Date: Fri, 29 Sep 2006 01:55:59 +1000 (EST) From: Bruce Evans <bde@zeta.org.au> To: Ruslan Ermilov <ru@FreeBSD.org> Cc: amd64@FreeBSD.org, current@FreeBSD.org, John Baldwin <jhb@FreeBSD.org> Subject: Re: [head tinderbox] failure on amd64/amd64 Message-ID: <20060929014648.T2802@epsplex.bde.org> In-Reply-To: <20060928101535.GB4708@rambler-co.ru> References: <20060928001028.A6D907302F@freebsd-current.sentex.ca> <20060928101535.GB4708@rambler-co.ru>
next in thread | previous in thread | raw e-mail | index | archive | help
On Thu, 28 Sep 2006, Ruslan Ermilov wrote: > I.e., by default, -m32 on amd64 still tunes for k8. I don't > know what others think about it (perhaps it would still be > a good idea to tune for k8 on amd64 even in the boot code), No, speed is unimportant and tuning for Athlons generally gives larger code (though it probably shouldn't with -Os). However, tuning for i386 might not give smallest code. amd64 can also execute non-i386 instructions so it could use "arch"ing instead of tuning for Athlons. At least bswap would be smaller (probably not enough other instructions to matter). I don't know how to use non-i386 instructions without losing tuning for i386's. > but for now this looked a good work-around to me, and it > definitely takes less bytes than the k8-tuned version. Bruce
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20060929014648.T2802>