Date: Fri, 15 Mar 2002 13:31:08 -0800 From: Kent Stewart <kstewart@owt.com> To: Jonathan Chen <jonc@chen.org.nz> Cc: Jesse Geddis <sgeine@yahoo.com>, Kris Kennaway <kris@obsecurity.org>, stable@FreeBSD.ORG Subject: Re: top wrong again? Message-ID: <3C92681C.1030409@owt.com> References: <20020315153245.B1282@grimoire.chen.org.nz> <NGBBKILMGLGEDIHMGJANOEDICBAA.sgeine@yahoo.com> <20020316090353.A27877@grimoire.chen.org.nz>
next in thread | previous in thread | raw e-mail | index | archive | help
Jonathan Chen wrote: > On Thu, Mar 14, 2002 at 09:38:49PM -0800, Jesse Geddis wrote: > >>here, different machine while compiling the kernel. maybe this is more >>along the lines of the original email: >> > > As another poster has said, the problem with running buildworld is > that processes get created and finish so quickly, that you rarely get > to see the process that's hogging the processor. If you hit the > <space> bar heaps of times, you may see the "cc/as/cc1" processes hit > the chart. The best is to run the buildworld on a *slow* box (eg a > 486) and you will see all the CPU hoggers hit the top(1) charts as it > struggles through the build. I kind of smiled here. I have setiathome running and the cpu is basically always at 99%. The load averages climb above 1.0 but never too far above 2.4-5. When I do a buildworld, I have found that the fastest wall clock time is achived with no -j specified. On an AMD 1600+ XP, the difference is several minutes faster for no -j than -j4. It isn't much but 21 minutes becomes 23 and that is about 10% slower for the -j4. The system and user times vary; however, the overhead from switching tasks costs wall clock time on a single cpu. Kent -- Kent Stewart Richland, WA http://users.owt.com/kstewart/index.html To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?3C92681C.1030409>