From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Message-ID: <9e4733910712101454l39285585j55e4d72ae4329cf9@mail.gmail.com> Date: Mon, 10 Dec 2007 17:54:42 -0500 From: "Jon Smirl" To: "Grant Likely" Subject: Re: [PATCH] Addition to the i2c series, copy the ppc mpc-i2c driver before changing it on powerpc In-Reply-To: MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 References: <9e4733910712101420o15b5e908jc7291a0086832467@mail.gmail.com> Cc: PowerPC dev list List-Id: Linux on PowerPC Developers Mail List List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , On 12/10/07, Grant Likely wrote: > On 12/10/07, Jon Smirl wrote: > > Copy mpc-i2c to preserve support for ARCH=ppc and allow changes on ARCH=powerpc > > > > Temporarily copy the mpc-i2c driver to continue support for the ppc > > architecture until it is removed in mid-2008. This file should be > > deleted as part of ppc's final removal. > > > > Signed-off-by: Jon Smirl > > For the record; I'm not fond of this approach. Supporting both bus > bindings in the single driver is simple and results in less code churn > when arch/ppc is removed, and encourages separation between the driver > proper and the bus bindings which is just a good idea for all drivers > in general. But it also triggers a testing burden on a bunch of hardware that I don't own. By copying off the known working ppc driver the testing burden is avoided. A subject for later discussion is whether platform bus should even exist when of_platform_bus is in use. I have removed platform_bus for the mpc5200 in my local builds. Removing platform bus exposed a bunch of junk from other platofrms that had inadvertently accumulated into the mpc5200 build. -- Jon Smirl jonsmirl@gmail.com