linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: "Henrik Rydberg" <rydberg@euromail.se>
To: Kevin McNeely <kev@cypress.com>
Cc: Dmitry Torokhov <dmitry.torokhov@gmail.com>,
	David Brown <davidb@codeaurora.org>,
	Trilok Soni <tsoni@codeaurora.org>,
	Samuel Ortiz <sameo@linux.intel.com>,
	Eric Miao <eric.y.miao@gmail.com>,
	Mike Frysinger <vapier@gentoo.org>,
	Alan Cox <alan@linux.intel.com>,
	linux-input@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: Re: [v3 1/3] 1/3 Touchscreen: Cypress TTSP G3 MTDEV Core Driver
Date: Fri, 31 Dec 2010 12:53:23 +0100	[thread overview]
Message-ID: <20101231115323.GA9709@polaris.bitmath.org> (raw)
In-Reply-To: <1293650268-1561-1-git-send-email-kev@cypress.com>

Hi Kevin,

Thanks for the changes, it looks much better now. Some comments on the MT
part inline.

> diff --git a/drivers/input/touchscreen/Kconfig b/drivers/input/touchscreen/Kconfig
> index 06ea8da..7d886bc 100644
> --- a/drivers/input/touchscreen/Kconfig
> +++ b/drivers/input/touchscreen/Kconfig
> @@ -124,6 +124,11 @@ config TOUCHSCREEN_CY8CTMG110
>  	  To compile this driver as a module, choose M here: the
>  	  module will be called cy8ctmg110_ts.
>  
> +config TOUCHSCREEN_CYTTSP_CORE
> +	bool "Cypress TTSP touchscreen core"
> +	help
> +	  Always activated for Cypress TTSP touchscreen
> +

Tristate please.

[...]
> +
> +struct cyttsp_tch {
> +	u16 x;
> +	u16 y;
> +	u8 z;
> +};

Does not appear to be used anywhere.

[...]
> +struct cyttsp {
> +	struct device *dev;
> +	int irq;
> +	struct input_dev *input;
> +	struct mutex mutex;

Is this really used?

> +	char phys[32];
> +	const struct bus_type *bus_type;
> +	const struct cyttsp_platform_data *platform_data;
> +	struct cyttsp_bus_ops *bus_ops;
> +	struct cyttsp_xydata xy_data;
> +	struct cyttsp_bootloader_data bl_data;
> +	struct cyttsp_sysinfo_data sysinfo_data;
> +	struct completion bl_ready;
> +	enum cyttsp_powerstate power_state;
> +};
> +
> +static const u8 bl_command[] = {
> +	CY_BL_FILE0, CY_BL_CMD, CY_BL_EXIT,
> +	CY_BL_KEY0, CY_BL_KEY1, CY_BL_KEY2,
> +	CY_BL_KEY3, CY_BL_KEY4, CY_BL_KEY5,
> +	CY_BL_KEY6, CY_BL_KEY7
> +};

If the CY_BL* values are only used here, how about putting the numbers here directly?

[...]
> +static int ttsp_read_block_data(struct cyttsp *ts, u8 command,
> +	u8 length, void *buf)
> +{
> +	int retval;
> +	int tries;
> +
> +	if (!buf || !length)
> +		return -EIO;

-EINVAL?

> +
> +	for (tries = 0, retval = -1;
> +		tries < CY_NUM_RETRY && (retval < 0);
> +		tries++)
> +		retval = ts->bus_ops->read(ts->bus_ops, command, length, buf);
> +
> +	return retval;
> +}

[...]
> +/* process current touches */
> +static int cyttsp_xy_worker(struct cyttsp *ts)
> +{
> +	u8 num_cur_tch = 0;

Does not appear to need initialization.

> +	u16 x;
> +	u16 y;
> +	u8 z;

Perhaps x[2], y[2] etc here.

> +
> +	/* Get touch data from CYTTSP device */
> +	if (ttsp_read_block_data(ts,
> +		CY_REG_BASE, sizeof(struct cyttsp_xydata), &ts->xy_data))
> +		return 0;
> +
> +	/* touch extension handling */
> +	if (ttsp_tch_ext(ts, &ts->xy_data))
> +		return 0;
> +
> +	/* provide flow control handshake */
> +	if (ts->platform_data->use_hndshk)
> +		if (cyttsp_hndshk(ts, ts->xy_data.hst_mode))
> +			return 0;
> +
> +	/* determine number of currently active touches */
> +	num_cur_tch = GET_NUM_TOUCHES(ts->xy_data.tt_stat);
> +
> +	/* check for any error conditions */
> +	if (ts->power_state == CY_IDLE_STATE)
> +		return 0;
> +	else if (GET_BOOTLOADERMODE(ts->xy_data.tt_mode)) {
> +		return -1;
> +	} else if (IS_LARGE_AREA(ts->xy_data.tt_stat) == 1) {
> +		/* terminate all active tracks */
> +		num_cur_tch = 0;
> +		dev_dbg(ts->dev, "%s: Large area detected\n", __func__);
> +	} else if (num_cur_tch > 2) {
> +		/* terminate all active tracks */
> +		num_cur_tch = 0;
> +		dev_dbg(ts->dev, "%s: Num touch error detected\n", __func__);
> +	} else if (IS_BAD_PKT(ts->xy_data.tt_mode)) {
> +		/* terminate all active tracks */
> +		num_cur_tch = 0;
> +		dev_dbg(ts->dev, "%s: Invalid buffer detected\n", __func__);
> +	}
> +
> +	/* send touches */
> +	if (!num_cur_tch)
> +		/* terminate previous active touch */
> +		input_mt_sync(ts->input);
> +
> +	if (num_cur_tch) {
> +		/* send touch 1 */
> +		/*
> +		 * If there is only one current active touch,
> +		 * it will be reported in the touch 1 regardless
> +		 * if it was reported in the touch 2 previously
> +		 */

Perhaps "touch" should be "track" or "slot" here?

> +		x = (ts->xy_data.tch1_xhi << 8) + ts->xy_data.tch1_xlo;
> +		y = (ts->xy_data.tch1_yhi << 8) + ts->xy_data.tch1_ylo;
> +		z = ts->xy_data.tch1_z;

Perhaps x[0] = ... here

> +		input_report_abs(ts->input, ABS_MT_POSITION_X, x);
> +		input_report_abs(ts->input, ABS_MT_POSITION_Y, y);
> +		input_report_abs(ts->input, ABS_MT_TOUCH_MAJOR, z);
> +		input_mt_sync(ts->input);

Then these could still be collected in a loop.

> +	}
> +
> +	if (num_cur_tch > 1) {
> +		/* send touch 2 */
> +		x = (ts->xy_data.tch2_xhi << 8) + ts->xy_data.tch2_xlo;
> +		y = (ts->xy_data.tch2_yhi << 8) + ts->xy_data.tch2_ylo;
> +		z = ts->xy_data.tch2_z;

And x[1] = ...

> +		input_report_abs(ts->input, ABS_MT_POSITION_X, x);
> +		input_report_abs(ts->input, ABS_MT_POSITION_Y, y);
> +		input_report_abs(ts->input, ABS_MT_TOUCH_MAJOR, z);
> +		input_mt_sync(ts->input);
> +	}

Ditto.

> +
> +	input_sync(ts->input);
> +
> +	return 0;
> +}

If the in-kernel tracking module was in place, would you consider switching to the slots protocol?

Thanks,
Henrik

  parent reply	other threads:[~2010-12-31 11:58 UTC|newest]

Thread overview: 70+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
     [not found] <Kevin McNeely <kev@cypress.com>
2010-07-12 20:56 ` [PATCH] i2c: cyttsp i2c touchscreen driver init submit Kevin McNeely
2010-07-13  2:34   ` Christoph Fritz
2010-08-04 16:30     ` Kevin McNeely
2010-07-13  6:48   ` Henrik Rydberg
2010-08-04 16:38     ` Kevin McNeely
2010-07-13  7:31   ` Trilok Soni
2010-07-13  7:55     ` Dmitry Torokhov
2010-07-13  8:42       ` Trilok Soni
2010-07-22 10:33         ` Trilok Soni
2010-07-27 15:20           ` Kevin McNeely
2010-08-04 17:27       ` Kevin McNeely
2010-07-19  9:28     ` Jean Delvare
2010-08-04 17:22     ` Kevin McNeely
2010-08-05 18:12 ` [PATCH] i2c: cyttsp i2c and spi " Kevin McNeely
2010-08-05 20:45   ` Trilok Soni
2010-08-05 21:07     ` Dmitry Torokhov
2010-08-07  0:39       ` Kevin McNeely
2010-08-07  0:52     ` Kevin McNeely
2010-08-05 23:06   ` Henrik Rydberg
2010-08-07  0:32     ` Kevin McNeely
2010-08-07  0:49       ` Henrik Rydberg
2010-08-10  0:51         ` Kevin McNeely
2010-08-06  9:06   ` Trilok Soni
2010-08-10  0:49     ` Kevin McNeely
2010-11-09 18:25 ` [PATCH] touchscreen: Cypress TTSP G3 MTDEV Core Driver Kevin McNeely
2010-11-15 16:46   ` Henrik Rydberg
2010-11-19 17:39     ` Kevin McNeely
2010-12-01  7:22       ` Trilok Soni
2010-12-01 14:38         ` Henrik Rydberg
2010-12-01 23:59           ` Kevin McNeely
2010-12-02  0:01             ` Henrik Rydberg
2010-12-02  0:34   ` Dmitry Torokhov
2010-11-09 18:25 ` [PATCH] i2c: Cypress TTSP G3 MTDEV I2C Device Driver Kevin McNeely
2010-11-09 18:25 ` [PATCH] spi: Cypress TTSP G3 MTDEV SPI " Kevin McNeely
2010-12-04  2:06 ` [v2] touchscreen Cypress TTSP G3 MTDEV Core Driver Kevin McNeely
2010-12-05  9:11   ` Henrik Rydberg
2010-12-04  2:06 ` [v2] 2/3 i2c: Cypress TTSP G3 MTDEV I2C Device Driver Kevin McNeely
2010-12-04  2:06 ` [v2] 3/3 spi: Cypress TTSP G3 MTDEV SPI " Kevin McNeely
2010-12-29 19:17 ` [v3 1/3] 1/3 Touchscreen: Cypress TTSP G3 MTDEV Core Driver Kevin McNeely
2010-12-30  6:04   ` Shubhrajyoti Datta
2011-01-05  0:45     ` Kevin McNeely
2010-12-31 11:53   ` Henrik Rydberg [this message]
2010-12-31 12:55     ` Trilok Soni
2010-12-31 13:58       ` Henrik Rydberg
2011-01-03  9:44         ` Trilok Soni
2011-01-03 17:03     ` Kevin McNeely
2011-01-03 18:45       ` Henrik Rydberg
2011-01-03 20:50         ` Kevin McNeely
2011-01-04  1:50   ` Hong Liu
2011-01-05  0:38     ` Kevin McNeely
2010-12-29 19:17 ` [v3 2/3] 2/3 i2c: Cypress TTSP G3 MTDEV I2C Device Driver Kevin McNeely
2011-01-04  1:45   ` Hong Liu
2011-01-05  0:37     ` Kevin McNeely
2010-12-29 19:17 ` [v3 3/3] 3/3 spi: Cypress TTSP G3 MTDEV SPI " Kevin McNeely
2011-01-05  0:54 ` [v4 1/3] 1/3 Touchscreen: Cypress TTSP G3 Core Driver Kevin McNeely
2011-01-05  8:59   ` Henrik Rydberg
2011-01-05 17:07     ` Kevin McNeely
2011-01-05 17:34       ` Henrik Rydberg
2011-01-10 19:27         ` Kevin McNeely
2011-01-10 21:11           ` Dmitry Torokhov
2011-01-10 21:17             ` Kevin McNeely
2011-02-24 18:31         ` Kevin McNeely
2011-02-27 12:34           ` Henrik Rydberg
2011-01-05  0:54 ` [v4 2/3] 2/3 i2c: Cypress TTSP G3 I2C Device Driver Kevin McNeely
2011-01-05  0:54 ` [v4 3/3] 3/3 spi: Cypress TTSP G3 SPI " Kevin McNeely
2011-01-12 18:45   ` Dmitry Torokhov
2011-01-12 19:02     ` Kevin McNeely
2011-01-20 11:10       ` Trilok Soni
2011-01-21  9:27         ` Dmitry Torokhov
2011-01-21 22:14           ` Kevin McNeely

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20101231115323.GA9709@polaris.bitmath.org \
    --to=rydberg@euromail.se \
    --cc=alan@linux.intel.com \
    --cc=davidb@codeaurora.org \
    --cc=dmitry.torokhov@gmail.com \
    --cc=eric.y.miao@gmail.com \
    --cc=kev@cypress.com \
    --cc=linux-input@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=sameo@linux.intel.com \
    --cc=tsoni@codeaurora.org \
    --cc=vapier@gentoo.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).