Date: Sat, 12 Jul 2014 16:23:23 -0400 From: George Rosamond <george@ceetonetechnology.com> To: freebsd-arm@freebsd.org Subject: Re: Impressive CPU usage Message-ID: <53C1993B.7040900@ceetonetechnology.com> In-Reply-To: <53C18DC9.2020205@fgznet.ch> References: <64FC570A-8FAF-4366-A9CA-6F91EBB6739B@freebsd.org> <53C18DC9.2020205@fgznet.ch>
next in thread | previous in thread | raw e-mail | index | archive | help
Andreas Tobler: > On 12.07.14 21:27, Tim Kientzle wrote: >> While building the git port on my BeagleBone Black: >> >> >> last pid: 87645; load averages: 1.05, 0.93, 0.72 up 0+13:15:47 >> 19:24:42 19 processes: 2 running, 17 sleeping CPU: 68.4% user, 0.0% >> nice, 28.5% system, 3.0% interrupt, 0.0% idle Mem: 57M Active, 352M >> Inact, 77M Wired, 1804K Cache, 67M Buf, 4280K Free Swap: >> >> PID USERNAME THR PRI NICE SIZE RES STATE TIME WCPU >> COMMAND 87645 root 1 69 0 44792K 33288K RUN 0:02 >> 1777.13% cc 87644 root 1 8 0 35336K 22832K wait >> 0:00 107.14% cc 87633 root 1 40 0 11568K 3056K RUN >> 0:00 0.84% top 87564 root 1 8 0 10304K 1932K wait >> 0:01 0.53% gmake 677 root 1 40 0 18880K 5136K select >> 0:16 0.04% sshd >> >> >> For the record: >> >> root@beaglebone:~ # uname -a FreeBSD beaglebone 11.0-CURRENT FreeBSD >> 11.0-CURRENT #0 r266929: Sun Jun 1 00:29:41 PDT 2014 >> root@thinkpad:/usr/home/tim/crochet-freebsd/work/obj/arm.armv6/usr/src/sys/BEAGLEBONE >> >> arm > > > +1: > > andreast@wandquad:~ % uname -ra > FreeBSD wandquad 11.0-CURRENT FreeBSD 11.0-CURRENT #12 r268501M: Thu Jul > 10 21:02:32 CEST 2014 > andreast@tcx58.andreas.nets:/usr/obj/arm.armv6/export/devel/fbsd/src/sys/WANDQUAD > arm > > last pid: 30275; load averages: 8.51, 7.73, 5.25 up 2+00:19:10 > 21:32:22 > 50 processes: 9 running, 41 sleeping > CPU: 71.2% user, 0.0% nice, 18.8% system, 9.9% interrupt, 0.2% idle > Mem: 233M Active, 260M Inact, 159M Wired, 244K Cache, 114M Buf, 1349M Free Great stuff. This could allow me to start up that selection of ARMv6 pkgs I had on mirrors.nycbug.org, since I was doing native, sans-poudriere builds. What exactly is enabling this performance increase? Has/can it be MFC'd to 10.x? g
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?53C1993B.7040900>