Path: csiph.com!usenet.pasdenom.info!news.albasani.net!fu-berlin.de!uni-berlin.de!individual.net!not-for-mail From: vandys@vsta.org Newsgroups: comp.lang.forth Subject: Re: Power of two bucket allocator Date: 3 Aug 2012 16:07:39 GMT Lines: 15 Message-ID: References: <2012Aug3.165537@mips.complang.tuwien.ac.at> <2203270b-5b74-4f5a-9fa9-cca0e40c6318@j11g2000vbc.googlegroups.com> X-Trace: individual.net TiuSM4FeZicBDqfqPvi9CAMWsEgHwfA1Ndzkft+AWFRFp9j0/aEcMGDA88159ZQkNr X-Orig-Path: not-for-mail Cancel-Lock: sha1:rpShkZoLl+LhL32MXoyu8Ves1CM= User-Agent: tin/1.8.3-20070201 ("Scotasay") (UNIX) (Linux/3.2.0-27-generic-pae (i686)) Xref: csiph.com comp.lang.forth:14696 Alex McDonald wrote: >> I'm not familiar with the term? > The allocated but unused and unusable space in the returned bucket. > Think Swiss cheese; most of it is holes. I see. Powers of two allocated out of page quanta really bound this in practice--McKusick's experience has been born out in my experience, both in actual UNIX kernels (HP-UX, PTX) as well as in ForthOS. Of course, the resource efficiency of Forth (and ForthOS) intersecting with modern bounty of RAM means it's easy to have "plenty". -- Andy Valencia Home page: http://www.vsta.org/andy/ To contact me: http://www.vsta.org/contact/andy.html