Re: [PATCH 5/5] ACPI: Add support for platform bus type

On Thu, Nov 1, 2012 at 7:35 AM, Rafael J. Wysocki [off-list ref] wrote:

quoted

quoted

@@ -1544,8 +1553,13 @@ static acpi_status acpi_bus_check_add(ac
         */
        device = NULL;
        acpi_bus_get_device(handle, &device);
-       if (ops->acpi_op_add && !device)
+       if (ops->acpi_op_add && !device) {
                acpi_add_single_object(&device, handle, type, sta, ops);
+               /* Is the device a known good platform device? */
+               if (device
+                   && !acpi_match_device_ids(device, acpi_platform_device_ids))
+                       acpi_create_platform_device(device);
+       }

That is ugly! any reason for not using acpi_driver for them.

Yes, a couple of reasons.  First off, there are existing platform drivers for
these things already and there's no reason for creating a "glue" layer between
those drivers and struct acpi_device objects.  Second, we're going to get rid
of acpi_driver things entirely going forward (the existing ones will be replaced
by platform drivers or included into the ACPI core).

that should be glue between acpi_device and platform_device.

how are you going to handle removing path ? aka when acpi_device get
trimed, how those platform get deleted?

quoted

there is not reason to treat those platform_device different from pci
root device and other pnp device.

Well, I agree. :-)

So those things you're talking about we'll be platform devices too in the
future.

quoted

something like attached patch. .add of acpi_driver ops could call
acpi_create_platform_device.

Been there, decided to do it differently this time.

So you are going to replace acpi_device/acpi_driver with
platform_device/platform_driver ?

Did you ever try to start to the work to see if it is doable? aka do
you have draft version to do that?

Yinghai