Date: Mon, 19 May 2003 15:58:24 +0200 From: Joerg Wunsch <j@uriah.heep.sax.de> To: Alexander Leidinger <Alexander@Leidinger.net> Cc: kris@obsecurity.org Subject: Re: cvs commit: ports/devel/avr-libc Makefile Message-ID: <20030519155824.M2937@uriah.heep.sax.de> In-Reply-To: <20030519153809.1b01f51b.Alexander@Leidinger.net>; from Alexander@Leidinger.net on Mon, May 19, 2003 at 03:38:09PM %2B0200 References: <200305190828.h4J8S8l0069634@repoman.freebsd.org> <20030519094115.GC1035@sunbay.com> <20030519115435.C2937@uriah.heep.sax.de> <20030519104604.GB17366@sunbay.com> <20030519125334.D2937@uriah.heep.sax.de> <20030519114332.GG17366@sunbay.com> <20030519115339.GI17366@sunbay.com> <20030519121239.GA6087@rot13.obsecurity.org> <20030519143325.H2937@uriah.heep.sax.de> <20030519153809.1b01f51b.Alexander@Leidinger.net>
next in thread | previous in thread | raw e-mail | index | archive | help
As Alexander Leidinger wrote: > People which don't care about some few percent more performance > don't need the default CPUTYPE, those who care will figure out how > to do it. That's exactly my point. Heck, when the default kernel was released of the burden to still support 80386 CPUs (where it was known that a kernel that must support these CPUs is less efficient than one that can concentrate on 80486+), the overall "make world" speedup gained by it was reported to be as few as 2 %. So i don't see why to bother with any -mcpu stuff by default. Except for few compute-intensive tasks, it will only gain a few percent at best. For populare AMD CPUs, the speedup is even likely to become negative... -- cheers, J"org .-.-. --... ...-- -.. . DL8DTL http://www.sax.de/~joerg/ NIC: JW11-RIPE Never trust an operating system you don't have sources for. ;-)
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20030519155824.M2937>