This is the mail archive of the
mailing list for the Cygwin project.
Re: gcc3.2 vs gcc2.95 - devolution never stops
- From: "Dockeen" <dockeen at mchsi dot com>
- To: <cygwin at cygwin dot com>
- Date: Thu, 29 Aug 2002 22:12:31 -0300
- Subject: Re: gcc3.2 vs gcc2.95 - devolution never stops
I think I mumbled something about this a few months ago.
I found that the execution speed for an imaging program
that we work on in our shop went up by a noticable amount,
anywhere from 15% - 25%
Because I am not compiling a ton of SLOC, the increased
compile time was not a big deal. I can see (and have
read) where it is for people pushing a lot of SLOC.
One nice thing about the 3.1+ family of gcc is that you
can set flags for optimization for the P3 and P4 processors.
(Yep, Intel around our shop). This helps the speed.
The code passed my regression testing...
The increased exe file size was there, stripping brought it
back close to where it was before.
Have a good night,
Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
Bug reporting: http://cygwin.com/bugs.html