[PATCH 0/3] hwmon: (pmbus) Add support for additional Flex BMR converters to the pmbus driver

[PATCH 0/3] hwmon: (pmbus) Add support for additional Flex BMR converters to the pmbus driver

[Erik Rosen]

Some Flex PMBus converters (BMR310, BMR458, BMR80, BMR490 and BMR491)
end up in an undefined state when trying to read an register that does
not exist in the chip. This causes the following chip access to also
fail even if it is a valid register read. This will mess up the pmbus
driver auto-detection process.

One way to reset the pmbus state machine to a known state is to read
a register that is known to exist. This read will fail but will also
reset the chip into a known state.

For such chips we suggest adding a new pmbus flag:
PMBUS_READ_STATUS_AFTER_FAILED_CHECK
By setting this flag the driver will try to read the STATUS register
after each failed register check. This read may fail, but it will put
the chip into a known state so that the auto-detection process can
proceed correctly.

Add support for Flex BMR310, BMR456, BMR457, BMR458, BMR480, BMR490,
BMR491 and BMR492 to the pmbus driver

This patch has been tested with Flex BMR310, BMR456, BMR457, BMR458,
BMR480, BMR490, BMR491 and BMR492 converter modules

Erik Rosen (3):
  Add new flag PMBUS_READ_STATUS_AFTER_FAILED_CHECK
  Add documentation for new flags
  Add support for Flex BMR310, BMR456, BMR457, BMR458, BMR480, BMR490,
    BMR491 and BMR492 to the pmbus driver

 Documentation/hwmon/pmbus-core.rst | 24 ++++++++++++++++++++++--
 Documentation/hwmon/pmbus.rst      | 11 +++++++----
 drivers/hwmon/pmbus/Kconfig        |  7 ++++---
 drivers/hwmon/pmbus/pmbus.c        | 24 ++++++++++++++++++++++--
 drivers/hwmon/pmbus/pmbus_core.c   |  2 ++
 include/linux/pmbus.h              | 13 +++++++++++++
 6 files changed, 70 insertions(+), 11 deletions(-)


base-commit: 1e28eed17697bcf343c6743f0028cc3b5dd88bf0
-- 
2.20.1

[PATCH 1/3] hwmon: (pmbus) Add new flag PMBUS_READ_STATUS_AFTER_FAILED_CHECK

[Erik Rosen]

Some PMBus chips end up in an undefined state when trying to read an
unsupported register. For such chips, it is necessary to reset the
chip pmbus controller to a known state after a failed register check.
This can be done by reading a known register. By setting this flag the
driver will try to read the STATUS register after each failed
register check. This read may fail, but it will put the chip into a
known state.

Signed-off-by: Erik Rosen <erik.rosen@metormote.com>
---
 drivers/hwmon/pmbus/pmbus_core.c |  2 ++
 include/linux/pmbus.h            | 13 +++++++++++++
 2 files changed, 15 insertions(+)

diff --git a/drivers/hwmon/pmbus/pmbus_core.c b/drivers/hwmon/pmbus/pmbus_core.c
index aadea85fe630..cb0b3c7c3434 100644
--- a/drivers/hwmon/pmbus/pmbus_core.c
+++ b/drivers/hwmon/pmbus/pmbus_core.c
@@ -512,6 +512,8 @@ static bool pmbus_check_register(struct i2c_client *client,
 	rv = func(client, page, reg);
 	if (rv >= 0 && !(data->flags & PMBUS_SKIP_STATUS_CHECK))
 		rv = pmbus_check_status_cml(client);
+	if (rv < 0 && (data->flags & PMBUS_READ_STATUS_AFTER_FAILED_CHECK))
+		data->read_status(client, -1);
 	pmbus_clear_fault_page(client, -1);
 	return rv >= 0;
 }
diff --git a/include/linux/pmbus.h b/include/linux/pmbus.h
index 12cbbf305969..edd7c84fef65 100644
--- a/include/linux/pmbus.h
+++ b/include/linux/pmbus.h
@@ -43,6 +43,19 @@
  */
 #define PMBUS_NO_CAPABILITY			BIT(2)
 
+/*
+ * PMBUS_READ_STATUS_AFTER_FAILED_CHECK
+ *
+ * Some PMBus chips end up in an undefined state when trying to read an
+ * unsupported register. For such chips, it is necessary to reset the
+ * chip pmbus controller to a known state after a failed register check.
+ * This can be done by reading a known register. By setting this flag the
+ * driver will try to read the STATUS register after each failed
+ * register check. This read may fail, but it will put the chip in a
+ * known state.
+ */
+#define PMBUS_READ_STATUS_AFTER_FAILED_CHECK	BIT(3)
+
 struct pmbus_platform_data {
 	u32 flags;		/* Device specific flags */
 
-- 
2.20.1

[PATCH 2/3] hwmon: (pmbus) Add documentation for new flags

[Erik Rosen]

Add documentation for the new pmbus flags PMBUS_WRITE_PROTECTED and
PMBUS_READ_STATUS_AFTER_FAILED_CHECK

Signed-off-by: Erik Rosen <erik.rosen@metormote.com>
---
 Documentation/hwmon/pmbus-core.rst | 24 ++++++++++++++++++++++--
 1 file changed, 22 insertions(+), 2 deletions(-)

diff --git a/Documentation/hwmon/pmbus-core.rst b/Documentation/hwmon/pmbus-core.rst
index 73e23ab42cc3..001c64f0f8c0 100644
--- a/Documentation/hwmon/pmbus-core.rst
+++ b/Documentation/hwmon/pmbus-core.rst
@@ -289,9 +289,13 @@ PMBus driver platform data
 ==========================
 
 PMBus platform data is defined in include/linux/pmbus.h. Platform data
-currently only provides a flag field with a single bit used::
+currently only provides a flag field with three bits used::
 
-	#define PMBUS_SKIP_STATUS_CHECK (1 << 0)
+	#define PMBUS_SKIP_STATUS_CHECK  BIT(0)
+
+  #define PMBUS_WRITE_PROTECTED BIT(1)
+
+  #define PMBUS_READ_STATUS_AFTER_FAILED_CHECK  BIT(2)
 
 	struct pmbus_platform_data {
 		u32 flags;              /* Device specific flags */
@@ -315,3 +319,19 @@ status register must be disabled.
 Some i2c controllers do not support single-byte commands (write commands with
 no data, i2c_smbus_write_byte()). With such controllers, clearing the status
 register is impossible, and the PMBUS_SKIP_STATUS_CHECK flag must be set.
+
+PMBUS_WRITE_PROTECTED
+
+Set if the chip is write protected and write protection is not determined
+by the standard WRITE_PROTECT command.
+
+PMBUS_READ_STATUS_AFTER_FAILED_CHECK
+  Read the STATUS register after each failed register check.
+
+Some PMBus chips end up in an undefined state when trying to read an
+unsupported register. For such chips, it is necessary to reset the
+chip pmbus controller to a known state after a failed register check.
+This can be done by reading a known register. By setting this flag the
+driver will try to read the STATUS register after each failed
+register check. This read may fail, but it will put the chip into a
+known state.
\ No newline at end of file
-- 
2.20.1

[PATCH 3/3] hwmon: (pmbus) Add support for additional Flex BMR converters to pmbus

[Erik Rosen]

Add support for Flex BMR310, BMR456, BMR457, BMR458, BMR480, BMR490,
BMR491 and BMR492 to the pmbus driver

Signed-off-by: Erik Rosen <erik.rosen@metormote.com>
---
 Documentation/hwmon/pmbus.rst | 11 +++++++----
 drivers/hwmon/pmbus/Kconfig   |  7 ++++---
 drivers/hwmon/pmbus/pmbus.c   | 24 ++++++++++++++++++++++--
 3 files changed, 33 insertions(+), 9 deletions(-)

diff --git a/Documentation/hwmon/pmbus.rst b/Documentation/hwmon/pmbus.rst
index c44f14115413..0514c3052d4a 100644
--- a/Documentation/hwmon/pmbus.rst
+++ b/Documentation/hwmon/pmbus.rst
@@ -3,15 +3,18 @@ Kernel driver pmbus
 
 Supported chips:
 
-  * Ericsson BMR453, BMR454
+  * Flex BMR453, BMR454, BMR456, BMR457, BMR458, BMR480,
+    BMR490, BMR491, BMR310, BMR492
 
-    Prefixes: 'bmr453', 'bmr454'
+    Prefixes: 'bmr453', 'bmr454', 'bmr456', 'bmr457', 'bmr458', 'bmr480',
+    'bmr490', 'bmr491', 'bmr310', 'bmr492'
 
     Addresses scanned: -
 
-    Datasheet:
+    Datasheets:
+
+ 	https://flexpowermodules.com/products
 
- http://archive.ericsson.net/service/internet/picov/get?DocNo=28701-EN/LZT146395
 
   * ON Semiconductor ADP4000, NCP4200, NCP4208
 
diff --git a/drivers/hwmon/pmbus/Kconfig b/drivers/hwmon/pmbus/Kconfig
index 32d2fc850621..59080d142bf7 100644
--- a/drivers/hwmon/pmbus/Kconfig
+++ b/drivers/hwmon/pmbus/Kconfig
@@ -19,9 +19,10 @@ config SENSORS_PMBUS
 	default y
 	help
 	  If you say yes here you get hardware monitoring support for generic
-	  PMBus devices, including but not limited to ADP4000, BMR453, BMR454,
-	  MAX20796, MDT040, NCP4200, NCP4208, PDT003, PDT006, PDT012, TPS40400,
-	  TPS544B20, TPS544B25, TPS544C20, TPS544C25, and UDT020.
+	  PMBus devices, including but not limited to ADP4000, BMR310, BMR453,
+	  BMR454, BMR456, BMR457, BMR458, BMR480, BMR490, BMR491, BMR492,
+	  MAX20796, MDT040, NCP4200, NCP4208, PDT003, PDT006, PDT012,
+	  TPS40400, TPS544B20, TPS544B25, TPS544C20, TPS544C25, and UDT020.
 
 	  This driver can also be built as a module. If so, the module will
 	  be called pmbus.
diff --git a/drivers/hwmon/pmbus/pmbus.c b/drivers/hwmon/pmbus/pmbus.c
index a1b4260e75b2..688c4a3a87e0 100644
--- a/drivers/hwmon/pmbus/pmbus.c
+++ b/drivers/hwmon/pmbus/pmbus.c
@@ -173,13 +173,18 @@ static int pmbus_probe(struct i2c_client *client)
 		return -ENOMEM;
 
 	device_info = (struct pmbus_device_info *)i2c_match_id(pmbus_id, client)->driver_data;
-	if (device_info->flags & PMBUS_SKIP_STATUS_CHECK) {
+	if (device_info->flags & PMBUS_SKIP_STATUS_CHECK ||
+	    device_info->flags & PMBUS_READ_STATUS_AFTER_FAILED_CHECK) {
 		pdata = devm_kzalloc(dev, sizeof(struct pmbus_platform_data),
 				     GFP_KERNEL);
 		if (!pdata)
 			return -ENOMEM;
 
-		pdata->flags = PMBUS_SKIP_STATUS_CHECK;
+		pdata->flags = 0;
+		if (device_info->flags & PMBUS_SKIP_STATUS_CHECK)
+			pdata->flags |= PMBUS_SKIP_STATUS_CHECK;
+		if (device_info->flags & PMBUS_READ_STATUS_AFTER_FAILED_CHECK)
+			pdata->flags |= PMBUS_READ_STATUS_AFTER_FAILED_CHECK;
 	}
 
 	info->pages = device_info->pages;
@@ -193,22 +198,37 @@ static const struct pmbus_device_info pmbus_info_one = {
 	.pages = 1,
 	.flags = 0
 };
+
 static const struct pmbus_device_info pmbus_info_zero = {
 	.pages = 0,
 	.flags = 0
 };
+
 static const struct pmbus_device_info pmbus_info_one_skip = {
 	.pages = 1,
 	.flags = PMBUS_SKIP_STATUS_CHECK
 };
 
+static const struct pmbus_device_info pmbus_info_one_status = {
+	.pages = 1,
+	.flags = PMBUS_READ_STATUS_AFTER_FAILED_CHECK
+};
+
 /*
  * Use driver_data to set the number of pages supported by the chip.
  */
 static const struct i2c_device_id pmbus_id[] = {
 	{"adp4000", (kernel_ulong_t)&pmbus_info_one},
+	{"bmr310", (kernel_ulong_t)&pmbus_info_one_status},
 	{"bmr453", (kernel_ulong_t)&pmbus_info_one},
 	{"bmr454", (kernel_ulong_t)&pmbus_info_one},
+	{"bmr456", (kernel_ulong_t)&pmbus_info_one},
+	{"bmr457", (kernel_ulong_t)&pmbus_info_one},
+	{"bmr458", (kernel_ulong_t)&pmbus_info_one_status},
+	{"bmr480", (kernel_ulong_t)&pmbus_info_one_status},
+	{"bmr490", (kernel_ulong_t)&pmbus_info_one_status},
+	{"bmr491", (kernel_ulong_t)&pmbus_info_one_status},
+	{"bmr492", (kernel_ulong_t)&pmbus_info_one},
 	{"dps460", (kernel_ulong_t)&pmbus_info_one_skip},
 	{"dps650ab", (kernel_ulong_t)&pmbus_info_one_skip},
 	{"dps800", (kernel_ulong_t)&pmbus_info_one_skip},
-- 
2.20.1

Re: [PATCH 3/3] hwmon: (pmbus) Add support for additional Flex BMR converters to pmbus

[Guenter Roeck]

On 5/5/21 11:32 AM, Erik Rosen wrote:
> Add support for Flex BMR310, BMR456, BMR457, BMR458, BMR480, BMR490,
> BMR491 and BMR492 to the pmbus driver
> 
> Signed-off-by: Erik Rosen <erik.rosen@metormote.com>
> ---
>  Documentation/hwmon/pmbus.rst | 11 +++++++----
>  drivers/hwmon/pmbus/Kconfig   |  7 ++++---
>  drivers/hwmon/pmbus/pmbus.c   | 24 ++++++++++++++++++++++--
>  3 files changed, 33 insertions(+), 9 deletions(-)
> 
> diff --git a/Documentation/hwmon/pmbus.rst b/Documentation/hwmon/pmbus.rst
> index c44f14115413..0514c3052d4a 100644
> --- a/Documentation/hwmon/pmbus.rst
> +++ b/Documentation/hwmon/pmbus.rst
> @@ -3,15 +3,18 @@ Kernel driver pmbus
>  
>  Supported chips:
>  
> -  * Ericsson BMR453, BMR454
> +  * Flex BMR453, BMR454, BMR456, BMR457, BMR458, BMR480,
> +    BMR490, BMR491, BMR310, BMR492
>  
> -    Prefixes: 'bmr453', 'bmr454'
> +    Prefixes: 'bmr453', 'bmr454', 'bmr456', 'bmr457', 'bmr458', 'bmr480',
> +    'bmr490', 'bmr491', 'bmr310', 'bmr492'
>  
>      Addresses scanned: -
>  
> -    Datasheet:
> +    Datasheets:
> +
> + 	https://flexpowermodules.com/products
>  
> - http://archive.ericsson.net/service/internet/picov/get?DocNo=28701-EN/LZT146395
>  
>    * ON Semiconductor ADP4000, NCP4200, NCP4208
>  
> diff --git a/drivers/hwmon/pmbus/Kconfig b/drivers/hwmon/pmbus/Kconfig
> index 32d2fc850621..59080d142bf7 100644
> --- a/drivers/hwmon/pmbus/Kconfig
> +++ b/drivers/hwmon/pmbus/Kconfig
> @@ -19,9 +19,10 @@ config SENSORS_PMBUS
>  	default y
>  	help
>  	  If you say yes here you get hardware monitoring support for generic
> -	  PMBus devices, including but not limited to ADP4000, BMR453, BMR454,
> -	  MAX20796, MDT040, NCP4200, NCP4208, PDT003, PDT006, PDT012, TPS40400,
> -	  TPS544B20, TPS544B25, TPS544C20, TPS544C25, and UDT020.
> +	  PMBus devices, including but not limited to ADP4000, BMR310, BMR453,
> +	  BMR454, BMR456, BMR457, BMR458, BMR480, BMR490, BMR491, BMR492,
> +	  MAX20796, MDT040, NCP4200, NCP4208, PDT003, PDT006, PDT012,
> +	  TPS40400, TPS544B20, TPS544B25, TPS544C20, TPS544C25, and UDT020.
>  
>  	  This driver can also be built as a module. If so, the module will
>  	  be called pmbus.
> diff --git a/drivers/hwmon/pmbus/pmbus.c b/drivers/hwmon/pmbus/pmbus.c
> index a1b4260e75b2..688c4a3a87e0 100644
> --- a/drivers/hwmon/pmbus/pmbus.c
> +++ b/drivers/hwmon/pmbus/pmbus.c
> @@ -173,13 +173,18 @@ static int pmbus_probe(struct i2c_client *client)
>  		return -ENOMEM;
>  
>  	device_info = (struct pmbus_device_info *)i2c_match_id(pmbus_id, client)->driver_data;
> -	if (device_info->flags & PMBUS_SKIP_STATUS_CHECK) {
> +	if (device_info->flags & PMBUS_SKIP_STATUS_CHECK ||
> +	    device_info->flags & PMBUS_READ_STATUS_AFTER_FAILED_CHECK) {

I don't think it makes sense to skip WRITE_PROTECT here.
Just make this
	if (device_info->flags) {

>  		pdata = devm_kzalloc(dev, sizeof(struct pmbus_platform_data),
>  				     GFP_KERNEL);
>  		if (!pdata)
>  			return -ENOMEM;
>  
> -		pdata->flags = PMBUS_SKIP_STATUS_CHECK;
> +		pdata->flags = 0;
> +		if (device_info->flags & PMBUS_SKIP_STATUS_CHECK)
> +			pdata->flags |= PMBUS_SKIP_STATUS_CHECK;
> +		if (device_info->flags & PMBUS_READ_STATUS_AFTER_FAILED_CHECK)
> +			pdata->flags |= PMBUS_READ_STATUS_AFTER_FAILED_CHECK;

and this
		pdata->flags = device_info->flags;

Guenter

>  	}
>  
>  	info->pages = device_info->pages;
> @@ -193,22 +198,37 @@ static const struct pmbus_device_info pmbus_info_one = {
>  	.pages = 1,
>  	.flags = 0
>  };
> +
>  static const struct pmbus_device_info pmbus_info_zero = {
>  	.pages = 0,
>  	.flags = 0
>  };
> +
>  static const struct pmbus_device_info pmbus_info_one_skip = {
>  	.pages = 1,
>  	.flags = PMBUS_SKIP_STATUS_CHECK
>  };
>  
> +static const struct pmbus_device_info pmbus_info_one_status = {
> +	.pages = 1,
> +	.flags = PMBUS_READ_STATUS_AFTER_FAILED_CHECK
> +};
> +
>  /*
>   * Use driver_data to set the number of pages supported by the chip.
>   */
>  static const struct i2c_device_id pmbus_id[] = {
>  	{"adp4000", (kernel_ulong_t)&pmbus_info_one},
> +	{"bmr310", (kernel_ulong_t)&pmbus_info_one_status},
>  	{"bmr453", (kernel_ulong_t)&pmbus_info_one},
>  	{"bmr454", (kernel_ulong_t)&pmbus_info_one},
> +	{"bmr456", (kernel_ulong_t)&pmbus_info_one},
> +	{"bmr457", (kernel_ulong_t)&pmbus_info_one},
> +	{"bmr458", (kernel_ulong_t)&pmbus_info_one_status},
> +	{"bmr480", (kernel_ulong_t)&pmbus_info_one_status},
> +	{"bmr490", (kernel_ulong_t)&pmbus_info_one_status},
> +	{"bmr491", (kernel_ulong_t)&pmbus_info_one_status},
> +	{"bmr492", (kernel_ulong_t)&pmbus_info_one},
>  	{"dps460", (kernel_ulong_t)&pmbus_info_one_skip},
>  	{"dps650ab", (kernel_ulong_t)&pmbus_info_one_skip},
>  	{"dps800", (kernel_ulong_t)&pmbus_info_one_skip},
> 

Re: [PATCH 2/3] hwmon: (pmbus) Add documentation for new flags

[Guenter Roeck]

On 5/5/21 11:32 AM, Erik Rosen wrote:
> Add documentation for the new pmbus flags PMBUS_WRITE_PROTECTED and
> PMBUS_READ_STATUS_AFTER_FAILED_CHECK
> 
> Signed-off-by: Erik Rosen <erik.rosen@metormote.com>
> ---
>  Documentation/hwmon/pmbus-core.rst | 24 ++++++++++++++++++++++--
>  1 file changed, 22 insertions(+), 2 deletions(-)
> 
> diff --git a/Documentation/hwmon/pmbus-core.rst b/Documentation/hwmon/pmbus-core.rst
> index 73e23ab42cc3..001c64f0f8c0 100644
> --- a/Documentation/hwmon/pmbus-core.rst
> +++ b/Documentation/hwmon/pmbus-core.rst
> @@ -289,9 +289,13 @@ PMBus driver platform data
>  ==========================
>  
>  PMBus platform data is defined in include/linux/pmbus.h. Platform data
> -currently only provides a flag field with a single bit used::
> +currently only provides a flag field with three bits used::
>  
> -	#define PMBUS_SKIP_STATUS_CHECK (1 << 0)
> +	#define PMBUS_SKIP_STATUS_CHECK  BIT(0)
> +
> +  #define PMBUS_WRITE_PROTECTED BIT(1)
> +
> +  #define PMBUS_READ_STATUS_AFTER_FAILED_CHECK  BIT(2)
>  
>  	struct pmbus_platform_data {
>  		u32 flags;              /* Device specific flags */
> @@ -315,3 +319,19 @@ status register must be disabled.
>  Some i2c controllers do not support single-byte commands (write commands with
>  no data, i2c_smbus_write_byte()). With such controllers, clearing the status
>  register is impossible, and the PMBUS_SKIP_STATUS_CHECK flag must be set.
> +
> +PMBUS_WRITE_PROTECTED
> +
> +Set if the chip is write protected and write protection is not determined
> +by the standard WRITE_PROTECT command.
> +
> +PMBUS_READ_STATUS_AFTER_FAILED_CHECK
> +  Read the STATUS register after each failed register check.
> +
> +Some PMBus chips end up in an undefined state when trying to read an
> +unsupported register. For such chips, it is necessary to reset the
> +chip pmbus controller to a known state after a failed register check.
> +This can be done by reading a known register. By setting this flag the
> +driver will try to read the STATUS register after each failed
> +register check. This read may fail, but it will put the chip into a
> +known state.
> \ No newline at end of file
> 

Formatting looks way off in this file. Please update to have
it look nice.

Thanks,
Guenter

Re: [PATCH 1/3] hwmon: (pmbus) Add new flag PMBUS_READ_STATUS_AFTER_FAILED_CHECK

[Guenter Roeck]

On 5/5/21 11:32 AM, Erik Rosen wrote:
> Some PMBus chips end up in an undefined state when trying to read an
> unsupported register. For such chips, it is necessary to reset the
> chip pmbus controller to a known state after a failed register check.
> This can be done by reading a known register. By setting this flag the
> driver will try to read the STATUS register after each failed
> register check. This read may fail, but it will put the chip into a
> known state.
> 
> Signed-off-by: Erik Rosen <erik.rosen@metormote.com>

For my reference:

Reviewed-by: Guenter Roeck <linux@roeck-us.net>

> ---
>  drivers/hwmon/pmbus/pmbus_core.c |  2 ++
>  include/linux/pmbus.h            | 13 +++++++++++++
>  2 files changed, 15 insertions(+)
> 
> diff --git a/drivers/hwmon/pmbus/pmbus_core.c b/drivers/hwmon/pmbus/pmbus_core.c
> index aadea85fe630..cb0b3c7c3434 100644
> --- a/drivers/hwmon/pmbus/pmbus_core.c
> +++ b/drivers/hwmon/pmbus/pmbus_core.c
> @@ -512,6 +512,8 @@ static bool pmbus_check_register(struct i2c_client *client,
>  	rv = func(client, page, reg);
>  	if (rv >= 0 && !(data->flags & PMBUS_SKIP_STATUS_CHECK))
>  		rv = pmbus_check_status_cml(client);
> +	if (rv < 0 && (data->flags & PMBUS_READ_STATUS_AFTER_FAILED_CHECK))
> +		data->read_status(client, -1);
>  	pmbus_clear_fault_page(client, -1);
>  	return rv >= 0;
>  }
> diff --git a/include/linux/pmbus.h b/include/linux/pmbus.h
> index 12cbbf305969..edd7c84fef65 100644
> --- a/include/linux/pmbus.h
> +++ b/include/linux/pmbus.h
> @@ -43,6 +43,19 @@
>   */
>  #define PMBUS_NO_CAPABILITY			BIT(2)
>  
> +/*
> + * PMBUS_READ_STATUS_AFTER_FAILED_CHECK
> + *
> + * Some PMBus chips end up in an undefined state when trying to read an
> + * unsupported register. For such chips, it is necessary to reset the
> + * chip pmbus controller to a known state after a failed register check.
> + * This can be done by reading a known register. By setting this flag the
> + * driver will try to read the STATUS register after each failed
> + * register check. This read may fail, but it will put the chip in a
> + * known state.
> + */
> +#define PMBUS_READ_STATUS_AFTER_FAILED_CHECK	BIT(3)
> +
>  struct pmbus_platform_data {
>  	u32 flags;		/* Device specific flags */
>  
>