From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1750894AbaKWFGT (ORCPT ); Sun, 23 Nov 2014 00:06:19 -0500 Received: from mga11.intel.com ([192.55.52.93]:64707 "EHLO mga11.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750702AbaKWFGS (ORCPT ); Sun, 23 Nov 2014 00:06:18 -0500 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.07,442,1413270000"; d="scan'208";a="636638555" Date: Sat, 22 Nov 2014 21:06:17 -0800 From: Fengguang Wu To: Steven Rostedt Cc: LKP , linux-kernel@vger.kernel.org Subject: Re: [ftrace/x86/extable] WARNING: CPU: 0 PID: 1 at init/main.c:803 do_one_initcall() Message-ID: <20141123050617.GA9218@wfg-t540p.sh.intel.com> References: <20141122184349.GB29361@wfg-t540p.sh.intel.com> <20141122141139.78927371@gandalf.local.home> <20141122195926.GA1707@wfg-t540p.sh.intel.com> <20141122165302.3df22b18@gandalf.local.home> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20141122165302.3df22b18@gandalf.local.home> User-Agent: Mutt/1.5.23 (2014-03-12) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Sat, Nov 22, 2014 at 04:53:02PM -0500, Steven Rostedt wrote: > On Sat, 22 Nov 2014 11:59:26 -0800 > Fengguang Wu wrote: > > > > > > Is it possible for your tests to see if a branch was rebased before > > > posting bug reports to LKML? This gave me a scare that my new code had > > > the same brain fart in it. > > > > Sure and good idea! I'll implement it right now. > > > > It should also make sure that the rebase does not include the offending > commit before not sending out the email. A rebase alone may still have > the bug. > > And if you really want to be fancy, you can check to see if the code > itself changed. The attached script takes two git releases (SHA1s, > tags, etc) or one git release and uses HEAD, and finds where the two > are different. Not just by commit id, but also by code. I use this > because I'll modify a branch to add an acked-by and not change the code, > but still want to know the last commit that I tested. It may not have > the same commit id. I definitely would like to be fancy. I think I'll not only check if the branch has been rebased, but also test the rebased branch out to make sure the problem is still there. I've been doing this check for the build bisects. When doing so, it no longer matters how the rebased branch differs from the original branch in source code level. For example, even the simple rebase to a new RC base might make a problem disappear. Thanks, Fengguang