[PATCH v5 0/6] FUSB302 USB-C controller support

[Soeren Moch]

Hi Sebastian,

On 04.09.24 02:03, Sebastian Reichel wrote:
> Hi Soeren,
>
> On Wed, Sep 04, 2024 at 12:26:43AM GMT, Soeren Moch wrote:
>>>    * non PD capable (reports 5V 0A)
>> With a dumb power supply (actually 12V w/o USB-PD) I see a error message
>> (not really sure if that makes sense, but no strong opinion from my side)
>> and 5V/1.5A reported with 'tcpm info', see below.
>> Can we expect 1.5A as fallback supply w/o USB-PD negotiation? 5V OK, but
>> I'm not sure if USB2/USB3 type A to type C cables are allowed according
>> to the USB spec, with lower possible currents.
>>
>> I'm not sure what the 'right' thing is to report in this situation
>> (also cable orientation makes no sense).
>> Just want to mention that I see something different here.
> USB-C sources, which are not PD capable can use pull-up resistors to
> change source advertisement as follows:
>
>      | advert.   | pull-up to 5V | pull-up to 3.3V |
>      | --------- | ------------- | --------------- |
>      | default   | 56K           | 36K             |
>      | 1.5A @ 5V | 22K           | 12K             |
>      | 3.0A @ 5V | 10K           | 4.7K            |
>
> default means 500mA for USB2 or 900mA for USB3, which is reported as
> 0 in this series.
Ah, OK. Thanks for your explanation. This makes sense now.
> Your setup seems to have pull resistors advertising
> the 1.5A profile. If your setup includes an USB-A to USB-C cable the
> pull resistors are in the cable, since the pull resistors are on the
> CC pins, which do not exist for USB-A.
This special setup is non-standard anyway, with 12V only save to use
for rock-5b.
The "3A-Cable" is advertising 1.5A, OK. Similar to my "35W USB-PD
power supply" advertising 12V/2.5A, hooray.
> The same pull-up resistors should give away the cable orientation :)
Should have tried this.
>
>>>    * RavPower 90W
>>>    * UGREEN 100W
>>>    * Anker 45W
>>>    * RavPower PB
>>>
>>> Changes since PATCHv4:
>>>    * Rebase to latest master (v2024.10-rc4)
>>>    * Fix regression introduced in PATCHv3: The exit code for fusb302_set_vconn()
>>>      when vconn stays the same should be 0 instead of undefined
>>>    * Fix regression introduced in PATCHv3: tcpm_set_polarity() needs to update
>>>      port->polarity when there is no driver specific set_polarity function
>> I can confirm that these fixes indeed fix my reported regression from v2.
>> Booting to Linux works perfectly fine when powering from simple USB-PD
>> supplies (12V, 20V), as well as with power from my notebook computer.
>>
>> So for the v5 series:
>> Tested-by: Soeren Moch <smoch at web.de>
>>
>> I also tested powering from a UGREEN USB-C dock (see boot log below).
>> This also worked well. Only the data role was not switched correctly
>> to host as required for the dock. At least we got this 'not supported'
>> message probably related to that. Maybe that can be fixed later in
>> follow-up patches.
> Yes, this series focuses on getting sink power role working as this
> is needed to enable USB-C support in Linux (or any other OS). Handling
> for data role swapping can happen on top of this. Linux should be able
> to swap the data role if it had been configured the wrong way around in
> the bootloader without triggering a hard reset.
In a quick test this had not worked. Maybe something wrong with dts?
Needs more investigation.
> So this should only be
> needed to have the dock's USB ports working in U-Boot itself.
Would be nice to have. Not that important for me.
>
> Also acting as USB-PD Source is not fully implemented and needs to be
> added on top of this once it is required by anyone.
Good to know. But not testable on rock-5b.

Regards,
Soeren
>
> -- Sebastian