This is the mail archive of the
mailing list for the GCC project.
Re: Faster compilation speed
- From: dewar at gnat dot com (Robert Dewar)
- To: gcc at gcc dot gnu dot org, robertlipe at usa dot net
- Date: Tue, 13 Aug 2002 15:48:59 -0400 (EDT)
- Subject: Re: Faster compilation speed
<<Now will you please quit arguing with Apple that GCC is not really too
slow for them today based solely on counterarguments that either some
other compiler for some other language was fast or that processors will
be faster sooner than the compiler can be made faster?
You miss my point. Which is that it is only worth doing things that have a
really substantial impact and can be done on a reasonably short time scale.
You are simply not going to get anywhere by, for example, worrying about
avoiding refolding expressions.
I would guess the two big opportunites are the persistent front end and PCH,
but from what I understand Apple has already done these two steps, so the
question is where to go from there, and that is far from clear.
I have always found GCC awfully slow myself. Remember that I am used to
using compilers that are far far faster than the code warrior compilers :-)
The thing to avoid is putting in a large amount of work that results in little
real speed up, at the expense of reliability and other improvements.
One thing that would be interesting is to know, for one of these giant OS
projects (which I assume are in the million line but not dozens of million
line range) what the division between front end time and back end time is.
In the case of Ada most of the time is spent in the back end for large programs
so there is not much we can do in the front end if optimization is turned on.