Date: Tue, 19 Apr 2005 12:23:38 -0600 From: Scott Long <scottl@samsco.org> To: obrien@freebsd.org Cc: freebsd-amd64@freebsd.org Subject: Re: Recommendations? Message-ID: <42654CAA.10809@samsco.org> In-Reply-To: <20050419181831.GA27679@dragon.NUXI.org> References: <200504131714.20356.jkim@niksun.com> <86fyxt1ovu.fsf@xps.des.no> <20050415005850.GB80903@dragon.NUXI.org> <2fd864e050414180911e4e5be@mail.gmail.com> <20050415211600.GA1473@dragon.NUXI.org> <1113603823.91832.7.camel@cream.xbsd.org> <86ekdbdnqh.fsf@xps.des.no> <4260F103.5050309@mail.uni-mainz.de> <20050419173247.GB26152@dragon.NUXI.org> <42654719.7030607@samsco.org> <20050419181831.GA27679@dragon.NUXI.org>
next in thread | previous in thread | raw e-mail | index | archive | help
David O'Brien wrote: > On Tue, Apr 19, 2005 at 11:59:53AM -0600, Scott Long wrote: > >>David O'Brien wrote: >> >>>On Sat, Apr 16, 2005 at 01:03:31PM +0200, O. Hartmann wrote: >>> >>> >>>>I have ASUS A8N-SLI Deluxe mobo. BIOS revision is 1006. I have no >>>>problems with >>>>ACPI. APIC works with NO_MIXED_MODE commented out in kernel. >>> >>> >>>Disable the NO_MIXED_MODE option is a performance degradation. >>>Bending over and taking it up the a** across all motherboards (which DES >>>has forced on 5.4-RELEASE); makes the ASUS A8N-SLI Deluxe not "perfectly >>>working". >>> >> >>According to the APIC maintainer, and according to some quick benchmarks >>myself, mixed mode has almost no performance impact. > > > Maybe with the latest code - if you're talking about before the very > recent jhb commits. Peter has said several times there is(was) a > performance issue. 'mixed mode' also is (was?) expressly prohibited by > the ACPI spec. > Mixed mode had a performance impact back in the 5.1 or 5.2 timeframe when the PIC/APIC code wasn't abstracted like it is now. This applies to 5.3 and 5.4. 6.0 will of course be better since the factors that make mixed mode a consideration are now solved a different way. Scott
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?42654CAA.10809>