From: Arnd Bergmann <arnd@arndb.de>
To: Nick Bowler <nbowler@elliptictech.com>
Subject: Re: [PATCH 0/3] Replace bmp085 with bmp18x
Date: Wed, 29 Feb 2012 21:26:55 +0000
User-Agent: KMail/1.12.2 (Linux/3.3.0-rc1; KDE/4.3.2; x86_64; ; )
Cc: Eric Andersson <eric.andersson@unixphere.com>,
        linux-kernel@vger.kernel.org, gregkh@linuxfoundation.org,
        christoph.mair@gmail.com, stefan.nilsson@unixphere.com,
        zhengguang.guo@bosch-sensortec.com, peter.moeller@cn.bosch.com,
        Jonathan Cameron <jic23@cam.ac.uk>
References: <1330545489-31169-1-git-send-email-eric.andersson@unixphere.com> <201202292018.45557.arnd@arndb.de> <20120229203743.GA21671@elliptictech.com>
In-Reply-To: <20120229203743.GA21671@elliptictech.com>
MIME-Version: 1.0
Content-Type: Text/Plain;
  charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
Message-Id: <201202292126.56083.arnd@arndb.de>
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 2135
Lines: 47

On Wednesday 29 February 2012, Nick Bowler wrote:
> On 2012-02-29 20:18 +0000, Arnd Bergmann wrote:
> > On Wednesday 29 February 2012, Eric Andersson wrote:
> > > This patch-set replaces the BMP085 driver with a driver for Bosch Sensortec's new
> > > generation of pressure sensors called BMP18x. These pressure sensors can be
> > > connected on I2C but also on SPI as a variant, so the driver implements both.
> > > Register-wise they are fully compatible with the older BMP085, so the driver will
> > > support those chips as well.
> > > 
> > > The driver is based on bmp085.c by Christoph Mair.
> > 
> > Hmm,  the implementation looks fine, but I don't think we should add support
> > for new devices with a one-off interface that we know is getting replaced by
> > something generic.
> 
> Notwithstanding the value of using newer generic interfaces, from what I
> can tell, there are no new sysfs attributes added by this driver over
> the one it replaces.  If this driver is going to supersede the old one,
> it has to retain the existing interface.  Otherwise, we must keep both
> drivers in the tree to avoid regressions.

Well, the driver is presented as a new one ;-)

I also think it would be nicer generally speaking to do this as incremental
updates, which could be as simple as

1. code cleanups in bmp085 without functional changes
2. add support for bmp18x
3. rename bmp085 to bmp18x, with no other changes

That would take care of the argument of adding a new driver with the "wrong"
interface as well as make the history of that driver cleaner, e.g. for
purposes of walking through the patches using 'git log --follow'.

> > Why is this not using IIO for its user interface?
> 
> Because IIO is currently in staging, and this driver is not?

Yes, this is of course a big limitation still and I agree that we
cannot hold up progress on drivers that don't use it yet.

	Arnd
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/