Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S934625AbcCPFEP (ORCPT ); Wed, 16 Mar 2016 01:04:15 -0400 Received: from mail-io0-f182.google.com ([209.85.223.182]:34123 "EHLO mail-io0-f182.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751672AbcCPFEM (ORCPT ); Wed, 16 Mar 2016 01:04:12 -0400 MIME-Version: 1.0 In-Reply-To: <56E7BEE8.2000004@samsung.com> References: <1457951097-27079-1-git-send-email-j.anaszewski@samsung.com> <20160315103613.3eb6347e@canb.auug.org.au> <56E7BEE8.2000004@samsung.com> Date: Tue, 15 Mar 2016 22:04:10 -0700 X-Google-Sender-Auth: kqu5Fe2sK5v4Pwh9OU1epJgVf-k Message-ID: Subject: Re: [GIT PULL] LED subsystem updates for 4.6 From: Linus Torvalds To: Jacek Anaszewski Cc: Stephen Rothwell , Linux Kernel Mailing List , Linux LED Subsystem , amitoj1606@gmail.com, ao2@ao2.it, drivshin@allworx.com, Geert Uytterhoeven , hkallweit1@gmail.com, stefan.wahren@i2se.com, Wei Yongjun Content-Type: text/plain; charset=UTF-8 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1090 Lines: 27 On Tue, Mar 15, 2016 at 12:51 AM, Jacek Anaszewski wrote: > > I just wanted to make sure that no unexpected problem has occurred > after rebasing onto 4.5 release. Is it in some way more advantageous to > base a pull request on rc7, than on a final release? I'd rather see the pull request based on whatever it has been tested on, and just keep it that way. Any rebasing will inevitably mean that you are basically throwing all previous testing out the window (or at least make it dubious). Rebasing also makes it much harder to see the history (for example, compare it against previous linux-next trees), so the rule really should be that you should never rebase unless you have a major reason to do so. So for example, if you actually find a problem, and you notice that that problem comes not from your own changes, but from the base you picked - *then* you'd want to rebase to a more stable base. But a rebase "just because" is not a good idea. I'll pull it, since it looks fairly harmless, but basically please don't do that again. Linus