i2c driver develpment with diolan DLN-2 / ACPI?

i2c driver develpment with diolan DLN-2 / ACPI?

[Martin Kepplinger]

Hi roberta or anyone who uses a diolan i2c/spi bridge device,

I try to do get the kernel driver for the connected chip loaded, but I'm
having some trouble getting it right.

Cables are connected to the device (i2c), and the (mainline) kernel
driver for the diolan DLN-2 device is compiled and loaded on plugin.

I don't use any of their (userspace) software.

Now I get a new i2c bus and can manually instanciate my driver

root at laptop:/sys/bus/i2c/devices/i2c-7# echo mma8452 0x1c > new_device

and the driver loads. (Now for development, later maybe for merging) I
added acpi support, I guess I'd need it, but ACPI_HANDLE(&client->dev)
isn't true (during probe) hence we don't continue. I append my
adding-acpi-patch below.

How do you do it?

Do *you* manually instantiate your i2c-driver during development with
your diolan device or should it somehow even get loaded automatically?

thanks so much for any help,

                     martin


--- a/drivers/iio/accel/mma8452.c
+++ b/drivers/iio/accel/mma8452.c
@@ -33,6 +33,7 @@
 #include <linux/of_device.h>
 #include <linux/of_irq.h>
 #include <linux/pm_runtime.h>
+#include <linux/acpi.h>

 #define MMA8452_STATUS				0x00
 #define  MMA8452_STATUS_DRDY			(BIT(2) | BIT(1) | BIT(0))
@@ -1434,6 +1435,19 @@ static const struct of_device_id mma8452_dt_ids[] = {
 };
 MODULE_DEVICE_TABLE(of, mma8452_dt_ids);

+static const char *mma8452_match_acpi_device(struct device *dev, int *data)
+{
+        const struct acpi_device_id *id;
+
+        id = acpi_match_device(dev->driver->acpi_match_table, dev);
+        if (!id)
+                return NULL;
+
+        *data = (int)id->driver_data;
+
+        return dev_name(dev);
+}
+
 static int mma8452_probe(struct i2c_client *client,
 			 const struct i2c_device_id *id)
 {
@@ -1441,21 +1455,35 @@ static int mma8452_probe(struct i2c_client *client,
 	struct iio_dev *indio_dev;
 	int ret;
 	const struct of_device_id *match;
+	const char *name = NULL;
+	int dev_id = 0;

-	match = of_match_device(mma8452_dt_ids, &client->dev);
-	if (!match) {
-		dev_err(&client->dev, "unknown device model\n");
-		return -ENODEV;
-	}
+pr_info("mma probe running!!\n");

 	indio_dev = devm_iio_device_alloc(&client->dev, sizeof(*data));
 	if (!indio_dev)
 		return -ENOMEM;

 	data = iio_priv(indio_dev);
+	i2c_set_clientdata(client, indio_dev);
 	data->client = client;
+
+	match = of_match_device(mma8452_dt_ids, &client->dev);
+	if (match) {
+		if (id)
+			name = id->name;
+		data->chip_info = match->data;
+	} else if (ACPI_HANDLE(&client->dev)) {
+		name = mma8452_match_acpi_device(&client->dev, &dev_id);
+		pr_info("mma acpi ok. found id %d\n", dev_id);
+		/* TODO chip_info setting*/
+		data->chip_info = &mma_chip_info_table[dev_id];
+	} else {
+		dev_err(&client->dev, "unknown device model\n");
+		return -ENODEV;
+	}
+
 	mutex_init(&data->lock);
-	data->chip_info = match->data;

 	ret = i2c_smbus_read_byte_data(client, MMA8452_WHO_AM_I);
 	if (ret < 0)
@@ -1475,11 +1503,10 @@ static int mma8452_probe(struct i2c_client *client,
 	}

 	dev_info(&client->dev, "registering %s accelerometer; ID 0x%x\n",
-		 match->compatible, data->chip_info->chip_id);
+		 name, data->chip_info->chip_id);

-	i2c_set_clientdata(client, indio_dev);
 	indio_dev->info = &mma8452_info;
-	indio_dev->name = id->name;
+	indio_dev->name = name;
 	indio_dev->dev.parent = &client->dev;
 	indio_dev->modes = INDIO_DIRECT_MODE;
 	indio_dev->channels = data->chip_info->channels;
@@ -1669,6 +1696,17 @@ static const struct dev_pm_ops mma8452_pm_ops = {
 			   mma8452_runtime_resume, NULL)
 };

+static const struct acpi_device_id mma8452_acpi_match[] = {
+	{ "MMA8451", mma8451 },
+	{ "MMA8452", mma8452 },
+	{ "MMA8453", mma8453 },
+	{ "MMA8652", mma8652 },
+	{ "MMA8653", mma8653 },
+	{ "FXLS8471", fxls8471 },
+	{ },
+};
+MODULE_DEVICE_TABLE(acpi, mma8452_acpi_match);
+
 static const struct i2c_device_id mma8452_id[] = {
 	{ "mma8451", mma8451 },
 	{ "mma8452", mma8452 },
@@ -1684,6 +1722,7 @@ static struct i2c_driver mma8452_driver = {
 	.driver = {
 		.name	= "mma8452",
 		.of_match_table = of_match_ptr(mma8452_dt_ids),
+		.acpi_match_table = ACPI_PTR(mma8452_acpi_match),
 		.pm	= &mma8452_pm_ops,
 	},
 	.probe = mma8452_probe,
-- 
2.1.4

i2c driver develpment with diolan DLN-2 / ACPI?

[Daniel Baluta]

On Wed, Apr 27, 2016 at 2:36 PM, Martin Kepplinger <martink@posteo.de> wrote:
> Hi roberta or anyone who uses a diolan i2c/spi bridge device,
>
> I try to do get the kernel driver for the connected chip loaded, but I'm
> having some trouble getting it right.
>
> Cables are connected to the device (i2c), and the (mainline) kernel
> driver for the diolan DLN-2 device is compiled and loaded on plugin.
>
> I don't use any of their (userspace) software.
>
> Now I get a new i2c bus and can manually instanciate my driver
>
> root at laptop:/sys/bus/i2c/devices/i2c-7# echo mma8452 0x1c > new_device
>
> and the driver loads. (Now for development, later maybe for merging) I
> added acpi support, I guess I'd need it, but ACPI_HANDLE(&client->dev)
> isn't true (during probe) hence we don't continue. I append my
> adding-acpi-patch below.
>
> How do you do it?
>
> Do *you* manually instantiate your i2c-driver during development with
> your diolan device or should it somehow even get loaded automatically?
>

Hi Martin,

Please also use the following two patches:

https://lkml.org/lkml/2014/12/16/290
https://lkml.org/lkml/2014/12/16/289

How does your setup looks like?

> thanks so much for any help,
>
>                      martin
>
>
> --- a/drivers/iio/accel/mma8452.c
> +++ b/drivers/iio/accel/mma8452.c
> @@ -33,6 +33,7 @@
>  #include <linux/of_device.h>
>  #include <linux/of_irq.h>
>  #include <linux/pm_runtime.h>
> +#include <linux/acpi.h>
>
>  #define MMA8452_STATUS                         0x00
>  #define  MMA8452_STATUS_DRDY                   (BIT(2) | BIT(1) | BIT(0))
> @@ -1434,6 +1435,19 @@ static const struct of_device_id mma8452_dt_ids[] = {
>  };
>  MODULE_DEVICE_TABLE(of, mma8452_dt_ids);
>
> +static const char *mma8452_match_acpi_device(struct device *dev, int *data)
> +{
> +        const struct acpi_device_id *id;
> +
> +        id = acpi_match_device(dev->driver->acpi_match_table, dev);
> +        if (!id)
> +                return NULL;
> +
> +        *data = (int)id->driver_data;
> +
> +        return dev_name(dev);
> +}
> +
>  static int mma8452_probe(struct i2c_client *client,
>                          const struct i2c_device_id *id)
>  {
> @@ -1441,21 +1455,35 @@ static int mma8452_probe(struct i2c_client *client,
>         struct iio_dev *indio_dev;
>         int ret;
>         const struct of_device_id *match;
> +       const char *name = NULL;
> +       int dev_id = 0;
>
> -       match = of_match_device(mma8452_dt_ids, &client->dev);
> -       if (!match) {
> -               dev_err(&client->dev, "unknown device model\n");
> -               return -ENODEV;
> -       }
> +pr_info("mma probe running!!\n");
>
>         indio_dev = devm_iio_device_alloc(&client->dev, sizeof(*data));
>         if (!indio_dev)
>                 return -ENOMEM;
>
>         data = iio_priv(indio_dev);
> +       i2c_set_clientdata(client, indio_dev);
>         data->client = client;
> +
> +       match = of_match_device(mma8452_dt_ids, &client->dev);
> +       if (match) {
> +               if (id)
> +                       name = id->name;
> +               data->chip_info = match->data;
> +       } else if (ACPI_HANDLE(&client->dev)) {
> +               name = mma8452_match_acpi_device(&client->dev, &dev_id);
> +               pr_info("mma acpi ok. found id %d\n", dev_id);
> +               /* TODO chip_info setting*/
> +               data->chip_info = &mma_chip_info_table[dev_id];
> +       } else {
> +               dev_err(&client->dev, "unknown device model\n");
> +               return -ENODEV;
> +       }
> +
>         mutex_init(&data->lock);
> -       data->chip_info = match->data;
>
>         ret = i2c_smbus_read_byte_data(client, MMA8452_WHO_AM_I);
>         if (ret < 0)
> @@ -1475,11 +1503,10 @@ static int mma8452_probe(struct i2c_client *client,
>         }
>
>         dev_info(&client->dev, "registering %s accelerometer; ID 0x%x\n",
> -                match->compatible, data->chip_info->chip_id);
> +                name, data->chip_info->chip_id);
>
> -       i2c_set_clientdata(client, indio_dev);
>         indio_dev->info = &mma8452_info;
> -       indio_dev->name = id->name;
> +       indio_dev->name = name;
>         indio_dev->dev.parent = &client->dev;
>         indio_dev->modes = INDIO_DIRECT_MODE;
>         indio_dev->channels = data->chip_info->channels;
> @@ -1669,6 +1696,17 @@ static const struct dev_pm_ops mma8452_pm_ops = {
>                            mma8452_runtime_resume, NULL)
>  };
>
> +static const struct acpi_device_id mma8452_acpi_match[] = {
> +       { "MMA8451", mma8451 },
> +       { "MMA8452", mma8452 },
> +       { "MMA8453", mma8453 },
> +       { "MMA8652", mma8652 },
> +       { "MMA8653", mma8653 },
> +       { "FXLS8471", fxls8471 },
> +       { },
> +};
> +MODULE_DEVICE_TABLE(acpi, mma8452_acpi_match);
> +
>  static const struct i2c_device_id mma8452_id[] = {
>         { "mma8451", mma8451 },
>         { "mma8452", mma8452 },
> @@ -1684,6 +1722,7 @@ static struct i2c_driver mma8452_driver = {
>         .driver = {
>                 .name   = "mma8452",
>                 .of_match_table = of_match_ptr(mma8452_dt_ids),
> +               .acpi_match_table = ACPI_PTR(mma8452_acpi_match),
>                 .pm     = &mma8452_pm_ops,
>         },
>         .probe = mma8452_probe,

This looks fine to me.

Daniel.

i2c driver develpment with diolan DLN-2 / ACPI?

[Martin Kepplinger]

Am 2016-04-27 um 13:51 schrieb Daniel Baluta:
> On Wed, Apr 27, 2016 at 2:36 PM, Martin Kepplinger <martink@posteo.de> wrote:
>> Hi roberta or anyone who uses a diolan i2c/spi bridge device,
>>
>> I try to do get the kernel driver for the connected chip loaded, but I'm
>> having some trouble getting it right.
>>
>> Cables are connected to the device (i2c), and the (mainline) kernel
>> driver for the diolan DLN-2 device is compiled and loaded on plugin.
>>
>> I don't use any of their (userspace) software.
>>
>> Now I get a new i2c bus and can manually instanciate my driver
>>
>> root at laptop:/sys/bus/i2c/devices/i2c-7# echo mma8452 0x1c > new_device
>>
>> and the driver loads. (Now for development, later maybe for merging) I
>> added acpi support, I guess I'd need it, but ACPI_HANDLE(&client->dev)
>> isn't true (during probe) hence we don't continue. I append my
>> adding-acpi-patch below.
>>
>> How do you do it?
>>
>> Do *you* manually instantiate your i2c-driver during development with
>> your diolan device or should it somehow even get loaded automatically?
>>
> 
> Hi Martin,
> 
> Please also use the following two patches:
> 
> https://lkml.org/lkml/2014/12/16/290
> https://lkml.org/lkml/2014/12/16/289
> 
> How does your setup looks like?
> 

Those patches don't apply anymore, also because of API changes. I save
them but I'll ignore the ACPI problem for now. Oh well, for development
only, I skip it and go ahead and try to write and read registers.

Different problem:

i2c_smbus_read_byte_data always returns with "Protocol Error".

DLN-2 jumpers are correct. Is there anything special to consider, using
the DLN-2, regardless of which chip you are talking to?

thanks
                      martin


>> thanks so much for any help,
>>
>>                      martin
>>
>>
>> --- a/drivers/iio/accel/mma8452.c
>> +++ b/drivers/iio/accel/mma8452.c
>> @@ -33,6 +33,7 @@
>>  #include <linux/of_device.h>
>>  #include <linux/of_irq.h>
>>  #include <linux/pm_runtime.h>
>> +#include <linux/acpi.h>
>>
>>  #define MMA8452_STATUS                         0x00
>>  #define  MMA8452_STATUS_DRDY                   (BIT(2) | BIT(1) | BIT(0))
>> @@ -1434,6 +1435,19 @@ static const struct of_device_id mma8452_dt_ids[] = {
>>  };
>>  MODULE_DEVICE_TABLE(of, mma8452_dt_ids);
>>
>> +static const char *mma8452_match_acpi_device(struct device *dev, int *data)
>> +{
>> +        const struct acpi_device_id *id;
>> +
>> +        id = acpi_match_device(dev->driver->acpi_match_table, dev);
>> +        if (!id)
>> +                return NULL;
>> +
>> +        *data = (int)id->driver_data;
>> +
>> +        return dev_name(dev);
>> +}
>> +
>>  static int mma8452_probe(struct i2c_client *client,
>>                          const struct i2c_device_id *id)
>>  {
>> @@ -1441,21 +1455,35 @@ static int mma8452_probe(struct i2c_client *client,
>>         struct iio_dev *indio_dev;
>>         int ret;
>>         const struct of_device_id *match;
>> +       const char *name = NULL;
>> +       int dev_id = 0;
>>
>> -       match = of_match_device(mma8452_dt_ids, &client->dev);
>> -       if (!match) {
>> -               dev_err(&client->dev, "unknown device model\n");
>> -               return -ENODEV;
>> -       }
>> +pr_info("mma probe running!!\n");
>>
>>         indio_dev = devm_iio_device_alloc(&client->dev, sizeof(*data));
>>         if (!indio_dev)
>>                 return -ENOMEM;
>>
>>         data = iio_priv(indio_dev);
>> +       i2c_set_clientdata(client, indio_dev);
>>         data->client = client;
>> +
>> +       match = of_match_device(mma8452_dt_ids, &client->dev);
>> +       if (match) {
>> +               if (id)
>> +                       name = id->name;
>> +               data->chip_info = match->data;
>> +       } else if (ACPI_HANDLE(&client->dev)) {
>> +               name = mma8452_match_acpi_device(&client->dev, &dev_id);
>> +               pr_info("mma acpi ok. found id %d\n", dev_id);
>> +               /* TODO chip_info setting*/
>> +               data->chip_info = &mma_chip_info_table[dev_id];
>> +       } else {
>> +               dev_err(&client->dev, "unknown device model\n");
>> +               return -ENODEV;
>> +       }
>> +
>>         mutex_init(&data->lock);
>> -       data->chip_info = match->data;
>>
>>         ret = i2c_smbus_read_byte_data(client, MMA8452_WHO_AM_I);
>>         if (ret < 0)
>> @@ -1475,11 +1503,10 @@ static int mma8452_probe(struct i2c_client *client,
>>         }
>>
>>         dev_info(&client->dev, "registering %s accelerometer; ID 0x%x\n",
>> -                match->compatible, data->chip_info->chip_id);
>> +                name, data->chip_info->chip_id);
>>
>> -       i2c_set_clientdata(client, indio_dev);
>>         indio_dev->info = &mma8452_info;
>> -       indio_dev->name = id->name;
>> +       indio_dev->name = name;
>>         indio_dev->dev.parent = &client->dev;
>>         indio_dev->modes = INDIO_DIRECT_MODE;
>>         indio_dev->channels = data->chip_info->channels;
>> @@ -1669,6 +1696,17 @@ static const struct dev_pm_ops mma8452_pm_ops = {
>>                            mma8452_runtime_resume, NULL)
>>  };
>>
>> +static const struct acpi_device_id mma8452_acpi_match[] = {
>> +       { "MMA8451", mma8451 },
>> +       { "MMA8452", mma8452 },
>> +       { "MMA8453", mma8453 },
>> +       { "MMA8652", mma8652 },
>> +       { "MMA8653", mma8653 },
>> +       { "FXLS8471", fxls8471 },
>> +       { },
>> +};
>> +MODULE_DEVICE_TABLE(acpi, mma8452_acpi_match);
>> +
>>  static const struct i2c_device_id mma8452_id[] = {
>>         { "mma8451", mma8451 },
>>         { "mma8452", mma8452 },
>> @@ -1684,6 +1722,7 @@ static struct i2c_driver mma8452_driver = {
>>         .driver = {
>>                 .name   = "mma8452",
>>                 .of_match_table = of_match_ptr(mma8452_dt_ids),
>> +               .acpi_match_table = ACPI_PTR(mma8452_acpi_match),
>>                 .pm     = &mma8452_pm_ops,
>>         },
>>         .probe = mma8452_probe,
> 
> This looks fine to me.
> 
> Daniel.
> 
> _______________________________________________
> Kernelnewbies mailing list
> Kernelnewbies at kernelnewbies.org
> http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
> 

i2c driver develpment with diolan DLN-2 / ACPI?

[Daniel Baluta]

On Sat, Apr 30, 2016 at 6:10 PM, Martin Kepplinger <martink@posteo.de> wrote:
> Am 2016-04-27 um 13:51 schrieb Daniel Baluta:
>> On Wed, Apr 27, 2016 at 2:36 PM, Martin Kepplinger <martink@posteo.de> wrote:
>>> Hi roberta or anyone who uses a diolan i2c/spi bridge device,
>>>
>>> I try to do get the kernel driver for the connected chip loaded, but I'm
>>> having some trouble getting it right.
>>>
>>> Cables are connected to the device (i2c), and the (mainline) kernel
>>> driver for the diolan DLN-2 device is compiled and loaded on plugin.
>>>
>>> I don't use any of their (userspace) software.
>>>
>>> Now I get a new i2c bus and can manually instanciate my driver
>>>
>>> root at laptop:/sys/bus/i2c/devices/i2c-7# echo mma8452 0x1c > new_device
>>>
>>> and the driver loads. (Now for development, later maybe for merging) I
>>> added acpi support, I guess I'd need it, but ACPI_HANDLE(&client->dev)
>>> isn't true (during probe) hence we don't continue. I append my
>>> adding-acpi-patch below.
>>>
>>> How do you do it?
>>>
>>> Do *you* manually instantiate your i2c-driver during development with
>>> your diolan device or should it somehow even get loaded automatically?
>>>
>>
>> Hi Martin,
>>
>> Please also use the following two patches:
>>
>> https://lkml.org/lkml/2014/12/16/290
>> https://lkml.org/lkml/2014/12/16/289
>>
>> How does your setup looks like?
>>
>
> Those patches don't apply anymore, also because of API changes. I save
> them but I'll ignore the ACPI problem for now. Oh well, for development
> only, I skip it and go ahead and try to write and read registers.
>
> Different problem:
>
> i2c_smbus_read_byte_data always returns with "Protocol Error".
>
> DLN-2 jumpers are correct. Is there anything special to consider, using
> the DLN-2, regardless of which chip you are talking to?
>

Please check the wiring. Also make sure you run at the correct
voltage for your specific chip.

One other thing, is the firmware version.

http://dlnware.com/Downloads

AFAIK Diolan by default comes in U2C compatiblity mode
which will not work with DLN2 drivers.

Daniel.

i2c driver develpment with diolan DLN-2 / ACPI?

[Martin Kepplinger]

Am 2016-04-30 um 18:48 schrieb Daniel Baluta:
> On Sat, Apr 30, 2016 at 6:10 PM, Martin Kepplinger <martink@posteo.de> wrote:
>> Am 2016-04-27 um 13:51 schrieb Daniel Baluta:
>>> On Wed, Apr 27, 2016 at 2:36 PM, Martin Kepplinger <martink@posteo.de> wrote:
>>>> Hi roberta or anyone who uses a diolan i2c/spi bridge device,
>>>>
>>>> I try to do get the kernel driver for the connected chip loaded, but I'm
>>>> having some trouble getting it right.
>>>>
>>>> Cables are connected to the device (i2c), and the (mainline) kernel
>>>> driver for the diolan DLN-2 device is compiled and loaded on plugin.
>>>>
>>>> I don't use any of their (userspace) software.
>>>>
>>>> Now I get a new i2c bus and can manually instanciate my driver
>>>>
>>>> root at laptop:/sys/bus/i2c/devices/i2c-7# echo mma8452 0x1c > new_device
>>>>
>>>> and the driver loads. (Now for development, later maybe for merging) I
>>>> added acpi support, I guess I'd need it, but ACPI_HANDLE(&client->dev)
>>>> isn't true (during probe) hence we don't continue. I append my
>>>> adding-acpi-patch below.
>>>>
>>>> How do you do it?
>>>>
>>>> Do *you* manually instantiate your i2c-driver during development with
>>>> your diolan device or should it somehow even get loaded automatically?
>>>>
>>>
>>> Hi Martin,
>>>
>>> Please also use the following two patches:
>>>
>>> https://lkml.org/lkml/2014/12/16/290
>>> https://lkml.org/lkml/2014/12/16/289
>>>
>>> How does your setup looks like?
>>>
>>
>> Those patches don't apply anymore, also because of API changes. I save
>> them but I'll ignore the ACPI problem for now. Oh well, for development
>> only, I skip it and go ahead and try to write and read registers.
>>
>> Different problem:
>>
>> i2c_smbus_read_byte_data always returns with "Protocol Error".
>>
>> DLN-2 jumpers are correct. Is there anything special to consider, using
>> the DLN-2, regardless of which chip you are talking to?
>>
> 
> Please check the wiring. Also make sure you run at the correct
> voltage for your specific chip.

Although I accidently used 5V before, I use 3,3V now, which should be
ok. I doubt 5V killed it, and "Protocol Error" suggests it is alive.

> 
> One other thing, is the firmware version.
> 
> http://dlnware.com/Downloads
> 
> AFAIK Diolan by default comes in U2C compatiblity mode
> which will not work with DLN2 drivers.

They sell the device with their different firmware seperately:
https://diolan.com/dln2 what I have, and
https://diolan.com/dln2u2c  in U2C compatibility.

> 
> Daniel.
> 
> _______________________________________________
> Kernelnewbies mailing list
> Kernelnewbies at kernelnewbies.org
> http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
>