From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1755250AbYKLXdh (ORCPT ); Wed, 12 Nov 2008 18:33:37 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1752246AbYKLXd2 (ORCPT ); Wed, 12 Nov 2008 18:33:28 -0500 Received: from smtp.polymtl.ca ([132.207.4.11]:48275 "EHLO smtp.polymtl.ca" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752245AbYKLXd2 (ORCPT ); Wed, 12 Nov 2008 18:33:28 -0500 Message-Id: <20081112231551.473569400@polymtl.ca> User-Agent: quilt/0.46-1 Date: Wed, 12 Nov 2008 18:15:51 -0500 From: Mathieu Desnoyers To: Linus Torvalds , akpm@linux-foundation.org, Ingo Molnar , Peter Zijlstra , Steven Rostedt , linux-kernel@vger.kernel.org Subject: [patch 00/17] Trace Clock v3 X-Poly-FromMTA: (test.casi.polymtl.ca [132.207.72.60]) at Wed, 12 Nov 2008 23:33:01 +0000 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Hi, In this new version, I have integrated the changes I did following the comments I received for v2. I also reimplemented the "generic" trace clock so it is less intrusive and simply use a standard timer. I is not as bound to the xtime_lock as it previously was and it does not have to modify kernel/time.c anymore. I think this is pretty close to a mergeable state. I plan to keep more exotic features, such as dealing with non-synchronized TSCs without cache-line bouncing, for future improvement. As a reminder, when the trace clock detects an unsynchronized TSC on the machine, it prints the information about which kernel command line arguments must be used so the user can have synchronized timestamp counters. However, if the user wants to use his system with non-synchronized TSCs, the cache-line bouncing fallback is used. This patchset applies on top of 2.6.28-rc4. Mathieu -- Mathieu Desnoyers OpenPGP key fingerprint: 8CD5 52C3 8E3C 4140 715F BA06 3F25 A8FE 3BAE 9A68