Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1635432 > unrolled thread

[PATCH] Input: elan_i2c - check if device is there before really probing

Started byDmitry Torokhov <dmitry.torokhov@gmail.com>
First post2017-05-04 08:40 +0200
Last post2017-05-09 00:50 +0200
Articles 5 — 2 participants

Back to article view | Back to linux.kernel

Contents

  [PATCH] Input: elan_i2c - check if device is there before really  probing Dmitry Torokhov <dmitry.torokhov@gmail.com> - 2017-05-04 08:40 +0200
    Re: [PATCH] Input: elan_i2c - check if device is there before really  probing Guenter Roeck <linux@roeck-us.net> - 2017-05-04 08:50 +0200
      Re: [PATCH] Input: elan_i2c - check if device is there before really  probing Dmitry Torokhov <dmitry.torokhov@gmail.com> - 2017-05-08 23:20 +0200
        Re: [PATCH] Input: elan_i2c - check if device is there before really  probing Guenter Roeck <linux@roeck-us.net> - 2017-05-08 23:50 +0200
          Re: [PATCH] Input: elan_i2c - check if device is there before really  probing Dmitry Torokhov <dmitry.torokhov@gmail.com> - 2017-05-09 00:50 +0200

#1635432 — [PATCH] Input: elan_i2c - check if device is there before really probing

FromDmitry Torokhov <dmitry.torokhov@gmail.com>
Date2017-05-04 08:40 +0200
Subject[PATCH] Input: elan_i2c - check if device is there before really probing
Message-ID<tDiAF-8dQ-1@gated-at.bofh.it>
Before trying to properly initialize the touchpad and generate bunch of
errors, let's first see it there is anything at the given address. If we
get error, fail silently with -ENXIO.

Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
---
 drivers/input/mouse/elan_i2c_core.c | 9 +++++++++
 1 file changed, 9 insertions(+)

diff --git a/drivers/input/mouse/elan_i2c_core.c b/drivers/input/mouse/elan_i2c_core.c
index 5d3b53dd2fa2..0479730e7c5c 100644
--- a/drivers/input/mouse/elan_i2c_core.c
+++ b/drivers/input/mouse/elan_i2c_core.c
@@ -1040,6 +1040,7 @@ static int elan_probe(struct i2c_client *client,
 	const struct elan_transport_ops *transport_ops;
 	struct device *dev = &client->dev;
 	struct elan_tp_data *data;
+	union i2c_smbus_data dummy;
 	unsigned long irqflags;
 	int error;
 
@@ -1091,6 +1092,14 @@ static int elan_probe(struct i2c_client *client,
 		return error;
 	}
 
+	/* Make sure there is something at this address */
+	error = i2c_smbus_xfer(client->adapter, client->addr, 0,
+			       I2C_SMBUS_READ, 0, I2C_SMBUS_BYTE, &dummy);
+	if (error) {
+		dev_dbg(&client->dev, "nothing at this address: %d\n", error);
+		return -ENXIO;
+	}
+
 	/* Initialize the touchpad. */
 	error = elan_initialize(data);
 	if (error)
-- 
2.13.0.rc1.294.g07d810a77f-goog


-- 
Dmitry

[toc] | [next] | [standalone]

#1635436

FromGuenter Roeck <linux@roeck-us.net>
Date2017-05-04 08:50 +0200
Message-ID<tDiKl-8hi-3@gated-at.bofh.it>
In reply to#1635432
On 05/03/2017 11:29 PM, Dmitry Torokhov wrote:
> Before trying to properly initialize the touchpad and generate bunch of
> errors, let's first see it there is anything at the given address. If we
> get error, fail silently with -ENXIO.
>
> Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
> ---
>  drivers/input/mouse/elan_i2c_core.c | 9 +++++++++
>  1 file changed, 9 insertions(+)
>
> diff --git a/drivers/input/mouse/elan_i2c_core.c b/drivers/input/mouse/elan_i2c_core.c
> index 5d3b53dd2fa2..0479730e7c5c 100644
> --- a/drivers/input/mouse/elan_i2c_core.c
> +++ b/drivers/input/mouse/elan_i2c_core.c
> @@ -1040,6 +1040,7 @@ static int elan_probe(struct i2c_client *client,
>  	const struct elan_transport_ops *transport_ops;
>  	struct device *dev = &client->dev;
>  	struct elan_tp_data *data;
> +	union i2c_smbus_data dummy;
>  	unsigned long irqflags;
>  	int error;
>
> @@ -1091,6 +1092,14 @@ static int elan_probe(struct i2c_client *client,
>  		return error;
>  	}
>
> +	/* Make sure there is something at this address */
> +	error = i2c_smbus_xfer(client->adapter, client->addr, 0,
> +			       I2C_SMBUS_READ, 0, I2C_SMBUS_BYTE, &dummy);

Can't you just call i2c_smbus_read_byte_data(client, 0); and check if it returns
an error ?

Thanks,
Guenter

> +	if (error) {
> +		dev_dbg(&client->dev, "nothing at this address: %d\n", error);
> +		return -ENXIO;
> +	}
> +
>  	/* Initialize the touchpad. */
>  	error = elan_initialize(data);
>  	if (error)
>

[toc] | [prev] | [next] | [standalone]

#1637727

FromDmitry Torokhov <dmitry.torokhov@gmail.com>
Date2017-05-08 23:20 +0200
Message-ID<tEYet-Lz-7@gated-at.bofh.it>
In reply to#1635436
On Wed, May 03, 2017 at 11:42:54PM -0700, Guenter Roeck wrote:
> On 05/03/2017 11:29 PM, Dmitry Torokhov wrote:
> >Before trying to properly initialize the touchpad and generate bunch of
> >errors, let's first see it there is anything at the given address. If we
> >get error, fail silently with -ENXIO.
> >
> >Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
> >---
> > drivers/input/mouse/elan_i2c_core.c | 9 +++++++++
> > 1 file changed, 9 insertions(+)
> >
> >diff --git a/drivers/input/mouse/elan_i2c_core.c b/drivers/input/mouse/elan_i2c_core.c
> >index 5d3b53dd2fa2..0479730e7c5c 100644
> >--- a/drivers/input/mouse/elan_i2c_core.c
> >+++ b/drivers/input/mouse/elan_i2c_core.c
> >@@ -1040,6 +1040,7 @@ static int elan_probe(struct i2c_client *client,
> > 	const struct elan_transport_ops *transport_ops;
> > 	struct device *dev = &client->dev;
> > 	struct elan_tp_data *data;
> >+	union i2c_smbus_data dummy;
> > 	unsigned long irqflags;
> > 	int error;
> >
> >@@ -1091,6 +1092,14 @@ static int elan_probe(struct i2c_client *client,
> > 		return error;
> > 	}
> >
> >+	/* Make sure there is something at this address */
> >+	error = i2c_smbus_xfer(client->adapter, client->addr, 0,
> >+			       I2C_SMBUS_READ, 0, I2C_SMBUS_BYTE, &dummy);
> 
> Can't you just call i2c_smbus_read_byte_data(client, 0); and check if it returns
> an error ?

Are we guaranteed to get a result from command "0"? I am not sure, and
looking at least at i2c_smbus_xfer_emulated() I2C_SMBUS_READ with
I2C_SMBUS_BYTE modifier works differently form
I2C_SMBUS_READ/I2C_SMBUS_BYTE_DATA (the former uses a single i2c message
vs 2 for the latter).

Also the form above is what i2c core uses to do detection when adapter
does not implement I2C_SMBUS_QUICK, so I'd rather leave it as it was.

Thanks.

-- 
Dmitry

[toc] | [prev] | [next] | [standalone]

#1637749

FromGuenter Roeck <linux@roeck-us.net>
Date2017-05-08 23:50 +0200
Message-ID<tEYHw-Wn-17@gated-at.bofh.it>
In reply to#1637727
On Mon, May 08, 2017 at 02:10:31PM -0700, Dmitry Torokhov wrote:
> On Wed, May 03, 2017 at 11:42:54PM -0700, Guenter Roeck wrote:
> > On 05/03/2017 11:29 PM, Dmitry Torokhov wrote:
> > >Before trying to properly initialize the touchpad and generate bunch of
> > >errors, let's first see it there is anything at the given address. If we
> > >get error, fail silently with -ENXIO.
> > >
> > >Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
> > >---
> > > drivers/input/mouse/elan_i2c_core.c | 9 +++++++++
> > > 1 file changed, 9 insertions(+)
> > >
> > >diff --git a/drivers/input/mouse/elan_i2c_core.c b/drivers/input/mouse/elan_i2c_core.c
> > >index 5d3b53dd2fa2..0479730e7c5c 100644
> > >--- a/drivers/input/mouse/elan_i2c_core.c
> > >+++ b/drivers/input/mouse/elan_i2c_core.c
> > >@@ -1040,6 +1040,7 @@ static int elan_probe(struct i2c_client *client,
> > > 	const struct elan_transport_ops *transport_ops;
> > > 	struct device *dev = &client->dev;
> > > 	struct elan_tp_data *data;
> > >+	union i2c_smbus_data dummy;
> > > 	unsigned long irqflags;
> > > 	int error;
> > >
> > >@@ -1091,6 +1092,14 @@ static int elan_probe(struct i2c_client *client,
> > > 		return error;
> > > 	}
> > >
> > >+	/* Make sure there is something at this address */
> > >+	error = i2c_smbus_xfer(client->adapter, client->addr, 0,
> > >+			       I2C_SMBUS_READ, 0, I2C_SMBUS_BYTE, &dummy);
> > 
> > Can't you just call i2c_smbus_read_byte_data(client, 0); and check if it returns
> > an error ?
> 
> Are we guaranteed to get a result from command "0"? I am not sure, and
> looking at least at i2c_smbus_xfer_emulated() I2C_SMBUS_READ with
> I2C_SMBUS_BYTE modifier works differently form
> I2C_SMBUS_READ/I2C_SMBUS_BYTE_DATA (the former uses a single i2c message
> vs 2 for the latter).
> 
> Also the form above is what i2c core uses to do detection when adapter
> does not implement I2C_SMBUS_QUICK, so I'd rather leave it as it was.
> 

How about i2c_smbus_read_byte() ?

	status = i2c_smbus_xfer(client->adapter, client->addr, client->flags,
	                        I2C_SMBUS_READ, 0,
				I2C_SMBUS_BYTE, &data);
	return (status < 0) ? status : data.byte;

Guenter

[toc] | [prev] | [next] | [standalone]

#1637779

FromDmitry Torokhov <dmitry.torokhov@gmail.com>
Date2017-05-09 00:50 +0200
Message-ID<tEZDz-1zC-1@gated-at.bofh.it>
In reply to#1637749
On Mon, May 08, 2017 at 02:44:25PM -0700, Guenter Roeck wrote:
> On Mon, May 08, 2017 at 02:10:31PM -0700, Dmitry Torokhov wrote:
> > On Wed, May 03, 2017 at 11:42:54PM -0700, Guenter Roeck wrote:
> > > On 05/03/2017 11:29 PM, Dmitry Torokhov wrote:
> > > >Before trying to properly initialize the touchpad and generate bunch of
> > > >errors, let's first see it there is anything at the given address. If we
> > > >get error, fail silently with -ENXIO.
> > > >
> > > >Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
> > > >---
> > > > drivers/input/mouse/elan_i2c_core.c | 9 +++++++++
> > > > 1 file changed, 9 insertions(+)
> > > >
> > > >diff --git a/drivers/input/mouse/elan_i2c_core.c b/drivers/input/mouse/elan_i2c_core.c
> > > >index 5d3b53dd2fa2..0479730e7c5c 100644
> > > >--- a/drivers/input/mouse/elan_i2c_core.c
> > > >+++ b/drivers/input/mouse/elan_i2c_core.c
> > > >@@ -1040,6 +1040,7 @@ static int elan_probe(struct i2c_client *client,
> > > > 	const struct elan_transport_ops *transport_ops;
> > > > 	struct device *dev = &client->dev;
> > > > 	struct elan_tp_data *data;
> > > >+	union i2c_smbus_data dummy;
> > > > 	unsigned long irqflags;
> > > > 	int error;
> > > >
> > > >@@ -1091,6 +1092,14 @@ static int elan_probe(struct i2c_client *client,
> > > > 		return error;
> > > > 	}
> > > >
> > > >+	/* Make sure there is something at this address */
> > > >+	error = i2c_smbus_xfer(client->adapter, client->addr, 0,
> > > >+			       I2C_SMBUS_READ, 0, I2C_SMBUS_BYTE, &dummy);
> > > 
> > > Can't you just call i2c_smbus_read_byte_data(client, 0); and check if it returns
> > > an error ?
> > 
> > Are we guaranteed to get a result from command "0"? I am not sure, and
> > looking at least at i2c_smbus_xfer_emulated() I2C_SMBUS_READ with
> > I2C_SMBUS_BYTE modifier works differently form
> > I2C_SMBUS_READ/I2C_SMBUS_BYTE_DATA (the former uses a single i2c message
> > vs 2 for the latter).
> > 
> > Also the form above is what i2c core uses to do detection when adapter
> > does not implement I2C_SMBUS_QUICK, so I'd rather leave it as it was.
> > 
> 
> How about i2c_smbus_read_byte() ?
> 
> 	status = i2c_smbus_xfer(client->adapter, client->addr, client->flags,
> 	                        I2C_SMBUS_READ, 0,
> 				I2C_SMBUS_BYTE, &data);
> 	return (status < 0) ? status : data.byte;

Yeah, I think this will work.

Thanks.

-- 
Dmitry

[toc] | [prev] | [standalone]

Back to top | Article view | linux.kernel

csiph-web