Re: [PATCH v2 1/2] platform: Fix device check for surfacepro3_button

From: Maximilian Luz
Date: Mon Jul 01 2019 - 21:33:26 EST

On 7/2/19 3:25 AM, Maximilian Luz wrote:
On 7/2/19 3:14 AM, Yu Chen wrote:
On Tue, Jul 02, 2019 at 02:37:39AM +0200, Maximilian Luz wrote:
+ * Surface Pro 4 and Surface Book 2 / Surface Pro 2017 use the same device
+ * ID (MSHW0040) for the power/volume buttons. Make sure this is the right
+ * device by checking for the _DSM method and OEM Platform Revision.
+ */
+static int surface_button_check_MSHW0040(struct acpi_device *dev)
+ acpi_handle handle = dev->handle;
+ union acpi_object *result;
+ u64 oem_platform_rev = 0;
+ // get OEM platform revision
+ result = acpi_evaluate_dsm_typed(handle, &MSHW0040_DSM_UUID,
Does it mean, only 5th, 6th and newer platforms have OEM platform revision?
3rd/4th will get NULL result? Or the opposite?

Correct, from my testing (with limited sample size) and AML code: 5th
and 6th generation devices have a non-zero OEM platform revision,
whereas 3rd and 4th gen. devices do not have any (i.e. result will be

+ if (result) {
+ oem_platform_rev = result->integer.value;
+ ACPI_FREE(result);
+ }
+ dev_dbg(&dev->dev, "OEM Platform Revision %llu\n", oem_platform_rev);
+ return oem_platform_rev == 0 ? 0 : -ENODEV;
if 3rd/4th do not have this oem rev information while 5th/newer have,
why the latter returns NODEV(it actually has this info)?

Since we always expect a non-zero platform revision (for 5th/6th gen.),
we can initialize it to zero and use that as "unknown"/"not available".
So if it can not be determined, we return NODEV.

Sorry, small mistake here: If it can be determined (i.e. is 5th or 6th
gen.) then we return NODEV. Not the other way around.

Also to clarify on your last question:

On 7/2/19 3:14 AM, Yu Chen wrote:
static int surface_button_add(struct acpi_device *device)
struct surface_button *button;
@@ -154,6 +188,10 @@ static int surface_button_add(struct acpi_device *device)
return -ENODEV;
+ error = surface_button_check_MSHW0040(device);
+ if (error)
+ return error;
ditto, 3rd/4th get error=0?

You are correct.