Return to BSD News archive
Newsgroups: comp.os.386bsd.questions Path: sserve!newshost.anu.edu.au!harbinger.cc.monash.edu.au!msuinfo!agate!howland.reston.ans.net!pipex!sunic!trane.uninett.no!eunet.no!nuug!EU.net!sun4nl!news.nic.surfnet.nl!eur.nl!cs.few.eur.nl!pk From: pk@cs.few.eur.nl (Paul Kranenburg) Subject: Re: NetBSD--which Suns can use it?? Message-ID: <1994Jun19.132149.24937@cs.few.eur.nl> Sender: news@cs.few.eur.nl Reply-To: pk@cs.few.eur.nl Organization: Erasmus University Rotterdam References: <2tbhn3$dgm@Venus.mcs.com> <DERAADT.94Jun12234937@newt.fsa.ca> <CrCD2F.9v2@cogsci.ed.ac.uk> Date: Sun, 19 Jun 1994 13:21:49 GMT Lines: 25 In <CrCD2F.9v2@cogsci.ed.ac.uk> richard@cogsci.ed.ac.uk (Richard Tobin) writes: >In article <DERAADT.94Jun12234937@newt.fsa.ca> deraadt@fsa.ca (Theo de Raadt) writes: >>Sun4c line: >> sparcstation 1, 1+, 2, IPC, IPX. >> It's very stable and reliable. It feels slower than SunOS. >Do you know why it feels slower? Is it the shared library >implementation? I don't think so (disclaimer: since I wrote the implementation, you may want to seek a second opinion), as the amount of overhead associated with loading shared libraries is essentially the same as in SunOS. Over the last few weeks I have done some work to reduce the overhead that comes with "context stealing" in the sun4c's MMU, but the effect of this was rather less than I had hoped for on my IPC (8 hardware contextses; not exactly too many when running X Windows). Also, the kernels I have built so far had -DDEBUG on, which carries extra overhead in locore.s when entering eg. trap handlers, but I cannot quantify its effects at the moment. -pk