From owner-freebsd-hackers Wed Jun 26 10:13:58 2002 Delivered-To: freebsd-hackers@freebsd.org Received: from apollo.backplane.com (apollo.backplane.com [216.240.41.2]) by hub.freebsd.org (Postfix) with ESMTP id 479C437B62A for ; Wed, 26 Jun 2002 10:11:55 -0700 (PDT) Received: from apollo.backplane.com (localhost [127.0.0.1]) by apollo.backplane.com (8.12.3/8.12.3) with ESMTP id g5QHBsl1021463; Wed, 26 Jun 2002 10:11:54 -0700 (PDT) (envelope-from dillon@apollo.backplane.com) Received: (from dillon@localhost) by apollo.backplane.com (8.12.3/8.12.3/Submit) id g5QHBsFH021462; Wed, 26 Jun 2002 10:11:54 -0700 (PDT) (envelope-from dillon) Date: Wed, 26 Jun 2002 10:11:54 -0700 (PDT) From: Matthew Dillon Message-Id: <200206261711.g5QHBsFH021462@apollo.backplane.com> To: Terry Lambert Cc: Peter Wemm , Alfred Perlstein , Patrick Thomas , freebsd-hackers@FreeBSD.ORG Subject: Re: tunings for many httpds... References: <20020625222632.B7C7D3811@overcee.wemm.org> <3D1970E7.697D4A49@mindspring.com> <200206260758.g5Q7wPgZ019100@apollo.backplane.com> <3D198D71.86779927@mindspring.com> Sender: owner-freebsd-hackers@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG : http://old.lwn.net/2002/0124/kernel.php3 : "What Rik van Riel is up to." : :FWIW: In the original Article: : : http://lwn.net/Articles/3327/ : :they say they've gon to a 3 level page table scheme for the Hammer :port. : :-- Terry They are still using a 2 level page table scheme in the linux kernel. That is, they didn't try to integrate the 3-level hardware in the hammer with the 2 level kernel representation. What they did was create a fixed third level (or a fixed first level depending on how you look at it). As far as Linux is concerned it is still a two-level page table. This did simplify other areas of the kernel, e.g. no need for 'himem' because all of physical memory is linearly mapped into KVM, but the actual mechanism is not all that complex. I think that in general the linux folks have found the copy-on-write page table sharing to be less effective then they hoped. The only advantage seems to occur with fork(). The programs that really need it, like databases, do not share data via fork() but instead use mmap() or SysV shared memory. Both Linux and FreeBSD wind up being in the same boat. Being able to proactively share page table pages will solve the problem for both systems. -Matt To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message