Re: [PATCH 3/3] i2c: add ACPI support for i2c-piix4

From: Andy Shevchenko
Date: Tue Jan 30 2018 - 13:22:42 EST


On Tue, Jan 30, 2018 at 3:43 AM, Andrew Cooks <andrew.cooks@xxxxxxxxxxxx> wrote:
> On 29/01/18 23:30, Andy Shevchenko wrote:
>> On Mon, Jan 29, 2018 at 5:54 AM, Andrew Cooks <andrew.cooks@xxxxxxxxxxxx> wrote:
>>> This enables the i2c-piix4 SMBus controller driver to enumerate I2C
>>> slave devices using ACPI. It builds on the related I2C mux device work
>>> in commit 8eb5c87a92c0 ("i2c: add ACPI support for I2C mux ports")
>>>
>>> In the i2c-piix4 driver the adapters are enumerated as:
>>> Main SMBus adapter Port 0, Port 2, ..., aux port (i.e., ASF adapter)
>>>
>>> However, in the AMD BKDG documentation[1], the implied order of ports is:
>>> Main SMBus adapter Port 0, ASF adapter, Port 2, Port 3, ...
>>>
>>> This ordering difference is unfortunate, and we assume that ACPI
>>> developers will use the Linux ordering.
>>>
>>> [1] 52740 BIOS and Kernel Developer's Guide (BKDG) for AMD Family 16h
>>> Models 30h-3Fh Processors
>>
>>> + if (has_acpi_companion(&dev->dev)) {
>>> + acpi_preset_companion(&adap->dev,
>>> + ACPI_COMPANION(&dev->dev),
>>> + piix4_adapter_count++);
>>> + }
>>
>> Wouldn't be enough just following:
>>
>> ACPI_COMPANION_SET(&adap->dev, ACPI_COMPANION(&dev->dev));
>
> In the ACPI DSDT, I've modeled the topology as follows:
>
> Device(SBUS) {
> Name(_ADR, 0x00140000)
> Name(_HID, "SMB0001") // SMBus CMI spec uses (invalid hid) "SMBUS01"
>
> Device(CH00) {
> Name(_ADR, 0) // main SMBus adapter, Port 0
> Name (_STR, Unicode ("Main SMBus adapter, Port 0"))
> }
>
> Device(CH01) {
> Name(_ADR, 1) // main SMBus Port 2 (Port 1 label is undefined)
> Name (_STR, Unicode ("Main SMBus adapter, Port 2"))
> }
>
> Device (CH02) {
> Name(_ADR, 2) // ASF adapter (what Linux calls 'aux port')
> Name (_STR, Unicode ("AUX / ASF SMBus adapter"))
> }
> }
>
> That is why each piix4 adapter matches a CH0? device that is descendant from the SBUS device in ACPI.
>
> It is similar to the topology shown in commit 8eb5c87a ("i2c: add ACPI support for I2C mux ports")

Ah, I see now. Your initial approach seems fine.

Perhaps even new helper can be made for users (currently i2c mux,
sdio, libata) in the future.

--
With Best Regards,
Andy Shevchenko