Re: cdc_acm and sam_ba support the same device

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Am 16.03.2011 22:05, schrieb Greg KH:
> On Wed, Mar 16, 2011 at 09:49:26PM +0100, Sven KÃhler wrote:
>> Am 16.03.2011 20:34, schrieb Greg KH:
>>> On Wed, Mar 16, 2011 at 07:28:25PM +0100, Johan Hovold wrote:
>>>> On Wed, Mar 16, 2011 at 08:23:11AM -0700, Greg KH wrote:
>>>>> On Wed, Mar 16, 2011 at 03:48:53PM +0100, Sven KÃhler wrote:
>>>>>> Hi,
>>>>>>
>>>>>> I'm not sure, whether you consider this is an issue. However, the device
>>>>>> I have (Lego NXT in SAM-BA mode, lsusb shows it as 03eb:6124 Atmel Corp.
>>>>>> at91sam SAMBA bootloader) is supported by both cdc_acm and sam_ba kernel
>>>>>> modules, which leads to unpredictable behaviour.
>>>>>> (Tested 2.6.38 - same issue in 2.6.37)
>>>>>
>>>>> I think Johan is the best one to answer this, as he submitted the sam-ba
>>>>> driver.
>>>>>
>>>>> Johan, any thoughts on the below info?
>>>>>
>>>>>
>>>>>>
>>>>>> If I plug the device in, the following happens if cdc_acm is already loaded:
>>>>>>> usb 4-1: new full speed USB device using uhci_hcd and address 3
>>>>>>> cdc_acm 4-1:1.0: This device cannot do calls on its own. It is not a modem.
>>>>>>> cdc_acm 4-1:1.0: ttyACM0: USB ACM device
>>>>>>> usbcore: registered new interface driver cdc_acm
>>>>>>> cdc_acm: v0.26:USB Abstract Control Model driver for USB modems and ISDN adapters
>>>>>>> USB Serial support registered for sam-ba
>>>>>>> usbcore: registered new interface driver sam-ba
>>>>>>> sam_ba: v1.0: Atmel SAM Boot Assistant (SAM-BA) driver
>>>>>>
>>>>>> After unloading both cdc_acm and sam_ba, the following happends:
>>>>>>> usb 4-1: new full speed USB device using uhci_hcd and address 4
>>>>>>> USB Serial support registered for sam-ba
>>>>>>> sam-ba 4-1:1.1: sam-ba converter detected
>>>>>>> usb 4-1: sam-ba converter now attached to ttyUSB0
>>>>>>> usbcore: registered new interface driver sam-ba
>>>>>>> sam_ba: v1.0: Atmel SAM Boot Assistant (SAM-BA) driver
>>>>>>> cdc_acm 4-1:1.0: This device cannot do calls on its own. It is not a modem.
>>>>>>> cdc_acm: probe of 4-1:1.0 failed with error -16
>>>>>>> usbcore: registered new interface driver cdc_acm
>>>>>>> cdc_acm: v0.26:USB Abstract Control Model driver for USB modems and ISDN adapters
>>>>>>
>>>>>> As you can see, the device is recognized by cdc_acm as well as sam_ba.
>>>>>> Depending on which module is loaded, the device is either claimed by
>>>>>> cdc_acm or sam_ba. The device name may be either ttyACM* or ttyUSB*
>>>>
>>>> Tradionally the generic usb-serial driver was used to access the SAM-BA
>>>> boot application; this is documented in the Atmel tools, which also
>>>> expect the device to show up as a ttyUSBx node. Due to changes in the
>>>> generic driver the tools stopped working from 2.6.32 and a dedicated
>>>> driver was later written (sam-ba).
>>>>
>>>> A couple of weeks after the sam-ba driver was submitted, however,
>>>> 5b239f0aebd4dd6f85b13decf5e18e86e35d57f0 (USB: cdc-acm: Add pseudo modem
>>>> without AT command capabilities) was submitted, which causes cdc_acm to
>>>> bind to these devices as well.
>>>>
>>>>>> I have no clue, whether there are Atmel SAM-BA implementations that do
>>>>>> not support the cdc profile.
>>>>>>
>>>>>> I'm not sure, whether the device should be blacklisted in cdc_acm in
>>>>>> favor of sam_ba. However, I have the feeling that if sam_ba is actually
>>>>>> installed, sam_ba should claim the device even if cdc_acm is already loaded.
>>>>>>
>>>>>> Last time I tested, the device was working fine with cdc_acm, though.
>>>>
>>>> I just connected a at91sam9260-ek board and can confirm that the samba
>>>> tools do seem to work with the cdc-acm driver, although it does require
>>>> a ttyUSBx symlink for the device to be recognised by the user-space tools
>>>> (latest v2.10).
>>>>
>>>> I'm not sure about how to proceed with this. What do you say, Greg?
>>>
>>> If the sam-ba driver is not needed, we can just remove it.
>>>
>>> The symlink issue should be fixable in the userspace tools to just get
>>> it to look for the different port (it should never have hard-coded
>>> ttyUSB0 anyway, what if it was on ttyUSB3?).
>>
>> It's an UI issue, basically. They show a list of devices and they only
>> show the ones matching the pattern ttyUSB*. In addition, you can't enter
>> a custom device name. That's why you have to create a symlink, for
>> example from ttyACM3 to ttyUSB7.
> 
> Wow, what forsight... :)
> 
> Have a pointer to the source code for this so I could fix it up?

The source is not available, I believe.
However, here's the download link for the software:
http://www.atmel.com/dyn/products/tools_card.asp?tool_id=3883

Actually, a few months ago (end of october 2010), I wrote an email to
Atmel, explaining the situation. They should be aware of the
ttyACM-issue. However, they have not released an updated version so far.
The only response I got from Atmel was "I will forward you comment". I'm
of course hoping, that it has been forwarded to the right people :-)


Regards,
  Sven

--
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


[Index of Archives]     [Linux Media]     [Linux Input]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]     [Old Linux USB Devel Archive]

  Powered by Linux