[RFC PATCH 0/2] AM335x-ICE: Add support for rotary-encoder

This series adds support for rotary-encoder on AM335x-ICE that is
connected to TI PCA9536 I2C GPIO expander.

The rotary encoder is different from the incremental encoders in the
sense that GPIO line status directly reflect the position(number)
pointed by the dial of the encoder. So, there is no need to count steps
or know the direction of rotation. Patch 1/2 adds DT property
"rotary-encoder,absolute-encoder" to support this encoder. Since, such
encoders do not actually require IRQ lines (as the encoder state can
always be known by reading GPIO lines at that moment), support for
polling based input is added.
The second patch adds DT entry for the rotary-encoder present on
AM335x ICE board.

Tested on AM335x ICE board that has above described rotary-encoder
without IRQs (polling mode).
I don't have hardware with incremental encoder or encoder with IRQ,
please give it a spin on incremental encoders for sanity.

Vignesh R (2):
  Input: rotary-encoder- Add support for absolute encoder
  ARM: dts: am335x-icev2: Add rotary-encoder node

 .../devicetree/bindings/input/rotary-encoder.txt   |   4 +
 Documentation/input/rotary-encoder.txt             |   9 ++
 arch/arm/boot/dts/am335x-icev2.dts                 |  11 ++
 drivers/input/misc/Kconfig                         |  11 ++
 drivers/input/misc/rotary_encoder.c                | 165 ++++++++++++++++-----
 5 files changed, 166 insertions(+), 34 deletions(-)

-- 
2.8.2

There are rotary-encoders where GPIO lines reflect the actual position
of the rotary encoder dial. For example, if dial points to 9, then four
GPIO lines connected to the rotary encoder will read HLLH(1001b = 9).
Add support for such rotary-encoder.
The driver relies on rotary-encoder,absolute-encoder DT property to
detect such encoders.
Since, GPIO IRQs are not necessary to work with
such encoders, optional polling mode support is added using
input_poll_dev skeleton. This is can be used by enabling
CONFIG_INPUT_GPIO_ROTARY_ENCODER_POLL_MODE_SUPPORT.

Signed-off-by: Vignesh R <vigneshr@ti.com>
---
 .../devicetree/bindings/input/rotary-encoder.txt   |   4 +
 Documentation/input/rotary-encoder.txt             |   9 ++
 drivers/input/misc/Kconfig                         |  11 ++
 drivers/input/misc/rotary_encoder.c                | 165 ++++++++++++++++-----
 4 files changed, 155 insertions(+), 34 deletions(-)

diff --git a/Documentation/devicetree/bindings/input/rotary-encoder.txt b/Documentation/devicetree/bindings/input/rotary-encoder.txt
index 6c9f0c8a846c..9c928dbd1500 100644
--- a/Documentation/devicetree/bindings/input/rotary-encoder.txt
+++ b/Documentation/devicetree/bindings/input/rotary-encoder.txt

@@ -12,6 +12,10 @@ Optional properties:
 - rotary-encoder,relative-axis: register a relative axis rather than an
   absolute one. Relative axis will only generate +1/-1 events on the input
   device, hence no steps need to be passed.
+- rotary-encoder,absolute-encoder: support encoders where GPIO lines
+  reflect the actual position of the rotary encoder dial. For example,
+  if dial points to 9, then four GPIO lines read HLLH(1001b = 9).
+  In this case, rotary-encoder,steps-per-period needed not be defined.
 - rotary-encoder,rollover: Automatic rollove when the rotary value becomes
   greater than the specified steps or smaller than 0. For absolute axis only.
 - rotary-encoder,steps-per-period: Number of steps (stable states) per period.

diff --git a/Documentation/input/rotary-encoder.txt b/Documentation/input/rotary-encoder.txt
index 46a74f0c551a..dc76092f5618 100644
--- a/Documentation/input/rotary-encoder.txt
+++ b/Documentation/input/rotary-encoder.txt

@@ -36,6 +36,15 @@ The phase diagram of these two outputs look like this:
                 |<>|
 	          one step (quarter-period mode)
 
+
+In some encoders, the value pointed by rotary encoders switch is directly
+reflected by status of the GPIOs connecting rotary encoder to CPU. For
+example, if rotary encoder dial points to 9, then the four GPIOs
+conncting rotary encoder will read HLLH(1001b = 9). Such encoders are
+supported by defining devictree binding "rotary-encoder,absolute-encoder".
+These encoders are also supported without need for event(IRQ) generation
+using Input subsytem's polling mode support.
+
 For more information, please see
 	https://en.wikipedia.org/wiki/Rotary_encoder
 

diff --git a/drivers/input/misc/Kconfig b/drivers/input/misc/Kconfig
index 1f2337abcf2f..81805f9a534c 100644
--- a/drivers/input/misc/Kconfig
+++ b/drivers/input/misc/Kconfig

@@ -566,6 +566,17 @@ config INPUT_GPIO_ROTARY_ENCODER
 	  To compile this driver as a module, choose M here: the
 	  module will be called rotary_encoder.
 
+config INPUT_GPIO_ROTARY_ENCODER_POLL_MODE_SUPPORT
+	bool "Polling mode support for rotary encoder"
+	depends on INPUT_GPIO_ROTARY_ENCODER
+	select INPUT_POLLDEV
+	help
+	  Say Y here to enable polling mode support for rotary encoders
+	  connected to GPIO lines that do have interrupt capability and
+	  report rotary encoder position as absolute value over GPIO
+	  lines. Check file: Documentation/input/rotary-encoder.txt for
+	  more information.
+
 config INPUT_RB532_BUTTON
 	tristate "Mikrotik Routerboard 532 button interface"
 	depends on MIKROTIK_RB532

diff --git a/drivers/input/misc/rotary_encoder.c b/drivers/input/misc/rotary_encoder.c
index c7fc8d4fb080..8f60289fb6cd 100644
--- a/drivers/input/misc/rotary_encoder.c
+++ b/drivers/input/misc/rotary_encoder.c

@@ -25,24 +25,29 @@
 #include <linux/of.h>
 #include <linux/pm.h>
 #include <linux/property.h>
+#include <linux/input-polldev.h>
 
 #define DRV_NAME "rotary-encoder"
 
 struct rotary_encoder {
 	struct input_dev *input;
-
+#ifdef CONFIG_INPUT_GPIO_ROTARY_ENCODER_POLL_MODE_SUPPORT
+	struct input_polled_dev *poll_dev;
+#endif
 	struct mutex access_mutex;
 
 	u32 steps;
 	u32 axis;
 	bool relative_axis;
 	bool rollover;
+	bool absolute_encoder;
 
 	unsigned int pos;
 
 	struct gpio_descs *gpios;
+	struct device *dev;
 
-	unsigned int *irq;
+	int *irq;
 
 	bool armed;
 	signed char dir;	/* 1 - clockwise, -1 - CCW */

@@ -67,6 +72,21 @@ static unsigned int rotary_encoder_get_state(struct rotary_encoder *encoder)
 	return ret & 3;
 }
 
+static unsigned int rotary_encoder_get_gpios_state(struct rotary_encoder
+						   *encoder)
+{
+	int i;
+	unsigned int ret = 0;
+
+	for (i = 0; i < encoder->gpios->ndescs; ++i) {
+		int val = gpiod_get_value_cansleep(encoder->gpios->desc[i]);
+
+		ret = ret << 1 | val;
+	}
+
+	return ret;
+}
+
 static void rotary_encoder_report_event(struct rotary_encoder *encoder)
 {
 	if (encoder->relative_axis) {

@@ -178,6 +198,72 @@ out:
 	return IRQ_HANDLED;
 }
 
+static void rotary_encoder_setup_input_params(struct rotary_encoder  *encoder)
+{
+	struct input_dev *input = encoder->input;
+	struct platform_device *pdev = to_platform_device(encoder->dev);
+
+	input->name = pdev->name;
+	input->id.bustype = BUS_HOST;
+	input->dev.parent = encoder->dev;
+
+	if (encoder->relative_axis)
+		input_set_capability(input, EV_REL, encoder->axis);
+	else
+		input_set_abs_params(input,
+				     encoder->axis, 0, encoder->steps, 0, 1);
+}
+
+static irqreturn_t rotary_absolute_encoder_irq(int irq, void *dev_id)
+{
+	struct rotary_encoder *encoder = dev_id;
+	unsigned int state;
+
+	mutex_lock(&encoder->access_mutex);
+
+	state = rotary_encoder_get_gpios_state(encoder);
+	if (state != encoder->last_stable) {
+		input_report_abs(encoder->input, encoder->axis, state);
+		input_sync(encoder->input);
+		encoder->last_stable = state;
+	}
+
+	mutex_lock(&encoder->access_mutex);
+
+	return IRQ_HANDLED;
+}
+
+#ifdef CONFIG_INPUT_GPIO_ROTARY_ENCODER_POLL_MODE_SUPPORT
+static void rotary_encoder_poll_gpios(struct input_polled_dev *poll_dev)
+{
+	struct rotary_encoder *encoder = poll_dev->private;
+	unsigned int state = rotary_encoder_get_gpios_state(encoder);
+
+	if (state != encoder->last_stable) {
+		input_report_abs(encoder->input, encoder->axis, state);
+		input_sync(encoder->input);
+		encoder->last_stable = state;
+	}
+}
+
+static int rotary_encoder_register_poll_device(struct rotary_encoder
+		*encoder)
+{
+	struct input_polled_dev *poll_dev =
+		devm_input_allocate_polled_device(encoder->dev);
+
+	if (!poll_dev)
+		return -ENOMEM;
+	poll_dev->private = encoder;
+	poll_dev->poll = rotary_encoder_poll_gpios;
+	encoder->input = poll_dev->input;
+	rotary_encoder_setup_input_params(encoder);
+	encoder->poll_dev = poll_dev;
+
+	return input_register_polled_device(poll_dev);
+}
+#endif
+
 static int rotary_encoder_probe(struct platform_device *pdev)
 {
 	struct device *dev = &pdev->dev;

@@ -227,38 +313,32 @@ static int rotary_encoder_probe(struct platform_device *pdev)
 		return -EINVAL;
 	}
 
-	input = devm_input_allocate_device(dev);
-	if (!input)
-		return -ENOMEM;
-
-	encoder->input = input;
-
-	input->name = pdev->name;
-	input->id.bustype = BUS_HOST;
-	input->dev.parent = dev;
-
-	if (encoder->relative_axis)
-		input_set_capability(input, EV_REL, encoder->axis);
-	else
-		input_set_abs_params(input,
-				     encoder->axis, 0, encoder->steps, 0, 1);
-
-	switch (steps_per_period >> (encoder->gpios->ndescs - 2)) {
-	case 4:
-		handler = &rotary_encoder_quarter_period_irq;
-		encoder->last_stable = rotary_encoder_get_state(encoder);
-		break;
-	case 2:
-		handler = &rotary_encoder_half_period_irq;
-		encoder->last_stable = rotary_encoder_get_state(encoder);
-		break;
-	case 1:
-		handler = &rotary_encoder_irq;
-		break;
-	default:
-		dev_err(dev, "'%d' is not a valid steps-per-period value\n",
-			steps_per_period);
-		return -EINVAL;
+	encoder->dev = dev;
+	encoder->absolute_encoder =
+		device_property_read_bool(dev,
+					  "rotary-encoder,absolute-encoder");
+	if (encoder->absolute_encoder) {
+		handler = rotary_absolute_encoder_irq;
+	} else {
+		switch (steps_per_period >> (encoder->gpios->ndescs - 2)) {
+		case 4:
+			handler = &rotary_encoder_quarter_period_irq;
+			encoder->last_stable =
+				rotary_encoder_get_state(encoder);
+			break;
+		case 2:
+			handler = &rotary_encoder_half_period_irq;
+			encoder->last_stable =
+				rotary_encoder_get_state(encoder);
+			break;
+		case 1:
+			handler = &rotary_encoder_irq;
+			break;
+		default:
+			dev_err(dev, "'%d' is not a valid steps-per-period value\n",
+				steps_per_period);
+			return -EINVAL;
+		}
 	}
 
 	encoder->irq =

@@ -271,6 +351,18 @@ static int rotary_encoder_probe(struct platform_device *pdev)
 	for (i = 0; i < encoder->gpios->ndescs; ++i) {
 		encoder->irq[i] = gpiod_to_irq(encoder->gpios->desc[i]);
 
+#ifdef CONFIG_INPUT_GPIO_ROTARY_ENCODER_POLL_MODE_SUPPORT
+		if (encoder->irq[i] < 0 && encoder->absolute_encoder) {
+			dev_info(dev, "Using poll mode\n");
+			err = rotary_encoder_register_poll_device(encoder);
+			if (err) {
+				dev_err(dev, "failed to register poll dev\n");
+				return err;
+			}
+			platform_set_drvdata(pdev, encoder);
+			return 0;
+		}
+#endif
 		err = devm_request_threaded_irq(dev, encoder->irq[i],
 				NULL, handler,
 				IRQF_TRIGGER_RISING | IRQF_TRIGGER_FALLING |

@@ -283,6 +375,11 @@ static int rotary_encoder_probe(struct platform_device *pdev)
 		}
 	}
 
+	input = devm_input_allocate_device(dev);
+	if (!input)
+		return -ENOMEM;
+	encoder->input = input;
+	rotary_encoder_setup_input_params(encoder);
 	err = input_register_device(input);
 	if (err) {
 		dev_err(dev, "failed to register input device\n");

-- 
2.8.2

AM335x ICE board has a rotary-encoder connected to PCA9536 I2C GPIO
expander. Add DT entry for the same.

Signed-off-by: Vignesh R <vigneshr@ti.com>
---

depends on patches to Add TI PCA9536 node posted to linux-gpio:
http://www.spinics.net/lists/arm-kernel/msg505194.html

 arch/arm/boot/dts/am335x-icev2.dts | 11 +++++++++++
 1 file changed, 11 insertions(+)

diff --git a/arch/arm/boot/dts/am335x-icev2.dts b/arch/arm/boot/dts/am335x-icev2.dts
index 7d8b8fefdf08..672dccbf203d 100644
--- a/arch/arm/boot/dts/am335x-icev2.dts
+++ b/arch/arm/boot/dts/am335x-icev2.dts

@@ -139,6 +139,17 @@
 			default-state = "off";
 		};
 	};
+
+	rotary-in0 {
+		compatible = "rotary-encoder";
+		gpios = <&pca9536 3 GPIO_ACTIVE_HIGH>,
+			<&pca9536 2 GPIO_ACTIVE_HIGH>,
+			<&pca9536 1 GPIO_ACTIVE_HIGH>,
+			<&pca9536 0 GPIO_ACTIVE_HIGH>;
+		linux,axis = <0>; /* ABS_X */
+		rotary-encoder,steps = <10>;
+		rotary-encoder,absolute-encoder;
+	};
 };
 
 &am33xx_pinmux {

-- 
2.8.2

On 05/19/2016 11:04 AM, Vignesh R wrote:

Ifdefs look dirty.

1. Double mutex_lock()? Oh, that is weird. Please document it at least
in form of lockdep annotations.

2. Maybe that should be unlock? Did you test this code?

+
+	return IRQ_HANDLED;
+}
+
+#ifdef CONFIG_INPUT_GPIO_ROTARY_ENCODER_POLL_MODE_SUPPORT

Same here and later in the code, ifdefs look dirty.

Best regards,
Krzysztof

Hi,

On 05/19/2016 04:55 PM, Krzysztof Kozlowski wrote:
[...]

On 05/19/2016 11:04 AM, Vignesh R wrote:

quoted

diff --git a/drivers/input/misc/rotary_encoder.c b/drivers/input/misc/rotary_encoder.c
index c7fc8d4fb080..8f60289fb6cd 100644
--- a/drivers/input/misc/rotary_encoder.c
+++ b/drivers/input/misc/rotary_encoder.c

@@ -25,24 +25,29 @@
 #include <linux/of.h>
 #include <linux/pm.h>
 #include <linux/property.h>
+#include <linux/input-polldev.h>
 
 #define DRV_NAME "rotary-encoder"
 
 struct rotary_encoder {
 	struct input_dev *input;
-
+#ifdef CONFIG_INPUT_GPIO_ROTARY_ENCODER_POLL_MODE_SUPPORT

Ifdefs look dirty.

I will drop this.

quoted

+	struct input_polled_dev *poll_dev;
+#endif
 	struct mutex access_mutex;
 
 	u32 steps;
 	u32 axis;
 	bool relative_axis;
 	bool rollover;
+	bool absolute_encoder;
 
 	unsigned int pos;
 
 	struct gpio_descs *gpios;
+	struct device *dev;
 
-	unsigned int *irq;
+	int *irq;
 
 	bool armed;
 	signed char dir;	/* 1 - clockwise, -1 - CCW */

@@ -67,6 +72,21 @@ static unsigned int rotary_encoder_get_state(struct rotary_encoder *encoder)
 	return ret & 3;
 }
 
+static unsigned int rotary_encoder_get_gpios_state(struct rotary_encoder
+						   *encoder)
+{
+	int i;
+	unsigned int ret = 0;
+
+	for (i = 0; i < encoder->gpios->ndescs; ++i) {
+		int val = gpiod_get_value_cansleep(encoder->gpios->desc[i]);
+
+		ret = ret << 1 | val;
+	}
+
+	return ret;
+}
+
 static void rotary_encoder_report_event(struct rotary_encoder *encoder)
 {
 	if (encoder->relative_axis) {

@@ -178,6 +198,72 @@ out:
 	return IRQ_HANDLED;
 }
 
+static void rotary_encoder_setup_input_params(struct rotary_encoder  *encoder)
+{
+	struct input_dev *input = encoder->input;
+	struct platform_device *pdev = to_platform_device(encoder->dev);
+
+	input->name = pdev->name;
+	input->id.bustype = BUS_HOST;
+	input->dev.parent = encoder->dev;
+
+	if (encoder->relative_axis)
+		input_set_capability(input, EV_REL, encoder->axis);
+	else
+		input_set_abs_params(input,
+				     encoder->axis, 0, encoder->steps, 0, 1);
+}
+
+static irqreturn_t rotary_absolute_encoder_irq(int irq, void *dev_id)
+{
+	struct rotary_encoder *encoder = dev_id;
+	unsigned int state;
+
+	mutex_lock(&encoder->access_mutex);
+
+	state = rotary_encoder_get_gpios_state(encoder);
+	if (state != encoder->last_stable) {
+		input_report_abs(encoder->input, encoder->axis, state);
+		input_sync(encoder->input);
+		encoder->last_stable = state;
+	}
+
+	mutex_lock(&encoder->access_mutex);

1. Double mutex_lock()? Oh, that is weird. Please document it at least
in form of lockdep annotations.

2. Maybe that should be unlock? Did you test this code?

Sorry... this should have been unlock.
I dont have a h/w with rotary-encoder hooked to IRQ line. It would be
much appreciated if someone with rotary-encoder hooked up to h/w IRQ
line test this patch.

quoted

+
+	return IRQ_HANDLED;
+}
+
+#ifdef CONFIG_INPUT_GPIO_ROTARY_ENCODER_POLL_MODE_SUPPORT

Same here and later in the code, ifdefs look dirty.

Ok, I will remove #ifdefs and directly select INPUT_POLLDEV when
INPUT_GPIO_ROTARY_ENCODER is selected.

-- 
Regards
Vignesh

On Thu, May 19, 2016 at 02:34:00PM +0530, Vignesh R wrote:

There are rotary-encoders where GPIO lines reflect the actual position
of the rotary encoder dial. For example, if dial points to 9, then four
GPIO lines connected to the rotary encoder will read HLLH(1001b = 9).
Add support for such rotary-encoder.
The driver relies on rotary-encoder,absolute-encoder DT property to
detect such encoders.
Since, GPIO IRQs are not necessary to work with
such encoders, optional polling mode support is added using
input_poll_dev skeleton. This is can be used by enabling
CONFIG_INPUT_GPIO_ROTARY_ENCODER_POLL_MODE_SUPPORT.

Does this really belong to a rotary encoder and not a new driver that
simply translates gpio-encoded value into ABS* event?

-- 
Dmitry

On Thu, May 19, 2016 at 02:34:00PM +0530, Vignesh R wrote:

I think this should be done as a new compatible string rather than a 
property. It seems like fundamentally different h/w.

Rob

Hello,

On Thu, May 19, 2016 at 02:34:00PM +0530, Vignesh R wrote:

There are rotary-encoders where GPIO lines reflect the actual position
of the rotary encoder dial. For example, if dial points to 9, then four
GPIO lines connected to the rotary encoder will read HLLH(1001b = 9).
Add support for such rotary-encoder.
The driver relies on rotary-encoder,absolute-encoder DT property to
detect such encoders.
Since, GPIO IRQs are not necessary to work with
such encoders, optional polling mode support is added using

I don't understand this. It's necessary in the same way as with the
already supported devices. I.e. you want to trigger an irq when the
encoder is moved and then check for it's position in the handler.

IMHO this is wrong, I'd formalize this device as:

	{
		compatible = "rotary-encoder";
		gpios = <&gpio 19 1>, <&gpio 20 0>, <...>, <...>;
		rotary-encoder,encoding = "binary";
		rotary-encoder,steps = <16>;
		rotary-encoder,steps-per-period = <16>;
		rotary-encoder,rollover;
	}

and support this with a v4 of

	http://thread.gmane.org/gmane.linux.kernel.input/48892

.

Best regards
Uwe

-- 
Pengutronix e.K.                           | Uwe Kleine-König            |
Industrial Linux Solutions                 | http://www.pengutronix.de/  |
--
To unsubscribe from this list: send the line "unsubscribe linux-input" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

On 5/20/2016 10:04 PM, Dmitry Torokhov wrote:

On Thu, May 19, 2016 at 02:34:00PM +0530, Vignesh R wrote:

quoted

There are rotary-encoders where GPIO lines reflect the actual position
of the rotary encoder dial. For example, if dial points to 9, then four
GPIO lines connected to the rotary encoder will read HLLH(1001b = 9).
Add support for such rotary-encoder.
The driver relies on rotary-encoder,absolute-encoder DT property to
detect such encoders.
Since, GPIO IRQs are not necessary to work with
such encoders, optional polling mode support is added using
input_poll_dev skeleton. This is can be used by enabling
CONFIG_INPUT_GPIO_ROTARY_ENCODER_POLL_MODE_SUPPORT.

Does this really belong to a rotary encoder and not a new driver that
simply translates gpio-encoded value into ABS* event?

Currently rotary encoder driver only supports incremental/step counting
rotary devices. However, the device that is there on am335x-ice is an
absolute encoder but, IMO, nevertheless a kind of rotary encoder. The
only difference is that there is no need to count steps and the absolute
position value is always available as binary encoded state of connected
GPIOs.
The hardware on am335x-ice is a mechanical rotary encoder switch
connected over 4 GPIOs. It is same as binary encoder described at [1]
(except there are 4 GPIO lines), so this lead me to add support in
rotary-encoder.

[1]https://en.wikipedia.org/wiki/Rotary_encoder#Standard_binary_encoding

On 5/22/2016 3:56 PM, Uwe Kleine-König wrote:

Hello,

On Thu, May 19, 2016 at 02:34:00PM +0530, Vignesh R wrote:

quoted

There are rotary-encoders where GPIO lines reflect the actual position
of the rotary encoder dial. For example, if dial points to 9, then four
GPIO lines connected to the rotary encoder will read HLLH(1001b = 9).
Add support for such rotary-encoder.
The driver relies on rotary-encoder,absolute-encoder DT property to
detect such encoders.
Since, GPIO IRQs are not necessary to work with
such encoders, optional polling mode support is added using

I don't understand this. It's necessary in the same way as with the
already supported devices. I.e. you want to trigger an irq when the
encoder is moved and then check for it's position in the handler.

Unlike already supported device, there is no need to count steps or
determine the direction of rotation. Hence, IRQ is not a requirement,
periodically polling the gpio lines is more than sufficient. With
absolute encoder, you are able to determine the current position at any
time just by looking at the gpio inputs.
Suppose we poll device at t=0ms and see gpio values are LLLH(1), if we
again poll device at t=500ms(which is what input_poll_dev helps to do)
and see that gpio values is LLHH(3), then we know that rotary encoder
has changed to 3. This can be done w/o IRQ for absolute encoders.

quoted

diff --git a/Documentation/devicetree/bindings/input/rotary-encoder.txt b/Documentation/devicetree/bindings/input/rotary-encoder.txt
index 6c9f0c8a846c..9c928dbd1500 100644
--- a/Documentation/devicetree/bindings/input/rotary-encoder.txt
+++ b/Documentation/devicetree/bindings/input/rotary-encoder.txt

@@ -12,6 +12,10 @@ Optional properties:
 - rotary-encoder,relative-axis: register a relative axis rather than an
   absolute one. Relative axis will only generate +1/-1 events on the input
   device, hence no steps need to be passed.
+- rotary-encoder,absolute-encoder: support encoders where GPIO lines
+  reflect the actual position of the rotary encoder dial. For example,
+  if dial points to 9, then four GPIO lines read HLLH(1001b = 9).
+  In this case, rotary-encoder,steps-per-period needed not be defined.

IMHO this is wrong, I'd formalize this device as:

	{
		compatible = "rotary-encoder";
		gpios = <&gpio 19 1>, <&gpio 20 0>, <...>, <...>;
		rotary-encoder,encoding = "binary";
		rotary-encoder,steps = <16>;
		rotary-encoder,steps-per-period = <16>;

The above bindings essential means quarter_period device. I would not
like to bother with all the logic in rotary_encoder_quarter_period_irq()
when we can know encoder->pos by directly reading state of gpio lines.

		rotary-encoder,rollover;
	}

and support this with a v4 of

	http://thread.gmane.org/gmane.linux.kernel.input/48892

IMHO, there needs to be separate IRQ handler for absolute encoders so
that the driver logic is greatly simplified.

--
To unsubscribe from this list: send the line "unsubscribe linux-input" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Hello,

On Mon, May 23, 2016 at 04:48:40PM +0530, R, Vignesh wrote:

On 5/22/2016 3:56 PM, Uwe Kleine-König wrote:

quoted

Hello,

On Thu, May 19, 2016 at 02:34:00PM +0530, Vignesh R wrote:

quoted

There are rotary-encoders where GPIO lines reflect the actual position
of the rotary encoder dial. For example, if dial points to 9, then four
GPIO lines connected to the rotary encoder will read HLLH(1001b = 9).
Add support for such rotary-encoder.
The driver relies on rotary-encoder,absolute-encoder DT property to
detect such encoders.
Since, GPIO IRQs are not necessary to work with
such encoders, optional polling mode support is added using

I don't understand this. It's necessary in the same way as with the
already supported devices. I.e. you want to trigger an irq when the
encoder is moved and then check for it's position in the handler.

Unlike already supported device, there is no need to count steps or
determine the direction of rotation. Hence, IRQ is not a requirement,
periodically polling the gpio lines is more than sufficient. With
absolute encoder, you are able to determine the current position at any
time just by looking at the gpio inputs.

The timing might not be that critical, but there is no reason to operate
this device without irqs, is there?

Suppose we poll device at t=0ms and see gpio values are LLLH(1), if we
again poll device at t=500ms(which is what input_poll_dev helps to do)
and see that gpio values is LLHH(3), then we know that rotary encoder
has changed to 3. This can be done w/o IRQ for absolute encoders.

quoted

quoted

diff --git a/Documentation/devicetree/bindings/input/rotary-encoder.txt b/Documentation/devicetree/bindings/input/rotary-encoder.txt
index 6c9f0c8a846c..9c928dbd1500 100644
--- a/Documentation/devicetree/bindings/input/rotary-encoder.txt
+++ b/Documentation/devicetree/bindings/input/rotary-encoder.txt

@@ -12,6 +12,10 @@ Optional properties:
 - rotary-encoder,relative-axis: register a relative axis rather than an
   absolute one. Relative axis will only generate +1/-1 events on the input
   device, hence no steps need to be passed.
+- rotary-encoder,absolute-encoder: support encoders where GPIO lines
+  reflect the actual position of the rotary encoder dial. For example,
+  if dial points to 9, then four GPIO lines read HLLH(1001b = 9).
+  In this case, rotary-encoder,steps-per-period needed not be defined.

IMHO this is wrong, I'd formalize this device as:

	{
		compatible = "rotary-encoder";
		gpios = <&gpio 19 1>, <&gpio 20 0>, <...>, <...>;
		rotary-encoder,encoding = "binary";
		rotary-encoder,steps = <16>;
		rotary-encoder,steps-per-period = <16>;

The above bindings essential means quarter_period device. I would not
like to bother with all the logic in rotary_encoder_quarter_period_irq()
when we can know encoder->pos by directly reading state of gpio lines.

OK, we have code that is more complex than it needs to be for your
device. But your device is a special case of the supported devices, so
I'd say don't bother that there is more logic in the driver than you
need and be lucky.

quoted

		rotary-encoder,rollover;
	}

and support this with a v4 of

	http://thread.gmane.org/gmane.linux.kernel.input/48892

IMHO, there needs to be separate IRQ handler for absolute encoders so
that the driver logic is greatly simplified.

As above, the added complexity can handle your case just fine. So just
use it.

One thing I see to improve is to make use of the additional GPIOs, but
this is orthogonal to the different encoding and the question if polling
is enough here.

Best regards
Uwe

-- 
Pengutronix e.K.                           | Uwe Kleine-König            |
Industrial Linux Solutions                 | http://www.pengutronix.de/  |
--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Hi,

On 05/23/2016 06:48 PM, Uwe Kleine-König wrote:

Hello,

On Mon, May 23, 2016 at 04:48:40PM +0530, R, Vignesh wrote:

quoted

On 5/22/2016 3:56 PM, Uwe Kleine-König wrote:

quoted

Hello,

On Thu, May 19, 2016 at 02:34:00PM +0530, Vignesh R wrote:

quoted

There are rotary-encoders where GPIO lines reflect the actual position
of the rotary encoder dial. For example, if dial points to 9, then four
GPIO lines connected to the rotary encoder will read HLLH(1001b = 9).
Add support for such rotary-encoder.
The driver relies on rotary-encoder,absolute-encoder DT property to
detect such encoders.
Since, GPIO IRQs are not necessary to work with
such encoders, optional polling mode support is added using

I don't understand this. It's necessary in the same way as with the
already supported devices. I.e. you want to trigger an irq when the
encoder is moved and then check for it's position in the handler.

Unlike already supported device, there is no need to count steps or
determine the direction of rotation. Hence, IRQ is not a requirement,
periodically polling the gpio lines is more than sufficient. With
absolute encoder, you are able to determine the current position at any
time just by looking at the gpio inputs.

The timing might not be that critical, but there is no reason to operate
this device without irqs, is there?

No, I am not against the use of IRQs. In fact, my patches add support
for both IRQ based(default) and polling based operation(optional). If
rotary-encoder does not have interrupt capable gpio lines connected to
it (like am335x-ice) then one can make use of the polling mode support.

quoted

Suppose we poll device at t=0ms and see gpio values are LLLH(1), if we
again poll device at t=500ms(which is what input_poll_dev helps to do)
and see that gpio values is LLHH(3), then we know that rotary encoder
has changed to 3. This can be done w/o IRQ for absolute encoders.

quoted

quoted

diff --git a/Documentation/devicetree/bindings/input/rotary-encoder.txt b/Documentation/devicetree/bindings/input/rotary-encoder.txt
index 6c9f0c8a846c..9c928dbd1500 100644
--- a/Documentation/devicetree/bindings/input/rotary-encoder.txt
+++ b/Documentation/devicetree/bindings/input/rotary-encoder.txt

@@ -12,6 +12,10 @@ Optional properties:
 - rotary-encoder,relative-axis: register a relative axis rather than an
   absolute one. Relative axis will only generate +1/-1 events on the input
   device, hence no steps need to be passed.
+- rotary-encoder,absolute-encoder: support encoders where GPIO lines
+  reflect the actual position of the rotary encoder dial. For example,
+  if dial points to 9, then four GPIO lines read HLLH(1001b = 9).
+  In this case, rotary-encoder,steps-per-period needed not be defined.

IMHO this is wrong, I'd formalize this device as:

	{
		compatible = "rotary-encoder";
		gpios = <&gpio 19 1>, <&gpio 20 0>, <...>, <...>;
		rotary-encoder,encoding = "binary";
		rotary-encoder,steps = <16>;
		rotary-encoder,steps-per-period = <16>;

The above bindings essential means quarter_period device. I would not
like to bother with all the logic in rotary_encoder_quarter_period_irq()
when we can know encoder->pos by directly reading state of gpio lines.

OK, we have code that is more complex than it needs to be for your
device. But your device is a special case of the supported devices, so
I'd say don't bother that there is more logic in the driver than you
need and be lucky.

More complexity is just a overhead. Since, encoder can be turned at a
rate faster than handling of IRQs (rotary_encoder_quarter_period_irq()
is threaded IRQ hence, priority is not close to real time), some states
can be missed. rotary_encoder_quarter_period_irq() is not robust in this
case, reading gpios directly is more suitable option. I see similar
views expressed in previously[1]

[1]http://lists.infradead.org/pipermail/linux-arm-kernel/2015-December/391196.html

-- 
Regards
Vignesh
--
To unsubscribe from this list: send the line "unsubscribe linux-input" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Hello,

On Tue, May 24, 2016 at 10:39:26AM +0530, Vignesh R wrote:

On 05/23/2016 06:48 PM, Uwe Kleine-König wrote:

quoted

On Mon, May 23, 2016 at 04:48:40PM +0530, R, Vignesh wrote:

quoted

On 5/22/2016 3:56 PM, Uwe Kleine-König wrote:

quoted

On Thu, May 19, 2016 at 02:34:00PM +0530, Vignesh R wrote:

quoted

+- rotary-encoder,absolute-encoder: support encoders where GPIO lines
+  reflect the actual position of the rotary encoder dial. For example,
+  if dial points to 9, then four GPIO lines read HLLH(1001b = 9).
+  In this case, rotary-encoder,steps-per-period needed not be defined.

IMHO this is wrong, I'd formalize this device as:

	{
		compatible = "rotary-encoder";
		gpios = <&gpio 19 1>, <&gpio 20 0>, <...>, <...>;
		rotary-encoder,encoding = "binary";
		rotary-encoder,steps = <16>;
		rotary-encoder,steps-per-period = <16>;

The above bindings essential means quarter_period device. I would not
like to bother with all the logic in rotary_encoder_quarter_period_irq()
when we can know encoder->pos by directly reading state of gpio lines.

OK, we have code that is more complex than it needs to be for your
device. But your device is a special case of the supported devices, so
I'd say don't bother that there is more logic in the driver than you
need and be lucky.

More complexity is just a overhead. Since, encoder can be turned at a
rate faster than handling of IRQs (rotary_encoder_quarter_period_irq()
is threaded IRQ hence, priority is not close to real time), some states

This problem isn't unique to your hardware. An "ordinary" encoder with
just two GPIOs and more than one period can be rotated faster than
1/irq_latency, too. There are two things that can be done:

 - undo the conversion to threaded irqs; or
 - read out the gpios in the fast handler and only delay decoding and
   reporting of the event

Both approaches have their disadvantages.

can be missed. rotary_encoder_quarter_period_irq() is not robust in this
case, reading gpios directly is more suitable option. I see similar
views expressed in previously[1]

[1]http://lists.infradead.org/pipermail/linux-arm-kernel/2015-December/391196.html

IMHO the right thing to do is to improve
rotary_encoder_quarter_period_irq (and also the other handlers for full
and half period mode) to make use of additional GPIOs. This way all
types of devices benefit and more code is shared.

Best regards
Uwe

-- 
Pengutronix e.K.                           | Uwe Kleine-König            |
Industrial Linux Solutions                 | http://www.pengutronix.de/  |
--
To unsubscribe from this list: send the line "unsubscribe linux-input" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

On 05/24/2016 01:50 PM, Uwe Kleine-König wrote:

Hello,

On Tue, May 24, 2016 at 10:39:26AM +0530, Vignesh R wrote:

quoted

On 05/23/2016 06:48 PM, Uwe Kleine-König wrote:

quoted

On Mon, May 23, 2016 at 04:48:40PM +0530, R, Vignesh wrote:

quoted

On 5/22/2016 3:56 PM, Uwe Kleine-König wrote:

quoted

On Thu, May 19, 2016 at 02:34:00PM +0530, Vignesh R wrote:

[...]

quoted

quoted

OK, we have code that is more complex than it needs to be for your
device. But your device is a special case of the supported devices, so
I'd say don't bother that there is more logic in the driver than you
need and be lucky.

More complexity is just a overhead. Since, encoder can be turned at a
rate faster than handling of IRQs (rotary_encoder_quarter_period_irq()
is threaded IRQ hence, priority is not close to real time), some states

This problem isn't unique to your hardware. An "ordinary" encoder with
just two GPIOs and more than one period can be rotated faster than
1/irq_latency, too.

But my hardware should not be affected by this problem. The whole point
of absolute encoder is to overcome the difficulty of software keeping
track of steps, one can read the gpios state anytime and find out what
value is the encoder pointing to.

There are two things that can be done:

 - undo the conversion to threaded irqs; or
 - read out the gpios in the fast handler and only delay decoding and
   reporting of the event

Both approaches have their disadvantages.

And both cannot guarantee that an event is not missed (on a loaded
system) and all of the state logic goes for a toss. For binary encoding
multiple IRQs(thats why incremental encoders are usually gray coded)
will fire at same time and handling all of the them is a considerable
overhead, there is good chance that some events are missed.

quoted

can be missed. rotary_encoder_quarter_period_irq() is not robust in this
case, reading gpios directly is more suitable option. I see similar
views expressed in previously[1]

[1]http://lists.infradead.org/pipermail/linux-arm-kernel/2015-December/391196.html

IMHO the right thing to do is to improve
rotary_encoder_quarter_period_irq (and also the other handlers for full
and half period mode) to make use of additional GPIOs. 

I doubt there exists any incremental encoder with more than 2
gpios(except for the extra reference GPIO or Z signal). So modifying
full/half period handlers maybe unnecessary.

This way all types of devices benefit and more code is shared.

Sorry.. but IMHO, there is little code sharing and more complexity. So,
I will leave it to the maintainer to decide whats the best approach here.

-- 
Regards
Vignesh

Hi Dmitry,

On 05/23/2016 02:48 PM, R, Vignesh wrote:

On 5/20/2016 10:04 PM, Dmitry Torokhov wrote:

quoted

On Thu, May 19, 2016 at 02:34:00PM +0530, Vignesh R wrote:

quoted

There are rotary-encoders where GPIO lines reflect the actual position
of the rotary encoder dial. For example, if dial points to 9, then four
GPIO lines connected to the rotary encoder will read HLLH(1001b = 9).
Add support for such rotary-encoder.
The driver relies on rotary-encoder,absolute-encoder DT property to
detect such encoders.
Since, GPIO IRQs are not necessary to work with
such encoders, optional polling mode support is added using
input_poll_dev skeleton. This is can be used by enabling
CONFIG_INPUT_GPIO_ROTARY_ENCODER_POLL_MODE_SUPPORT.

Does this really belong to a rotary encoder and not a new driver that
simply translates gpio-encoded value into ABS* event?

Currently rotary encoder driver only supports incremental/step counting
rotary devices. However, the device that is there on am335x-ice is an
absolute encoder but, IMO, nevertheless a kind of rotary encoder. The
only difference is that there is no need to count steps and the absolute
position value is always available as binary encoded state of connected
GPIOs.
The hardware on am335x-ice is a mechanical rotary encoder switch
connected over 4 GPIOs. It is same as binary encoder described at [1]
(except there are 4 GPIO lines), so this lead me to add support in
rotary-encoder.

[1]https://en.wikipedia.org/wiki/Rotary_encoder#Standard_binary_encoding

Could you please comment on how would you like to support above
described encoder: As a new driver or with existing driver with new
compatible/mode setting via DT or as suggest by Uwe in another reply?
IMHO, supporting using existing driver with new mode/compatible string
looks a better option as the hardware is a kind of rotary-encoder.

-- 
Regards
Vignesh

Hi Dmitry,

On Wednesday 25 May 2016 02:14 PM, Vignesh R wrote:

Hi Dmitry,

On 05/23/2016 02:48 PM, R, Vignesh wrote:

quoted

On 5/20/2016 10:04 PM, Dmitry Torokhov wrote:

quoted

On Thu, May 19, 2016 at 02:34:00PM +0530, Vignesh R wrote:

quoted

There are rotary-encoders where GPIO lines reflect the actual position
of the rotary encoder dial. For example, if dial points to 9, then four
GPIO lines connected to the rotary encoder will read HLLH(1001b = 9).
Add support for such rotary-encoder.
The driver relies on rotary-encoder,absolute-encoder DT property to
detect such encoders.
Since, GPIO IRQs are not necessary to work with
such encoders, optional polling mode support is added using
input_poll_dev skeleton. This is can be used by enabling
CONFIG_INPUT_GPIO_ROTARY_ENCODER_POLL_MODE_SUPPORT.

Does this really belong to a rotary encoder and not a new driver that
simply translates gpio-encoded value into ABS* event?

Currently rotary encoder driver only supports incremental/step counting
rotary devices. However, the device that is there on am335x-ice is an
absolute encoder but, IMO, nevertheless a kind of rotary encoder. The
only difference is that there is no need to count steps and the absolute
position value is always available as binary encoded state of connected
GPIOs.
The hardware on am335x-ice is a mechanical rotary encoder switch
connected over 4 GPIOs. It is same as binary encoder described at [1]
(except there are 4 GPIO lines), so this lead me to add support in
rotary-encoder.

[1]https://en.wikipedia.org/wiki/Rotary_encoder#Standard_binary_encoding

Could you please comment on how would you like to support above
described encoder: As a new driver or with existing driver with new
compatible/mode setting via DT or as suggest by Uwe in another reply?
IMHO, supporting using existing driver with new mode/compatible string
looks a better option as the hardware is a kind of rotary-encoder.

It would be great if you could comment on how would you like to see
support for absolute rotary-encoder(the one I described above)? As a new
driver or handle it by adding new compatible to existing rotary-encoder
driver?


-- 
Regards
Vignesh

Dmitry,

On 6/16/2016 4:17 PM, Vignesh R wrote:
[...]

quoted

quoted

On 5/20/2016 10:04 PM, Dmitry Torokhov wrote:

quoted

On Thu, May 19, 2016 at 02:34:00PM +0530, Vignesh R wrote:

quoted

There are rotary-encoders where GPIO lines reflect the actual position
of the rotary encoder dial. For example, if dial points to 9, then four
GPIO lines connected to the rotary encoder will read HLLH(1001b = 9).
Add support for such rotary-encoder.
The driver relies on rotary-encoder,absolute-encoder DT property to
detect such encoders.
Since, GPIO IRQs are not necessary to work with
such encoders, optional polling mode support is added using
input_poll_dev skeleton. This is can be used by enabling
CONFIG_INPUT_GPIO_ROTARY_ENCODER_POLL_MODE_SUPPORT.

Does this really belong to a rotary encoder and not a new driver that
simply translates gpio-encoded value into ABS* event?

Currently rotary encoder driver only supports incremental/step counting
rotary devices. However, the device that is there on am335x-ice is an
absolute encoder but, IMO, nevertheless a kind of rotary encoder. The
only difference is that there is no need to count steps and the absolute
position value is always available as binary encoded state of connected
GPIOs.
The hardware on am335x-ice is a mechanical rotary encoder switch
connected over 4 GPIOs. It is same as binary encoder described at [1]
(except there are 4 GPIO lines), so this lead me to add support in
rotary-encoder.

[1]https://en.wikipedia.org/wiki/Rotary_encoder#Standard_binary_encoding

Could you please comment on how would you like to support above
described encoder: As a new driver or with existing driver with new
compatible/mode setting via DT or as suggest by Uwe in another reply?
IMHO, supporting using existing driver with new mode/compatible string
looks a better option as the hardware is a kind of rotary-encoder.

Could you please suggest the right solution here? Do you think new
driver is need to support above described absolute encoder h/w? It would
be great to have your input now, so that I can work on the new set of
patches and submit them for v4.9 merge window as soon as v4.8 merge
window closes?

Regards
Vignesh