Re: [PATCH v2 1/3] usb: gadget: ccid: add support for USB CCID Gadget Device

[Andrzej Pietrasiewicz]

W dniu 28.05.2018 oÂ11:32, Marcus Folkesson pisze:
> Hi Andrzej,
> 
> Thank you for reviewing.
> 
> On Mon, May 28, 2018 at 11:12:27AM +0200, Andrzej Pietrasiewicz wrote:
>> W dniu 28.05.2018 oÂ10:38, Marcus Folkesson pisze:
>>> Hi Andrzej,
>>>
>>> On Mon, May 28, 2018 at 09:04:51AM +0200, Andrzej Pietrasiewicz wrote:
>>>> Mi Marcus,
>>>>
>>>> W dniu 26.05.2018 oÂ23:19, Marcus Folkesson pisze:
>>>>> Chip Card Interface Device (CCID) protocol is a USB protocol that
>>>>> allows a smartcard device to be connected to a computer via a card
>>>>> reader using a standard USB interface, without the need for each manufacturer
>>>>> of smartcards to provide its own reader or protocol.
>>>>>
>>>>> This gadget driver makes Linux show up as a CCID device to the host and let a
>>>>> userspace daemon act as the smartcard.
>>>>>
>>>>> This is useful when the Linux gadget itself should act as a cryptographic
>>>>> device or forward APDUs to an embedded smartcard device.
>>>>>
>>>>> Signed-off-by: Marcus Folkesson <marcus.folkesson@xxxxxxxxx>
>>>>> ---
>>>>
>>>>>     
>>>>> +config USB_CONFIGFS_CCID
>>>>> +	bool "Chip Card Interface Device (CCID)"
>>>>> +	depends on USB_CONFIGFS
>>>>> +	select USB_F_CCID
>>>>> +	help
>>>>> +	  The CCID function driver provides generic emulation of a
>>>>> +	  Chip Card Interface Device (CCID).
>>>>> +
>>>>> +	  You will need a user space server talking to /dev/ccidg*,
>>>>> +	  since the kernel itself does not implement CCID/TPDU/APDU
>>>>> +	  protocol.
>>>>
>>>> Your function needs a userspace daemon to work.
>>>> It seems you want to use FunctionFS for such a purpose
>>>> instead of creating a new function.
>>>>
>>>> Andrzej
>>>
>>>>> +	  since the kernel itself does not implement CCID/TPDU/APDU
>>> Oops, the driver does handle CCID.
>>
>> Which parts of code do this handling?
> 
> My bad, I was thinking about the USB descriptors and endpoints setup.
> That is of cause not part of the CCID protocol.
> 
>>
>> Is there any kind of state machine usual for protocols?
>> If the protocol is stateless then isn't it just a data format then?
> 
> The protocol is stateless.
> 
>>
>> Which part of this handling must be done in kernel and why?
>>
>> Does the said handling do anything other than forwarding the
>> traffic between USB and a character device?
> 
> No, it forward the CCID messages to the character device to be handled
> by the application.
> 
>>

My opinion is: this wants to be done with FunctionFS.

Andrzej