*BSD News Article 73896


Return to BSD News archive

Path: euryale.cc.adfa.oz.au!newshost.anu.edu.au!harbinger.cc.monash.edu.au!munnari.OZ.AU!news.ecn.uoknor.edu!solace!eru.mt.luth.se!bloom-beacon.mit.edu!news.mathworks.com!fu-berlin.de!irz401!orion.sax.de!uriah.heep!news
From: j@uriah.heep.sax.de (J Wunsch)
Newsgroups: comp.os.linux.networking,comp.unix.bsd.netbsd.misc,comp.unix.bsd.freebsd.misc
Subject: Re: TCP latency
Date: 16 Jul 1996 10:14:59 GMT
Organization: Private BSD site, Dresden
Lines: 36
Message-ID: <4sfq33$meq@uriah.heep.sax.de>
References: <4paedl$4bm@engnews2.eng.sun.com> <4s6k8o$4o0@fox.ksu.ksu.edu>
  <31E6FD92.41C67EA6@dyson.iquest.net> <4s8cuq$ljd@bertrand.ccs.carleton.ca>
  <31E7C0DD.41C67EA6@dyson.iquest.net> <4s8tcn$jsh@fido.asd.sgi.com>
  <31E80ACA.167EB0E7@dyson.iquest.net> <4sadde$qsv@linux.cs.helsinki.fi>
  <31E9E3A7.41C67EA6@dyson.iquest.net> <4sefde$f0l@fido.asd.sgi.com>
Reply-To: joerg_wunsch@uriah.heep.sax.de (Joerg Wunsch)
NNTP-Posting-Host: localhost.heep.sax.de
Mime-Version: 1.0
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 8bit
X-Newsreader: knews 0.9.6
X-Phone: +49-351-2012 669
X-PGP-Fingerprint: DC 47 E6 E4 FF A6 E9 8F  93 21 E0 7D F9 12 D6 4E
Xref: euryale.cc.adfa.oz.au comp.os.linux.networking:45409 comp.unix.bsd.netbsd.misc:4074 comp.unix.bsd.freebsd.misc:23725

lm@neteng.engr.sgi.com (Larry McVoy) wrote:

> Maybe gettimeofday() would have been a better null syscall, what do
you think?

No.  If you de-optimized your gettimeofday() in order to give as
accurate information as possible, but have poor hardware (i.e.,
everything less than i586 on the iNtel front), gettimeofday() requires
IO operations to the ISA bus, and this will make it _very_ slow, for
the benefit of giving accurate information (i.e., the returned values
have a much better resolution than ``HZ'').

So you will essentially measure IO overhead on an obsolete bus system.

Looking at syscalls.master...

	kill(1, 0)  - should just return you success or EPERM,
	              depending on your effective UID), has to walk
	              the process table however, so it depends on the
	              organization of this table

	dup(0)      - might have some VFS overhead

	getppid()   - cannot be cached in userland, the parent might
	              have died since the previous call, so the process
	              got a new parent

	umask(umask(0)) - does modify the process context in the kernel, but
	              the modification is of a very simple nature

-- 
cheers, J"org

joerg_wunsch@uriah.heep.sax.de -- http://www.sax.de/~joerg/ -- NIC: JW11-RIPE
Never trust an operating system you don't have sources for. ;-)