Re: [v3 PATCH 3/5] phy: Add USB Type-C PHY driver for rk3399

[Guenter Roeck]

On Sun, Jun 26, 2016 at 7:19 PM, Chris Zhong <zyw@xxxxxxxxxxxxxx> wrote:
> Hi Heiko
>
>
> On 06/25/2016 03:39 AM, Heiko Stuebner wrote:
>>
>> Am Donnerstag, 23. Juni 2016, 18:27:52 schrieb Kishon Vijay Abraham I:
>>>
>>> Hi,
>>>
>>> On Thursday 23 June 2016 06:21 PM, Chris Zhong wrote:
>>>>
>>>> Add a PHY provider driver for the rk3399 SoC Type-c PHY. The USB
>>>> Type-C PHY is designed to support the USB3 and DP applications. The
>>>> PHY basically has two main components: USB3 and DisplyPort. USB3
>>>> operates in SuperSpeed mode and the DP can operate at RBR, HBR and
>>>> HBR2 data rates.
>>>>
>>>> Signed-off-by: Chris Zhong <zyw@xxxxxxxxxxxxxx>
>>>> Signed-off-by: Kever Yang <kever.yang@xxxxxxxxxxxxxx>
>>>>
>>>> ---
>>>>
>>>> Changes in v3:
>>>> - remove the phy framework(Kishon Vijay Abraham I)
>>>> - add parentheses around the macro
>>>> - use a single space between type and name
>>>> - add spaces after opening and before closing braces.
>>>> - use u16 for register value
>>>> - remove type-c phy header file
>>>> - CodingStyle optimization
>>>> - use some cable extcon to get type-c port information
>>>> - add a extcon to notify Display Port
>>>>
>>>> Changes in v2:
>>>> - select RESET_CONTROLLER
>>>> - alphabetic order
>>>> - modify some spelling mistakes
>>>> - make mode cleaner
>>>> - use bool for enable/disable
>>>> - check all of the return value
>>>> - return a better err number
>>>> - use more readx_poll_timeout()
>>>> - clk_disable_unprepare(tcphy->clk_ref);
>>>> - remove unuse functions, rockchip_typec_phy_power_on/off
>>>> - remove unnecessary typecast from void *
>>>> - use dts node to distinguish between phys.
>>>>
>>>> Changes in v1:
>>>> - update the licence note
>>>> - init core clock to 50MHz
>>>> - use extcon API
>>>> - remove unused global
>>>> - add some comments for magic num
>>>> - change usleep_range(1000, 2000) tousleep_range(1000, 1050)
>>>> - remove __func__ from dev_err
>>>> - return err number when get clk failed
>>>> - remove ADDR_ADJ define
>>>> - use devm_clk_get(&pdev->dev, "tcpdcore")
>>>>
>>>>   drivers/phy/Kconfig              |    8 +
>>>>   drivers/phy/Makefile             |    1 +
>>>>   drivers/phy/phy-rockchip-typec.c | 1027
>>>>   ++++++++++++++++++++++++++++++++++++++ 3 files changed, 1036
>>>>   insertions(+)
>>>>   create mode 100644 drivers/phy/phy-rockchip-typec.c
>>>>
>>>> diff --git a/drivers/phy/Kconfig b/drivers/phy/Kconfig
>>>> index 26566db..ec87b3a 100644
>>>> --- a/drivers/phy/Kconfig
>>>> +++ b/drivers/phy/Kconfig
>>>> @@ -351,6 +351,14 @@ config PHY_ROCKCHIP_DP
>>>>
>>>>         help
>>>>
>>>>           Enable this to support the Rockchip Display Port PHY.
>>>>
>>>> +config PHY_ROCKCHIP_TYPEC
>>>> +       tristate "Rockchip TYPEC PHY Driver"
>>>> +       depends on ARCH_ROCKCHIP && OF
>>>> +       select GENERIC_PHY
>>>
>>> Why? None of the generic PHY API's are used here. Why do you want select
>>> generic PHY?
>>
>> I'm actually wondering, why there are no phy ops to start and stop the py.
>> Right now the device seems to be on and handling all the extcon notifies
>> all
>> the time even if no-one is using the phy.
>>
>> There are two users of this phy, the dp-controller as well as some usb
>> component. The phy framework is nicely refcounted, so can handle any
>> number
>> of phy users and somehow I'd expect the phy to do nothing (and try to not
>> consume power) if neither the dp-controller nor the usb-part is actually
>> running.
>>
>> It may very well be my ignorance, but Chris could you explain, if there is
>> a
>> reason for this?
>>
>>
>> Thanks
>> Heiko
>>
>>
> It is good idea, The USB3 and DP controller detect the extcon cable state:
> USB/USB HOST/DP. If a Type-C device plugged, call phy power on, the phy
> driver get all the state of extcon: dfp, ufp, dp, flip, pin assignment, and
> then finish the phy init. So the phy driver need not register extcon
> notification.
> But this mechanism allows phy driver only focus plug/unplug event, if some
> other things happen, such as data role change, the phy will not be notified.
> I'm not sure if this situation exists.

Sorry, I think I am lost (again).

All roles can change on the fly. Role changes can be triggered from
user space, or by the remote partner. If we restrict such role
changes, we would have to reject all locally triggered role change
requests, as well as all role change requests from the remote, after
the initial connection was established. I don't really see the point
of doing that, though. Wasn't this what the notifications were all
about ?

Guenter