[prev in list] [next in list] [prev in thread] [next in thread] 

List:       gentoo-user
Subject:    Re: [gentoo-user] CHOST Optimizations
From:       Sami =?iso-8859-15?q?N=E4=E4t=E4nen?= <sn.ml () bayminer ! com>
Date:       2004-02-29 23:53:15
Message-ID: 200403010153.15597.sn.ml () bayminer ! com
[Download RAW message or body]

On Monday 01 March 2004 00:31, Peter Ruskin wrote:
> On Sunday 29 Feb 2004 22:10, Tony Clark wrote:
> > > -fomit-frame-pointer is redundant - -O? implies that flag.
> > >
> > > I'm currently using:
> > > CFLAGS="-march=athlon-xp -O2 -mmmx -m3dnow -msse -mfpmath=sse
> > > -ftracer -fprefetch-loop-arrays -pipe"
> > >
> > > Peter
> >
> > Not too sure that is correct on i386 hardware as it will interfere
> > with debuging.
> > -O also turns on -fomit-frame-pointer on machines where doing so
> > does not interfere with debugging.
>
> That's how I understood the gcc manpage.  Why explicitely use
> -fomit-frame-pointer when "-O also turns on -fomit-frame-pointer on
> machines where doing so does not interfere with debugging" and "-O2
> turns on all optimization flags specified by -O" and "-O3 turns on
> all optimizations specified by -O2"?
>
> My CFLAGS above are an experiment.  At the first sign of trouble they
> go back to CFLAGS="-march=athlon-xp -O2 -pipe".

You will need to add the -fomit-frame-pointer in x86 archs, because it 
is not turned on by the optimization. This is so, because x86 code is 
not debugable if it doesn't provide frame pointers.


--
gentoo-user@gentoo.org mailing list

[prev in list] [next in list] [prev in thread] [next in thread]