Date: Sun, 5 Jan 2003 19:15:55 -0800 From: David Schultz <dschultz@uclink.Berkeley.EDU> To: Brett Glass <brett@lariat.org> Cc: Rahul Siddharthan <rsidd@online.fr>, chat@FreeBSD.ORG Subject: Re: Bystander shot by a spam filter Message-ID: <20030106031555.GB1938@HAL9000.homeunix.com> In-Reply-To: <4.3.2.7.2.20030104201251.029387d0@localhost> References: <4.3.2.7.2.20030104112015.026a5530@localhost> <4.3.2.7.2.20030104201251.029387d0@localhost>
next in thread | previous in thread | raw e-mail | index | archive | help
Thus spake Brett Glass <brett@lariat.org>: > At 04:25 PM 1/4/2003, Rahul Siddharthan wrote: > > >GCC 3.2.1 seems to perform around as well, on my code, as Intel's > >compiler. > > Depends on your code. A program consisting mostly of function calls > isn't going to be much of a challenge for any compiler. But try some > serious nested loops, or floating point, and GCC generates about the > most naive code you could imagine. You could do better dashing it off > in assembly language. My own experience has been that recent versions of GCC 2 botch floating point horribly. Any time the compiler encounters code containing floating point, it starts managing the stack and registers poorly, it doesn't find loop invariants anymore, and other optimiztions go out the window. However, GCC 3 fixes all of these problems in the examples I have tried, so evidently the developers did something right. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-chat" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20030106031555.GB1938>