From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S262118AbTEHUeZ (ORCPT ); Thu, 8 May 2003 16:34:25 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S262127AbTEHUeZ (ORCPT ); Thu, 8 May 2003 16:34:25 -0400 Received: from fmr04.intel.com ([143.183.121.6]:27895 "EHLO caduceus.sc.intel.com") by vger.kernel.org with ESMTP id S262118AbTEHUeV convert rfc822-to-8bit (ORCPT ); Thu, 8 May 2003 16:34:21 -0400 Message-ID: From: "Perez-Gonzalez, Inaky" To: "'Ming Lei'" , "'linux-kernel@vger.kernel.org'" Subject: RE: can linux RT thread corrupt global variable? Date: Thu, 8 May 2003 13:46:27 -0700 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2653.19) Content-Type: text/plain; charset="ISO-8859-1" Content-Transfer-Encoding: 8BIT Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org > From: Ming Lei [mailto:lei.ming@attbi.com] > > ... > > There is a global variable I define as 'int cpl'. All the threads and main > process may alter cpl at any time. cpl may have one of these values {0, > 0xf000006e, 0xf0000068, 0xe0000000, 0xe0000060}. > > at some point of execution which cpl should be a value say > e0000060, but the actual value retained at cpl is another say e0000000; that > is, the value is changed without the program actually done anything on it. > The retained value I observed is kind of historic value(one of these value > in the above set), not the arbituary value. The problem had occured just > after context switch, also occured during a thread execution. Probably GCC is keeping the value of the variable around in some register. Did you declare the global variable as 'volatile'? [as in "it can be modified by entities external to the flow of the code"]. Try that. On the other hand - being an integer, access should be atomic, most of the time, but I would not really trust that, at least portability wise. Protecting access to it with a mutex would be a good idea ... > Is linux kernel 2.4.10 considered strictly preemptive such as VxWorks or > other RTOS? Nope > I guess 2.4.10 may simulate preemptive with running scheduler on > every syscall or interrupt returns. Am I right? You need the preemption patches or 2.5 I hope you are not trying to do serious real-time w/ Linux ... are you? > Is printf() real-time priority thread safe? Dunno - you'd have to check the internal implementation in glibc. Last time I heard anything about it, it had some mutex dangling around, but I could be wrong. Iñaky Pérez-González -- Not speaking for Intel -- all opinions are my own (and my fault)