From owner-freebsd-hackers Sun Jan 26 14:10:03 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id OAA27554 for hackers-outgoing; Sun, 26 Jan 1997 14:10:03 -0800 (PST) Received: from root.com (implode.root.com [198.145.90.17]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id OAA27538 for ; Sun, 26 Jan 1997 14:09:55 -0800 (PST) Received: from localhost (localhost [127.0.0.1]) by root.com (8.7.6/8.6.5) with SMTP id OAA08360; Sun, 26 Jan 1997 14:07:28 -0800 (PST) Message-Id: <199701262207.OAA08360@root.com> X-Authentication-Warning: implode.root.com: Host localhost [127.0.0.1] didn't use HELO protocol To: Terry Lambert cc: michaelh@cet.co.jp, proff@suburbia.net, hackers@FreeBSD.ORG Subject: Re: SLAB stuff, and applications to current net code (fwd) In-reply-to: Your message of "Sun, 26 Jan 1997 14:18:45 MST." <199701262118.OAA02324@phaeton.artisoft.com> From: David Greenman Reply-To: dg@root.com Date: Sun, 26 Jan 1997 14:07:28 -0800 Sender: owner-hackers@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk >> The allocator in BSD is designed to be as fast as possible and trades >> space efficiency for performance. I'm very skeptical that a SLAB allocator >> would be any faster than the current allocation algorithm, although it >> would likely be more space efficient. > >Are you looking at the TLB overhead for SLAB? See my other reply regarding ordering of allocations. Memory that was recently freed is re-used before memory that was freed long ago, which should have both better cache effects and less TLB overhead (than the original BSD allocator). This should compare well to the slab allocator, but it would perhaps be interesting to actually do a formal comparison. -DG David Greenman Core-team/Principal Architect, The FreeBSD Project