From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S261538AbTJCXhE (ORCPT ); Fri, 3 Oct 2003 19:37:04 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S261546AbTJCXhE (ORCPT ); Fri, 3 Oct 2003 19:37:04 -0400 Received: from [205.180.85.17] ([205.180.85.17]:45744 "EHLO mail.fastclick.net") by vger.kernel.org with ESMTP id S261538AbTJCXet (ORCPT ); Fri, 3 Oct 2003 19:34:49 -0400 Message-ID: <3F7E0704.4020009@fastclick.com> Date: Fri, 03 Oct 2003 16:32:20 -0700 From: Brett Reply-To: brettspamacct@fastclick.com User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.4) Gecko/20030624 X-Accept-Language: en-us, en MIME-Version: 1.0 To: Kevin Kahley CC: linux-kernel@vger.kernel.org Subject: Re: exclusive memory held by a process References: In-Reply-To: Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org I asked the same question yesterday. If you look in /prod/[pid]/maps you can see the memory layout per process. Maybe you could find this information out by looking through the maps files? I didn't get a response about this but I'm still going to try this. Good luck, Brett Kevin Kahley wrote: > I am working with a pc system that has no swap space and no hard disk. > I have direct control over 3 processes on this system and need to > guarantee that they never use more than 190 MB of memory. I have been > getting the information about memory usage from /proc/*pid*/status but can > not seem to make exact sense of what it is reporting. Here are the values > I am getting: > > Name: g > State: S (sleeping) > Tgid: 292 > Pid: 292 > PPid: 284 > TracerPid: 0 > Uid: 0 0 0 0 > Gid: 0 0 0 0 > FDSize: 256 > Groups: > VmSize: 7772 kB > VmLck: 7772 kB > VmRSS: 7772 kB > VmData: 4284 kB > VmStk: 32 kB > VmExe: 1028 kB > VmLib: 1760 kB > SigPnd: 0000000000000000 > SigBlk: 0000000080000000 > SigIgn: 0000000000010000 > SigCgt: 0000000380000000 > CapInh: 0000000000000000 > CapPrm: 00000000fffffeff > CapEff: 00000000fffffeff > > > Name: s > State: S (sleeping) > Tgid: 288 > Pid: 288 > PPid: 287 > TracerPid: 0 > Uid: 0 0 0 0 > Gid: 0 0 0 0 > FDSize: 32 > Groups: > VmSize: 83248 kB > VmLck: 83248 kB > VmRSS: 83244 kB > VmData: 81256 kB > VmStk: 20 kB > VmExe: 84 kB > VmLib: 1760 kB > SigPnd: 0000000000000000 > SigBlk: 0000000080000000 > SigIgn: 0000000000000000 > SigCgt: 7ffffffffffbfeff > CapInh: 0000000000000000 > CapPrm: 00000000fffffeff > CapEff: 00000000fffffeff > > > Name: v > State: S (sleeping) > Tgid: 248 > Pid: 248 > PPid: 1 > TracerPid: 0 > Uid: 0 0 0 0 > Gid: 0 0 0 0 > FDSize: 32 > Groups: > VmSize: 85576 kB > VmLck: 85576 kB > VmRSS: 77436 kB > VmData: 74780 kB > VmStk: 20 kB > VmExe: 476 kB > VmLib: 2452 kB > SigPnd: 0000000000000000 > SigBlk: 0000000080000000 > SigIgn: 0000000000000000 > SigCgt: 00000003e78074ff > CapInh: 0000000000000000 > CapPrm: 00000000fffffeff > CapEff: 00000000fffffeff > > > I have read in many places that VmSize is the total in-memory size of the > running process, but does this include memory that is being shared? I > thought about adding VmRSS, VmData, and VmStk, but that is greater than > VmSize? Can anyone tell me what is the amount of memory solely used by a > process? It's my understanding that my system should crash if these three > processes exceed 190 MB, but using /proc/*pid* values does not confirm > this... > > please CC me on any responses: kkahley@cs.uic.edu > > Thank you much in advance. > > - > To unsubscribe from this list: send the line "unsubscribe linux-kernel" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html > Please read the FAQ at http://www.tux.org/lkml/ >