Re: egcs-optimized GNOME -- WOW!

[Kevin Miller <sar hempseed com>]

>From what I've heard and read from the egcs docs.. egcs is supposed to
have a possible speedup in compiling over gcc, thats one reason why its
being worked on (just what I read)...  

I dont think the 3 step sounds bad, compile a compiler with yours, compile
the stage 1 with the first, then compile the installable compiler with the
stage1 compiler.. gets it really close to your machine I think...

I've been recompiling gnome with the list of flags sent out.. for those of
you who havent tried it.. it rox, and those of you who havent tried gimp
w/ optimizations like that, it flies.  YMMV

sar

On Tue, 13 Oct 1998, Spauldo da Hippie Elf wrote:

> I could be very mistaken here, but if I am, blame sar (hi kevin!)
> 
> gcc was prolly installed on yer machine by yer distro.  
> 
> egcs does a 3-step bootstrap process during compilation, so that it can
> compile itself with itself (sounds odd, don't it?) and set proper
> optimizations for your particular machine.
> 
> So if you compiled egcs, then that could be why it's faster.  Other
> reasons could exist of course, I'm just goin' what sar told me earlier
> today when I compiled it.
> 
> Soren Harward wrote:
> > 
> > I finally got everything I needed to compiled with egcs and the
> > optimization flags '-O6 -mpentiumpro -march=pentiumpro' and MAN DOES GNOME
> > FLY!!  It's great!!  Thanks a lot to Mr. Hult, everything is beautiful.  I
> > think even egcs is faster than gcc at compiling the stuff in the first
> > place (no definitive tests done yet).
> > 
> > Now if I could just get StarOffice running this fast...
> > 
> > --
> >  Soren Harward <soren@cinternet.net> | Windows DOES come with a tool
> >  http://www.cinternet.net/~soren/    | to restore a corrupt Registry.
> >  Internet Information Systems Admin  |
> >  Cinternet, Inc.  (513) 891-1228     | It's called FDISK.
> > 
> > --
> >          To unsubscribe: mail gnome-list-request@gnome.org with
> >                        "unsubscribe" as the Subject.
> 
> 
> -- 
>          To unsubscribe: mail gnome-list-request@gnome.org with 
>                        "unsubscribe" as the Subject.
> 

------------------------------------------
>From linux/kernel/panic.c :

int bad_user_access_length(void)
{
        panic("bad_user_access_length executed (not cool, dude)");
}

Something tells me this isn't a good thing.