From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Rafael J. Wysocki" Subject: Re: [PATCHv2 3/4] clk: new basic clk type for fractional divider Date: Sat, 17 May 2014 00:38:05 +0200 Message-ID: <1454520.Q33DvSH0ns@vostro.rjw.lan> References: <1400161226-24067-1-git-send-email-heikki.krogerus@linux.intel.com> <1400161226-24067-4-git-send-email-heikki.krogerus@linux.intel.com> <20140515165349.9521.75672@quantum> Mime-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: 7Bit Return-path: Received: from v094114.home.net.pl ([79.96.170.134]:51519 "HELO v094114.home.net.pl" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with SMTP id S1755314AbaEPWVO (ORCPT ); Fri, 16 May 2014 18:21:14 -0400 In-Reply-To: <20140515165349.9521.75672@quantum> Sender: linux-acpi-owner@vger.kernel.org List-Id: linux-acpi@vger.kernel.org To: Mike Turquette Cc: Heikki Krogerus , Mika Westerberg , Jin Yao , Li Aubrey , Andy Shevchenko , linux-acpi@vger.kernel.org, linux-kernel@vger.kernel.org On Thursday, May 15, 2014 09:53:49 AM Mike Turquette wrote: > Quoting Heikki Krogerus (2014-05-15 06:40:25) > > Fractional divider clocks are fairly common. This adds basic > > type for them. > > > > Signed-off-by: Heikki Krogerus > > Taken into clk-next. > > Just FYI, there was some talk at Embedded Linux Conference on providing > a better abstraction layer for some of these "basic" clock types. This > abstraction would allow the basic clock types to implement the > machine-agnostic logic (e.g. an incoming rate is divided by a value) and > then platforms and drivers could plug in the machine-specific parts > (e.g. divider is made up of m/n, or divider is power-of-two, or divider > is a simple integer with min == 1 and max == 5). > > All of that is to say that in time this fractional divider could go away > once the abstraction layer allows us to fold the m/n divider stuff into > a core divider implementation. > > Nothing wrong with the patch for now, so I've taken it for 3.16. Well, OK, but I guess [4/4] depends on it and [4/4] also depends on [1-2/4], so how you're proposing to resolve this? Rafael