Date: Mon, 23 Apr 2001 10:54:40 -0700 (PDT) From: John Baldwin <jhb@FreeBSD.org> To: Andrew Gallatin <gallatin@cs.duke.edu> Cc: alpha@FreeBSD.org Subject: Re: SMP is stable now. Whee! :) Message-ID: <XFMail.010423105440.jhb@FreeBSD.org> In-Reply-To: <15074.2220.727775.544765@grasshopper.cs.duke.edu>
next in thread | previous in thread | raw e-mail | index | archive | help
On 21-Apr-01 Andrew Gallatin wrote: > > John Baldwin writes: > > > > > > I'm happy to report that with this patchset, plus the > > > cricical_enter/exit changes I just committed, a dual cpu 2100 made it > > > through a buildworld last night just fine ;) > > > > Woo! The dual 4100 here is still cranking through a -j 8 world. > > What's the appropriate -j level for an MP box, anyway? > > I just naively used -j 2. Should I go higher? I realize going higher > will be a better stress test, where does the overhead from cache > pollution & extra scheduling outwheigh the possability of having > another job whose I/O has completed & is ready to run? > At least on a modern UP, like my tbird, omitting the -j seems to speed > things up slightly.. Hrm, I've heard the magic formula 4 * NCPU before. I.e. -j 4 should be better on UP even so that there is somthing for the CPU to do while waiting on I/O. > Drew -- John Baldwin <jhb@FreeBSD.org> -- http://www.FreeBSD.org/~jhb/ PGP Key: http://www.baldwin.cx/~john/pgpkey.asc "Power Users Use the Power to Serve!" - http://www.FreeBSD.org/ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?XFMail.010423105440.jhb>