Date: Mon, 30 Jan 2006 11:51:21 -0600 (CST) From: Randy Schultz <schultz@sgi.com> To: freebsd-questions@freebsd.org Subject: Re: did something change in SMP scheduling from 5.4 to 6.0? Message-ID: <Pine.BSF.4.63.0601301137520.41670@tdream.americas.sgi.com> In-Reply-To: <44u0bpznc2.fsf@be-well.ilk.org> References: <006001c61d47$973bccc0$1225a8c0@kittycat> <Pine.BSF.4.63.0601261042150.41670@tdream.americas.sgi.com> <44u0bpznc2.fsf@be-well.ilk.org>
next in thread | previous in thread | raw e-mail | index | archive | help
Tnx for the response. On Fri, 27 Jan 2006, Lowell Gilbert spaketh thusly: -} -}What was the actual workload at that point? It looks like there's Not much actually. ~.3 or .4. -}only one runnable process, and it's running, so there is no reason to -}care which processor it's on. Yeah that's what I was thinking as well but I have been pushing it a fair bit and still see everything on the first proc. Here is 1 example: last pid: 82302; load averages: 2.20, 0.82, 0.41 up 6+23:37:10 11:24:50 127 processes: 6 running, 121 sleeping CPU states: 16.1% user, 0.0% nice, 2.1% system, 0.0% interrupt, 81.9% idle Mem: 295M Active, 1050M Inact, 294M Wired, 72K Cache, 112M Buf, 364M Free Swap: 4096M Total, 4096M Free PID USERNAME THR PRI NICE SIZE RES STATE C TIME WCPU COMMAND 655 schultz 1 76 0 30024K 29276K RUN 0 50:32 2.25% Xorg 82301 root 1 139 0 10784K 10200K RUN 0 0:00 2.25% cc1 12185 schultz 1 76 0 76532K 75704K select 0 22:37 0.00% opera 14583 schultz 1 76 0 5512K 4548K select 0 5:49 0.00% xterm-static 3013 schultz 8 20 0 91760K 77840K kserel 0 4:57 0.00% evolution-2.2 10846 schultz 1 76 0 5452K 4460K select 0 3:24 0.00% xterm-static 41670 schultz 1 76 0 10236K 8572K select 0 2:23 0.00% pine 667 schultz 1 76 0 20844K 10208K select 0 1:20 0.00% xchat Now when running the compile it seemed to make no difference whether or not I was running 1 or 3 compiles, the system still hugged cpu0. I did see occasional hops to the second cpu but they were rare. -} -}I'm not sure there's an issue here. Spreading the work among two Me either but I thought I'ld ask those who know more. -}processors might make slightly better use of cache space, but would -}make power-saving modes less useful. Now if you run two completely -}independent CPU-intensive processes, you'll see both CPUs in use. Well, I have and have seen them running all on cpu0 and have seen some spread out. What I cannot be sure of is when multiple compiles are running on the same cpu what exactly each was doing. For example I've kicked off several portupgrades and they all run mostly on cpu0 however with a portupgrade there is so much more going. <shrug> If I can get more data I'll post back here. Tnx again for the help. -- Randy (schultz@sgi.com) 715-726-2832 email bodhisattva <*> "There is no fire like passion, there is no shark like hatred, there is no snare like folly, there is no torrent like greed."
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.BSF.4.63.0601301137520.41670>