From owner-freebsd-questions Fri Sep 6 11:30:11 1996 Return-Path: owner-questions Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id LAA18402 for questions-outgoing; Fri, 6 Sep 1996 11:30:11 -0700 (PDT) Received: from dyson.iquest.net (dyson.iquest.net [198.70.144.127]) by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id LAA18373 for ; Fri, 6 Sep 1996 11:29:59 -0700 (PDT) Received: (from root@localhost) by dyson.iquest.net (8.7.5/8.6.9) id NAA03415; Fri, 6 Sep 1996 13:29:40 -0500 (EST) From: John Dyson Message-Id: <199609061829.NAA03415@dyson.iquest.net> Subject: Re: SYSV Shared Memory Question To: rwm@MPGN.COM (Rob Miracle) Date: Fri, 6 Sep 1996 13:29:40 -0500 (EST) Cc: freebsd-questions@FreeBSD.ORG In-Reply-To: <3.0b11.32.19960906110535.00727e2c@central.TanSoft.COM> from "Rob Miracle" at Sep 6, 96 11:05:37 am Reply-To: dyson@FreeBSD.ORG X-Mailer: ELM [version 2.4 PL24 ME8] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-questions@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > We need a boat load of SYSV shared memory and we get a Kernel panic when we > try to request it. Here are the specifics: > > On our FreeBSD 2.x.x (we have tried 2.1.0, 2.1.5, and 2.2.snap (most recent)) > when we set SHMMAXPGS to 22000 and SHMALL to something over 30000, we get a > kernel panic in kmem_suballoc of type 3. > The problem is fixable, but... > > My questions are: > > Is there a limit to the amount of SYSV shared memory in FreeBSD(any version)? > What are the real meanings of SHMMAXPGS, SHMALL, SHMMAX, and other shared memory > options? > Unfortunately, the code allocates Kernel Virtual space for the shared region. That is TOTALLY unnecessary, but that is how it was implemented. I can probably fix it, but I was planning on a total rewrite. Of course, a rewrite would be in a longer timeframe than you are interested in. > > We really want to move away from our DecStations and move to PCs but this is > a big sticky issue. > I will try to see if there is a work-around that I can create. The code is calling for a rewrite though. (Even though FreeBSD has LOTS of Kernel virtual space: a little less than 256Mbytes, the code still wastes it unnecessarily.) I will try to see what I can do this weekend -- I have some ideas. I don't have any SV shared mem test tools, so if I get anything implemented -- you'll have to test it :-)!!! John dyson@freebsd.org