From mboxrd@z Thu Jan 1 00:00:00 1970 From: David Brownell Subject: Re: Common code for TSC 2101 and 2102 Date: Mon, 13 Aug 2007 22:34:41 -0700 Message-ID: <200708132234.41538.david-b@pacbell.net> References: <200708102004.59747.david-b@pacbell.net> <20070813070512.GC13948@atomide.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <20070813070512.GC13948@atomide.com> Content-Disposition: inline List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: linux-omap-open-source-bounces@linux.omap.com Errors-To: linux-omap-open-source-bounces@linux.omap.com To: Tony Lindgren Cc: Linux OMAP List-Id: linux-omap@vger.kernel.org On Monday 13 August 2007, Tony Lindgren wrote: > * David Brownell [070810 20:05]: > > On Friday 10 August 2007, Tony Lindgren wrote: > > > I'll be pushing Andrzej's patch today. > > > > I saw it ... it was after I've finally started to dig > > out from under my pile of patches going upstream, and > > even fired up an OMAP build again! :) > > Cool :) Following this will be several patches, two related to this driver: - H4 board support, using TSC2101 (not TSC2102) - A biggish tsc210x cleanup patch, with some bugfixes The main thing I'd call a bug in the tsc210x code is the hwmon support not reporting milliVolts. Easily enough solved after passing some more info down the hwmon code via platform_data: actual VREF, chip type, precision. For boards using external VREF, that value should be in the tsc210x chip platform data. (And Andrzej, I'm curious why you settled on VREF=1.25V vs 2.50V ... not that it's necessarily an issue, just "why". A comment in the code might be good.) Somewhat related, I think the TSC setup code should likely hold off until it can fetch calibration data from EEPROM or somesuch ... temperature readings are not all that useful until that data is available. Alternatively, do like I did in ads7846: punt it all to userspace, just provide the raw milliVolt readings and expect someone else to compute temperature. I only did really basic touchscreen testing, to verify that data came out of /dev/input/event1. Not to check that it was "good" data. I suspect there's some work yet to be done there. - Dave