From owner-freebsd-hackers Sat Jul 13 22:34:34 1996 Return-Path: owner-hackers Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id WAA10318 for hackers-outgoing; Sat, 13 Jul 1996 22:34:34 -0700 (PDT) Received: from rover.village.org (rover.village.org [204.144.255.49]) by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id WAA10074 for ; Sat, 13 Jul 1996 22:31:58 -0700 (PDT) Received: from localhost (localhost [127.0.0.1]) by rover.village.org (8.7.5/8.6.6) with SMTP id XAA07492; Sat, 13 Jul 1996 23:29:09 -0600 (MDT) Message-Id: <199607140529.XAA07492@rover.village.org> To: Bruce Evans Subject: Re: Some interesting papers on BSD ... Cc: matt@lkg.dec.com, freebsd-hackers@FreeBSD.ORG, tech-kern@netbsd.org In-reply-to: Your message of Tue, 09 Jul 1996 11:31:59 +1000 Date: Sat, 13 Jul 1996 23:29:08 -0600 From: Warner Losh Sender: owner-hackers@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk : spl is probably fundamentally wrong for SMP. I haven't thought much : about what to use instead. The Solbourne people might disagree with you. As might the VMS Digitial people. Both groups used a scheme where you would raise the spl *AND* grab access locks to data structures (I think the latter was optional on VMS depending on the, as I recall it, NCPU SYSGEN parameter). Both these systems used fairly fine grained locking and took a lot of time to get right and robust. Sun, when they were writing Solaris, is rumored to have produced "warlock." This program would go through all your sources and warn of potential deadlock situations in them. They never released it as they considered it a competitive advantage to them... There is no other way to do MP synchronization than using some kind of explicit locking that doesn't rely on the interrupt level.[*] Warner [*] Or I'd like ot be pointed at something that proves this statement wrong.