From mboxrd@z Thu Jan 1 00:00:00 1970 From: Josh Cartwright Subject: Re: [PATCH v2 2/6] rtc: pm8xxx: use regmap API for register accesses Date: Mon, 10 Mar 2014 16:35:24 -0500 Message-ID: <20140310213524.GL18529@joshc.qualcomm.com> References: <1394477089-3996-1-git-send-email-joshc@codeaurora.org> <1394477089-3996-3-git-send-email-joshc@codeaurora.org> <20140310134741.e5b97d12d47e1b990501f8bc@linux-foundation.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Content-Disposition: inline In-Reply-To: <20140310134741.e5b97d12d47e1b990501f8bc@linux-foundation.org> Sender: linux-kernel-owner@vger.kernel.org To: Andrew Morton Cc: Alessandro Zummo , linux-arm-msm@vger.kernel.org, rtc-linux@googlegroups.com, linux-kernel@vger.kernel.org, Mark Brown List-Id: linux-arm-msm@vger.kernel.org On Mon, Mar 10, 2014 at 01:47:41PM -0700, Andrew Morton wrote: > On Mon, 10 Mar 2014 13:44:45 -0500 Josh Cartwright wrote: > > Now that the parent mfd driver has been made to work again, and has been > > reworked to create a regmap instance intended for it's children to use, > > rework the pm8xxx driver to use the regmap API for it's register > > accesses. > > Do we need a > > depends on REGMAP > > in Kconfig? Good question. This driver already gets the REGMAP dependency through it's parent (MFD_PM8XXX), which does 'select REGMAP'. I recall there being some history as to why individual drivers shouldn't explicitly depend on REGMAP, as it's the parent drivers/subsystem responsibility to select it when needed, however I can't recall details. Perhaps it's due to REGMAP not being a user-selectable option. (FWIW, I did force a build of this driver with !REGMAP, and it builds just fine with the regmap_write()/_read() stubs, but of course it wouldn't do anything meaningful at runtime) Thanks, Josh -- Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, hosted by The Linux Foundation