Date: Tue, 17 Jul 2007 15:18:43 +0200 From: "Claus Guttesen" <kometen@gmail.com> To: "Eric Anderson" <anderson@freebsd.org> Cc: current@freebsd.org Subject: Re: ULE/SCHED_SMP diff for 7.0 Message-ID: <b41c75520707170618o4106de94g57e60d2c93a68329@mail.gmail.com> In-Reply-To: <469CACEC.1000103@freebsd.org> References: <20070716233030.D92541@10.0.0.1> <b41c75520707170318r2152b9f0l8d2ec7ea592fe450@mail.gmail.com> <469CACEC.1000103@freebsd.org>
next in thread | previous in thread | raw e-mail | index | archive | help
> >> This patch is scheduled for inclusion in 7.0. I would like anyone who > >> cares to run it to validate that it does not create any stability or > >> performance regression over the existing ULE. This patch replaces ULE > >> with SCHED_SMP, which will now no longer exist as a seperate fork of ULE. > > > > Not very scientific nor precise but using 4bsd as scheduler 'make -j 3 > > buildkernel' completed in 11 min. 58 secs. and ule did the same in 13 > > min. 26 secs. So ule seems slower. This is on a dual zeon @ 3.2 Ghz > > (the first 64-bit from Intel, not very fast but hot) and 3 GB ram and > > 15 RPM scsi-disk with /usr on zfs. > > > > Ahah! 15 RPM drives, no wonder! :) > > On a serious note, can you do that same test, with '-j 4' or higher? I > think you can easily do two per processor, at least that's what I do on > a Core 2 Duo. Shure: sched_ule: -j 3 buildkernel: 13:23 -j 4 buildkernel: 12:38 -j 5 buildkernel: 12:41 -j 6 buildkernel: 12:47 sched_4bsd: -j 3 buildkernel: 11:43 -j 4 buildkernel: 12:02 So sched_ule seems to handle more processes slightly better than 4bsd albeit it does it slower. ule's sweet spot is -j 4 and 4bsd is -j 3. -- regards Claus When lenity and cruelty play for a kingdom, the gentlest gamester is the soonest winner. Shakespeare
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?b41c75520707170618o4106de94g57e60d2c93a68329>