From mboxrd@z Thu Jan 1 00:00:00 1970 From: Kevin Hilman Subject: Re: [PATCH 00/10 v3] omap3: pm: introduce support for 3630 OPPs Date: Wed, 25 Nov 2009 07:24:15 -0800 Message-ID: <877htemxuo.fsf@deeprootsystems.com> References: <1259122159-1583-1-git-send-email-nm@ti.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Received: from mail-px0-f180.google.com ([209.85.216.180]:58959 "EHLO mail-px0-f180.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753306AbZKYPYL (ORCPT ); Wed, 25 Nov 2009 10:24:11 -0500 Received: by pxi10 with SMTP id 10so5823326pxi.33 for ; Wed, 25 Nov 2009 07:24:17 -0800 (PST) In-Reply-To: <1259122159-1583-1-git-send-email-nm@ti.com> (Nishanth Menon's message of "Tue\, 24 Nov 2009 22\:09\:09 -0600") Sender: linux-omap-owner@vger.kernel.org List-Id: linux-omap@vger.kernel.org To: Nishanth Menon Cc: linux-omap Nishanth Menon writes: > Hi, > Here is V3 of the patch series. What changed in V3: > * Major rewrite of OPP APIs. This includes changes from offline > discussions with lots of folks and taking in l-o comments - all > patches have changed as a result. > * rebased to latest lo-pm > > V2: http://marc.info/?l=linux-omap&m=125809232732700&w=2 > V1: http://marc.info/?l=linux-omap&m=125800488923479&w=2 > > NOTE: After a bunch of frustrating efforts, current OPP accessor > implementation is constrained by two files: > smartreflex.c and resource34xx.c. I agree that the current opp > accessor implementation leaves lot to be desired, but the nature of > struct omap_opp cannot change unless we essentially rewrite the > implementation of the mentioned files. > > Ref: http://elinux.org/OMAP_Power_Management#Future_directions > > I propose we could take this in in three stages: > Stage 1: Introduce this series in as starters if there are no major > issues > Stage 2: We cleanup/replace resource34xx and smartreflex.c > Stage 3: Go through a second round of optimization/cleanup of opp > accessors implementation (this time for good hopefully) I'm open to this 3-step approach assuming that stage 2 and 3 remain a priority. This "stage 1" series will not go upstream. I will only push upstream after getting to stage 3. After some review and discussion, I can pull Stage 1 into the PM branch for broader testing and as a base for more people to help on stage 2 and stage 3. Review comments on individual patches to follow. Kevin