[RFC v2 0/2] platform/x86: Allow retrieving the number of WMI object instances

[RFC v2 0/2] platform/x86: Allow retrieving the number of WMI object instances

[Armin Wolf]

This experimental patch series allows WMI drivers to retrieve
the number of WMI object instances. This functionallity benefits
several current and upcoming WMI drivers, the dell-wmi-sysman driver
is converted to use this new API as an example.

The changes are compile-tested only, the change to the dell-wmi-sysman
driver also needs to be tested on real hardware.

Armin Wolf (2):
  platform/x86: wmi: Allow retrieving the number of WMI object instances
  platform/x86: dell-sysman: Improve instance detection

 .../x86/dell/dell-wmi-sysman/sysman.c         | 15 ++++---
 drivers/platform/x86/wmi.c                    | 40 +++++++++++++++++++
 include/linux/acpi.h                          |  2 +
 include/linux/wmi.h                           |  2 +
 4 files changed, 51 insertions(+), 8 deletions(-)

--
2.30.2

[RFC v2 1/2] platform/x86: wmi: Allow retrieving the number of WMI object instances

[Armin Wolf]

Currently, the WMI driver core knows how many instances of a given
WMI object exist, but WMI drivers cannot access this information.
At the same time, some current and upcoming WMI drivers want to
have access to this information. Add wmi_instance_count() and
wmidev_instance_count() to allow WMI drivers to get the number of
WMI object instances.

Signed-off-by: Armin Wolf <W_Armin@gmx.de>
---
 drivers/platform/x86/wmi.c | 40 ++++++++++++++++++++++++++++++++++++++
 include/linux/acpi.h       |  2 ++
 include/linux/wmi.h        |  2 ++
 3 files changed, 44 insertions(+)

diff --git a/drivers/platform/x86/wmi.c b/drivers/platform/x86/wmi.c
index c226dd4163a1..7c1a904dec5f 100644
--- a/drivers/platform/x86/wmi.c
+++ b/drivers/platform/x86/wmi.c
@@ -263,6 +263,46 @@ int set_required_buffer_size(struct wmi_device *wdev, u64 length)
 }
 EXPORT_SYMBOL_GPL(set_required_buffer_size);

+/**
+ * wmi_instance_count - Get number of WMI object instances
+ * @guid_string: 36 char string of the form fa50ff2b-f2e8-45de-83fa-65417f2f49ba
+ * @instance_count: variable to hold the instance count
+ *
+ * Get the number of WMI object instances.
+ *
+ * Returns: acpi_status signaling success or error.
+ */
+acpi_status wmi_instance_count(const char *guid_string, u8 *instance_count)
+{
+	struct wmi_block *wblock;
+	acpi_status status;
+
+	status = find_guid(guid_string, &wblock);
+	if (ACPI_FAILURE(status))
+		return status;
+
+	*instance_count = wmidev_instance_count(&wblock->dev);
+
+	return AE_OK;
+}
+EXPORT_SYMBOL_GPL(wmi_instance_count);
+
+/**
+ * wmidev_instance_count - Get number of WMI object instances
+ * @wdev: A wmi bus device from a driver
+ *
+ * Get the number of WMI object instances.
+ *
+ * Returns: Number of WMI object instances.
+ */
+u8 wmidev_instance_count(struct wmi_device *wdev)
+{
+	struct wmi_block *wblock = container_of(wdev, struct wmi_block, dev);
+
+	return wblock->gblock.instance_count;
+}
+EXPORT_SYMBOL_GPL(wmidev_instance_count);
+
 /**
  * wmi_evaluate_method - Evaluate a WMI method (deprecated)
  * @guid_string: 36 char string of the form fa50ff2b-f2e8-45de-83fa-65417f2f49ba
diff --git a/include/linux/acpi.h b/include/linux/acpi.h
index efff750f326d..ab2a4b23e7a3 100644
--- a/include/linux/acpi.h
+++ b/include/linux/acpi.h
@@ -412,6 +412,8 @@ extern bool acpi_is_pnp_device(struct acpi_device *);

 typedef void (*wmi_notify_handler) (u32 value, void *context);

+acpi_status wmi_instance_count(const char *guid, u8 *instance_count);
+
 extern acpi_status wmi_evaluate_method(const char *guid, u8 instance,
 					u32 method_id,
 					const struct acpi_buffer *in,
diff --git a/include/linux/wmi.h b/include/linux/wmi.h
index c1a3bd4e4838..763bd382cf2d 100644
--- a/include/linux/wmi.h
+++ b/include/linux/wmi.h
@@ -35,6 +35,8 @@ extern acpi_status wmidev_evaluate_method(struct wmi_device *wdev,
 extern union acpi_object *wmidev_block_query(struct wmi_device *wdev,
 					     u8 instance);

+u8 wmidev_instance_count(struct wmi_device *wdev);
+
 extern int set_required_buffer_size(struct wmi_device *wdev, u64 length);

 /**
--
2.30.2

Re: [RFC v2 1/2] platform/x86: wmi: Allow retrieving the number of WMI object instances

[James Seo]

On Wed, Apr 26, 2023 at 11:28:47PM +0200, Armin Wolf wrote:
> Currently, the WMI driver core knows how many instances of a given
> WMI object exist, but WMI drivers cannot access this information.
> At the same time, some current and upcoming WMI drivers want to
> have access to this information. Add wmi_instance_count() and
> wmidev_instance_count() to allow WMI drivers to get the number of
> WMI object instances.
> 
> Signed-off-by: Armin Wolf <W_Armin@gmx.de>

Tested-by: James Seo <james@equiv.tech>

Re: [RFC v2 1/2] platform/x86: wmi: Allow retrieving the number of WMI object instances

[James Seo]

On Wed, Apr 26, 2023 at 11:28:47PM +0200, Armin Wolf wrote:
> Currently, the WMI driver core knows how many instances of a given
> WMI object exist, but WMI drivers cannot access this information.
> At the same time, some current and upcoming WMI drivers want to
> have access to this information. Add wmi_instance_count() and
> wmidev_instance_count() to allow WMI drivers to get the number of
> WMI object instances.
> 
> Signed-off-by: Armin Wolf <W_Armin@gmx.de>

Tested-by: James Seo <james@equiv.tech>

Re: [RFC v2 1/2] platform/x86: wmi: Allow retrieving the number of WMI object instances

[Hans de Goede]

Hi Armin,

Thank you for your work on this.

On 4/26/23 23:28, Armin Wolf wrote:
> Currently, the WMI driver core knows how many instances of a given
> WMI object exist, but WMI drivers cannot access this information.
> At the same time, some current and upcoming WMI drivers want to
> have access to this information. Add wmi_instance_count() and
> wmidev_instance_count() to allow WMI drivers to get the number of
> WMI object instances.
> 
> Signed-off-by: Armin Wolf <W_Armin@gmx.de>
> ---
>  drivers/platform/x86/wmi.c | 40 ++++++++++++++++++++++++++++++++++++++
>  include/linux/acpi.h       |  2 ++
>  include/linux/wmi.h        |  2 ++
>  3 files changed, 44 insertions(+)
> 
> diff --git a/drivers/platform/x86/wmi.c b/drivers/platform/x86/wmi.c
> index c226dd4163a1..7c1a904dec5f 100644
> --- a/drivers/platform/x86/wmi.c
> +++ b/drivers/platform/x86/wmi.c
> @@ -263,6 +263,46 @@ int set_required_buffer_size(struct wmi_device *wdev, u64 length)
>  }
>  EXPORT_SYMBOL_GPL(set_required_buffer_size);
> 
> +/**
> + * wmi_instance_count - Get number of WMI object instances
> + * @guid_string: 36 char string of the form fa50ff2b-f2e8-45de-83fa-65417f2f49ba
> + * @instance_count: variable to hold the instance count
> + *
> + * Get the number of WMI object instances.
> + *
> + * Returns: acpi_status signaling success or error.
> + */
> +acpi_status wmi_instance_count(const char *guid_string, u8 *instance_count)
> +{
> +	struct wmi_block *wblock;
> +	acpi_status status;
> +
> +	status = find_guid(guid_string, &wblock);
> +	if (ACPI_FAILURE(status))
> +		return status;
> +
> +	*instance_count = wmidev_instance_count(&wblock->dev);
> +
> +	return AE_OK;
> +}
> +EXPORT_SYMBOL_GPL(wmi_instance_count);

I would prefer this to have a normal kernel function prototype
which returns -errno rather then returning an acpi_status. E.g. :

/**
 * wmi_instance_count - Get number of WMI object instances
 * @guid_string: 36 char string of the form fa50ff2b-f2e8-45de-83fa-65417f2f49ba
 *
 * Get the number of WMI object instances.
 *
 * Returns: The number of WMI object instances, 0 if the GUID is not found.
 */
int wmi_instance_count(const char *guid_string)
{
	struct wmi_block *wblock;
	acpi_status status;

	status = find_guid(guid_string, &wblock);
	if (ACPI_FAILURE(status))
		return 0;

	return wmidev_instance_count(&wblock->dev);
}
EXPORT_SYMBOL_GPL(wmi_instance_count);

This will also allow this to completely replace
the get_instance_count() function in dell-wmi-sysman.

Note I have just gone with always returning 0 here
on error. I guess you could look at the status and
return 0 for not-found and -errno for other errors
but I don't think any callers will care for the difference,
so just always returning 0 seems easier for callers to
deal with.

As always this is just a suggestion, let me know if
you think this is a bad idea.

Regards,

Hans










> +
> +/**
> + * wmidev_instance_count - Get number of WMI object instances
> + * @wdev: A wmi bus device from a driver
> + *
> + * Get the number of WMI object instances.
> + *
> + * Returns: Number of WMI object instances.
> + */
> +u8 wmidev_instance_count(struct wmi_device *wdev)
> +{
> +	struct wmi_block *wblock = container_of(wdev, struct wmi_block, dev);
> +
> +	return wblock->gblock.instance_count;
> +}
> +EXPORT_SYMBOL_GPL(wmidev_instance_count);
> +
>  /**
>   * wmi_evaluate_method - Evaluate a WMI method (deprecated)
>   * @guid_string: 36 char string of the form fa50ff2b-f2e8-45de-83fa-65417f2f49ba
> diff --git a/include/linux/acpi.h b/include/linux/acpi.h
> index efff750f326d..ab2a4b23e7a3 100644
> --- a/include/linux/acpi.h
> +++ b/include/linux/acpi.h
> @@ -412,6 +412,8 @@ extern bool acpi_is_pnp_device(struct acpi_device *);
> 
>  typedef void (*wmi_notify_handler) (u32 value, void *context);
> 
> +acpi_status wmi_instance_count(const char *guid, u8 *instance_count);
> +
>  extern acpi_status wmi_evaluate_method(const char *guid, u8 instance,
>  					u32 method_id,
>  					const struct acpi_buffer *in,
> diff --git a/include/linux/wmi.h b/include/linux/wmi.h
> index c1a3bd4e4838..763bd382cf2d 100644
> --- a/include/linux/wmi.h
> +++ b/include/linux/wmi.h
> @@ -35,6 +35,8 @@ extern acpi_status wmidev_evaluate_method(struct wmi_device *wdev,
>  extern union acpi_object *wmidev_block_query(struct wmi_device *wdev,
>  					     u8 instance);
> 
> +u8 wmidev_instance_count(struct wmi_device *wdev);
> +
>  extern int set_required_buffer_size(struct wmi_device *wdev, u64 length);
> 
>  /**
> --
> 2.30.2
> 

Re: [RFC v2 1/2] platform/x86: wmi: Allow retrieving the number of WMI object instances

[Armin Wolf]

Am 27.04.23 um 11:43 schrieb Hans de Goede:

> Hi Armin,
>
> Thank you for your work on this.
>
> On 4/26/23 23:28, Armin Wolf wrote:
>> Currently, the WMI driver core knows how many instances of a given
>> WMI object exist, but WMI drivers cannot access this information.
>> At the same time, some current and upcoming WMI drivers want to
>> have access to this information. Add wmi_instance_count() and
>> wmidev_instance_count() to allow WMI drivers to get the number of
>> WMI object instances.
>>
>> Signed-off-by: Armin Wolf <W_Armin@gmx.de>
>> ---
>>   drivers/platform/x86/wmi.c | 40 ++++++++++++++++++++++++++++++++++++++
>>   include/linux/acpi.h       |  2 ++
>>   include/linux/wmi.h        |  2 ++
>>   3 files changed, 44 insertions(+)
>>
>> diff --git a/drivers/platform/x86/wmi.c b/drivers/platform/x86/wmi.c
>> index c226dd4163a1..7c1a904dec5f 100644
>> --- a/drivers/platform/x86/wmi.c
>> +++ b/drivers/platform/x86/wmi.c
>> @@ -263,6 +263,46 @@ int set_required_buffer_size(struct wmi_device *wdev, u64 length)
>>   }
>>   EXPORT_SYMBOL_GPL(set_required_buffer_size);
>>
>> +/**
>> + * wmi_instance_count - Get number of WMI object instances
>> + * @guid_string: 36 char string of the form fa50ff2b-f2e8-45de-83fa-65417f2f49ba
>> + * @instance_count: variable to hold the instance count
>> + *
>> + * Get the number of WMI object instances.
>> + *
>> + * Returns: acpi_status signaling success or error.
>> + */
>> +acpi_status wmi_instance_count(const char *guid_string, u8 *instance_count)
>> +{
>> +	struct wmi_block *wblock;
>> +	acpi_status status;
>> +
>> +	status = find_guid(guid_string, &wblock);
>> +	if (ACPI_FAILURE(status))
>> +		return status;
>> +
>> +	*instance_count = wmidev_instance_count(&wblock->dev);
>> +
>> +	return AE_OK;
>> +}
>> +EXPORT_SYMBOL_GPL(wmi_instance_count);
> I would prefer this to have a normal kernel function prototype
> which returns -errno rather then returning an acpi_status. E.g. :
>
> /**
>   * wmi_instance_count - Get number of WMI object instances
>   * @guid_string: 36 char string of the form fa50ff2b-f2e8-45de-83fa-65417f2f49ba
>   *
>   * Get the number of WMI object instances.
>   *
>   * Returns: The number of WMI object instances, 0 if the GUID is not found.
>   */
> int wmi_instance_count(const char *guid_string)
> {
> 	struct wmi_block *wblock;
> 	acpi_status status;
>
> 	status = find_guid(guid_string, &wblock);
> 	if (ACPI_FAILURE(status))
> 		return 0;
>
> 	return wmidev_instance_count(&wblock->dev);
> }
> EXPORT_SYMBOL_GPL(wmi_instance_count);
>
> This will also allow this to completely replace
> the get_instance_count() function in dell-wmi-sysman.
>
> Note I have just gone with always returning 0 here
> on error. I guess you could look at the status and
> return 0 for not-found and -errno for other errors
> but I don't think any callers will care for the difference,
> so just always returning 0 seems easier for callers to
> deal with.
>
> As always this is just a suggestion, let me know if
> you think this is a bad idea.
>
> Regards,
>
> Hans
>
I like this idea. Returning a negative errno on error would allow drivers to
distinguish between "WMI object not found" and "zero instances found", which
might be useful for some drivers.

Maybe a function for converting ACPI errors to POSIX errors already exists,
otherwise i will just write one myself.

Armin Wolf

>
>> +
>> +/**
>> + * wmidev_instance_count - Get number of WMI object instances
>> + * @wdev: A wmi bus device from a driver
>> + *
>> + * Get the number of WMI object instances.
>> + *
>> + * Returns: Number of WMI object instances.
>> + */
>> +u8 wmidev_instance_count(struct wmi_device *wdev)
>> +{
>> +	struct wmi_block *wblock = container_of(wdev, struct wmi_block, dev);
>> +
>> +	return wblock->gblock.instance_count;
>> +}
>> +EXPORT_SYMBOL_GPL(wmidev_instance_count);
>> +
>>   /**
>>    * wmi_evaluate_method - Evaluate a WMI method (deprecated)
>>    * @guid_string: 36 char string of the form fa50ff2b-f2e8-45de-83fa-65417f2f49ba
>> diff --git a/include/linux/acpi.h b/include/linux/acpi.h
>> index efff750f326d..ab2a4b23e7a3 100644
>> --- a/include/linux/acpi.h
>> +++ b/include/linux/acpi.h
>> @@ -412,6 +412,8 @@ extern bool acpi_is_pnp_device(struct acpi_device *);
>>
>>   typedef void (*wmi_notify_handler) (u32 value, void *context);
>>
>> +acpi_status wmi_instance_count(const char *guid, u8 *instance_count);
>> +
>>   extern acpi_status wmi_evaluate_method(const char *guid, u8 instance,
>>   					u32 method_id,
>>   					const struct acpi_buffer *in,
>> diff --git a/include/linux/wmi.h b/include/linux/wmi.h
>> index c1a3bd4e4838..763bd382cf2d 100644
>> --- a/include/linux/wmi.h
>> +++ b/include/linux/wmi.h
>> @@ -35,6 +35,8 @@ extern acpi_status wmidev_evaluate_method(struct wmi_device *wdev,
>>   extern union acpi_object *wmidev_block_query(struct wmi_device *wdev,
>>   					     u8 instance);
>>
>> +u8 wmidev_instance_count(struct wmi_device *wdev);
>> +
>>   extern int set_required_buffer_size(struct wmi_device *wdev, u64 length);
>>
>>   /**
>> --
>> 2.30.2
>>

Re: [RFC v2 1/2] platform/x86: wmi: Allow retrieving the number of WMI object instances

[Hans de Goede]

Hi,

On 4/27/23 18:26, Armin Wolf wrote:
> Am 27.04.23 um 11:43 schrieb Hans de Goede:
> 
>> Hi Armin,
>>
>> Thank you for your work on this.
>>
>> On 4/26/23 23:28, Armin Wolf wrote:
>>> Currently, the WMI driver core knows how many instances of a given
>>> WMI object exist, but WMI drivers cannot access this information.
>>> At the same time, some current and upcoming WMI drivers want to
>>> have access to this information. Add wmi_instance_count() and
>>> wmidev_instance_count() to allow WMI drivers to get the number of
>>> WMI object instances.
>>>
>>> Signed-off-by: Armin Wolf <W_Armin@gmx.de>
>>> ---
>>>   drivers/platform/x86/wmi.c | 40 ++++++++++++++++++++++++++++++++++++++
>>>   include/linux/acpi.h       |  2 ++
>>>   include/linux/wmi.h        |  2 ++
>>>   3 files changed, 44 insertions(+)
>>>
>>> diff --git a/drivers/platform/x86/wmi.c b/drivers/platform/x86/wmi.c
>>> index c226dd4163a1..7c1a904dec5f 100644
>>> --- a/drivers/platform/x86/wmi.c
>>> +++ b/drivers/platform/x86/wmi.c
>>> @@ -263,6 +263,46 @@ int set_required_buffer_size(struct wmi_device *wdev, u64 length)
>>>   }
>>>   EXPORT_SYMBOL_GPL(set_required_buffer_size);
>>>
>>> +/**
>>> + * wmi_instance_count - Get number of WMI object instances
>>> + * @guid_string: 36 char string of the form fa50ff2b-f2e8-45de-83fa-65417f2f49ba
>>> + * @instance_count: variable to hold the instance count
>>> + *
>>> + * Get the number of WMI object instances.
>>> + *
>>> + * Returns: acpi_status signaling success or error.
>>> + */
>>> +acpi_status wmi_instance_count(const char *guid_string, u8 *instance_count)
>>> +{
>>> +    struct wmi_block *wblock;
>>> +    acpi_status status;
>>> +
>>> +    status = find_guid(guid_string, &wblock);
>>> +    if (ACPI_FAILURE(status))
>>> +        return status;
>>> +
>>> +    *instance_count = wmidev_instance_count(&wblock->dev);
>>> +
>>> +    return AE_OK;
>>> +}
>>> +EXPORT_SYMBOL_GPL(wmi_instance_count);
>> I would prefer this to have a normal kernel function prototype
>> which returns -errno rather then returning an acpi_status. E.g. :
>>
>> /**
>>   * wmi_instance_count - Get number of WMI object instances
>>   * @guid_string: 36 char string of the form fa50ff2b-f2e8-45de-83fa-65417f2f49ba
>>   *
>>   * Get the number of WMI object instances.
>>   *
>>   * Returns: The number of WMI object instances, 0 if the GUID is not found.
>>   */
>> int wmi_instance_count(const char *guid_string)
>> {
>>     struct wmi_block *wblock;
>>     acpi_status status;
>>
>>     status = find_guid(guid_string, &wblock);
>>     if (ACPI_FAILURE(status))
>>         return 0;
>>
>>     return wmidev_instance_count(&wblock->dev);
>> }
>> EXPORT_SYMBOL_GPL(wmi_instance_count);
>>
>> This will also allow this to completely replace
>> the get_instance_count() function in dell-wmi-sysman.
>>
>> Note I have just gone with always returning 0 here
>> on error. I guess you could look at the status and
>> return 0 for not-found and -errno for other errors
>> but I don't think any callers will care for the difference,
>> so just always returning 0 seems easier for callers to
>> deal with.
>>
>> As always this is just a suggestion, let me know if
>> you think this is a bad idea.
>>
>> Regards,
>>
>> Hans
>>
> I like this idea. Returning a negative errno on error would allow drivers to
> distinguish between "WMI object not found" and "zero instances found", which
> might be useful for some drivers.
> 
> Maybe a function for converting ACPI errors to POSIX errors already exists,
> otherwise i will just write one myself.

Ok, that sounds good to me. I'm looking forward to a non RFC submission
of these changes.

Regards,

Hans

[RFC v2 2/2] platform/x86: dell-sysman: Improve instance detection

[Armin Wolf]

The WMI driver core already knows how many WMI object instances
are available, use this information instead of probing the WMI object
manually.

Compile-tested only.

Signed-off-by: Armin Wolf <W_Armin@gmx.de>
---
 .../platform/x86/dell/dell-wmi-sysman/sysman.c    | 15 +++++++--------
 1 file changed, 7 insertions(+), 8 deletions(-)

diff --git a/drivers/platform/x86/dell/dell-wmi-sysman/sysman.c b/drivers/platform/x86/dell/dell-wmi-sysman/sysman.c
index 0285b47d99d1..526d60b510bb 100644
--- a/drivers/platform/x86/dell/dell-wmi-sysman/sysman.c
+++ b/drivers/platform/x86/dell/dell-wmi-sysman/sysman.c
@@ -7,6 +7,7 @@

 #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt

+#include <linux/acpi.h>
 #include <linux/fs.h>
 #include <linux/dmi.h>
 #include <linux/module.h>
@@ -303,16 +304,14 @@ union acpi_object *get_wmiobj_pointer(int instance_id, const char *guid_string)
  */
 int get_instance_count(const char *guid_string)
 {
-	union acpi_object *wmi_obj = NULL;
-	int i = 0;
+	acpi_status status;
+	u8 instance_count;

-	do {
-		kfree(wmi_obj);
-		wmi_obj = get_wmiobj_pointer(i, guid_string);
-		i++;
-	} while (wmi_obj);
+	status = wmi_instance_count(guid_string, &instance_count);
+	if (ACPI_FAILURE(status))
+		return 0;

-	return (i-1);
+	return instance_count;
 }

 /**
--
2.30.2