From mboxrd@z Thu Jan 1 00:00:00 1970 From: Vladimir Zapolskiy Subject: Re: [PATCH] i2c: pnx: fix warnings caused by enabling unprepared clock Date: Tue, 20 Oct 2015 13:58:28 +0300 Message-ID: <56261E54.7040808@mleia.com> References: <1445107947-2888-1-git-send-email-vz@mleia.com> <5625343A.90407@mleia.com> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 7bit Return-path: Received: from li271-223.members.linode.com ([178.79.152.223]:38253 "EHLO mail.mleia.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750805AbbJTK6a (ORCPT ); Tue, 20 Oct 2015 06:58:30 -0400 In-Reply-To: Sender: linux-i2c-owner@vger.kernel.org List-Id: linux-i2c@vger.kernel.org To: Shubhrajyoti Datta Cc: Vitaly Wool , Wolfram Sang , linux-i2c@vger.kernel.org On 20.10.2015 11:54, Shubhrajyoti Datta wrote: > On Mon, Oct 19, 2015 at 11:49 PM, Vladimir Zapolskiy wrote: >> On 19.10.2015 19:21, Shubhrajyoti Datta wrote: >>> On Sun, Oct 18, 2015 at 12:22 AM, Vladimir Zapolskiy wrote: >>>> If common clock framework is configured, the driver generates a warning, >>>> which is fixed by this change: >>> Maybe just describe the issue and the fix. >>> Is it just a warning or the clk enable doesn't work ? >>> >>> I feel the crash log in the commit msg is not very informative. >> >> It is not a crash, it is a WARN_ON(1) backtrace. > > ok > >> >> The backtrace is informative enough IMHO, because if you check the code >> around given drivers/clk/clk.c:727 you may find the rootcause, the >> WARN_ON() and that the clock is not enabled as a result. >> >> CLK_COMMON selects HAVE_CLK_PREPARE and all drivers used on platforms >> with common clock framework must have clk_prepare/clk_unprepare, this >> information is omitted as well-known. >> >> But if you insist, I will improve the commit message, I'm interested in >> applying this trivial fix as soon as possible, then concentrate on doing >> something more fascinating. > > My request will be to describe the issue. > Like the probe fails or you access registers with clocks not enabled > > Or it is a warning fix and the controller works etc. If I change initial description to the one below, would you be satisfied with it? The driver can not be used on a platform with common clock framework until clk_prepare/clk_unprepare calls are added, otherwise clk_enable calls will fail and a warning is generated. -- With best wishes, Vladimir