* [PATCH] Input: add touchscreen driver for MELFAS MCS-5000 controller
@ 2009-05-11 12:28 Joonyoung Shim
2009-05-12 2:09 ` Dmitry Torokhov
0 siblings, 1 reply; 7+ messages in thread
From: Joonyoung Shim @ 2009-05-11 12:28 UTC (permalink / raw)
To: linux-input; +Cc: dmitry.torokhov, kyungmin.park, m.szyprowski
The MELPAS MCS-5000 is the touchscreen controller. The overview of this
controller can see at the following website:
http://www.melfas.com/product/product01.asp?k_r=eng_
This patch supports only the i2c interface.
Signed-off-by: Joonyoung Shim <jy0922.shim@samsung.com>
---
drivers/input/touchscreen/Kconfig | 12 +
drivers/input/touchscreen/Makefile | 1 +
drivers/input/touchscreen/mcs5000_ts.c | 377 ++++++++++++++++++++++++++++++++
include/linux/i2c/mcs5000_ts.h | 11 +
4 files changed, 401 insertions(+), 0 deletions(-)
create mode 100644 drivers/input/touchscreen/mcs5000_ts.c
create mode 100644 include/linux/i2c/mcs5000_ts.h
diff --git a/drivers/input/touchscreen/Kconfig b/drivers/input/touchscreen/Kconfig
index 82c388e..ac5ebb4 100644
--- a/drivers/input/touchscreen/Kconfig
+++ b/drivers/input/touchscreen/Kconfig
@@ -481,4 +481,16 @@ config TOUCHSCREEN_TSC2007
To compile this driver as a module, choose M here: the
module will be called tsc2007.
+config TOUCHSCREEN_MCS5000
+ tristate "MELFAS MCS-5000 touchscreen"
+ depends on I2C
+ help
+ Say Y here if you have the MELFAS MCS-5000 touchscreen controller
+ chip in your system.
+
+ If unsure, say N.
+
+ To compile this driver as a module, choose M here: the
+ module will be called mcs5000_ts.
+
endif
diff --git a/drivers/input/touchscreen/Makefile b/drivers/input/touchscreen/Makefile
index bef7415..5e4a9dd 100644
--- a/drivers/input/touchscreen/Makefile
+++ b/drivers/input/touchscreen/Makefile
@@ -16,6 +16,7 @@ obj-$(CONFIG_TOUCHSCREEN_GUNZE) += gunze.o
obj-$(CONFIG_TOUCHSCREEN_ELO) += elo.o
obj-$(CONFIG_TOUCHSCREEN_FUJITSU) += fujitsu_ts.o
obj-$(CONFIG_TOUCHSCREEN_INEXIO) += inexio.o
+obj-$(CONFIG_TOUCHSCREEN_MCS5000) += mcs5000_ts.o
obj-$(CONFIG_TOUCHSCREEN_MIGOR) += migor_ts.o
obj-$(CONFIG_TOUCHSCREEN_MTOUCH) += mtouch.o
obj-$(CONFIG_TOUCHSCREEN_MK712) += mk712.o
diff --git a/drivers/input/touchscreen/mcs5000_ts.c b/drivers/input/touchscreen/mcs5000_ts.c
new file mode 100644
index 0000000..c29c072
--- /dev/null
+++ b/drivers/input/touchscreen/mcs5000_ts.c
@@ -0,0 +1,377 @@
+/*
+ * mcs5000_ts.c - Touch screen driver for MELFAS MCS-5000 controller
+ *
+ * Copyright (C) 2008 Samsung Electronics Co.Ltd
+ * Author: Joonyoung Shim <jy0922.shim@samsung.com>
+ *
+ * Based on wm97xx-core.c
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the
+ * Free Software Foundation; either version 2 of the License, or (at your
+ * option) any later version.
+ *
+ */
+
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/i2c.h>
+#include <linux/i2c/mcs5000_ts.h>
+#include <linux/interrupt.h>
+#include <linux/input.h>
+#include <linux/irq.h>
+#include <linux/workqueue.h>
+
+#define DEFAULT_PRESSURE 0x100
+
+/* Registers */
+#define MCS5000_TS_STATUS 0x00
+#define STATUS_OFFSET 0
+#define STATUS_NO (0 << STATUS_OFFSET)
+#define STATUS_INIT (1 << STATUS_OFFSET)
+#define STATUS_SENSING (2 << STATUS_OFFSET)
+#define STATUS_COORD (3 << STATUS_OFFSET)
+#define STATUS_GESTURE (4 << STATUS_OFFSET)
+#define ERROR_OFFSET 4
+#define ERROR_NO (0 << ERROR_OFFSET)
+#define ERROR_POWER_ON_RESET (1 << ERROR_OFFSET)
+#define ERROR_INT_RESET (2 << ERROR_OFFSET)
+#define ERROR_EXT_RESET (3 << ERROR_OFFSET)
+#define ERROR_INVALID_REG_ADDRESS (8 << ERROR_OFFSET)
+#define ERROR_INVALID_REG_VALUE (9 << ERROR_OFFSET)
+
+#define MCS5000_TS_OP_MODE 0x01
+#define RESET_OFFSET 0
+#define RESET_NO (0 << RESET_OFFSET)
+#define RESET_EXT_SOFT (1 << RESET_OFFSET)
+#define OP_MODE_OFFSET 1
+#define OP_MODE_SLEEP (0 << OP_MODE_OFFSET)
+#define OP_MODE_ACTIVE (1 << OP_MODE_OFFSET)
+#define GESTURE_OFFSET 4
+#define GESTURE_DISABLE (0 << GESTURE_OFFSET)
+#define GESTURE_ENABLE (1 << GESTURE_OFFSET)
+#define PROXIMITY_OFFSET 5
+#define PROXIMITY_DISABLE (0 << PROXIMITY_OFFSET)
+#define PROXIMITY_ENABLE (1 << PROXIMITY_OFFSET)
+#define SCAN_MODE_OFFSET 6
+#define SCAN_MODE_INTERRUPT (0 << SCAN_MODE_OFFSET)
+#define SCAN_MODE_POLLING (1 << SCAN_MODE_OFFSET)
+#define REPORT_RATE_OFFSET 7
+#define REPORT_RATE_40 (0 << REPORT_RATE_OFFSET)
+#define REPORT_RATE_80 (1 << REPORT_RATE_OFFSET)
+
+#define MCS5000_TS_SENS_CTL 0x02
+#define MCS5000_TS_FILTER_CTL 0x03
+#define PRI_FILTER_OFFSET 0
+#define SEC_FILTER_OFFSET 4
+
+#define MCS5000_TS_X_SIZE_UPPER 0x08
+#define MCS5000_TS_X_SIZE_LOWER 0x09
+#define MCS5000_TS_Y_SIZE_UPPER 0x0A
+#define MCS5000_TS_Y_SIZE_LOWER 0x0B
+
+#define MCS5000_TS_INPUT_INFO 0x10
+#define INPUT_TYPE_OFFSET 0
+#define INPUT_TYPE_NONTOUCH (0 << INPUT_TYPE_OFFSET)
+#define INPUT_TYPE_SINGLE (1 << INPUT_TYPE_OFFSET)
+#define INPUT_TYPE_DUAL (2 << INPUT_TYPE_OFFSET)
+#define INPUT_TYPE_PALM (3 << INPUT_TYPE_OFFSET)
+#define INPUT_TYPE_PROXIMITY (7 << INPUT_TYPE_OFFSET)
+#define GESTURE_CODE_OFFSET 3
+#define GESTURE_CODE_NO (0 << GESTURE_CODE_OFFSET)
+
+#define MCS5000_TS_X_POS_UPPER 0x11
+#define MCS5000_TS_X_POS_LOWER 0x12
+#define MCS5000_TS_Y_POS_UPPER 0x13
+#define MCS5000_TS_Y_POS_LOWER 0x14
+#define MCS5000_TS_Z_POS 0x15
+#define MCS5000_TS_WIDTH 0x16
+#define MCS5000_TS_GESTURE_VAL 0x17
+#define MCS5000_TS_MODULE_REV 0x20
+#define MCS5000_TS_FIRMWARE_VER 0x21
+
+enum mcs5000_ts_read_offset {
+ READ_INPUT_INFO,
+ READ_X_POS_UPPER,
+ READ_X_POS_LOWER,
+ READ_Y_POS_UPPER,
+ READ_Y_POS_LOWER,
+ READ_BLOCK_SIZE,
+};
+
+/* Touchscreen absolute values */
+static int abs_x[3] = {0, 1024, 0};
+module_param_array(abs_x, int, NULL, 0);
+MODULE_PARM_DESC(abs_x, "Touchscreen absolute X min, max, fuzz");
+
+static int abs_y[3] = {0, 1024, 0};
+module_param_array(abs_y, int, NULL, 0);
+MODULE_PARM_DESC(abs_y, "Touchscreen absolute Y min, max, fuzz");
+
+static int abs_p[3] = {0, 256, 0};
+module_param_array(abs_p, int, NULL, 0);
+MODULE_PARM_DESC(abs_p, "Touchscreen absolute Pressure min, max, fuzz");
+
+/* Each client has this additional data */
+struct mcs5000_ts_data {
+ struct i2c_client *client;
+ struct input_dev *input_dev;
+ struct work_struct ts_event_work;
+ struct mcs5000_ts_platform_data *platform_data;
+
+ unsigned int irq;
+};
+
+static struct i2c_driver mcs5000_ts_driver;
+
+static void mcs5000_ts_input_read(struct mcs5000_ts_data *data)
+{
+ struct i2c_client *client = data->client;
+ u8 buffer[READ_BLOCK_SIZE];
+ int err;
+ int x;
+ int y;
+
+ err = i2c_smbus_read_i2c_block_data(client, MCS5000_TS_INPUT_INFO,
+ READ_BLOCK_SIZE, buffer);
+ if (err < 0) {
+ dev_err(&client->dev, "%s, err[%d]\n", __func__, err);
+ return;
+ }
+
+ switch (buffer[READ_INPUT_INFO]) {
+ case INPUT_TYPE_NONTOUCH:
+ input_report_abs(data->input_dev, ABS_PRESSURE, 0);
+ input_report_key(data->input_dev, BTN_TOUCH, 0);
+ input_sync(data->input_dev);
+ break;
+ case INPUT_TYPE_SINGLE:
+ x = (buffer[READ_X_POS_UPPER] << 8) | buffer[READ_X_POS_LOWER];
+ y = (buffer[READ_Y_POS_UPPER] << 8) | buffer[READ_Y_POS_LOWER];
+
+ input_report_key(data->input_dev, BTN_TOUCH, 1);
+ input_report_abs(data->input_dev, ABS_X, x);
+ input_report_abs(data->input_dev, ABS_Y, y);
+ input_report_abs(data->input_dev, ABS_PRESSURE,
+ DEFAULT_PRESSURE);
+ input_sync(data->input_dev);
+ break;
+ case INPUT_TYPE_DUAL:
+ /* TODO */
+ break;
+ case INPUT_TYPE_PALM:
+ /* TODO */
+ break;
+ case INPUT_TYPE_PROXIMITY:
+ /* TODO */
+ break;
+ default:
+ dev_err(&client->dev, "Unknown ts input type %d\n",
+ buffer[READ_INPUT_INFO]);
+ break;
+ }
+}
+
+static void mcs5000_ts_irq_worker(struct work_struct *work)
+{
+ struct mcs5000_ts_data *data = container_of(work,
+ struct mcs5000_ts_data, ts_event_work);
+
+ mcs5000_ts_input_read(data);
+
+ enable_irq(data->irq);
+}
+
+static irqreturn_t mcs5000_ts_interrupt(int irq, void *dev_id)
+{
+ struct mcs5000_ts_data *data = dev_id;
+
+ if (!work_pending(&data->ts_event_work)) {
+ disable_irq_nosync(data->irq);
+ schedule_work(&data->ts_event_work);
+ }
+
+ return IRQ_HANDLED;
+}
+
+static int mcs5000_ts_input_init(struct mcs5000_ts_data *data)
+{
+ struct input_dev *input_dev;
+ int ret = 0;
+
+ INIT_WORK(&data->ts_event_work, mcs5000_ts_irq_worker);
+
+ data->input_dev = input_allocate_device();
+ if (data->input_dev == NULL) {
+ ret = -ENOMEM;
+ goto err_input;
+ }
+
+ input_dev = data->input_dev;
+ input_dev->name = "MELPAS MCS-5000 Touchscreen";
+ set_bit(EV_ABS, input_dev->evbit);
+ set_bit(ABS_X, input_dev->absbit);
+ set_bit(ABS_Y, input_dev->absbit);
+ set_bit(ABS_PRESSURE, input_dev->absbit);
+ input_set_abs_params(input_dev, ABS_X, abs_x[0], abs_x[1], abs_x[2], 0);
+ input_set_abs_params(input_dev, ABS_Y, abs_y[0], abs_y[1], abs_y[2], 0);
+ input_set_abs_params(input_dev, ABS_PRESSURE, abs_p[0], abs_p[1],
+ abs_p[2], 0);
+ set_bit(EV_KEY, input_dev->evbit);
+ set_bit(BTN_TOUCH, input_dev->keybit);
+
+ ret = input_register_device(data->input_dev);
+ if (ret < 0)
+ goto err_register;
+
+ if (request_irq(data->irq, mcs5000_ts_interrupt,
+ IRQF_TRIGGER_LOW, "mcs5000_ts_input", data)) {
+ dev_err(&data->client->dev, "Failed to register interrupt\n");
+ ret = -EINVAL;
+ goto err_irq;
+ }
+
+ input_set_drvdata(input_dev, data);
+
+ return 0;
+err_irq:
+ input_unregister_device(data->input_dev);
+err_register:
+ input_free_device(data->input_dev);
+err_input:
+ return ret;
+}
+
+static void mcs5000_ts_phys_init(struct mcs5000_ts_data *data)
+{
+ struct i2c_client *client = data->client;
+ struct mcs5000_ts_platform_data *platform_data = data->platform_data;
+
+ /* Touch reset & sleep mode */
+ i2c_smbus_write_byte_data(client, MCS5000_TS_OP_MODE,
+ RESET_EXT_SOFT | OP_MODE_SLEEP);
+
+ /* Touch size */
+ i2c_smbus_write_byte_data(client, MCS5000_TS_X_SIZE_UPPER,
+ platform_data->x_size >> 8);
+ i2c_smbus_write_byte_data(client, MCS5000_TS_X_SIZE_LOWER,
+ platform_data->x_size & 0xff);
+ i2c_smbus_write_byte_data(client, MCS5000_TS_Y_SIZE_UPPER,
+ platform_data->y_size >> 8);
+ i2c_smbus_write_byte_data(client, MCS5000_TS_Y_SIZE_LOWER,
+ platform_data->y_size & 0xff);
+
+ /* Touch active mode & 80 report rate */
+ i2c_smbus_write_byte_data(data->client, MCS5000_TS_OP_MODE,
+ OP_MODE_ACTIVE | REPORT_RATE_80);
+}
+
+static int mcs5000_ts_probe(struct i2c_client *client,
+ const struct i2c_device_id *idp)
+{
+ struct mcs5000_ts_data *data;
+ int ret;
+
+ data = kzalloc(sizeof(struct mcs5000_ts_data), GFP_KERNEL);
+ if (!data)
+ dev_err(&client->dev, "Failed to allocate driver data\n");
+ ret = -ENOMEM;
+ goto exit;
+ }
+
+ data->client = client;
+ data->irq = client->irq;
+ data->platform_data = client->dev.platform_data;
+
+ i2c_set_clientdata(client, data);
+
+ if (data->platform_data->set_pin)
+ data->platform_data->set_pin();
+
+ ret = mcs5000_ts_input_init(data);
+ if (ret)
+ goto exit_free;
+
+ mcs5000_ts_phys_init(data);
+
+ return 0;
+
+exit_free:
+ kfree(data);
+ i2c_set_clientdata(client, NULL);
+exit:
+ return ret;
+}
+
+static int mcs5000_ts_remove(struct i2c_client *client)
+{
+ struct mcs5000_ts_data *data = i2c_get_clientdata(client);
+
+ cancel_work_sync(&data->ts_event_work);
+ free_irq(data->irq, data);
+ input_unregister_device(data->input_dev);
+ input_free_device(data->input_dev);
+ kfree(data);
+
+ i2c_set_clientdata(client, NULL);
+
+ return 0;
+}
+
+#ifdef CONFIG_PM
+static int mcs5000_ts_suspend(struct i2c_client *client, pm_message_t mesg)
+{
+ /* Touch sleep mode */
+ i2c_smbus_write_byte_data(client, MCS5000_TS_OP_MODE, OP_MODE_SLEEP);
+
+ return 0;
+}
+
+static int mcs5000_ts_resume(struct i2c_client *client)
+{
+ struct mcs5000_ts_data *data;
+
+ data = i2c_get_clientdata(client);
+ mcs5000_ts_phys_init(data);
+
+ return 0;
+}
+#else
+#define mcs5000_ts_suspend NULL
+#define mcs5000_ts_resume NULL
+#endif
+
+static const struct i2c_device_id mcs5000_ts_id[] = {
+ { "mcs5000_ts", 0 },
+ { }
+};
+MODULE_DEVICE_TABLE(i2c, mcs5000_ts_id);
+
+static struct i2c_driver mcs5000_ts_driver = {
+ .driver = {
+ .name = "mcs5000_ts",
+ },
+ .probe = mcs5000_ts_probe,
+ .remove = mcs5000_ts_remove,
+ .suspend = mcs5000_ts_suspend,
+ .resume = mcs5000_ts_resume,
+ .id_table = mcs5000_ts_id,
+};
+
+static int __init mcs5000_ts_init(void)
+{
+ return i2c_add_driver(&mcs5000_ts_driver);
+}
+
+static void __exit mcs5000_ts_exit(void)
+{
+ i2c_del_driver(&mcs5000_ts_driver);
+}
+
+module_init(mcs5000_ts_init);
+module_exit(mcs5000_ts_exit);
+
+/* Module information */
+MODULE_AUTHOR("Joonyoung Shim <jy0922.shim@samsung.com>");
+MODULE_DESCRIPTION("Touch screen driver for MELFAS MCS-5000 controller");
+MODULE_LICENSE("GPL");
diff --git a/include/linux/i2c/mcs5000_ts.h b/include/linux/i2c/mcs5000_ts.h
new file mode 100644
index 0000000..69d92b6
--- /dev/null
+++ b/include/linux/i2c/mcs5000_ts.h
@@ -0,0 +1,11 @@
+#ifndef __LINUX_MCS5000_TS_H
+#define __LINUX_MCS5000_TS_H
+
+/* platform data for the MELFAS MCS5000 touchscreen driver */
+struct mcs5000_ts_platform_data {
+ void (*set_pin)(void);
+ int x_size;
+ int y_size;
+};
+
+#endif /* __LINUX_MCS5000_TS_H */
--
1.6.0.4
^ permalink raw reply related [flat|nested] 7+ messages in thread
* Re: [PATCH] Input: add touchscreen driver for MELFAS MCS-5000 controller
2009-05-11 12:28 [PATCH] Input: add touchscreen driver for MELFAS MCS-5000 controller Joonyoung Shim
@ 2009-05-12 2:09 ` Dmitry Torokhov
2009-05-14 12:52 ` Joonyoung Shim
0 siblings, 1 reply; 7+ messages in thread
From: Dmitry Torokhov @ 2009-05-12 2:09 UTC (permalink / raw)
To: Joonyoung Shim; +Cc: linux-input, kyungmin.park, m.szyprowski
Hi,
On Mon, May 11, 2009 at 09:28:30PM +0900, Joonyoung Shim wrote:
> The MELPAS MCS-5000 is the touchscreen controller. The overview of this
> controller can see at the following website:
>
> http://www.melfas.com/product/product01.asp?k_r=eng_
>
> This patch supports only the i2c interface.
>
Thank you for the patch. Some comments below.
> +
> +/* Touchscreen absolute values */
> +static int abs_x[3] = {0, 1024, 0};
> +module_param_array(abs_x, int, NULL, 0);
> +MODULE_PARM_DESC(abs_x, "Touchscreen absolute X min, max, fuzz");
> +
> +static int abs_y[3] = {0, 1024, 0};
> +module_param_array(abs_y, int, NULL, 0);
> +MODULE_PARM_DESC(abs_y, "Touchscreen absolute Y min, max, fuzz");
> +
Do we need to have these as module parameters give the fact that we can
set them dynamically with EVIOCSABS? I'd =say they should go.
> +static int abs_p[3] = {0, 256, 0};
> +module_param_array(abs_p, int, NULL, 0);
> +MODULE_PARM_DESC(abs_p, "Touchscreen absolute Pressure min, max, fuzz");
> +
> +/* Each client has this additional data */
> +struct mcs5000_ts_data {
> + struct i2c_client *client;
> + struct input_dev *input_dev;
> + struct work_struct ts_event_work;
> + struct mcs5000_ts_platform_data *platform_data;
> +
> + unsigned int irq;
> +};
> +
> +static struct i2c_driver mcs5000_ts_driver;
> +
> +static void mcs5000_ts_input_read(struct mcs5000_ts_data *data)
> +{
> + struct i2c_client *client = data->client;
> + u8 buffer[READ_BLOCK_SIZE];
> + int err;
> + int x;
> + int y;
> +
> + err = i2c_smbus_read_i2c_block_data(client, MCS5000_TS_INPUT_INFO,
> + READ_BLOCK_SIZE, buffer);
> + if (err < 0) {
> + dev_err(&client->dev, "%s, err[%d]\n", __func__, err);
> + return;
> + }
> +
> + switch (buffer[READ_INPUT_INFO]) {
> + case INPUT_TYPE_NONTOUCH:
> + input_report_abs(data->input_dev, ABS_PRESSURE, 0);
> + input_report_key(data->input_dev, BTN_TOUCH, 0);
> + input_sync(data->input_dev);
> + break;
> + case INPUT_TYPE_SINGLE:
> + x = (buffer[READ_X_POS_UPPER] << 8) | buffer[READ_X_POS_LOWER];
> + y = (buffer[READ_Y_POS_UPPER] << 8) | buffer[READ_Y_POS_LOWER];
> +
> + input_report_key(data->input_dev, BTN_TOUCH, 1);
> + input_report_abs(data->input_dev, ABS_X, x);
> + input_report_abs(data->input_dev, ABS_Y, y);
> + input_report_abs(data->input_dev, ABS_PRESSURE,
> + DEFAULT_PRESSURE);
If the hardware does not support pressure reading don't fake it.
BTN_TOUCH should be enough to signal touch.
> + input_sync(data->input_dev);
> + break;
> + case INPUT_TYPE_DUAL:
> + /* TODO */
> + break;
> + case INPUT_TYPE_PALM:
> + /* TODO */
> + break;
> + case INPUT_TYPE_PROXIMITY:
> + /* TODO */
> + break;
> + default:
> + dev_err(&client->dev, "Unknown ts input type %d\n",
> + buffer[READ_INPUT_INFO]);
> + break;
> + }
> +}
> +
> +static void mcs5000_ts_irq_worker(struct work_struct *work)
> +{
> + struct mcs5000_ts_data *data = container_of(work,
> + struct mcs5000_ts_data, ts_event_work);
> +
> + mcs5000_ts_input_read(data);
> +
> + enable_irq(data->irq);
> +}
> +
> +static irqreturn_t mcs5000_ts_interrupt(int irq, void *dev_id)
> +{
> + struct mcs5000_ts_data *data = dev_id;
> +
> + if (!work_pending(&data->ts_event_work)) {
> + disable_irq_nosync(data->irq);
> + schedule_work(&data->ts_event_work);
> + }
> +
> + return IRQ_HANDLED;
> +}
> +
> +static int mcs5000_ts_input_init(struct mcs5000_ts_data *data)
> +{
> + struct input_dev *input_dev;
> + int ret = 0;
> +
> + INIT_WORK(&data->ts_event_work, mcs5000_ts_irq_worker);
> +
> + data->input_dev = input_allocate_device();
> + if (data->input_dev == NULL) {
> + ret = -ENOMEM;
> + goto err_input;
> + }
> +
> + input_dev = data->input_dev;
> + input_dev->name = "MELPAS MCS-5000 Touchscreen";
input_dev->id.bustype = BUS_I2C;
input_dev->dev.parent = <i2c device>;
> + set_bit(EV_ABS, input_dev->evbit);
> + set_bit(ABS_X, input_dev->absbit);
> + set_bit(ABS_Y, input_dev->absbit);
> + set_bit(ABS_PRESSURE, input_dev->absbit);
> + input_set_abs_params(input_dev, ABS_X, abs_x[0], abs_x[1], abs_x[2], 0);
> + input_set_abs_params(input_dev, ABS_Y, abs_y[0], abs_y[1], abs_y[2], 0);
> + input_set_abs_params(input_dev, ABS_PRESSURE, abs_p[0], abs_p[1],
> + abs_p[2], 0);
This line should go.
> + set_bit(EV_KEY, input_dev->evbit);
> + set_bit(BTN_TOUCH, input_dev->keybit);
> +
> + ret = input_register_device(data->input_dev);
> + if (ret < 0)
> + goto err_register;
> +
> + if (request_irq(data->irq, mcs5000_ts_interrupt,
> + IRQF_TRIGGER_LOW, "mcs5000_ts_input", data)) {
> + dev_err(&data->client->dev, "Failed to register interrupt\n");
> + ret = -EINVAL;
Why EINVAL and not EBUSY? Or better yet, why don't you propagate what
reqiest_irq returned?
> + goto err_irq;
> + }
> +
> + input_set_drvdata(input_dev, data);
> +
> + return 0;
> +err_irq:
> + input_unregister_device(data->input_dev);
> +err_register:
> + input_free_device(data->input_dev);
Do not call input_free_device() after input_unregister_device(), it will
cause double-free. EIther rearrange initialization order or do
data->input_dev = NULL; right after calling input_unregister_device().
> +err_input:
> + return ret;
> +}
> +
> +static void mcs5000_ts_phys_init(struct mcs5000_ts_data *data)
> +{
> + struct i2c_client *client = data->client;
> + struct mcs5000_ts_platform_data *platform_data = data->platform_data;
> +
> + /* Touch reset & sleep mode */
> + i2c_smbus_write_byte_data(client, MCS5000_TS_OP_MODE,
> + RESET_EXT_SOFT | OP_MODE_SLEEP);
> +
> + /* Touch size */
> + i2c_smbus_write_byte_data(client, MCS5000_TS_X_SIZE_UPPER,
> + platform_data->x_size >> 8);
> + i2c_smbus_write_byte_data(client, MCS5000_TS_X_SIZE_LOWER,
> + platform_data->x_size & 0xff);
> + i2c_smbus_write_byte_data(client, MCS5000_TS_Y_SIZE_UPPER,
> + platform_data->y_size >> 8);
> + i2c_smbus_write_byte_data(client, MCS5000_TS_Y_SIZE_LOWER,
> + platform_data->y_size & 0xff);
> +
> + /* Touch active mode & 80 report rate */
> + i2c_smbus_write_byte_data(data->client, MCS5000_TS_OP_MODE,
> + OP_MODE_ACTIVE | REPORT_RATE_80);
> +}
> +
> +static int mcs5000_ts_probe(struct i2c_client *client,
> + const struct i2c_device_id *idp)
This should be marked __devinit.
> +{
> + struct mcs5000_ts_data *data;
> + int ret;
> +
> + data = kzalloc(sizeof(struct mcs5000_ts_data), GFP_KERNEL);
> + if (!data)
> + dev_err(&client->dev, "Failed to allocate driver data\n");
> + ret = -ENOMEM;
> + goto exit;
> + }
> +
> + data->client = client;
> + data->irq = client->irq;
> + data->platform_data = client->dev.platform_data;
> +
> + i2c_set_clientdata(client, data);
> +
> + if (data->platform_data->set_pin)
> + data->platform_data->set_pin();
> +
> + ret = mcs5000_ts_input_init(data);
> + if (ret)
> + goto exit_free;
> +
> + mcs5000_ts_phys_init(data);
> +
> + return 0;
> +
> +exit_free:
> + kfree(data);
> + i2c_set_clientdata(client, NULL);
> +exit:
> + return ret;
> +}
> +
> +static int mcs5000_ts_remove(struct i2c_client *client)
This should be marked __devexit.
> +{
> + struct mcs5000_ts_data *data = i2c_get_clientdata(client);
> +
> + cancel_work_sync(&data->ts_event_work);
There is a race here, IRQ handler may resubmit work after
cancel_work_sync() returns. You need to make sure that
IRQ handler does not resubmit work while device is being shut down.
> + free_irq(data->irq, data);
> + input_unregister_device(data->input_dev);
> + input_free_device(data->input_dev);
Just input_unregister_device() is enough.
> + kfree(data);
> +
> + i2c_set_clientdata(client, NULL);
> +
> + return 0;
> +}
> +
> +#ifdef CONFIG_PM
> +static int mcs5000_ts_suspend(struct i2c_client *client, pm_message_t mesg)
> +{
> + /* Touch sleep mode */
> + i2c_smbus_write_byte_data(client, MCS5000_TS_OP_MODE, OP_MODE_SLEEP);
> +
> + return 0;
> +}
> +
> +static int mcs5000_ts_resume(struct i2c_client *client)
> +{
> + struct mcs5000_ts_data *data;
> +
> + data = i2c_get_clientdata(client);
> + mcs5000_ts_phys_init(data);
> +
> + return 0;
> +}
> +#else
> +#define mcs5000_ts_suspend NULL
> +#define mcs5000_ts_resume NULL
> +#endif
> +
> +static const struct i2c_device_id mcs5000_ts_id[] = {
> + { "mcs5000_ts", 0 },
> + { }
> +};
> +MODULE_DEVICE_TABLE(i2c, mcs5000_ts_id);
> +
> +static struct i2c_driver mcs5000_ts_driver = {
> + .driver = {
> + .name = "mcs5000_ts",
> + },
> + .probe = mcs5000_ts_probe,
> + .remove = mcs5000_ts_remove,
__devexit_p()
#ifdef CONFIG_PM
> + .suspend = mcs5000_ts_suspend,
> + .resume = mcs5000_ts_resume,
#endif
> + .id_table = mcs5000_ts_id,
> +};
> +
Thanks.
--
Dmitry
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [PATCH] Input: add touchscreen driver for MELFAS MCS-5000 controller
2009-05-12 2:09 ` Dmitry Torokhov
@ 2009-05-14 12:52 ` Joonyoung Shim
2009-05-14 15:09 ` Dmitry Torokhov
0 siblings, 1 reply; 7+ messages in thread
From: Joonyoung Shim @ 2009-05-14 12:52 UTC (permalink / raw)
To: Dmitry Torokhov; +Cc: linux-input, kyungmin.park, m.szyprowski
Hi,
Thank you for your review.
> Hi,
>
> On Mon, May 11, 2009 at 09:28:30PM +0900, Joonyoung Shim wrote:
>> The MELPAS MCS-5000 is the touchscreen controller. The overview of this
>> controller can see at the following website:
>>
>> http://www.melfas.com/product/product01.asp?k_r=eng_
>>
>> This patch supports only the i2c interface.
>>
>
> Thank you for the patch. Some comments below.
>
>> +
>> +/* Touchscreen absolute values */
>> +static int abs_x[3] = {0, 1024, 0};
>> +module_param_array(abs_x, int, NULL, 0);
>> +MODULE_PARM_DESC(abs_x, "Touchscreen absolute X min, max, fuzz");
>> +
>> +static int abs_y[3] = {0, 1024, 0};
>> +module_param_array(abs_y, int, NULL, 0);
>> +MODULE_PARM_DESC(abs_y, "Touchscreen absolute Y min, max, fuzz");
>> +
>
> Do we need to have these as module parameters give the fact that we can
> set them dynamically with EVIOCSABS? I'd =say they should go.
I think that this part is unnecessary, we can get resolution of
touchscreen from platform_data or i will use the defined values.
>
>> +static int abs_p[3] = {0, 256, 0};
>> +module_param_array(abs_p, int, NULL, 0);
>> +MODULE_PARM_DESC(abs_p, "Touchscreen absolute Pressure min, max, fuzz");
>> +
>> +/* Each client has this additional data */
>> +struct mcs5000_ts_data {
>> + struct i2c_client *client;
>> + struct input_dev *input_dev;
>> + struct work_struct ts_event_work;
>> + struct mcs5000_ts_platform_data *platform_data;
>> +
>> + unsigned int irq;
>> +};
>> +
>> +static struct i2c_driver mcs5000_ts_driver;
>> +
>> +static void mcs5000_ts_input_read(struct mcs5000_ts_data *data)
>> +{
>> + struct i2c_client *client = data->client;
>> + u8 buffer[READ_BLOCK_SIZE];
>> + int err;
>> + int x;
>> + int y;
>> +
>> + err = i2c_smbus_read_i2c_block_data(client, MCS5000_TS_INPUT_INFO,
>> + READ_BLOCK_SIZE, buffer);
>> + if (err < 0) {
>> + dev_err(&client->dev, "%s, err[%d]\n", __func__, err);
>> + return;
>> + }
>> +
>> + switch (buffer[READ_INPUT_INFO]) {
>> + case INPUT_TYPE_NONTOUCH:
>> + input_report_abs(data->input_dev, ABS_PRESSURE, 0);
>> + input_report_key(data->input_dev, BTN_TOUCH, 0);
>> + input_sync(data->input_dev);
>> + break;
>> + case INPUT_TYPE_SINGLE:
>> + x = (buffer[READ_X_POS_UPPER] << 8) | buffer[READ_X_POS_LOWER];
>> + y = (buffer[READ_Y_POS_UPPER] << 8) | buffer[READ_Y_POS_LOWER];
>> +
>> + input_report_key(data->input_dev, BTN_TOUCH, 1);
>> + input_report_abs(data->input_dev, ABS_X, x);
>> + input_report_abs(data->input_dev, ABS_Y, y);
>> + input_report_abs(data->input_dev, ABS_PRESSURE,
>> + DEFAULT_PRESSURE);
>
> If the hardware does not support pressure reading don't fake it.
> BTN_TOUCH should be enough to signal touch.
MCS-5000 supports pressure reading, but the value of pressure is unstable in
my target, so i used the static value defined.
I will add pressure reading after more test.
signal touch? Do you mean single touch?
>
>> + input_sync(data->input_dev);
>> + break;
>> + case INPUT_TYPE_DUAL:
>> + /* TODO */
>> + break;
>> + case INPUT_TYPE_PALM:
>> + /* TODO */
>> + break;
>> + case INPUT_TYPE_PROXIMITY:
>> + /* TODO */
>> + break;
>> + default:
>> + dev_err(&client->dev, "Unknown ts input type %d\n",
>> + buffer[READ_INPUT_INFO]);
>> + break;
>> + }
>> +}
>> +
>> +static void mcs5000_ts_irq_worker(struct work_struct *work)
>> +{
>> + struct mcs5000_ts_data *data = container_of(work,
>> + struct mcs5000_ts_data, ts_event_work);
>> +
>> + mcs5000_ts_input_read(data);
>> +
>> + enable_irq(data->irq);
>> +}
>> +
>> +static irqreturn_t mcs5000_ts_interrupt(int irq, void *dev_id)
>> +{
>> + struct mcs5000_ts_data *data = dev_id;
>> +
>> + if (!work_pending(&data->ts_event_work)) {
>> + disable_irq_nosync(data->irq);
>> + schedule_work(&data->ts_event_work);
>> + }
>> +
>> + return IRQ_HANDLED;
>> +}
>> +
>> +static int mcs5000_ts_input_init(struct mcs5000_ts_data *data)
>> +{
>> + struct input_dev *input_dev;
>> + int ret = 0;
>> +
>> + INIT_WORK(&data->ts_event_work, mcs5000_ts_irq_worker);
>> +
>> + data->input_dev = input_allocate_device();
>> + if (data->input_dev == NULL) {
>> + ret = -ENOMEM;
>> + goto err_input;
>> + }
>> +
>> + input_dev = data->input_dev;
>> + input_dev->name = "MELPAS MCS-5000 Touchscreen";
>
> input_dev->id.bustype = BUS_I2C;
> input_dev->dev.parent = <i2c device>;
ok, i will add it.
>
>
>> + set_bit(EV_ABS, input_dev->evbit);
>> + set_bit(ABS_X, input_dev->absbit);
>> + set_bit(ABS_Y, input_dev->absbit);
>> + set_bit(ABS_PRESSURE, input_dev->absbit);
>> + input_set_abs_params(input_dev, ABS_X, abs_x[0], abs_x[1], abs_x[2], 0);
>> + input_set_abs_params(input_dev, ABS_Y, abs_y[0], abs_y[1], abs_y[2], 0);
>> + input_set_abs_params(input_dev, ABS_PRESSURE, abs_p[0], abs_p[1],
>> + abs_p[2], 0);
>
> This line should go.
>
>> + set_bit(EV_KEY, input_dev->evbit);
>> + set_bit(BTN_TOUCH, input_dev->keybit);
>> +
>> + ret = input_register_device(data->input_dev);
>> + if (ret < 0)
>> + goto err_register;
>> +
>> + if (request_irq(data->irq, mcs5000_ts_interrupt,
>> + IRQF_TRIGGER_LOW, "mcs5000_ts_input", data)) {
>> + dev_err(&data->client->dev, "Failed to register interrupt\n");
>> + ret = -EINVAL;
>
> Why EINVAL and not EBUSY? Or better yet, why don't you propagate what
> reqiest_irq returned?
Hmm, EINVAL is used in wm97xx-core.c file, but you are right.
I will fix it using the latter.
>
>> + goto err_irq;
>> + }
>> +
>> + input_set_drvdata(input_dev, data);
>> +
>> + return 0;
>> +err_irq:
>> + input_unregister_device(data->input_dev);
>> +err_register:
>> + input_free_device(data->input_dev);
>
> Do not call input_free_device() after input_unregister_device(), it will
> cause double-free. EIther rearrange initialization order or do
> data->input_dev = NULL; right after calling input_unregister_device().
ok, i will fix it.
>
>> +err_input:
>> + return ret;
>> +}
>> +
>> +static void mcs5000_ts_phys_init(struct mcs5000_ts_data *data)
>> +{
>> + struct i2c_client *client = data->client;
>> + struct mcs5000_ts_platform_data *platform_data = data->platform_data;
>> +
>> + /* Touch reset & sleep mode */
>> + i2c_smbus_write_byte_data(client, MCS5000_TS_OP_MODE,
>> + RESET_EXT_SOFT | OP_MODE_SLEEP);
>> +
>> + /* Touch size */
>> + i2c_smbus_write_byte_data(client, MCS5000_TS_X_SIZE_UPPER,
>> + platform_data->x_size >> 8);
>> + i2c_smbus_write_byte_data(client, MCS5000_TS_X_SIZE_LOWER,
>> + platform_data->x_size & 0xff);
>> + i2c_smbus_write_byte_data(client, MCS5000_TS_Y_SIZE_UPPER,
>> + platform_data->y_size >> 8);
>> + i2c_smbus_write_byte_data(client, MCS5000_TS_Y_SIZE_LOWER,
>> + platform_data->y_size & 0xff);
>> +
>> + /* Touch active mode & 80 report rate */
>> + i2c_smbus_write_byte_data(data->client, MCS5000_TS_OP_MODE,
>> + OP_MODE_ACTIVE | REPORT_RATE_80);
>> +}
>> +
>> +static int mcs5000_ts_probe(struct i2c_client *client,
>> + const struct i2c_device_id *idp)
>
> This should be marked __devinit.
ok, i will add it.
>
>> +{
>> + struct mcs5000_ts_data *data;
>> + int ret;
>> +
>> + data = kzalloc(sizeof(struct mcs5000_ts_data), GFP_KERNEL);
>> + if (!data)
>> + dev_err(&client->dev, "Failed to allocate driver data\n");
>> + ret = -ENOMEM;
>> + goto exit;
>> + }
>> +
>> + data->client = client;
>> + data->irq = client->irq;
>> + data->platform_data = client->dev.platform_data;
>> +
>> + i2c_set_clientdata(client, data);
>> +
>> + if (data->platform_data->set_pin)
>> + data->platform_data->set_pin();
>> +
>> + ret = mcs5000_ts_input_init(data);
>> + if (ret)
>> + goto exit_free;
>> +
>> + mcs5000_ts_phys_init(data);
>> +
>> + return 0;
>> +
>> +exit_free:
>> + kfree(data);
>> + i2c_set_clientdata(client, NULL);
>> +exit:
>> + return ret;
>> +}
>> +
>> +static int mcs5000_ts_remove(struct i2c_client *client)
>
> This should be marked __devexit.
ok, i will add it.
>
>> +{
>> + struct mcs5000_ts_data *data = i2c_get_clientdata(client);
>> +
>> + cancel_work_sync(&data->ts_event_work);
>
> There is a race here, IRQ handler may resubmit work after
> cancel_work_sync() returns. You need to make sure that
> IRQ handler does not resubmit work while device is being shut down.
ok, how about doing free_irq() before cancel_work_sync() call?
>
>> + free_irq(data->irq, data);
>> + input_unregister_device(data->input_dev);
>> + input_free_device(data->input_dev);
>
> Just input_unregister_device() is enough.
ok.
>
>> + kfree(data);
>> +
>> + i2c_set_clientdata(client, NULL);
>> +
>> + return 0;
>> +}
>> +
>> +#ifdef CONFIG_PM
>> +static int mcs5000_ts_suspend(struct i2c_client *client, pm_message_t mesg)
>> +{
>> + /* Touch sleep mode */
>> + i2c_smbus_write_byte_data(client, MCS5000_TS_OP_MODE, OP_MODE_SLEEP);
>> +
>> + return 0;
>> +}
>> +
>> +static int mcs5000_ts_resume(struct i2c_client *client)
>> +{
>> + struct mcs5000_ts_data *data;
>> +
>> + data = i2c_get_clientdata(client);
>> + mcs5000_ts_phys_init(data);
>> +
>> + return 0;
>> +}
>> +#else
>> +#define mcs5000_ts_suspend NULL
>> +#define mcs5000_ts_resume NULL
>> +#endif
>> +
>> +static const struct i2c_device_id mcs5000_ts_id[] = {
>> + { "mcs5000_ts", 0 },
>> + { }
>> +};
>> +MODULE_DEVICE_TABLE(i2c, mcs5000_ts_id);
>> +
>> +static struct i2c_driver mcs5000_ts_driver = {
>> + .driver = {
>> + .name = "mcs5000_ts",
>> + },
>> + .probe = mcs5000_ts_probe,
>> + .remove = mcs5000_ts_remove,
>
> __devexit_p()
ok.
>
> #ifdef CONFIG_PM
>> + .suspend = mcs5000_ts_suspend,
>> + .resume = mcs5000_ts_resume,
> #endif
ok, i will add it.
>
>> + .id_table = mcs5000_ts_id,
>> +};
>> +
>
> Thanks.
>
Thanks, too.
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [PATCH] Input: add touchscreen driver for MELFAS MCS-5000 controller
2009-05-14 12:52 ` Joonyoung Shim
@ 2009-05-14 15:09 ` Dmitry Torokhov
2009-05-15 0:49 ` Joonyoung Shim
2009-06-03 5:16 ` Joonyoung Shim
0 siblings, 2 replies; 7+ messages in thread
From: Dmitry Torokhov @ 2009-05-14 15:09 UTC (permalink / raw)
To: Joonyoung Shim; +Cc: linux-input, kyungmin.park, m.szyprowski
On Thu, May 14, 2009 at 09:52:41PM +0900, Joonyoung Shim wrote:
> Hi,
>
> Thank you for your review.
>
> > Hi,
> >
> > On Mon, May 11, 2009 at 09:28:30PM +0900, Joonyoung Shim wrote:
> >> The MELPAS MCS-5000 is the touchscreen controller. The overview of this
> >> controller can see at the following website:
> >>
> >> http://www.melfas.com/product/product01.asp?k_r=eng_
> >>
> >> This patch supports only the i2c interface.
> >>
> >
> > Thank you for the patch. Some comments below.
> >
> >> +
> >> +/* Touchscreen absolute values */
> >> +static int abs_x[3] = {0, 1024, 0};
> >> +module_param_array(abs_x, int, NULL, 0);
> >> +MODULE_PARM_DESC(abs_x, "Touchscreen absolute X min, max, fuzz");
> >> +
> >> +static int abs_y[3] = {0, 1024, 0};
> >> +module_param_array(abs_y, int, NULL, 0);
> >> +MODULE_PARM_DESC(abs_y, "Touchscreen absolute Y min, max, fuzz");
> >> +
> >
> > Do we need to have these as module parameters give the fact that we can
> > set them dynamically with EVIOCSABS? I'd =say they should go.
>
> I think that this part is unnecessary, we can get resolution of
> touchscreen from platform_data or i will use the defined values.
>
Ok, that would work as well.
> >> + case INPUT_TYPE_SINGLE:
> >> + x = (buffer[READ_X_POS_UPPER] << 8) | buffer[READ_X_POS_LOWER];
> >> + y = (buffer[READ_Y_POS_UPPER] << 8) | buffer[READ_Y_POS_LOWER];
> >> +
> >> + input_report_key(data->input_dev, BTN_TOUCH, 1);
> >> + input_report_abs(data->input_dev, ABS_X, x);
> >> + input_report_abs(data->input_dev, ABS_Y, y);
> >> + input_report_abs(data->input_dev, ABS_PRESSURE,
> >> + DEFAULT_PRESSURE);
> >
> > If the hardware does not support pressure reading don't fake it.
> > BTN_TOUCH should be enough to signal touch.
>
> MCS-5000 supports pressure reading, but the value of pressure is unstable in
> my target, so i used the static value defined.
> I will add pressure reading after more test.
OK. Alternatively you may indicate in the platform data if pressure
reading is supported and set ABS_PRESSURE bit and report pressure only
when you know it works well.
>
> signal touch? Do you mean single touch?
I meant "signal" as "convey", "indicate".
> >> +
> >> + if (request_irq(data->irq, mcs5000_ts_interrupt,
> >> + IRQF_TRIGGER_LOW, "mcs5000_ts_input", data)) {
> >> + dev_err(&data->client->dev, "Failed to register interrupt\n");
> >> + ret = -EINVAL;
> >
> > Why EINVAL and not EBUSY? Or better yet, why don't you propagate what
> > reqiest_irq returned?
>
> Hmm, EINVAL is used in wm97xx-core.c file, but you are right.
I am always taking patches ;))
>
> >
> >> +{
> >> + struct mcs5000_ts_data *data = i2c_get_clientdata(client);
> >> +
> >> + cancel_work_sync(&data->ts_event_work);
> >
> > There is a race here, IRQ handler may resubmit work after
> > cancel_work_sync() returns. You need to make sure that
> > IRQ handler does not resubmit work while device is being shut down.
>
> ok, how about doing free_irq() before cancel_work_sync() call?
>
Then there is a risk that your work will try to enable_irq() on irq that
was freed. I am not sure if IRQ core will be happy with it,
--
Dmitry
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [PATCH] Input: add touchscreen driver for MELFAS MCS-5000 controller
2009-05-14 15:09 ` Dmitry Torokhov
@ 2009-05-15 0:49 ` Joonyoung Shim
2009-05-15 2:52 ` Dmitry Torokhov
2009-06-03 5:16 ` Joonyoung Shim
1 sibling, 1 reply; 7+ messages in thread
From: Joonyoung Shim @ 2009-05-15 0:49 UTC (permalink / raw)
To: Dmitry Torokhov; +Cc: linux-input, kyungmin.park, m.szyprowski
>>>> + case INPUT_TYPE_SINGLE:
>>>> + x = (buffer[READ_X_POS_UPPER] << 8) | buffer[READ_X_POS_LOWER];
>>>> + y = (buffer[READ_Y_POS_UPPER] << 8) | buffer[READ_Y_POS_LOWER];
>>>> +
>>>> + input_report_key(data->input_dev, BTN_TOUCH, 1);
>>>> + input_report_abs(data->input_dev, ABS_X, x);
>>>> + input_report_abs(data->input_dev, ABS_Y, y);
>>>> + input_report_abs(data->input_dev, ABS_PRESSURE,
>>>> + DEFAULT_PRESSURE);
>>> If the hardware does not support pressure reading don't fake it.
>>> BTN_TOUCH should be enough to signal touch.
>> MCS-5000 supports pressure reading, but the value of pressure is unstable in
>> my target, so i used the static value defined.
>> I will add pressure reading after more test.
>
> OK. Alternatively you may indicate in the platform data if pressure
> reading is supported and set ABS_PRESSURE bit and report pressure only
> when you know it works well.
OK, i will add the thing about pressure in the platform data.
>> signal touch? Do you mean single touch?
>
> I meant "signal" as "convey", "indicate".
I see :)
>
>>>> +
>>>> + if (request_irq(data->irq, mcs5000_ts_interrupt,
>>>> + IRQF_TRIGGER_LOW, "mcs5000_ts_input", data)) {
>>>> + dev_err(&data->client->dev, "Failed to register interrupt\n");
>>>> + ret = -EINVAL;
>>> Why EINVAL and not EBUSY? Or better yet, why don't you propagate what
>>> reqiest_irq returned?
>> Hmm, EINVAL is used in wm97xx-core.c file, but you are right.
>
> I am always taking patches ;))
OK, i will send the patch about it.
>
>>>> +{
>>>> + struct mcs5000_ts_data *data = i2c_get_clientdata(client);
>>>> +
>>>> + cancel_work_sync(&data->ts_event_work);
>>> There is a race here, IRQ handler may resubmit work after
>>> cancel_work_sync() returns. You need to make sure that
>>> IRQ handler does not resubmit work while device is being shut down.
>> ok, how about doing free_irq() before cancel_work_sync() call?
>>
>
> Then there is a risk that your work will try to enable_irq() on irq that
> was freed. I am not sure if IRQ core will be happy with it,
Oh, i didn't think about that, then, how about the following patch?
The work handler decides trying to enable_irq() by whether irq is NULL or
not.
@@ -179,7 +179,8 @@ static void mcs5000_ts_irq_worker(struct work_struct *work)
mcs5000_ts_input_read(data);
- enable_irq(data->irq);
+ if (data->irq)
+ enable_irq(data->irq);
}
static irqreturn_t mcs5000_ts_interrupt(int irq, void *dev_id)
@@ -307,8 +308,9 @@ static int mcs5000_ts_remove(struct i2c_client *client)
{
struct mcs5000_ts_data *data = i2c_get_clientdata(client);
- cancel_work_sync(&data->ts_event_work);
free_irq(data->irq, data);
+ data->irq = 0;
+ cancel_work_sync(&data->ts_event_work);
input_unregister_device(data->input_dev);
input_free_device(data->input_dev);
kfree(data);
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [PATCH] Input: add touchscreen driver for MELFAS MCS-5000 controller
2009-05-15 0:49 ` Joonyoung Shim
@ 2009-05-15 2:52 ` Dmitry Torokhov
0 siblings, 0 replies; 7+ messages in thread
From: Dmitry Torokhov @ 2009-05-15 2:52 UTC (permalink / raw)
To: Joonyoung Shim; +Cc: linux-input, kyungmin.park, m.szyprowski
On Thursday 14 May 2009 17:49:40 Joonyoung Shim wrote:
> >>>> + case INPUT_TYPE_SINGLE:
> >>>> + x = (buffer[READ_X_POS_UPPER] << 8) | buffer[READ_X_POS_LOWER];
> >>>> + y = (buffer[READ_Y_POS_UPPER] << 8) | buffer[READ_Y_POS_LOWER];
> >>>> +
> >>>> + input_report_key(data->input_dev, BTN_TOUCH, 1);
> >>>> + input_report_abs(data->input_dev, ABS_X, x);
> >>>> + input_report_abs(data->input_dev, ABS_Y, y);
> >>>> + input_report_abs(data->input_dev, ABS_PRESSURE,
> >>>> + DEFAULT_PRESSURE);
> >>>
> >>> If the hardware does not support pressure reading don't fake it.
> >>> BTN_TOUCH should be enough to signal touch.
> >>
> >> MCS-5000 supports pressure reading, but the value of pressure is
> >> unstable in my target, so i used the static value defined.
> >> I will add pressure reading after more test.
> >
> > OK. Alternatively you may indicate in the platform data if pressure
> > reading is supported and set ABS_PRESSURE bit and report pressure only
> > when you know it works well.
>
> OK, i will add the thing about pressure in the platform data.
>
> >> signal touch? Do you mean single touch?
> >
> > I meant "signal" as "convey", "indicate".
>
> I see :)
>
> >>>> +
> >>>> + if (request_irq(data->irq, mcs5000_ts_interrupt,
> >>>> + IRQF_TRIGGER_LOW, "mcs5000_ts_input", data)) {
> >>>> + dev_err(&data->client->dev, "Failed to register interrupt\n");
> >>>> + ret = -EINVAL;
> >>>
> >>> Why EINVAL and not EBUSY? Or better yet, why don't you propagate what
> >>> reqiest_irq returned?
> >>
> >> Hmm, EINVAL is used in wm97xx-core.c file, but you are right.
> >
> > I am always taking patches ;))
>
> OK, i will send the patch about it.
>
> >>>> +{
> >>>> + struct mcs5000_ts_data *data = i2c_get_clientdata(client);
> >>>> +
> >>>> + cancel_work_sync(&data->ts_event_work);
> >>>
> >>> There is a race here, IRQ handler may resubmit work after
> >>> cancel_work_sync() returns. You need to make sure that
> >>> IRQ handler does not resubmit work while device is being shut down.
> >>
> >> ok, how about doing free_irq() before cancel_work_sync() call?
> >
> > Then there is a risk that your work will try to enable_irq() on irq that
> > was freed. I am not sure if IRQ core will be happy with it,
>
> Oh, i didn't think about that, then, how about the following patch?
> The work handler decides trying to enable_irq() by whether irq is NULL or
> not.
>
> @@ -179,7 +179,8 @@ static void mcs5000_ts_irq_worker(struct work_struct
> *work)
>
> mcs5000_ts_input_read(data);
>
> - enable_irq(data->irq);
> + if (data->irq)
> + enable_irq(data->irq);
> }
>
> static irqreturn_t mcs5000_ts_interrupt(int irq, void *dev_id)
> @@ -307,8 +308,9 @@ static int mcs5000_ts_remove(struct i2c_client *client)
> {
> struct mcs5000_ts_data *data = i2c_get_clientdata(client);
>
> - cancel_work_sync(&data->ts_event_work);
> free_irq(data->irq, data);
> + data->irq = 0;
Not if you want to enable IRQ 0 every once in a while. I think a separate
flag, the check in the IRQ handler as opposed to workqueue handler and a
memory barrier should work though.
> + cancel_work_sync(&data->ts_event_work);
> input_unregister_device(data->input_dev);
> input_free_device(data->input_dev);
> kfree(data);
--
Dmitry
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [PATCH] Input: add touchscreen driver for MELFAS MCS-5000 controller
2009-05-14 15:09 ` Dmitry Torokhov
2009-05-15 0:49 ` Joonyoung Shim
@ 2009-06-03 5:16 ` Joonyoung Shim
1 sibling, 0 replies; 7+ messages in thread
From: Joonyoung Shim @ 2009-06-03 5:16 UTC (permalink / raw)
To: dmitry.torokhov; +Cc: linux-input, kyungmin.park, m.szyprowski
Hi, sorry for long time no reponse.
>>>> + case INPUT_TYPE_SINGLE:
>>>> + x = (buffer[READ_X_POS_UPPER] << 8) | buffer[READ_X_POS_LOWER];
>>>> + y = (buffer[READ_Y_POS_UPPER] << 8) | buffer[READ_Y_POS_LOWER];
>>>> +
>>>> + input_report_key(data->input_dev, BTN_TOUCH, 1);
>>>> + input_report_abs(data->input_dev, ABS_X, x);
>>>> + input_report_abs(data->input_dev, ABS_Y, y);
>>>> + input_report_abs(data->input_dev, ABS_PRESSURE,
>>>> + DEFAULT_PRESSURE);
>>> If the hardware does not support pressure reading don't fake it.
>>> BTN_TOUCH should be enough to signal touch.
>> MCS-5000 supports pressure reading, but the value of pressure is unstable in
>> my target, so i used the static value defined.
>> I will add pressure reading after more test.
>
> OK. Alternatively you may indicate in the platform data if pressure
> reading is supported and set ABS_PRESSURE bit and report pressure only
> when you know it works well.
This is my mistake, the MCS-5000 does not support pressure. I confused pressure
and proximity.
>>>> +{
>>>> + struct mcs5000_ts_data *data = i2c_get_clientdata(client);
>>>> +
>>>> + cancel_work_sync(&data->ts_event_work);
>>> There is a race here, IRQ handler may resubmit work after
>>> cancel_work_sync() returns. You need to make sure that
>>> IRQ handler does not resubmit work while device is being shut down.
>> ok, how about doing free_irq() before cancel_work_sync() call?
>>
>
> Then there is a risk that your work will try to enable_irq() on irq that
> was freed. I am not sure if IRQ core will be happy with it,
>
It is no problem to try to enable_irq() on irq that was freed, but a new
problem is balance of disable_irq() and enable_irq().
I refered drivers/net/phy/phy.c file.
I will post patch v2.
thanks.
^ permalink raw reply [flat|nested] 7+ messages in thread
end of thread, other threads:[~2009-06-03 5:16 UTC | newest]
Thread overview: 7+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2009-05-11 12:28 [PATCH] Input: add touchscreen driver for MELFAS MCS-5000 controller Joonyoung Shim
2009-05-12 2:09 ` Dmitry Torokhov
2009-05-14 12:52 ` Joonyoung Shim
2009-05-14 15:09 ` Dmitry Torokhov
2009-05-15 0:49 ` Joonyoung Shim
2009-05-15 2:52 ` Dmitry Torokhov
2009-06-03 5:16 ` Joonyoung Shim
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).