[PATCH 0/3] cap1106: add support for cap11xx variants

[PATCH 0/3] cap1106: add support for cap11xx variants

[Matt Ranostay]

These patches enable using cap11xx devices that have different
number of capacitance channels, and using active-high on the
interrupt output.

Matt Ranostay (3):
  cap1106: Add support for various cap11xx devices
  cap1106: support for active-high interrupt option
  dt: cap1106 active-high property addition

 .../devicetree/bindings/input/cap1106.txt          |  4 ++
 drivers/input/keyboard/cap1106.c                   | 58 ++++++++++++----------
 2 files changed, 35 insertions(+), 27 deletions(-)

-- 
1.9.1

[PATCH 1/3] cap1106: Add support for various cap11xx devices

[Matt Ranostay]

Several other variants of the cap11xx device exists with a varying
number of capacitance detection channels. Add support for creating
the channels dynamically.

Signed-off-by: Matt Ranostay <mranostay@gmail.com>
---
 drivers/input/keyboard/cap1106.c | 54 +++++++++++++++++++---------------------
 1 file changed, 26 insertions(+), 28 deletions(-)

diff --git a/drivers/input/keyboard/cap1106.c b/drivers/input/keyboard/cap1106.c
index d70b65a..b9c43b5 100644
--- a/drivers/input/keyboard/cap1106.c
+++ b/drivers/input/keyboard/cap1106.c
@@ -55,8 +55,6 @@
 #define CAP1106_REG_MANUFACTURER_ID	0xfe
 #define CAP1106_REG_REVISION		0xff
 
-#define CAP1106_NUM_CHN 6
-#define CAP1106_PRODUCT_ID	0x55
 #define CAP1106_MANUFACTURER_ID	0x5d
 
 struct cap1106_priv {
@@ -64,7 +62,8 @@ struct cap1106_priv {
 	struct input_dev *idev;
 
 	/* config */
-	unsigned short keycodes[CAP1106_NUM_CHN];
+	u32 *keycodes;
+	unsigned int num_channels;
 };
 
 static const struct reg_default cap1106_reg_defaults[] = {
@@ -151,7 +150,7 @@ static irqreturn_t cap1106_thread_func(int irq_num, void *data)
 	if (ret < 0)
 		goto out;
 
-	for (i = 0; i < CAP1106_NUM_CHN; i++)
+	for (i = 0; i < priv->num_channels; i++)
 		input_report_key(priv->idev, priv->keycodes[i],
 				 status & (1 << i));
 
@@ -189,27 +188,23 @@ static int cap1106_i2c_probe(struct i2c_client *i2c_client,
 	struct cap1106_priv *priv;
 	struct device_node *node;
 	int i, error, irq, gain = 0;
-	unsigned int val, rev;
-	u32 gain32, keycodes[CAP1106_NUM_CHN];
+	unsigned int val, prod, rev;
+	u32 gain32;
 
 	priv = devm_kzalloc(dev, sizeof(*priv), GFP_KERNEL);
 	if (!priv)
 		return -ENOMEM;
 
+	priv->num_channels = (unsigned long) id->driver_data;
+	priv->keycodes = devm_kzalloc(dev,
+		sizeof(u32) * priv->num_channels, GFP_KERNEL);
+	if (!priv->keycodes)
+		return -ENOMEM;
+
 	priv->regmap = devm_regmap_init_i2c(i2c_client, &cap1106_regmap_config);
 	if (IS_ERR(priv->regmap))
 		return PTR_ERR(priv->regmap);
 
-	error = regmap_read(priv->regmap, CAP1106_REG_PRODUCT_ID, &val);
-	if (error)
-		return error;
-
-	if (val != CAP1106_PRODUCT_ID) {
-		dev_err(dev, "Product ID: Got 0x%02x, expected 0x%02x\n",
-			val, CAP1106_PRODUCT_ID);
-		return -ENODEV;
-	}
-
 	error = regmap_read(priv->regmap, CAP1106_REG_MANUFACTURER_ID, &val);
 	if (error)
 		return error;
@@ -220,6 +215,10 @@ static int cap1106_i2c_probe(struct i2c_client *i2c_client,
 		return -ENODEV;
 	}
 
+	error = regmap_read(priv->regmap, CAP1106_REG_PRODUCT_ID, &prod);
+	if (error < 0)
+		return error;
+
 	error = regmap_read(priv->regmap, CAP1106_REG_REVISION, &rev);
 	if (error < 0)
 		return error;
@@ -235,17 +234,12 @@ static int cap1106_i2c_probe(struct i2c_client *i2c_client,
 			dev_err(dev, "Invalid sensor-gain value %d\n", gain32);
 	}
 
-	BUILD_BUG_ON(ARRAY_SIZE(keycodes) != ARRAY_SIZE(priv->keycodes));
-
 	/* Provide some useful defaults */
-	for (i = 0; i < ARRAY_SIZE(keycodes); i++)
-		keycodes[i] = KEY_A + i;
+	for (i = 0; i < priv->num_channels; i++)
+		priv->keycodes[i] = KEY_A + i;
 
 	of_property_read_u32_array(node, "linux,keycodes",
-				   keycodes, ARRAY_SIZE(keycodes));
-
-	for (i = 0; i < ARRAY_SIZE(keycodes); i++)
-		priv->keycodes[i] = keycodes[i];
+				  priv->keycodes, priv->num_channels);
 
 	error = regmap_update_bits(priv->regmap, CAP1106_REG_MAIN_CONTROL,
 				   CAP1106_REG_MAIN_CONTROL_GAIN_MASK,
@@ -269,17 +263,17 @@ static int cap1106_i2c_probe(struct i2c_client *i2c_client,
 	if (of_property_read_bool(node, "autorepeat"))
 		__set_bit(EV_REP, priv->idev->evbit);
 
-	for (i = 0; i < CAP1106_NUM_CHN; i++)
+	for (i = 0; i < priv->num_channels; i++)
 		__set_bit(priv->keycodes[i], priv->idev->keybit);
 
 	__clear_bit(KEY_RESERVED, priv->idev->keybit);
 
 	priv->idev->keycode = priv->keycodes;
 	priv->idev->keycodesize = sizeof(priv->keycodes[0]);
-	priv->idev->keycodemax = ARRAY_SIZE(priv->keycodes);
+	priv->idev->keycodemax = priv->num_channels;
 
 	priv->idev->id.vendor = CAP1106_MANUFACTURER_ID;
-	priv->idev->id.product = CAP1106_PRODUCT_ID;
+	priv->idev->id.product = prod;
 	priv->idev->id.version = rev;
 
 	priv->idev->open = cap1106_input_open;
@@ -313,12 +307,16 @@ static int cap1106_i2c_probe(struct i2c_client *i2c_client,
 
 static const struct of_device_id cap1106_dt_ids[] = {
 	{ .compatible = "microchip,cap1106", },
+	{ .compatible = "microchip,cap1126", },
+	{ .compatible = "microchip,cap1188", },
 	{}
 };
 MODULE_DEVICE_TABLE(of, cap1106_dt_ids);
 
 static const struct i2c_device_id cap1106_i2c_ids[] = {
-	{ "cap1106", 0 },
+	{ "cap1106", 6 },
+	{ "cap1126", 6 },
+	{ "cap1188", 8 },
 	{}
 };
 MODULE_DEVICE_TABLE(i2c, cap1106_i2c_ids);
-- 
1.9.1

[PATCH 2/3] cap1106: support for active-high interrupt option

[Matt Ranostay]

Some applications need to use the active-high push-pull interrupt
option. This allows it be enabled in the device tree child node.

Signed-off-by: Matt Ranostay <mranostay@gmail.com>
---
 drivers/input/keyboard/cap1106.c | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/drivers/input/keyboard/cap1106.c b/drivers/input/keyboard/cap1106.c
index b9c43b5..33e2590 100644
--- a/drivers/input/keyboard/cap1106.c
+++ b/drivers/input/keyboard/cap1106.c
@@ -234,6 +234,12 @@ static int cap1106_i2c_probe(struct i2c_client *i2c_client,
 			dev_err(dev, "Invalid sensor-gain value %d\n", gain32);
 	}
 
+	if (of_property_read_bool(node, "microchip,active-high")) {
+		error = regmap_write(priv->regmap, CAP1106_REG_CONFIG2, 0);
+		if (error)
+			return error;
+	}
+
 	/* Provide some useful defaults */
 	for (i = 0; i < priv->num_channels; i++)
 		priv->keycodes[i] = KEY_A + i;
-- 
1.9.1

[PATCH 3/3] dt: cap1106 active-high property addition

[Matt Ranostay]

Documenting the microchip,active-high property for interrupt pin
behavior.

Signed-off-by: Matt Ranostay <mranostay@gmail.com>
---
 Documentation/devicetree/bindings/input/cap1106.txt | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/Documentation/devicetree/bindings/input/cap1106.txt b/Documentation/devicetree/bindings/input/cap1106.txt
index 4b46390..7e64fd6 100644
--- a/Documentation/devicetree/bindings/input/cap1106.txt
+++ b/Documentation/devicetree/bindings/input/cap1106.txt
@@ -26,6 +26,10 @@ Optional properties:
 				Valid values are 1, 2, 4, and 8.
 				By default, a gain of 1 is set.
 
+	microchip,active-high:	By default the interrupt pin  is active low open
+				drain. This property allows using the active high
+				push-pull output.
+
 	linux,keycodes:		Specifies an array of numeric keycode values to
 				be used for the channels. If this property is
 				omitted, KEY_A, KEY_B, etc are used as
-- 
1.9.1

Re: [PATCH 1/3] cap1106: Add support for various cap11xx devices

[Daniel Mack]

Hi,

On 09/21/2014 05:01 AM, Matt Ranostay wrote:
> Several other variants of the cap11xx device exists with a varying
> number of capacitance detection channels. Add support for creating
> the channels dynamically.

Thanks for the patches!

> 
> Signed-off-by: Matt Ranostay <mranostay@gmail.com>
> ---
>  drivers/input/keyboard/cap1106.c | 54 +++++++++++++++++++---------------------

Please also add a patch to rename the file to cap11xx.c, and make sure
to export the patch with 'git format-patch -M' to detect the rename.

> diff --git a/drivers/input/keyboard/cap1106.c b/drivers/input/keyboard/cap1106.c
> index d70b65a..b9c43b5 100644
> --- a/drivers/input/keyboard/cap1106.c
> +++ b/drivers/input/keyboard/cap1106.c
> @@ -55,8 +55,6 @@
>  #define CAP1106_REG_MANUFACTURER_ID	0xfe
>  #define CAP1106_REG_REVISION		0xff
>  
> -#define CAP1106_NUM_CHN 6
> -#define CAP1106_PRODUCT_ID	0x55
>  #define CAP1106_MANUFACTURER_ID	0x5d
>  
>  struct cap1106_priv {
> @@ -64,7 +62,8 @@ struct cap1106_priv {
>  	struct input_dev *idev;
>  
>  	/* config */
> -	unsigned short keycodes[CAP1106_NUM_CHN];
> +	u32 *keycodes;

unsigned short *, please. See below.

> @@ -189,27 +188,23 @@ static int cap1106_i2c_probe(struct i2c_client *i2c_client,
>  	struct cap1106_priv *priv;
>  	struct device_node *node;
>  	int i, error, irq, gain = 0;
> -	unsigned int val, rev;
> -	u32 gain32, keycodes[CAP1106_NUM_CHN];
> +	unsigned int val, prod, rev;
> +	u32 gain32;
>  
>  	priv = devm_kzalloc(dev, sizeof(*priv), GFP_KERNEL);
>  	if (!priv)
>  		return -ENOMEM;
>  
> +	priv->num_channels = (unsigned long) id->driver_data;

priv->num_channels is unsigned int.

Also, a BUG_ON(!priv->num_channels) wouldn't harm.

> +	priv->keycodes = devm_kzalloc(dev,
> +		sizeof(u32) * priv->num_channels, GFP_KERNEL);

Use devm_kcalloc() to allocate an array.

> @@ -235,17 +234,12 @@ static int cap1106_i2c_probe(struct i2c_client *i2c_client,
>  			dev_err(dev, "Invalid sensor-gain value %d\n", gain32);
>  	}
>  
> -	BUILD_BUG_ON(ARRAY_SIZE(keycodes) != ARRAY_SIZE(priv->keycodes));
> -
>  	/* Provide some useful defaults */
> -	for (i = 0; i < ARRAY_SIZE(keycodes); i++)
> -		keycodes[i] = KEY_A + i;
> +	for (i = 0; i < priv->num_channels; i++)
> +		priv->keycodes[i] = KEY_A + i;
>  
>  	of_property_read_u32_array(node, "linux,keycodes",
> -				   keycodes, ARRAY_SIZE(keycodes));
> -
> -	for (i = 0; i < ARRAY_SIZE(keycodes); i++)
> -		priv->keycodes[i] = keycodes[i];
> +				  priv->keycodes, priv->num_channels);

Hmm, no. Internally, you have to store the keycodes as unsigned short,
otherwise EVIOC{G|S}KEYCODE from usespace doesn't work.

of_property_read_u16_array() should work here for unsigned short, I
guess. Otherwise, you'd have to open-code the routine.

> @@ -313,12 +307,16 @@ static int cap1106_i2c_probe(struct i2c_client *i2c_client,
>  
>  static const struct of_device_id cap1106_dt_ids[] = {
>  	{ .compatible = "microchip,cap1106", },
> +	{ .compatible = "microchip,cap1126", },
> +	{ .compatible = "microchip,cap1188", },

Hmm, how can that work unless you set .data to the number of channels
here? Did you test that with a DT-enabled board?


Thanks,
Daniel

Re: [PATCH 2/3] cap1106: support for active-high interrupt option

[Daniel Mack]

On 09/21/2014 05:01 AM, Matt Ranostay wrote:
> Some applications need to use the active-high push-pull interrupt
> option. This allows it be enabled in the device tree child node.
> 
> Signed-off-by: Matt Ranostay <mranostay@gmail.com>
> ---
>  drivers/input/keyboard/cap1106.c | 6 ++++++
>  1 file changed, 6 insertions(+)
> 
> diff --git a/drivers/input/keyboard/cap1106.c b/drivers/input/keyboard/cap1106.c
> index b9c43b5..33e2590 100644
> --- a/drivers/input/keyboard/cap1106.c
> +++ b/drivers/input/keyboard/cap1106.c
> @@ -234,6 +234,12 @@ static int cap1106_i2c_probe(struct i2c_client *i2c_client,
>  			dev_err(dev, "Invalid sensor-gain value %d\n", gain32);
>  	}
>  
> +	if (of_property_read_bool(node, "microchip,active-high")) {

I think the name of that property should make clear it's only changing
the interrupt output driver configuration. What about
"microchip,irq-active-high"?

Also, patch 3/3, which documents this new property, can be squashed into
this one.

> +		error = regmap_write(priv->regmap, CAP1106_REG_CONFIG2, 0);

This register controls a lot more details than that. Overriding it with
0 doesn't seem right. Please use regmap_update_bits() to just update
ALT_POL, and also add a #define for it, so the next reader knows what
the code is doing :)


Thanks,
Daniel

Re: [PATCH 1/3] cap1106: Add support for various cap11xx devices

[Daniel Mack]

Hi,

On 09/21/2014 05:01 AM, Matt Ranostay wrote:
>  	priv->regmap = devm_regmap_init_i2c(i2c_client, &cap1106_regmap_config);
>  	if (IS_ERR(priv->regmap))
>  		return PTR_ERR(priv->regmap);
>  
> -	error = regmap_read(priv->regmap, CAP1106_REG_PRODUCT_ID, &val);
> -	if (error)
> -		return error;
> -
> -	if (val != CAP1106_PRODUCT_ID) {
> -		dev_err(dev, "Product ID: Got 0x%02x, expected 0x%02x\n",
> -			val, CAP1106_PRODUCT_ID);
> -		return -ENODEV;
> -	}
> -

Btw - the purpose of this code was to detect board configuration
mismatch. After all, I2C lacks a way to properly identify peripherals,
so the more runtime checks we do at probe time, the more of a chance we
have to detect wrong setups. This device is actually well implemented
and tells us something about itself.

Hence, I'd propose to define a structure like this:

struct cap11xx_hw_model {
	uint8_t product_id;
	unsigned int num_channels;
};

... and attach instances of that to the members of cap1106_dt_ids[] and
cap1106_i2c_ids[]. In the probe function, check that the contents of
CAP1106_PRODUCT_ID match what is expected by the configured model.


Thanks,
Daniel

Re: [PATCH 1/3] cap1106: Add support for various cap11xx devices

[Matt Ranostay]

On Sun, Sep 21, 2014 at 2:58 AM, Daniel Mack <daniel@zonque.org> wrote:
>
> Hi,
>
> On 09/21/2014 05:01 AM, Matt Ranostay wrote:
> > Several other variants of the cap11xx device exists with a varying
> > number of capacitance detection channels. Add support for creating
> > the channels dynamically.
>
> Thanks for the patches!
>
> >
> > Signed-off-by: Matt Ranostay <mranostay@gmail.com>
> > ---
> >  drivers/input/keyboard/cap1106.c | 54 +++++++++++++++++++---------------------
>
> Please also add a patch to rename the file to cap11xx.c, and make sure
> to export the patch with 'git format-patch -M' to detect the rename.
>
> > diff --git a/drivers/input/keyboard/cap1106.c b/drivers/input/keyboard/cap1106.c
> > index d70b65a..b9c43b5 100644
> > --- a/drivers/input/keyboard/cap1106.c
> > +++ b/drivers/input/keyboard/cap1106.c
> > @@ -55,8 +55,6 @@
> >  #define CAP1106_REG_MANUFACTURER_ID  0xfe
> >  #define CAP1106_REG_REVISION         0xff
> >
> > -#define CAP1106_NUM_CHN 6
> > -#define CAP1106_PRODUCT_ID   0x55
> >  #define CAP1106_MANUFACTURER_ID      0x5d
> >
> >  struct cap1106_priv {
> > @@ -64,7 +62,8 @@ struct cap1106_priv {
> >       struct input_dev *idev;
> >
> >       /* config */
> > -     unsigned short keycodes[CAP1106_NUM_CHN];
> > +     u32 *keycodes;
>
> unsigned short *, please. See below.
>
> > @@ -189,27 +188,23 @@ static int cap1106_i2c_probe(struct i2c_client *i2c_client,
> >       struct cap1106_priv *priv;
> >       struct device_node *node;
> >       int i, error, irq, gain = 0;
> > -     unsigned int val, rev;
> > -     u32 gain32, keycodes[CAP1106_NUM_CHN];
> > +     unsigned int val, prod, rev;
> > +     u32 gain32;
> >
> >       priv = devm_kzalloc(dev, sizeof(*priv), GFP_KERNEL);
> >       if (!priv)
> >               return -ENOMEM;
> >
> > +     priv->num_channels = (unsigned long) id->driver_data;
>
> priv->num_channels is unsigned int.
>
> Also, a BUG_ON(!priv->num_channels) wouldn't harm.

Oops. Will fix..

>
> > +     priv->keycodes = devm_kzalloc(dev,
> > +             sizeof(u32) * priv->num_channels, GFP_KERNEL);
>
> Use devm_kcalloc() to allocate an array.
>
> > @@ -235,17 +234,12 @@ static int cap1106_i2c_probe(struct i2c_client *i2c_client,
> >                       dev_err(dev, "Invalid sensor-gain value %d\n", gain32);
> >       }
> >
> > -     BUILD_BUG_ON(ARRAY_SIZE(keycodes) != ARRAY_SIZE(priv->keycodes));
> > -
> >       /* Provide some useful defaults */
> > -     for (i = 0; i < ARRAY_SIZE(keycodes); i++)
> > -             keycodes[i] = KEY_A + i;
> > +     for (i = 0; i < priv->num_channels; i++)
> > +             priv->keycodes[i] = KEY_A + i;
> >
> >       of_property_read_u32_array(node, "linux,keycodes",
> > -                                keycodes, ARRAY_SIZE(keycodes));
> > -
> > -     for (i = 0; i < ARRAY_SIZE(keycodes); i++)
> > -             priv->keycodes[i] = keycodes[i];
> > +                               priv->keycodes, priv->num_channels);
>
> Hmm, no. Internally, you have to store the keycodes as unsigned short,
> otherwise EVIOC{G|S}KEYCODE from usespace doesn't work.
>
> of_property_read_u16_array() should work here for unsigned short, I
> guess. Otherwise, you'd have to open-code the routine.
>
Ok I'l test that.

> > @@ -313,12 +307,16 @@ static int cap1106_i2c_probe(struct i2c_client *i2c_client,
> >
> >  static const struct of_device_id cap1106_dt_ids[] = {
> >       { .compatible = "microchip,cap1106", },
> > +     { .compatible = "microchip,cap1126", },
> > +     { .compatible = "microchip,cap1188", },
>
> Hmm, how can that work unless you set .data to the number of channels
> here? Did you test that with a DT-enabled board?
>
Yes it was tested on a BBB.  The num_channels is set from cap1106_i2c_ids

>
> Thanks,
> Daniel

Re: [PATCH 1/3] cap1106: Add support for various cap11xx devices

[Matt Ranostay]

On Sun, Sep 21, 2014 at 2:58 AM, Daniel Mack <daniel@zonque.org> wrote:
> Hi,
>
> On 09/21/2014 05:01 AM, Matt Ranostay wrote:
>> Several other variants of the cap11xx device exists with a varying
>> number of capacitance detection channels. Add support for creating
>> the channels dynamically.
>
> Thanks for the patches!
>
>>
>> Signed-off-by: Matt Ranostay <mranostay@gmail.com>
>> ---
>>  drivers/input/keyboard/cap1106.c | 54 +++++++++++++++++++---------------------
>
> Please also add a patch to rename the file to cap11xx.c, and make sure
> to export the patch with 'git format-patch -M' to detect the rename.
>
>> diff --git a/drivers/input/keyboard/cap1106.c b/drivers/input/keyboard/cap1106.c
>> index d70b65a..b9c43b5 100644
>> --- a/drivers/input/keyboard/cap1106.c
>> +++ b/drivers/input/keyboard/cap1106.c
>> @@ -55,8 +55,6 @@
>>  #define CAP1106_REG_MANUFACTURER_ID  0xfe
>>  #define CAP1106_REG_REVISION         0xff
>>
>> -#define CAP1106_NUM_CHN 6
>> -#define CAP1106_PRODUCT_ID   0x55
>>  #define CAP1106_MANUFACTURER_ID      0x5d
>>
>>  struct cap1106_priv {
>> @@ -64,7 +62,8 @@ struct cap1106_priv {
>>       struct input_dev *idev;
>>
>>       /* config */
>> -     unsigned short keycodes[CAP1106_NUM_CHN];
>> +     u32 *keycodes;
>
> unsigned short *, please. See below.
>
>> @@ -189,27 +188,23 @@ static int cap1106_i2c_probe(struct i2c_client *i2c_client,
>>       struct cap1106_priv *priv;
>>       struct device_node *node;
>>       int i, error, irq, gain = 0;
>> -     unsigned int val, rev;
>> -     u32 gain32, keycodes[CAP1106_NUM_CHN];
>> +     unsigned int val, prod, rev;
>> +     u32 gain32;
>>
>>       priv = devm_kzalloc(dev, sizeof(*priv), GFP_KERNEL);
>>       if (!priv)
>>               return -ENOMEM;
>>
>> +     priv->num_channels = (unsigned long) id->driver_data;
>
> priv->num_channels is unsigned int.
>
> Also, a BUG_ON(!priv->num_channels) wouldn't harm.
>
>> +     priv->keycodes = devm_kzalloc(dev,
>> +             sizeof(u32) * priv->num_channels, GFP_KERNEL);
>
> Use devm_kcalloc() to allocate an array.
>
>> @@ -235,17 +234,12 @@ static int cap1106_i2c_probe(struct i2c_client *i2c_client,
>>                       dev_err(dev, "Invalid sensor-gain value %d\n", gain32);
>>       }
>>
>> -     BUILD_BUG_ON(ARRAY_SIZE(keycodes) != ARRAY_SIZE(priv->keycodes));
>> -
>>       /* Provide some useful defaults */
>> -     for (i = 0; i < ARRAY_SIZE(keycodes); i++)
>> -             keycodes[i] = KEY_A + i;
>> +     for (i = 0; i < priv->num_channels; i++)
>> +             priv->keycodes[i] = KEY_A + i;
>>
>>       of_property_read_u32_array(node, "linux,keycodes",
>> -                                keycodes, ARRAY_SIZE(keycodes));
>> -
>> -     for (i = 0; i < ARRAY_SIZE(keycodes); i++)
>> -             priv->keycodes[i] = keycodes[i];
>> +                               priv->keycodes, priv->num_channels);
>
> Hmm, no. Internally, you have to store the keycodes as unsigned short,
> otherwise EVIOC{G|S}KEYCODE from usespace doesn't work.
>
> of_property_read_u16_array() should work here for unsigned short, I
> guess. Otherwise, you'd have to open-code the routine.
>
Problem with u16 is you'll to mark it */bits/ 16* in the device tree
entry.. I doubt that is acceptable


>> @@ -313,12 +307,16 @@ static int cap1106_i2c_probe(struct i2c_client *i2c_client,
>>
>>  static const struct of_device_id cap1106_dt_ids[] = {
>>       { .compatible = "microchip,cap1106", },
>> +     { .compatible = "microchip,cap1126", },
>> +     { .compatible = "microchip,cap1188", },
>
> Hmm, how can that work unless you set .data to the number of channels
> here? Did you test that with a DT-enabled board?
>
>
> Thanks,
> Daniel

Re: [PATCH 2/3] cap1106: support for active-high interrupt option

[Dmitry Torokhov]

On Sun, Sep 21, 2014 at 12:06:30PM +0200, Daniel Mack wrote:
> On 09/21/2014 05:01 AM, Matt Ranostay wrote:
> > Some applications need to use the active-high push-pull interrupt
> > option. This allows it be enabled in the device tree child node.
> > 
> > Signed-off-by: Matt Ranostay <mranostay@gmail.com>
> > ---
> >  drivers/input/keyboard/cap1106.c | 6 ++++++
> >  1 file changed, 6 insertions(+)
> > 
> > diff --git a/drivers/input/keyboard/cap1106.c b/drivers/input/keyboard/cap1106.c
> > index b9c43b5..33e2590 100644
> > --- a/drivers/input/keyboard/cap1106.c
> > +++ b/drivers/input/keyboard/cap1106.c
> > @@ -234,6 +234,12 @@ static int cap1106_i2c_probe(struct i2c_client *i2c_client,
> >  			dev_err(dev, "Invalid sensor-gain value %d\n", gain32);
> >  	}
> >  
> > +	if (of_property_read_bool(node, "microchip,active-high")) {
> 
> I think the name of that property should make clear it's only changing
> the interrupt output driver configuration. What about
> "microchip,irq-active-high"?

Can we infer the setting from IRQ flags by chance?

Thanks.

-- 
Dmitry

Re: [PATCH 1/3] cap1106: Add support for various cap11xx devices

[Dmitry Torokhov]

On Sun, Sep 21, 2014 at 05:28:05PM -0700, Matt Ranostay wrote:
> On Sun, Sep 21, 2014 at 2:58 AM, Daniel Mack <daniel@zonque.org> wrote:
> > Hi,
> >
> > On 09/21/2014 05:01 AM, Matt Ranostay wrote:
> >> Several other variants of the cap11xx device exists with a varying
> >> number of capacitance detection channels. Add support for creating
> >> the channels dynamically.
> >
> > Thanks for the patches!
> >
> >>
> >> Signed-off-by: Matt Ranostay <mranostay@gmail.com>
> >> ---
> >>  drivers/input/keyboard/cap1106.c | 54 +++++++++++++++++++---------------------
> >
> > Please also add a patch to rename the file to cap11xx.c, and make sure
> > to export the patch with 'git format-patch -M' to detect the rename.
> >
> >> diff --git a/drivers/input/keyboard/cap1106.c b/drivers/input/keyboard/cap1106.c
> >> index d70b65a..b9c43b5 100644
> >> --- a/drivers/input/keyboard/cap1106.c
> >> +++ b/drivers/input/keyboard/cap1106.c
> >> @@ -55,8 +55,6 @@
> >>  #define CAP1106_REG_MANUFACTURER_ID  0xfe
> >>  #define CAP1106_REG_REVISION         0xff
> >>
> >> -#define CAP1106_NUM_CHN 6
> >> -#define CAP1106_PRODUCT_ID   0x55
> >>  #define CAP1106_MANUFACTURER_ID      0x5d
> >>
> >>  struct cap1106_priv {
> >> @@ -64,7 +62,8 @@ struct cap1106_priv {
> >>       struct input_dev *idev;
> >>
> >>       /* config */
> >> -     unsigned short keycodes[CAP1106_NUM_CHN];
> >> +     u32 *keycodes;
> >
> > unsigned short *, please. See below.
> >
> >> @@ -189,27 +188,23 @@ static int cap1106_i2c_probe(struct i2c_client *i2c_client,
> >>       struct cap1106_priv *priv;
> >>       struct device_node *node;
> >>       int i, error, irq, gain = 0;
> >> -     unsigned int val, rev;
> >> -     u32 gain32, keycodes[CAP1106_NUM_CHN];
> >> +     unsigned int val, prod, rev;
> >> +     u32 gain32;
> >>
> >>       priv = devm_kzalloc(dev, sizeof(*priv), GFP_KERNEL);
> >>       if (!priv)
> >>               return -ENOMEM;
> >>
> >> +     priv->num_channels = (unsigned long) id->driver_data;
> >
> > priv->num_channels is unsigned int.
> >
> > Also, a BUG_ON(!priv->num_channels) wouldn't harm.
> >
> >> +     priv->keycodes = devm_kzalloc(dev,
> >> +             sizeof(u32) * priv->num_channels, GFP_KERNEL);

You do not need to allocate keycodes separately if you make a flexible length
array at the end of the main structure.

> >
> > Use devm_kcalloc() to allocate an array.
> >
> >> @@ -235,17 +234,12 @@ static int cap1106_i2c_probe(struct i2c_client *i2c_client,
> >>                       dev_err(dev, "Invalid sensor-gain value %d\n", gain32);
> >>       }
> >>
> >> -     BUILD_BUG_ON(ARRAY_SIZE(keycodes) != ARRAY_SIZE(priv->keycodes));
> >> -
> >>       /* Provide some useful defaults */
> >> -     for (i = 0; i < ARRAY_SIZE(keycodes); i++)
> >> -             keycodes[i] = KEY_A + i;
> >> +     for (i = 0; i < priv->num_channels; i++)
> >> +             priv->keycodes[i] = KEY_A + i;
> >>
> >>       of_property_read_u32_array(node, "linux,keycodes",
> >> -                                keycodes, ARRAY_SIZE(keycodes));
> >> -
> >> -     for (i = 0; i < ARRAY_SIZE(keycodes); i++)
> >> -             priv->keycodes[i] = keycodes[i];
> >> +                               priv->keycodes, priv->num_channels);
> >
> > Hmm, no. Internally, you have to store the keycodes as unsigned short,
> > otherwise EVIOC{G|S}KEYCODE from usespace doesn't work.
> >
> > of_property_read_u16_array() should work here for unsigned short, I
> > guess. Otherwise, you'd have to open-code the routine.
> >
> Problem with u16 is you'll to mark it */bits/ 16* in the device tree
> entry.. I doubt that is acceptable

You can make keymap u32. As long as you set input->keycodesize appropriately
(and we do) everything should work just fine.

Thanks.

-- 
Dmitry

Re: [PATCH 1/3] cap1106: Add support for various cap11xx devices

[Daniel Mack]

On 09/22/2014 12:46 AM, Matt Ranostay wrote:
> On Sun, Sep 21, 2014 at 2:58 AM, Daniel Mack <daniel@zonque.org> wrote:
>> On 09/21/2014 05:01 AM, Matt Ranostay wrote:

>>> @@ -313,12 +307,16 @@ static int cap1106_i2c_probe(struct i2c_client *i2c_client,
>>>
>>>  static const struct of_device_id cap1106_dt_ids[] = {
>>>       { .compatible = "microchip,cap1106", },
>>> +     { .compatible = "microchip,cap1126", },
>>> +     { .compatible = "microchip,cap1188", },
>>
>> Hmm, how can that work unless you set .data to the number of channels
>> here? Did you test that with a DT-enabled board?
>>
> Yes it was tested on a BBB.  The num_channels is set from cap1106_i2c_ids

Ah ok. I forgot there's this fallback to the i2c ids. What others driver
do is to use of_match_device() in the probe function, and then access
->data of the returned match.

But I'm fine with falling back to cap1106_i2c_ids unless anyone else has
objections.


Thanks,
Daniel

Re: [PATCH 2/3] cap1106: support for active-high interrupt option

[Daniel Mack]

On 09/22/2014 07:56 AM, Dmitry Torokhov wrote:
> On Sun, Sep 21, 2014 at 12:06:30PM +0200, Daniel Mack wrote:
>> On 09/21/2014 05:01 AM, Matt Ranostay wrote:
>>> Some applications need to use the active-high push-pull interrupt
>>> option. This allows it be enabled in the device tree child node.
>>>
>>> Signed-off-by: Matt Ranostay <mranostay@gmail.com>
>>> ---
>>>  drivers/input/keyboard/cap1106.c | 6 ++++++
>>>  1 file changed, 6 insertions(+)
>>>
>>> diff --git a/drivers/input/keyboard/cap1106.c b/drivers/input/keyboard/cap1106.c
>>> index b9c43b5..33e2590 100644
>>> --- a/drivers/input/keyboard/cap1106.c
>>> +++ b/drivers/input/keyboard/cap1106.c
>>> @@ -234,6 +234,12 @@ static int cap1106_i2c_probe(struct i2c_client *i2c_client,
>>>  			dev_err(dev, "Invalid sensor-gain value %d\n", gain32);
>>>  	}
>>>  
>>> +	if (of_property_read_bool(node, "microchip,active-high")) {
>>
>> I think the name of that property should make clear it's only changing
>> the interrupt output driver configuration. What about
>> "microchip,irq-active-high"?
> 
> Can we infer the setting from IRQ flags by chance?

Hmm, I thought of that as well, but there could be electrical wiring
setups that want the CPU's hardware pin in push/pull mode but the one on
the sensor chip in open-drain. I'd rather not make the assuption the
pins are directly connected and have both sides individually configurable.


Thanks,
Daniel