[PATCH] Input: synaptics-rmi4: Support regulator supplies

[PATCH] Input: synaptics-rmi4: Support regulator supplies

[Bjorn Andersson]

From: Bjorn Andersson <bjorn.andersson@sonymobile.com>

Support the two supplies - vdd and vio - to make it possible to control
power to the Synaptics chip.

Signed-off-by: Bjorn Andersson <bjorn.andersson@sonymobile.com>
---
 .../devicetree/bindings/input/rmi4/rmi_i2c.txt     |  7 ++++
 drivers/input/rmi4/rmi_i2c.c                       | 45 ++++++++++++++++++++++
 2 files changed, 52 insertions(+)

diff --git a/Documentation/devicetree/bindings/input/rmi4/rmi_i2c.txt b/Documentation/devicetree/bindings/input/rmi4/rmi_i2c.txt
index 95fa715c6046..a8c31f40f816 100644
--- a/Documentation/devicetree/bindings/input/rmi4/rmi_i2c.txt
+++ b/Documentation/devicetree/bindings/input/rmi4/rmi_i2c.txt
@@ -22,6 +22,13 @@ See Documentation/devicetree/bindings/interrupt-controller/interrupts.txt
 - syna,reset-delay-ms: The number of milliseconds to wait after resetting the
 			device.
 
+- vdd-supply: VDD power supply.
+See Documentation/devicetree/bindings/regulator/regulator.txt
+
+- vio-supply: VIO power supply
+See Documentation/devicetree/bindings/regulator/regulator.txt
+
+
 Function Parameters:
 Parameters specific to RMI functions are contained in child nodes of the rmi device
  node. Documentation for the parameters of each function can be found in:
diff --git a/drivers/input/rmi4/rmi_i2c.c b/drivers/input/rmi4/rmi_i2c.c
index a96a326b53bd..a8c794daba04 100644
--- a/drivers/input/rmi4/rmi_i2c.c
+++ b/drivers/input/rmi4/rmi_i2c.c
@@ -11,6 +11,8 @@
 #include <linux/rmi.h>
 #include <linux/irq.h>
 #include <linux/of.h>
+#include <linux/delay.h>
+#include <linux/regulator/consumer.h>
 #include "rmi_driver.h"
 
 #define BUFFER_SIZE_INCREMENT 32
@@ -37,6 +39,8 @@ struct rmi_i2c_xport {
 
 	u8 *tx_buf;
 	size_t tx_buf_size;
+
+	struct regulator_bulk_data supplies[2];
 };
 
 #define RMI_PAGE_SELECT_REGISTER 0xff
@@ -246,6 +250,22 @@ static int rmi_i2c_probe(struct i2c_client *client,
 		return -ENODEV;
 	}
 
+	rmi_i2c->supplies[0].supply = "vdd";
+	rmi_i2c->supplies[1].supply = "vio";
+	retval = devm_regulator_bulk_get(&client->dev,
+					 ARRAY_SIZE(rmi_i2c->supplies),
+					 rmi_i2c->supplies);
+	if (retval < 0)
+		return retval;
+
+	retval = regulator_bulk_enable(ARRAY_SIZE(rmi_i2c->supplies),
+				       rmi_i2c->supplies);
+	if (retval < 0)
+		return retval;
+
+	/* Allow the firmware to get ready */
+	msleep(10);
+
 	rmi_i2c->client = client;
 	mutex_init(&rmi_i2c->page_mutex);
 
@@ -286,6 +306,8 @@ static int rmi_i2c_remove(struct i2c_client *client)
 	struct rmi_i2c_xport *rmi_i2c = i2c_get_clientdata(client);
 
 	rmi_unregister_transport_device(&rmi_i2c->xport);
+	regulator_bulk_disable(ARRAY_SIZE(rmi_i2c->supplies),
+			       rmi_i2c->supplies);
 
 	return 0;
 }
@@ -308,6 +330,10 @@ static int rmi_i2c_suspend(struct device *dev)
 			dev_warn(dev, "Failed to enable irq for wake: %d\n",
 				ret);
 	}
+
+	regulator_bulk_disable(ARRAY_SIZE(rmi_i2c->supplies),
+			       rmi_i2c->supplies);
+
 	return ret;
 }
 
@@ -317,6 +343,14 @@ static int rmi_i2c_resume(struct device *dev)
 	struct rmi_i2c_xport *rmi_i2c = i2c_get_clientdata(client);
 	int ret;
 
+	ret = regulator_bulk_enable(ARRAY_SIZE(rmi_i2c->supplies),
+				    rmi_i2c->supplies);
+	if (ret)
+		return ret;
+
+	/* Allow the firmware to get ready */
+	msleep(10);
+
 	enable_irq(rmi_i2c->irq);
 	if (device_may_wakeup(&client->dev)) {
 		ret = disable_irq_wake(rmi_i2c->irq);
@@ -346,6 +380,9 @@ static int rmi_i2c_runtime_suspend(struct device *dev)
 
 	disable_irq(rmi_i2c->irq);
 
+	regulator_bulk_disable(ARRAY_SIZE(rmi_i2c->supplies),
+			       rmi_i2c->supplies);
+
 	return 0;
 }
 
@@ -355,6 +392,14 @@ static int rmi_i2c_runtime_resume(struct device *dev)
 	struct rmi_i2c_xport *rmi_i2c = i2c_get_clientdata(client);
 	int ret;
 
+	ret = regulator_bulk_enable(ARRAY_SIZE(rmi_i2c->supplies),
+				    rmi_i2c->supplies);
+	if (ret)
+		return ret;
+
+	/* Allow the firmware to get ready */
+	msleep(10);
+
 	enable_irq(rmi_i2c->irq);
 
 	ret = rmi_driver_resume(rmi_i2c->xport.rmi_dev);
-- 
2.5.0

[PATCH] Input: synaptics-rmi4: Support regulator supplies

[Bjorn Andersson]

From: Bjorn Andersson <bjorn.andersson@sonymobile.com>

Support the two supplies - vdd and vio - to make it possible to control
power to the Synaptics chip.

Signed-off-by: Bjorn Andersson <bjorn.andersson@sonymobile.com>
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
---
 .../devicetree/bindings/input/rmi4/rmi_i2c.txt     |  7 ++++
 drivers/input/rmi4/rmi_i2c.c                       | 45 ++++++++++++++++++++++
 2 files changed, 52 insertions(+)

diff --git a/Documentation/devicetree/bindings/input/rmi4/rmi_i2c.txt b/Documentation/devicetree/bindings/input/rmi4/rmi_i2c.txt
index 95fa715c6046..a8c31f40f816 100644
--- a/Documentation/devicetree/bindings/input/rmi4/rmi_i2c.txt
+++ b/Documentation/devicetree/bindings/input/rmi4/rmi_i2c.txt
@@ -22,6 +22,13 @@ See Documentation/devicetree/bindings/interrupt-controller/interrupts.txt
 - syna,reset-delay-ms: The number of milliseconds to wait after resetting the
 			device.
 
+- vdd-supply: VDD power supply.
+See Documentation/devicetree/bindings/regulator/regulator.txt
+
+- vio-supply: VIO power supply
+See Documentation/devicetree/bindings/regulator/regulator.txt
+
+
 Function Parameters:
 Parameters specific to RMI functions are contained in child nodes of the rmi device
  node. Documentation for the parameters of each function can be found in:
diff --git a/drivers/input/rmi4/rmi_i2c.c b/drivers/input/rmi4/rmi_i2c.c
index a96a326b53bd..a8c794daba04 100644
--- a/drivers/input/rmi4/rmi_i2c.c
+++ b/drivers/input/rmi4/rmi_i2c.c
@@ -11,6 +11,8 @@
 #include <linux/rmi.h>
 #include <linux/irq.h>
 #include <linux/of.h>
+#include <linux/delay.h>
+#include <linux/regulator/consumer.h>
 #include "rmi_driver.h"
 
 #define BUFFER_SIZE_INCREMENT 32
@@ -37,6 +39,8 @@ struct rmi_i2c_xport {
 
 	u8 *tx_buf;
 	size_t tx_buf_size;
+
+	struct regulator_bulk_data supplies[2];
 };
 
 #define RMI_PAGE_SELECT_REGISTER 0xff
@@ -246,6 +250,22 @@ static int rmi_i2c_probe(struct i2c_client *client,
 		return -ENODEV;
 	}
 
+	rmi_i2c->supplies[0].supply = "vdd";
+	rmi_i2c->supplies[1].supply = "vio";
+	retval = devm_regulator_bulk_get(&client->dev,
+					 ARRAY_SIZE(rmi_i2c->supplies),
+					 rmi_i2c->supplies);
+	if (retval < 0)
+		return retval;
+
+	retval = regulator_bulk_enable(ARRAY_SIZE(rmi_i2c->supplies),
+				       rmi_i2c->supplies);
+	if (retval < 0)
+		return retval;
+
+	/* Allow the firmware to get ready */
+	msleep(10);
+
 	rmi_i2c->client = client;
 	mutex_init(&rmi_i2c->page_mutex);
 
@@ -286,6 +306,8 @@ static int rmi_i2c_remove(struct i2c_client *client)
 	struct rmi_i2c_xport *rmi_i2c = i2c_get_clientdata(client);
 
 	rmi_unregister_transport_device(&rmi_i2c->xport);
+	regulator_bulk_disable(ARRAY_SIZE(rmi_i2c->supplies),
+			       rmi_i2c->supplies);
 
 	return 0;
 }
@@ -308,6 +330,10 @@ static int rmi_i2c_suspend(struct device *dev)
 			dev_warn(dev, "Failed to enable irq for wake: %d\n",
 				ret);
 	}
+
+	regulator_bulk_disable(ARRAY_SIZE(rmi_i2c->supplies),
+			       rmi_i2c->supplies);
+
 	return ret;
 }
 
@@ -317,6 +343,14 @@ static int rmi_i2c_resume(struct device *dev)
 	struct rmi_i2c_xport *rmi_i2c = i2c_get_clientdata(client);
 	int ret;
 
+	ret = regulator_bulk_enable(ARRAY_SIZE(rmi_i2c->supplies),
+				    rmi_i2c->supplies);
+	if (ret)
+		return ret;
+
+	/* Allow the firmware to get ready */
+	msleep(10);
+
 	enable_irq(rmi_i2c->irq);
 	if (device_may_wakeup(&client->dev)) {
 		ret = disable_irq_wake(rmi_i2c->irq);
@@ -346,6 +380,9 @@ static int rmi_i2c_runtime_suspend(struct device *dev)
 
 	disable_irq(rmi_i2c->irq);
 
+	regulator_bulk_disable(ARRAY_SIZE(rmi_i2c->supplies),
+			       rmi_i2c->supplies);
+
 	return 0;
 }
 
@@ -355,6 +392,14 @@ static int rmi_i2c_runtime_resume(struct device *dev)
 	struct rmi_i2c_xport *rmi_i2c = i2c_get_clientdata(client);
 	int ret;
 
+	ret = regulator_bulk_enable(ARRAY_SIZE(rmi_i2c->supplies),
+				    rmi_i2c->supplies);
+	if (ret)
+		return ret;
+
+	/* Allow the firmware to get ready */
+	msleep(10);
+
 	enable_irq(rmi_i2c->irq);
 
 	ret = rmi_driver_resume(rmi_i2c->xport.rmi_dev);
-- 
2.5.0

Re: [PATCH] Input: synaptics-rmi4: Support regulator supplies

[Bjorn Andersson]

On Wed, Mar 30, 2016 at 9:57 AM, Bjorn Andersson
<bjorn.andersson@linaro.org> wrote:
> From: Bjorn Andersson <bjorn.andersson@sonymobile.com>
>
> Support the two supplies - vdd and vio - to make it possible to control
> power to the Synaptics chip.
>
> Signed-off-by: Bjorn Andersson <bjorn.andersson@sonymobile.com>

Sorry for the spam, ^C'ed this and resent it with a valid
Signed-off-by before I saw that this actually did hit the list.

Please ignore this and consider the same patch sent 20 seconds later.

Regards,
Bjorn

> ---
>  .../devicetree/bindings/input/rmi4/rmi_i2c.txt     |  7 ++++
>  drivers/input/rmi4/rmi_i2c.c                       | 45 ++++++++++++++++++++++
>  2 files changed, 52 insertions(+)
>
> diff --git a/Documentation/devicetree/bindings/input/rmi4/rmi_i2c.txt b/Documentation/devicetree/bindings/input/rmi4/rmi_i2c.txt
> index 95fa715c6046..a8c31f40f816 100644
> --- a/Documentation/devicetree/bindings/input/rmi4/rmi_i2c.txt
> +++ b/Documentation/devicetree/bindings/input/rmi4/rmi_i2c.txt
> @@ -22,6 +22,13 @@ See Documentation/devicetree/bindings/interrupt-controller/interrupts.txt
>  - syna,reset-delay-ms: The number of milliseconds to wait after resetting the
>                         device.
>
> +- vdd-supply: VDD power supply.
> +See Documentation/devicetree/bindings/regulator/regulator.txt
> +
> +- vio-supply: VIO power supply
> +See Documentation/devicetree/bindings/regulator/regulator.txt
> +
> +
>  Function Parameters:
>  Parameters specific to RMI functions are contained in child nodes of the rmi device
>   node. Documentation for the parameters of each function can be found in:
> diff --git a/drivers/input/rmi4/rmi_i2c.c b/drivers/input/rmi4/rmi_i2c.c
> index a96a326b53bd..a8c794daba04 100644
> --- a/drivers/input/rmi4/rmi_i2c.c
> +++ b/drivers/input/rmi4/rmi_i2c.c
> @@ -11,6 +11,8 @@
>  #include <linux/rmi.h>
>  #include <linux/irq.h>
>  #include <linux/of.h>
> +#include <linux/delay.h>
> +#include <linux/regulator/consumer.h>
>  #include "rmi_driver.h"
>
>  #define BUFFER_SIZE_INCREMENT 32
> @@ -37,6 +39,8 @@ struct rmi_i2c_xport {
>
>         u8 *tx_buf;
>         size_t tx_buf_size;
> +
> +       struct regulator_bulk_data supplies[2];
>  };
>
>  #define RMI_PAGE_SELECT_REGISTER 0xff
> @@ -246,6 +250,22 @@ static int rmi_i2c_probe(struct i2c_client *client,
>                 return -ENODEV;
>         }
>
> +       rmi_i2c->supplies[0].supply = "vdd";
> +       rmi_i2c->supplies[1].supply = "vio";
> +       retval = devm_regulator_bulk_get(&client->dev,
> +                                        ARRAY_SIZE(rmi_i2c->supplies),
> +                                        rmi_i2c->supplies);
> +       if (retval < 0)
> +               return retval;
> +
> +       retval = regulator_bulk_enable(ARRAY_SIZE(rmi_i2c->supplies),
> +                                      rmi_i2c->supplies);
> +       if (retval < 0)
> +               return retval;
> +
> +       /* Allow the firmware to get ready */
> +       msleep(10);
> +
>         rmi_i2c->client = client;
>         mutex_init(&rmi_i2c->page_mutex);
>
> @@ -286,6 +306,8 @@ static int rmi_i2c_remove(struct i2c_client *client)
>         struct rmi_i2c_xport *rmi_i2c = i2c_get_clientdata(client);
>
>         rmi_unregister_transport_device(&rmi_i2c->xport);
> +       regulator_bulk_disable(ARRAY_SIZE(rmi_i2c->supplies),
> +                              rmi_i2c->supplies);
>
>         return 0;
>  }
> @@ -308,6 +330,10 @@ static int rmi_i2c_suspend(struct device *dev)
>                         dev_warn(dev, "Failed to enable irq for wake: %d\n",
>                                 ret);
>         }
> +
> +       regulator_bulk_disable(ARRAY_SIZE(rmi_i2c->supplies),
> +                              rmi_i2c->supplies);
> +
>         return ret;
>  }
>
> @@ -317,6 +343,14 @@ static int rmi_i2c_resume(struct device *dev)
>         struct rmi_i2c_xport *rmi_i2c = i2c_get_clientdata(client);
>         int ret;
>
> +       ret = regulator_bulk_enable(ARRAY_SIZE(rmi_i2c->supplies),
> +                                   rmi_i2c->supplies);
> +       if (ret)
> +               return ret;
> +
> +       /* Allow the firmware to get ready */
> +       msleep(10);
> +
>         enable_irq(rmi_i2c->irq);
>         if (device_may_wakeup(&client->dev)) {
>                 ret = disable_irq_wake(rmi_i2c->irq);
> @@ -346,6 +380,9 @@ static int rmi_i2c_runtime_suspend(struct device *dev)
>
>         disable_irq(rmi_i2c->irq);
>
> +       regulator_bulk_disable(ARRAY_SIZE(rmi_i2c->supplies),
> +                              rmi_i2c->supplies);
> +
>         return 0;
>  }
>
> @@ -355,6 +392,14 @@ static int rmi_i2c_runtime_resume(struct device *dev)
>         struct rmi_i2c_xport *rmi_i2c = i2c_get_clientdata(client);
>         int ret;
>
> +       ret = regulator_bulk_enable(ARRAY_SIZE(rmi_i2c->supplies),
> +                                   rmi_i2c->supplies);
> +       if (ret)
> +               return ret;
> +
> +       /* Allow the firmware to get ready */
> +       msleep(10);
> +
>         enable_irq(rmi_i2c->irq);
>
>         ret = rmi_driver_resume(rmi_i2c->xport.rmi_dev);
> --
> 2.5.0
>

Re: [PATCH] Input: synaptics-rmi4: Support regulator supplies

[Dmitry Torokhov]

Hi Bjorn,

On Wed, Mar 30, 2016 at 09:57:29AM -0700, Bjorn Andersson wrote:
> From: Bjorn Andersson <bjorn.andersson-/MT0OVThwyLZJqsBc5GL+g@public.gmane.org>
> 
> Support the two supplies - vdd and vio - to make it possible to control
> power to the Synaptics chip.
> 
> Signed-off-by: Bjorn Andersson <bjorn.andersson-/MT0OVThwyLZJqsBc5GL+g@public.gmane.org>
> Signed-off-by: Bjorn Andersson <bjorn.andersson-QSEj5FYQhm4dnm+yROfE0A@public.gmane.org>
> ---
>  .../devicetree/bindings/input/rmi4/rmi_i2c.txt     |  7 ++++
>  drivers/input/rmi4/rmi_i2c.c                       | 45 ++++++++++++++++++++++

Would not we need pretty much the same changes for SPI devices? Can this
be done in core?

Thanks.

>  2 files changed, 52 insertions(+)
> 
> diff --git a/Documentation/devicetree/bindings/input/rmi4/rmi_i2c.txt b/Documentation/devicetree/bindings/input/rmi4/rmi_i2c.txt
> index 95fa715c6046..a8c31f40f816 100644
> --- a/Documentation/devicetree/bindings/input/rmi4/rmi_i2c.txt
> +++ b/Documentation/devicetree/bindings/input/rmi4/rmi_i2c.txt
> @@ -22,6 +22,13 @@ See Documentation/devicetree/bindings/interrupt-controller/interrupts.txt
>  - syna,reset-delay-ms: The number of milliseconds to wait after resetting the
>  			device.
>  
> +- vdd-supply: VDD power supply.
> +See Documentation/devicetree/bindings/regulator/regulator.txt
> +
> +- vio-supply: VIO power supply
> +See Documentation/devicetree/bindings/regulator/regulator.txt
> +
> +
>  Function Parameters:
>  Parameters specific to RMI functions are contained in child nodes of the rmi device
>   node. Documentation for the parameters of each function can be found in:
> diff --git a/drivers/input/rmi4/rmi_i2c.c b/drivers/input/rmi4/rmi_i2c.c
> index a96a326b53bd..a8c794daba04 100644
> --- a/drivers/input/rmi4/rmi_i2c.c
> +++ b/drivers/input/rmi4/rmi_i2c.c
> @@ -11,6 +11,8 @@
>  #include <linux/rmi.h>
>  #include <linux/irq.h>
>  #include <linux/of.h>
> +#include <linux/delay.h>
> +#include <linux/regulator/consumer.h>
>  #include "rmi_driver.h"
>  
>  #define BUFFER_SIZE_INCREMENT 32
> @@ -37,6 +39,8 @@ struct rmi_i2c_xport {
>  
>  	u8 *tx_buf;
>  	size_t tx_buf_size;
> +
> +	struct regulator_bulk_data supplies[2];
>  };
>  
>  #define RMI_PAGE_SELECT_REGISTER 0xff
> @@ -246,6 +250,22 @@ static int rmi_i2c_probe(struct i2c_client *client,
>  		return -ENODEV;
>  	}
>  
> +	rmi_i2c->supplies[0].supply = "vdd";
> +	rmi_i2c->supplies[1].supply = "vio";
> +	retval = devm_regulator_bulk_get(&client->dev,
> +					 ARRAY_SIZE(rmi_i2c->supplies),
> +					 rmi_i2c->supplies);
> +	if (retval < 0)
> +		return retval;
> +
> +	retval = regulator_bulk_enable(ARRAY_SIZE(rmi_i2c->supplies),
> +				       rmi_i2c->supplies);
> +	if (retval < 0)
> +		return retval;
> +
> +	/* Allow the firmware to get ready */
> +	msleep(10);
> +
>  	rmi_i2c->client = client;
>  	mutex_init(&rmi_i2c->page_mutex);
>  
> @@ -286,6 +306,8 @@ static int rmi_i2c_remove(struct i2c_client *client)
>  	struct rmi_i2c_xport *rmi_i2c = i2c_get_clientdata(client);
>  
>  	rmi_unregister_transport_device(&rmi_i2c->xport);
> +	regulator_bulk_disable(ARRAY_SIZE(rmi_i2c->supplies),
> +			       rmi_i2c->supplies);
>  
>  	return 0;
>  }
> @@ -308,6 +330,10 @@ static int rmi_i2c_suspend(struct device *dev)
>  			dev_warn(dev, "Failed to enable irq for wake: %d\n",
>  				ret);
>  	}
> +
> +	regulator_bulk_disable(ARRAY_SIZE(rmi_i2c->supplies),
> +			       rmi_i2c->supplies);
> +
>  	return ret;
>  }
>  
> @@ -317,6 +343,14 @@ static int rmi_i2c_resume(struct device *dev)
>  	struct rmi_i2c_xport *rmi_i2c = i2c_get_clientdata(client);
>  	int ret;
>  
> +	ret = regulator_bulk_enable(ARRAY_SIZE(rmi_i2c->supplies),
> +				    rmi_i2c->supplies);
> +	if (ret)
> +		return ret;
> +
> +	/* Allow the firmware to get ready */
> +	msleep(10);
> +
>  	enable_irq(rmi_i2c->irq);
>  	if (device_may_wakeup(&client->dev)) {
>  		ret = disable_irq_wake(rmi_i2c->irq);
> @@ -346,6 +380,9 @@ static int rmi_i2c_runtime_suspend(struct device *dev)
>  
>  	disable_irq(rmi_i2c->irq);
>  
> +	regulator_bulk_disable(ARRAY_SIZE(rmi_i2c->supplies),
> +			       rmi_i2c->supplies);
> +
>  	return 0;
>  }
>  
> @@ -355,6 +392,14 @@ static int rmi_i2c_runtime_resume(struct device *dev)
>  	struct rmi_i2c_xport *rmi_i2c = i2c_get_clientdata(client);
>  	int ret;
>  
> +	ret = regulator_bulk_enable(ARRAY_SIZE(rmi_i2c->supplies),
> +				    rmi_i2c->supplies);
> +	if (ret)
> +		return ret;
> +
> +	/* Allow the firmware to get ready */
> +	msleep(10);
> +
>  	enable_irq(rmi_i2c->irq);
>  
>  	ret = rmi_driver_resume(rmi_i2c->xport.rmi_dev);
> -- 
> 2.5.0
> 

-- 
Dmitry
--
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

Re: [PATCH] Input: synaptics-rmi4: Support regulator supplies

[Bjorn Andersson]

On Thu 31 Mar 11:19 PDT 2016, Dmitry Torokhov wrote:

> Hi Bjorn,
> 
> On Wed, Mar 30, 2016 at 09:57:29AM -0700, Bjorn Andersson wrote:
> > From: Bjorn Andersson <bjorn.andersson@sonymobile.com>
> > 
> > Support the two supplies - vdd and vio - to make it possible to control
> > power to the Synaptics chip.
> > 
> > Signed-off-by: Bjorn Andersson <bjorn.andersson@sonymobile.com>
> > Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
> > ---
> >  .../devicetree/bindings/input/rmi4/rmi_i2c.txt     |  7 ++++
> >  drivers/input/rmi4/rmi_i2c.c                       | 45 ++++++++++++++++++++++
> 
> Would not we need pretty much the same changes for SPI devices? Can this
> be done in core?
> 

Yes, I believe it needs the exact same steps.

I did a initial quick hack on v1 of the patchset and back then it was
possible, when I rebased it a few weeks back I kept ending up in getting
interrupts with the power off.

Looking at the code this is likely because in the resume paths the IRQ
is enabled before we jump to rmi_driver_resume(), so putting this in the
core I ended up calling rmi_process_interrupt_requests() before powering
up the chip.

I assume it's done this way to allow incoming interrupts while functions
are being resumed. Andrew, can you comment on this?

Regards,
Bjorn

> Thanks.
> 
> >  2 files changed, 52 insertions(+)
> > 
> > diff --git a/Documentation/devicetree/bindings/input/rmi4/rmi_i2c.txt b/Documentation/devicetree/bindings/input/rmi4/rmi_i2c.txt
> > index 95fa715c6046..a8c31f40f816 100644
> > --- a/Documentation/devicetree/bindings/input/rmi4/rmi_i2c.txt
> > +++ b/Documentation/devicetree/bindings/input/rmi4/rmi_i2c.txt
> > @@ -22,6 +22,13 @@ See Documentation/devicetree/bindings/interrupt-controller/interrupts.txt
> >  - syna,reset-delay-ms: The number of milliseconds to wait after resetting the
> >  			device.
> >  
> > +- vdd-supply: VDD power supply.
> > +See Documentation/devicetree/bindings/regulator/regulator.txt
> > +
> > +- vio-supply: VIO power supply
> > +See Documentation/devicetree/bindings/regulator/regulator.txt
> > +
> > +
> >  Function Parameters:
> >  Parameters specific to RMI functions are contained in child nodes of the rmi device
> >   node. Documentation for the parameters of each function can be found in:
> > diff --git a/drivers/input/rmi4/rmi_i2c.c b/drivers/input/rmi4/rmi_i2c.c
> > index a96a326b53bd..a8c794daba04 100644
> > --- a/drivers/input/rmi4/rmi_i2c.c
> > +++ b/drivers/input/rmi4/rmi_i2c.c
> > @@ -11,6 +11,8 @@
> >  #include <linux/rmi.h>
> >  #include <linux/irq.h>
> >  #include <linux/of.h>
> > +#include <linux/delay.h>
> > +#include <linux/regulator/consumer.h>
> >  #include "rmi_driver.h"
> >  
> >  #define BUFFER_SIZE_INCREMENT 32
> > @@ -37,6 +39,8 @@ struct rmi_i2c_xport {
> >  
> >  	u8 *tx_buf;
> >  	size_t tx_buf_size;
> > +
> > +	struct regulator_bulk_data supplies[2];
> >  };
> >  
> >  #define RMI_PAGE_SELECT_REGISTER 0xff
> > @@ -246,6 +250,22 @@ static int rmi_i2c_probe(struct i2c_client *client,
> >  		return -ENODEV;
> >  	}
> >  
> > +	rmi_i2c->supplies[0].supply = "vdd";
> > +	rmi_i2c->supplies[1].supply = "vio";
> > +	retval = devm_regulator_bulk_get(&client->dev,
> > +					 ARRAY_SIZE(rmi_i2c->supplies),
> > +					 rmi_i2c->supplies);
> > +	if (retval < 0)
> > +		return retval;
> > +
> > +	retval = regulator_bulk_enable(ARRAY_SIZE(rmi_i2c->supplies),
> > +				       rmi_i2c->supplies);
> > +	if (retval < 0)
> > +		return retval;
> > +
> > +	/* Allow the firmware to get ready */
> > +	msleep(10);
> > +
> >  	rmi_i2c->client = client;
> >  	mutex_init(&rmi_i2c->page_mutex);
> >  
> > @@ -286,6 +306,8 @@ static int rmi_i2c_remove(struct i2c_client *client)
> >  	struct rmi_i2c_xport *rmi_i2c = i2c_get_clientdata(client);
> >  
> >  	rmi_unregister_transport_device(&rmi_i2c->xport);
> > +	regulator_bulk_disable(ARRAY_SIZE(rmi_i2c->supplies),
> > +			       rmi_i2c->supplies);
> >  
> >  	return 0;
> >  }
> > @@ -308,6 +330,10 @@ static int rmi_i2c_suspend(struct device *dev)
> >  			dev_warn(dev, "Failed to enable irq for wake: %d\n",
> >  				ret);
> >  	}
> > +
> > +	regulator_bulk_disable(ARRAY_SIZE(rmi_i2c->supplies),
> > +			       rmi_i2c->supplies);
> > +
> >  	return ret;
> >  }
> >  
> > @@ -317,6 +343,14 @@ static int rmi_i2c_resume(struct device *dev)
> >  	struct rmi_i2c_xport *rmi_i2c = i2c_get_clientdata(client);
> >  	int ret;
> >  
> > +	ret = regulator_bulk_enable(ARRAY_SIZE(rmi_i2c->supplies),
> > +				    rmi_i2c->supplies);
> > +	if (ret)
> > +		return ret;
> > +
> > +	/* Allow the firmware to get ready */
> > +	msleep(10);
> > +
> >  	enable_irq(rmi_i2c->irq);
> >  	if (device_may_wakeup(&client->dev)) {
> >  		ret = disable_irq_wake(rmi_i2c->irq);
> > @@ -346,6 +380,9 @@ static int rmi_i2c_runtime_suspend(struct device *dev)
> >  
> >  	disable_irq(rmi_i2c->irq);
> >  
> > +	regulator_bulk_disable(ARRAY_SIZE(rmi_i2c->supplies),
> > +			       rmi_i2c->supplies);
> > +
> >  	return 0;
> >  }
> >  
> > @@ -355,6 +392,14 @@ static int rmi_i2c_runtime_resume(struct device *dev)
> >  	struct rmi_i2c_xport *rmi_i2c = i2c_get_clientdata(client);
> >  	int ret;
> >  
> > +	ret = regulator_bulk_enable(ARRAY_SIZE(rmi_i2c->supplies),
> > +				    rmi_i2c->supplies);
> > +	if (ret)
> > +		return ret;
> > +
> > +	/* Allow the firmware to get ready */
> > +	msleep(10);
> > +
> >  	enable_irq(rmi_i2c->irq);
> >  
> >  	ret = rmi_driver_resume(rmi_i2c->xport.rmi_dev);
> > -- 
> > 2.5.0
> > 
> 
> -- 
> Dmitry

Re: [PATCH] Input: synaptics-rmi4: Support regulator supplies

[Andrew Duggan]

On 03/31/2016 12:14 PM, Bjorn Andersson wrote:
> On Thu 31 Mar 11:19 PDT 2016, Dmitry Torokhov wrote:
>
>> Hi Bjorn,
>>
>> On Wed, Mar 30, 2016 at 09:57:29AM -0700, Bjorn Andersson wrote:
>>> From: Bjorn Andersson <bjorn.andersson@sonymobile.com>
>>>
>>> Support the two supplies - vdd and vio - to make it possible to control
>>> power to the Synaptics chip.
>>>
>>> Signed-off-by: Bjorn Andersson <bjorn.andersson@sonymobile.com>
>>> Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
>>> ---
>>>   .../devicetree/bindings/input/rmi4/rmi_i2c.txt     |  7 ++++
>>>   drivers/input/rmi4/rmi_i2c.c                       | 45 ++++++++++++++++++++++
>> Would not we need pretty much the same changes for SPI devices? Can this
>> be done in core?
>>
> Yes, I believe it needs the exact same steps.
>
> I did a initial quick hack on v1 of the patchset and back then it was
> possible, when I rebased it a few weeks back I kept ending up in getting
> interrupts with the power off.
>
> Looking at the code this is likely because in the resume paths the IRQ
> is enabled before we jump to rmi_driver_resume(), so putting this in the
> core I ended up calling rmi_process_interrupt_requests() before powering
> up the chip.

Actually, I don't think the irq needs to be enabled before calling 
rmi_driver_resume(). Typically, the functions are just reading and 
writing to registers and do not need to handle interrupts. We could 
probably call to rmi_driver_resume() before enabling the irq. I can 
double check that there are not any exceptions to this.

I have also considered adding a power callback to the core so that the 
transport drivers can set the power independently of suspend and resume. 
One example would be to shut off power to a touchpad if a mouse is 
connected. If we do need to have the irq enabled before calling 
rmi_driver_resume() we could still move regulator support to the core 
and call the power callback from the transport drivers.

Andrew

> I assume it's done this way to allow incoming interrupts while functions
> are being resumed. Andrew, can you comment on this?
>
> Regards,
> Bjorn
>
>> Thanks.
>>
>>>   2 files changed, 52 insertions(+)
>>>
>>> diff --git a/Documentation/devicetree/bindings/input/rmi4/rmi_i2c.txt b/Documentation/devicetree/bindings/input/rmi4/rmi_i2c.txt
>>> index 95fa715c6046..a8c31f40f816 100644
>>> --- a/Documentation/devicetree/bindings/input/rmi4/rmi_i2c.txt
>>> +++ b/Documentation/devicetree/bindings/input/rmi4/rmi_i2c.txt
>>> @@ -22,6 +22,13 @@ See Documentation/devicetree/bindings/interrupt-controller/interrupts.txt
>>>   - syna,reset-delay-ms: The number of milliseconds to wait after resetting the
>>>   			device.
>>>   
>>> +- vdd-supply: VDD power supply.
>>> +See Documentation/devicetree/bindings/regulator/regulator.txt
>>> +
>>> +- vio-supply: VIO power supply
>>> +See Documentation/devicetree/bindings/regulator/regulator.txt
>>> +
>>> +
>>>   Function Parameters:
>>>   Parameters specific to RMI functions are contained in child nodes of the rmi device
>>>    node. Documentation for the parameters of each function can be found in:
>>> diff --git a/drivers/input/rmi4/rmi_i2c.c b/drivers/input/rmi4/rmi_i2c.c
>>> index a96a326b53bd..a8c794daba04 100644
>>> --- a/drivers/input/rmi4/rmi_i2c.c
>>> +++ b/drivers/input/rmi4/rmi_i2c.c
>>> @@ -11,6 +11,8 @@
>>>   #include <linux/rmi.h>
>>>   #include <linux/irq.h>
>>>   #include <linux/of.h>
>>> +#include <linux/delay.h>
>>> +#include <linux/regulator/consumer.h>
>>>   #include "rmi_driver.h"
>>>   
>>>   #define BUFFER_SIZE_INCREMENT 32
>>> @@ -37,6 +39,8 @@ struct rmi_i2c_xport {
>>>   
>>>   	u8 *tx_buf;
>>>   	size_t tx_buf_size;
>>> +
>>> +	struct regulator_bulk_data supplies[2];
>>>   };
>>>   
>>>   #define RMI_PAGE_SELECT_REGISTER 0xff
>>> @@ -246,6 +250,22 @@ static int rmi_i2c_probe(struct i2c_client *client,
>>>   		return -ENODEV;
>>>   	}
>>>   
>>> +	rmi_i2c->supplies[0].supply = "vdd";
>>> +	rmi_i2c->supplies[1].supply = "vio";
>>> +	retval = devm_regulator_bulk_get(&client->dev,
>>> +					 ARRAY_SIZE(rmi_i2c->supplies),
>>> +					 rmi_i2c->supplies);
>>> +	if (retval < 0)
>>> +		return retval;
>>> +
>>> +	retval = regulator_bulk_enable(ARRAY_SIZE(rmi_i2c->supplies),
>>> +				       rmi_i2c->supplies);
>>> +	if (retval < 0)
>>> +		return retval;
>>> +
>>> +	/* Allow the firmware to get ready */
>>> +	msleep(10);
>>> +
>>>   	rmi_i2c->client = client;
>>>   	mutex_init(&rmi_i2c->page_mutex);
>>>   
>>> @@ -286,6 +306,8 @@ static int rmi_i2c_remove(struct i2c_client *client)
>>>   	struct rmi_i2c_xport *rmi_i2c = i2c_get_clientdata(client);
>>>   
>>>   	rmi_unregister_transport_device(&rmi_i2c->xport);
>>> +	regulator_bulk_disable(ARRAY_SIZE(rmi_i2c->supplies),
>>> +			       rmi_i2c->supplies);
>>>   
>>>   	return 0;
>>>   }
>>> @@ -308,6 +330,10 @@ static int rmi_i2c_suspend(struct device *dev)
>>>   			dev_warn(dev, "Failed to enable irq for wake: %d\n",
>>>   				ret);
>>>   	}
>>> +
>>> +	regulator_bulk_disable(ARRAY_SIZE(rmi_i2c->supplies),
>>> +			       rmi_i2c->supplies);
>>> +
>>>   	return ret;
>>>   }
>>>   
>>> @@ -317,6 +343,14 @@ static int rmi_i2c_resume(struct device *dev)
>>>   	struct rmi_i2c_xport *rmi_i2c = i2c_get_clientdata(client);
>>>   	int ret;
>>>   
>>> +	ret = regulator_bulk_enable(ARRAY_SIZE(rmi_i2c->supplies),
>>> +				    rmi_i2c->supplies);
>>> +	if (ret)
>>> +		return ret;
>>> +
>>> +	/* Allow the firmware to get ready */
>>> +	msleep(10);
>>> +
>>>   	enable_irq(rmi_i2c->irq);
>>>   	if (device_may_wakeup(&client->dev)) {
>>>   		ret = disable_irq_wake(rmi_i2c->irq);
>>> @@ -346,6 +380,9 @@ static int rmi_i2c_runtime_suspend(struct device *dev)
>>>   
>>>   	disable_irq(rmi_i2c->irq);
>>>   
>>> +	regulator_bulk_disable(ARRAY_SIZE(rmi_i2c->supplies),
>>> +			       rmi_i2c->supplies);
>>> +
>>>   	return 0;
>>>   }
>>>   
>>> @@ -355,6 +392,14 @@ static int rmi_i2c_runtime_resume(struct device *dev)
>>>   	struct rmi_i2c_xport *rmi_i2c = i2c_get_clientdata(client);
>>>   	int ret;
>>>   
>>> +	ret = regulator_bulk_enable(ARRAY_SIZE(rmi_i2c->supplies),
>>> +				    rmi_i2c->supplies);
>>> +	if (ret)
>>> +		return ret;
>>> +
>>> +	/* Allow the firmware to get ready */
>>> +	msleep(10);
>>> +
>>>   	enable_irq(rmi_i2c->irq);
>>>   
>>>   	ret = rmi_driver_resume(rmi_i2c->xport.rmi_dev);
>>> -- 
>>> 2.5.0
>>>
>> -- 
>> Dmitry

Re: [PATCH] Input: synaptics-rmi4: Support regulator supplies

[Bjorn Andersson]

On Thu 31 Mar 18:47 PDT 2016, Andrew Duggan wrote:

> On 03/31/2016 12:14 PM, Bjorn Andersson wrote:
> >On Thu 31 Mar 11:19 PDT 2016, Dmitry Torokhov wrote:
> >
> >>Hi Bjorn,
> >>
> >>On Wed, Mar 30, 2016 at 09:57:29AM -0700, Bjorn Andersson wrote:
> >>>From: Bjorn Andersson <bjorn.andersson@sonymobile.com>
> >>>
> >>>Support the two supplies - vdd and vio - to make it possible to control
> >>>power to the Synaptics chip.
> >>>
> >>>Signed-off-by: Bjorn Andersson <bjorn.andersson@sonymobile.com>
> >>>Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
> >>>---
> >>>  .../devicetree/bindings/input/rmi4/rmi_i2c.txt     |  7 ++++
> >>>  drivers/input/rmi4/rmi_i2c.c                       | 45 ++++++++++++++++++++++
> >>Would not we need pretty much the same changes for SPI devices? Can this
> >>be done in core?
> >>
> >Yes, I believe it needs the exact same steps.
> >
> >I did a initial quick hack on v1 of the patchset and back then it was
> >possible, when I rebased it a few weeks back I kept ending up in getting
> >interrupts with the power off.
> >
> >Looking at the code this is likely because in the resume paths the IRQ
> >is enabled before we jump to rmi_driver_resume(), so putting this in the
> >core I ended up calling rmi_process_interrupt_requests() before powering
> >up the chip.
> 
> Actually, I don't think the irq needs to be enabled before calling
> rmi_driver_resume(). Typically, the functions are just reading and writing
> to registers and do not need to handle interrupts. We could probably call to
> rmi_driver_resume() before enabling the irq. I can double check that there
> are not any exceptions to this.
> 

I finally got back to giving this a spin.

The problem is that we register the transport device with the driver,
which triggers the rmi_driver probe() which resolves the resources. We
then continue on and call rmi_i2c_init_irq() which will (implicitly)
enable the irq. So if the rmi_driver probe() does not finish in a serial
fashion we will enable interrupts before we have fully initialized the
core.

I don't know if this causes other issues, but with the required delay
after enabling the regulators we always get an interrupt before the
rmi_driver probe() function is finished.

> I have also considered adding a power callback to the core so that the
> transport drivers can set the power independently of suspend and resume. One
> example would be to shut off power to a touchpad if a mouse is connected. If
> we do need to have the irq enabled before calling rmi_driver_resume() we
> could still move regulator support to the core and call the power callback
> from the transport drivers.
> 

I see no (sane) way of waiting for the rmi_driver to finish probeing;
there could be cases where it's powered by a regulator (or reset gpio)
that is not yet probed. EPROBE_DEFER will handle this, but we can't wait
for it in the transport driver.


I therefor think these physical resources should be handled in the
context of the transport layer, to make sure we don't have temporal
dependencies to the other layers.

Or we should not have the rmi_driver as a separate device driver at all
- it could be a "library" that runs in the context of the transport
device.

Regards,
Bjorn

Re: [PATCH] Input: synaptics-rmi4: Support regulator supplies

[Andrew Duggan]

Hi Bjorn,

On 04/21/2016 03:37 PM, Bjorn Andersson wrote:
> On Thu 31 Mar 18:47 PDT 2016, Andrew Duggan wrote:
>
>> On 03/31/2016 12:14 PM, Bjorn Andersson wrote:
>>> On Thu 31 Mar 11:19 PDT 2016, Dmitry Torokhov wrote:
>>>
>>>> Hi Bjorn,
>>>>
>>>> On Wed, Mar 30, 2016 at 09:57:29AM -0700, Bjorn Andersson wrote:
>>>>> From: Bjorn Andersson <bjorn.andersson-/MT0OVThwyLZJqsBc5GL+g@public.gmane.org>
>>>>>
>>>>> Support the two supplies - vdd and vio - to make it possible to control
>>>>> power to the Synaptics chip.
>>>>>
>>>>> Signed-off-by: Bjorn Andersson <bjorn.andersson-/MT0OVThwyLZJqsBc5GL+g@public.gmane.org>
>>>>> Signed-off-by: Bjorn Andersson <bjorn.andersson-QSEj5FYQhm4dnm+yROfE0A@public.gmane.org>
>>>>> ---
>>>>>   .../devicetree/bindings/input/rmi4/rmi_i2c.txt     |  7 ++++
>>>>>   drivers/input/rmi4/rmi_i2c.c                       | 45 ++++++++++++++++++++++
>>>> Would not we need pretty much the same changes for SPI devices? Can this
>>>> be done in core?
>>>>
>>> Yes, I believe it needs the exact same steps.
>>>
>>> I did a initial quick hack on v1 of the patchset and back then it was
>>> possible, when I rebased it a few weeks back I kept ending up in getting
>>> interrupts with the power off.
>>>
>>> Looking at the code this is likely because in the resume paths the IRQ
>>> is enabled before we jump to rmi_driver_resume(), so putting this in the
>>> core I ended up calling rmi_process_interrupt_requests() before powering
>>> up the chip.
>> Actually, I don't think the irq needs to be enabled before calling
>> rmi_driver_resume(). Typically, the functions are just reading and writing
>> to registers and do not need to handle interrupts. We could probably call to
>> rmi_driver_resume() before enabling the irq. I can double check that there
>> are not any exceptions to this.
>>
> I finally got back to giving this a spin.
>
> The problem is that we register the transport device with the driver,
> which triggers the rmi_driver probe() which resolves the resources. We
> then continue on and call rmi_i2c_init_irq() which will (implicitly)
> enable the irq. So if the rmi_driver probe() does not finish in a serial
> fashion we will enable interrupts before we have fully initialized the
> core.
>
> I don't know if this causes other issues, but with the required delay
> after enabling the regulators we always get an interrupt before the
> rmi_driver probe() function is finished.

I have not observed any issues related to timing, but it looks like on 
the systems which I have tested on rmi_driver() seems to be completing 
synchronously before the init_irq() call. I was making the assumption 
that rmi_driver() would have completed by the time 
rmi_register_transport_device() returned. But, based on your description 
and looking into the base driver code I see that the probe can be 
deferred and that assumption isn't always true.

>> I have also considered adding a power callback to the core so that the
>> transport drivers can set the power independently of suspend and resume. One
>> example would be to shut off power to a touchpad if a mouse is connected. If
>> we do need to have the irq enabled before calling rmi_driver_resume() we
>> could still move regulator support to the core and call the power callback
>> from the transport drivers.
>>
> I see no (sane) way of waiting for the rmi_driver to finish probeing;
> there could be cases where it's powered by a regulator (or reset gpio)
> that is not yet probed. EPROBE_DEFER will handle this, but we can't wait
> for it in the transport driver.

Do we need to wait for rmi_driver to finish probing? What about setting 
a flag at the end of rmi_driver_probe() which 
rmi_process_interrupt_requests() can check before processing interrupts. 
If rmi_driver hasn't finished probing it could just return.

>
> I therefor think these physical resources should be handled in the
> context of the transport layer, to make sure we don't have temporal
> dependencies to the other layers.

I'm fine with enabling the regulators in the transport driver's probe 
function before calling rmi_register_transport_device() to make sure the 
device is powered on. What about exporting common functions from 
rmi_driver.c which implement common regulator functionality which can 
then be called by the transports? To avoid duplication between rmi_i2c 
and rmi_spi.

> Or we should not have the rmi_driver as a separate device driver at all
> - it could be a "library" that runs in the context of the transport
> device.

I would have to look into this further to understand the impact on the 
bus architecture of merging the physical and transport drivers. But, I 
don't think this particular issue warrants such a change. But, if having 
them as separate devices does cause a lot of other problems, it  might 
be possible to merge them.

Andrew

> Regards,
> Bjorn

--
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

Re: [PATCH] Input: synaptics-rmi4: Support regulator supplies

[Bjorn Andersson]

On Thu 05 May 13:55 PDT 2016, Andrew Duggan wrote:

> Hi Bjorn,
> 
> On 04/21/2016 03:37 PM, Bjorn Andersson wrote:
[..]
> >I see no (sane) way of waiting for the rmi_driver to finish probeing;
> >there could be cases where it's powered by a regulator (or reset gpio)
> >that is not yet probed. EPROBE_DEFER will handle this, but we can't wait
> >for it in the transport driver.
> 
> Do we need to wait for rmi_driver to finish probing? What about setting a
> flag at the end of rmi_driver_probe() which rmi_process_interrupt_requests()
> can check before processing interrupts. If rmi_driver hasn't finished
> probing it could just return.
> 

Note that the required resources could be provided by a kernel module
that is loaded at some future time, or never. So we can't really stall
the transport probe().

> >
> >I therefor think these physical resources should be handled in the
> >context of the transport layer, to make sure we don't have temporal
> >dependencies to the other layers.
> 
> I'm fine with enabling the regulators in the transport driver's probe
> function before calling rmi_register_transport_device() to make sure the
> device is powered on. What about exporting common functions from
> rmi_driver.c which implement common regulator functionality which can then
> be called by the transports? To avoid duplication between rmi_i2c and
> rmi_spi.
> 

For the DT binding documents I think it's best to just duplicate the
information, as long as we don't see more common properties between a
growing number of transports (unlikely).


For the implementation we need a context to operate on before probing
the common code, as far as I can see the two places are per transport or
in struct rmi_transport_dev.

The latter would allow us to provide a few common helper functions.


That part that I can see would make it worth adding this is the delay
(Tpowerup?), especially if we need to do something more advanced here.

I've found various numbers of Tpowerup (10ms to 150ms) and some
implementations out there leave the regulators on during sleep, while
others cut the power.

So it seems we will be up for some level of additional logic here, that
warrants the de-duplication.

> >Or we should not have the rmi_driver as a separate device driver at all
> >- it could be a "library" that runs in the context of the transport
> >device.
> 
> I would have to look into this further to understand the impact on the bus
> architecture of merging the physical and transport drivers. But, I don't
> think this particular issue warrants such a change. But, if having them as
> separate devices does cause a lot of other problems, it  might be possible
> to merge them.
> 

I get the feeling that there is an expected 1:1 relationship between the
transport and core driver, making the use of the driver model for
separation overkill and a potential cause of issues.

I agree this might not warrant the churn of a rewrite, but I'm concern
about the above helpers just working around an artificial issue.

Regards,
Bjorn
--
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