  man is how he treats  someone who can  do him
absolutely no good." -- Samuel Johnson, lexicographer (1709-1784)
--
The Toronto Linux Users Group.      Meetings: http://tlug.ss.org
TLUG requests: Linux topics, No HTML, wrap text below 80 columns
How to UNSUBSCRIBE: http://tlug.ss.org/subscribe.shtml