This is the mail archive of the gcc@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: gcc compilation speed


On Tue, 20 Jan 2004, Steven Bosscher wrote:

> On Tuesday 20 January 2004 07:30, Matt Austern wrote:
> > My main concern, actually, is that 3.5 may be worse.  It may not
> > be fair to talk about performance measurements this early, when
> > tree-SSA hasn't even been merged into mainline, but so far it
> > looks like it may be making things slower, even at low
> > optimization levels.  I don't think we have a very clear idea
> > yet of what's causing the slowdown; one disturbing possibility
> > is that it might be because we're carrying around twice as much
> > intermediate representation.
>
> *sigh*
>
> Numbers?  Bug reports?

I'm afraid Matt is right at least for our case of MICO project: C++ CORBA
implementation. I have tested tree-ssa twice and it was always slower
than 3.x and main-line that time. I will try to come with some numbers
with testing on MICO's ORB core, just to see if it is still right.

Cheers,

Karel
--
Karel Gardas                  kgardas@objectsecurity.com
ObjectSecurity Ltd.           http://www.objectsecurity.com


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]