Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752198AbaKQMTu (ORCPT ); Mon, 17 Nov 2014 07:19:50 -0500 Received: from bear.ext.ti.com ([192.94.94.41]:35667 "EHLO bear.ext.ti.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751978AbaKQMTr (ORCPT ); Mon, 17 Nov 2014 07:19:47 -0500 Message-ID: <5469E793.80909@ti.com> Date: Mon, 17 Nov 2014 17:48:27 +0530 From: Vignesh R User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:31.0) Gecko/20100101 Thunderbird/31.2.0 MIME-Version: 1.0 To: Richard Cochran CC: Rob Herring , Pawel Moll , Mark Rutland , Ian Campbell , Kumar Gala , Benoit Cousson , Tony Lindgren , Russell King , Jonathan Cameron , Hartmut Knaack , Dmitry Torokhov , Lee Jones , Sebastian Andrzej Siewior , Lars-Peter Clausen , Peter Meerwald , Samuel Ortiz , Felipe Balbi , Brad Griffis , Sanjeev Sharma , Paul Gortmaker , Jan Kardell , , , , , , Subject: Re: [PATCH v4 2/6] input: touchscreen: ti_am335x_tsc: Remove udelay in interrupt handler References: <1415941651-28962-1-git-send-email-vigneshr@ti.com> <1415941651-28962-3-git-send-email-vigneshr@ti.com> <20141115192739.GA18766@localhost.localdomain> <54697919.5030302@ti.com> <20141117082337.GA3946@localhost.localdomain> In-Reply-To: <20141117082337.GA3946@localhost.localdomain> Content-Type: text/plain; charset="windows-1252" Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Monday 17 November 2014 01:53 PM, Richard Cochran wrote: > On Mon, Nov 17, 2014 at 09:57:05AM +0530, Vignesh R wrote: >> My patches are based on v3.18rc2. I tested my patches on am335x-evm >> using tslib. > > No beaglebone + cape testing? > >> Please explain "touch is broken"? What is the behaviour of TSC? > > With plain v3.17 plus your series, the cursor is almost never near the > pen. Mostly it jitters around the right hand edge. > > My customer had already changed the step delay (I think by trial and > error, not sure) in order to get the cursor near the pen. I ported > this change onto your series (see patch, below), but still the pen up > event causes a huge cursor jump. > > (Again, I did solve the pen up issue, but in a totally different > way. I never posted the fix, because I could not be sure that it would > work on a wide variety of boards.) > >> Which ADC channels are being used for TSC? > > &tscadc { > status = "okay"; > tsc { > ti,wires = <4>; > ti,x-plate-resistance = <300>; > ti,coordinate-readouts = <5>; > ti,wire-config = <0x00 0x11 0x23 0x32>; > }; What is the ti,charge-delay value you have tried? Try setting it to 0xB000 and experiment. > > adc { > ti,adc-channels = <5 6 7>; > }; > }; > > So for this particular design, your series really does not help, not > even a little. You did not test the series on many boards. I am > concerned that this series only works on the one board you did test, > and that it may break functionality on other people's boards. > I just have one am335x-evm available. I will search for a beaglebone cape and test the patches. Let Sebastian, Johannes come back with their results. If the series works on multiple boards then, I believe, this series can go in. Regards Vignesh > Thanks, > Richard > > --- > diff --git a/drivers/input/touchscreen/ti_am335x_tsc.c b/drivers/input/touchscreen/ti_am335x_tsc.c > index b84493f..77a4883 100644 > --- a/drivers/input/touchscreen/ti_am335x_tsc.c > +++ b/drivers/input/touchscreen/ti_am335x_tsc.c > @@ -148,7 +148,7 @@ static void titsc_step_config(struct titsc *ts_dev) > end_step = first_step + tsc_steps; > for (i = end_step - ts_dev->coordinate_readouts; i < end_step; i++) { > titsc_writel(ts_dev, REG_STEPCONFIG(i), config); > - titsc_writel(ts_dev, REG_STEPDELAY(i), STEPCONFIG_OPENDLY); > + titsc_writel(ts_dev, REG_STEPDELAY(i), STEPCONFIG_OPENDLY | STEPDELAY_SAMPLE(20)); > } > > config = 0; > @@ -172,7 +172,7 @@ static void titsc_step_config(struct titsc *ts_dev) > end_step = first_step + ts_dev->coordinate_readouts; > for (i = first_step; i < end_step; i++) { > titsc_writel(ts_dev, REG_STEPCONFIG(i), config); > - titsc_writel(ts_dev, REG_STEPDELAY(i), STEPCONFIG_OPENDLY); > + titsc_writel(ts_dev, REG_STEPDELAY(i), STEPCONFIG_OPENDLY | STEPDELAY_SAMPLE(20)); > } > > /* Make CHARGECONFIG same as IDLECONFIG */ > @@ -188,13 +188,13 @@ static void titsc_step_config(struct titsc *ts_dev) > STEPCONFIG_INP(ts_dev->inp_xp); > titsc_writel(ts_dev, REG_STEPCONFIG(end_step), config); > titsc_writel(ts_dev, REG_STEPDELAY(end_step), > - STEPCONFIG_OPENDLY); > + STEPCONFIG_OPENDLY | STEPDELAY_SAMPLE(20)); > > end_step++; > config |= STEPCONFIG_INP(ts_dev->inp_yn); > titsc_writel(ts_dev, REG_STEPCONFIG(end_step), config); > titsc_writel(ts_dev, REG_STEPDELAY(end_step), > - STEPCONFIG_OPENDLY); > + STEPCONFIG_OPENDLY | STEPDELAY_SAMPLE(20)); > > /* The steps end ... end - readouts * 2 + 2 and bit 0 for TS_Charge */ > stepenable = 1; > -- 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/