Am 17.03.2016 18:04, schrieb Greg KH:
> On Thu, Mar 17, 2016 at 05:41:13PM +0100, walter harms wrote:
>>
>>
>> Am 29.02.2016 17:10, schrieb Greg KH:
>>> On Mon, Feb 29, 2016 at 02:47:45PM +0100, walter harms wrote:
>>>> Hello List,
>>>>
>>>> is someone aware of a problem with DTR in that driver ?
>>>>
>>>> I use: Linux version 3.4.53-MMI-SW.401_0.1-43
>>>
>>> That's a very old and obsolete kernel version. Please ask for support
>>> from the vendor that is forcing you to live with that kernel version,
>>> that is what you are paying for.
>>>
>>> We can help you here if you try the latest kernel release, can you try
>>> 4.4?
>>>
>>
>> Hi,
>> i would like to give same feedback.
>>
>> Using a newer version of linux caused unexpected problems unrelated to the interface
>> but we found a possible trouble spot.
>
> What problems did you have?
>
>> the chip cp2105 has a One Time Programmable register to define DTR etc.
>> we were not aware of that.
>>
>> I would suggest to add an ioctl to get the mode. (I did not see one).
>> so someone has a change to find the problem next time.
>
> The mode of what? The tty line? Something else?
>
> Please be more specific.
>
mea culpa, as usual i forgot that not everyone is working on the same problem ....
I will write that in the slight hope that the next one will not search for not
existing hardware bugs ...
Desc: the cp2105 is a USB->serial converter with GPIO option (!)
symptom: Devices attached to the serial port that require all Signals do not work
but hardware (wiring etc) is ok.
Tests showed that toggling DTR signal did not work for no obvious reason.
(RTS did work).
reason: (tested on cp2105 prototype board)
the chip cp2105 has a One Time Programmable register to define modem/gpio mode.
The configuration is normally done by the producer of the board.
If that register is not set (in our case) the DTR is not working.
Note the documentation says: "Up to 1024 Bytes of EEPROM or OTP ROM" but most times
the documentation refers to "EEPROM".
suggestion:
there should be an ioctl() related to the configuration in the eeprom.
Writing sounds a bit dangerous for OTP but otherwise seems reasonable.
I guess other chips are working like this also, so this could work as prototype
for that.
note: I have tested that with linux 3.4 (s.a) using a newer version of linux caused
unexpected problems (With our Prototype) unrelated to the interface. I looked
into the current driver and found no reason to believe that it would behave different.
We found this python script helpful, what we could use to test the related register
in the cp2105-prototype board. https://github.com/karlp/mixed-cygnals/blob/master/cp2105.py
/* things i found after i understood the problem ...
What may be helpful (untested, http://cp210x-program.sourceforge.net )
discussion/patch 02/2016 https://patchwork.ozlabs.org/patch/569553/
hope that helps,
re,
wh
--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
