From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S261737AbVDEPLs (ORCPT ); Tue, 5 Apr 2005 11:11:48 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S261773AbVDEPLs (ORCPT ); Tue, 5 Apr 2005 11:11:48 -0400 Received: from palrel13.hp.com ([156.153.255.238]:31958 "EHLO palrel13.hp.com") by vger.kernel.org with ESMTP id S261737AbVDEPLo (ORCPT ); Tue, 5 Apr 2005 11:11:44 -0400 Message-Id: <200504051521.UAA00681@harvest.india.hp.com> From: "Amanulla G" To: Cc: Subject: Re: /proc on 2.4.21 & 2.6 kernels.... Date: Tue, 5 Apr 2005 20:41:36 +0530 MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit X-Mailer: Microsoft Office Outlook, Build 11.0.5510 X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1441 Thread-Index: AcU58cHsqYoNJh4DQoiSD1Jc47iRcQ== Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Hi, I would like to know the information on /proc under 2.4.21 based kernels. On 2.4.21 based kernels, /proc has got two types of entries. /proc/pid & /proc/.tid The statistics under /proc/pid reflect just resource utilization of pid alone. They are not summarization of resource utilization of process and the threads it creates. I would like to know is there any way that we can make sure that statistics under /proc/pid reflect the resource utilization of threads it creates too? Here are the additional details. 1) One line summary of the problem: >> On 2.4.21 based kernels, /proc/pid doesn't include the consolidated resource utilization of the threads it has created. 2) Full description of the problem/report: >> Assume we have a process with pid = 1234 And this process creates two threads with thread IDs 1235 & 1236. We have three entries a) /proc/1234 b) /proc/.1235 c) /proc/.1236 The statistics under /proc/1234 doesn't reflect resource utilization of threads. 3) Keywords (i.e., modules, networking, kernel): >> /proc 4) Kernel version (from /proc/version): >> 2.4.21 5) Output of Oops.. message (if applicable) with symbolic information resolved (see Documentation/oops-tracing.txt) >> N/A 6) A small shell script or example program which triggers the problem (if possible) >> Can be easily seen with multi-threaded processes We would like to know 1) Is this a known problem? 2) Whether this problem has been fixed? If yes, in which version of the kernel this issue has been fixed? 3) In 2.6 based kernels we have different way of getting thread data. /proc//tasks/tid Does /proc//stat has got the summarized statistics of resource utilization of threads it has created ? or it just reflects the resource utilization of only? Thanks in advance! With Best Regards, Amanulla