Date: Sun, 9 Feb 2003 06:35:05 -0500 (EST) From: Jeff Roberson <jroberson@chesapeake.net> To: phk@phk.freebsd.dk Cc: Alfred Perlstein <bright@mu.org>, Peter Wemm <peter@wemm.org>, Jake Burkholder <jake@locore.ca>, <arch@FreeBSD.ORG> Subject: Re: New kernel allocation API Message-ID: <20030209063342.W77658-100000@mail.chesapeake.net> In-Reply-To: <25848.1044781098@critter.freebsd.dk>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sun, 9 Feb 2003 phk@phk.freebsd.dk wrote: > In message <20030209085317.GP88781@elvis.mu.org>, Alfred Perlstein writes: > >* phk@phk.freebsd.dk <phk@phk.freebsd.dk> [030209 00:31] wrote: > >> In message <20030209075706.GO88781@elvis.mu.org>, Alfred Perlstein writes: > >> > >> >I may see if I can convert some paths in the kernel to use zones. :) > >> > >> I have a number of places where zones can be used instead of malloc, > >> but since none of the places I have tried previously have resulted > >> in measurable performance differences it has not been high on my > >> agenda. > > > >Well, once converted to zones they could also be modded to take > >advantage of the init/fini processing which _should_ make things > >faster. Also don't forget even if you don't see a perf improvement > >you'll see siginifigant space savings if your size is not a power of > >two or multiple of page size. > > I'm aware of all this, but I guess most of the objects I deal with > don't really have enough trafic to come out measurable, probably > because they don't contain mutexes. > I did pipes once when I originally did UMA. The perf was measurable. It had carefully crafted init/fini ctor/dtor though. Also, creds could possibly benefit. UMA mbufs would rule too. :-) Cheers, Jeff To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-arch" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20030209063342.W77658-100000>