Return to BSD News archive
Path: euryale.cc.adfa.oz.au!newshost.anu.edu.au!harbinger.cc.monash.edu.au!munnari.OZ.AU!news.mel.connect.com.au!news.mira.net.au!vic.news.telstra.net!act.news.telstra.net!imci3!newsfeed.internetmci.com!news.inc.net!newspump.sol.net!newsfeeder.sdsu.edu!sgigate.sgi.com!timbuk.cray.com!walter.cray.com!fido.asd.sgi.com!neteng!lm From: lm@neteng.engr.sgi.com (Larry McVoy) Newsgroups: comp.os.linux.networking,comp.unix.bsd.netbsd.misc,comp.unix.bsd.freebsd.misc Subject: Re: TCP latency Followup-To: comp.os.linux.networking,comp.unix.bsd.netbsd.misc,comp.unix.bsd.freebsd.misc Date: 16 Jul 1996 21:46:01 GMT Organization: Silicon Graphics Inc., Mountain View, CA Lines: 45 Message-ID: <4sh2ip$2nk@fido.asd.sgi.com> References: <4paedl$4bm@engnews2.Eng.Sun.COM> <31E7C0DD.41C67EA6@dyson.iquest.net> <4s8tcn$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> <31EBB017.167EB0E7@dyson.iquest.net> Reply-To: lm@slovax.engr.sgi.com NNTP-Posting-Host: neteng.engr.sgi.com X-Newsreader: TIN [version 1.2 PL2] Xref: euryale.cc.adfa.oz.au comp.os.linux.networking:45465 comp.unix.bsd.netbsd.misc:4084 comp.unix.bsd.freebsd.misc:23767 John S. Dyson (toor@dyson.iquest.net) wrote: : Larry McVoy wrote: : > Well, I doubt you'll buy it, but I agree with Linus that your message : > certainly implied that Linux special cased the null syscal metric used : > in lmbench. IF that was your intended implication, you may not have been : > lying, but you certainly were implying something that simply isn't true. : > : Firstly, you are one of the first people that I have seen who apparently : understands that a lie requires intent. Additionally, I was speaking of : something inside of ME. I did not do it. Cool, we're moving towards a conversation, rather than yelling at each other. I like it. Could you see, in rereading your posting, that perhaps people might have seen implication that Linux had a special case for a null syscall? I'll grant you that it never helps to assume the worst, but I think a lot of objective observers would have read your posting the same way. It could have been an honest mistake, but why did you phrase it with the obvious implication that Linux had cheated/optimized that case, whatever? I think Linus was a little justified in pointing out that it was simply good coding, rather than a special case optimization, that made Linux fast. I imagine he felt pretty insulted that you suggested he'd special case to make Linux look better - for what it is worth, I would be amazed if he did that, he's far too concerned with "doing the right thing" (as are you, John, you guys are somewhat similar in that respect) to special case up his code. I'll bet if I suggested that FreeBSD had special cases in it for benchmarking, you would go ballistic. And if it was in an area where you had worked hard to solve the _general case_ and then I implied you had done some sleazy hack for benchmarking purposes, you would scream foul even more loudly. At any rate, I'm encouraged to see that we are all calming down a bit and maybe we could move this topic onto something a bit more interesting, like, say, the perfect loaded TCP latency benchmark? As Alan has brought up, it isn't fair to just do a benchmark with one client, because it doesn't stress the arp table enough. Does everyone buy that argument and what should be done about it? -- --- Larry McVoy lm@sgi.com http://reality.sgi.com/lm (415) 933-1804