This is the mail archive of the
mailing list for the GCC project.
Re: gcc compile-time performance
- From: Jan Hubicka <jh at suse dot cz>
- To: Robert Dewar <dewar at gnat dot com>
- Cc: neil at daikokuya dot demon dot co dot uk, aoliva at redhat dot com,chip dot cuntz at earthling dot net, davem at redhat dot com, gcc at gcc dot gnu dot org,jh at suse dot cz
- Date: Sun, 19 May 2002 13:41:04 +0200
- Subject: Re: gcc compile-time performance
- References: <20020519111849.D908FF28CC@nile.gnat.com>
> > Incidentally, if and when we get multibyte-char support done
> > properly, lexing could become quite expensive.
> I disagree, unless you have some perverse definiteion of "properly".
Unforutnately this is the case. Multiple byte char support is rather costy, as
you no longer can rely on the fixed size of character complicating everything
around. I've seen the progress of integrating multibyte support to different
preprocessor and the cost has been about twofold slowdown.
In common case preprocessor is not main issue for us, but there are definitly
marginal cases, where prerocessor abuse makes it extremly costy. It is
winderfull to have it optimized and clean, hope we will suceed to do the same
with other parts of compiler. We have long way to go.