Date: Tue, 22 Jul 2003 16:07:31 -0700 From: Marcel Moolenaar <marcel@xcllnt.net> To: Poul-Henning Kamp <phk@phk.freebsd.dk> Cc: cvs-all@FreeBSD.org Subject: Re: cvs commit: src/sys/kern init_main.c kern_malloc.c md5c.c subr_autoconf.c subr_mbuf.c subr_prf.c tty_subr.c vfs_cluster.c vfs_subr.c Message-ID: <20030722230731.GB61493@athlon.pn.xcllnt.net> In-Reply-To: <16119.1058914594@critter.freebsd.dk> References: <3F1DBD05.A4886D5E@imimic.com> <16119.1058914594@critter.freebsd.dk>
next in thread | previous in thread | raw e-mail | index | archive | help
On Wed, Jul 23, 2003 at 12:56:34AM +0200, Poul-Henning Kamp wrote: > > And the only two criteria I think are trivial to use for proving an > actual benefit is: > 1. less code is generated. > 2. it runs faster in tests. criterium 1 is the worst possible. Only criterium 2 makes sense. ia64 specifically moves all the hard work to the compiler. It's not unsurprising that a normal -O yields marginal performance. Only when one takes advantage of speculation, prefetching and optimization techniques that increase ILP (which most of the time imply code expansion -- loop unrolling, inlining, code duplication) will you see a performance increase. The old gcc metric of less code is better has been demonstrated to not work in general nowadays. -- Marcel Moolenaar USPA: A-39004 marcel@xcllnt.net
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20030722230731.GB61493>