[BUG] usb-serial-simple: device conflict with Owon VDS1022I

[BUG] usb-serial-simple: device conflict with Owon VDS1022I

[Amit]

Hi,

     PATCH 6.1 082/167 adds serial support for OWON HDS200 through the 
simple usb serial driver (). This causes issues with Owon VDS1022. The 
VID and PID is the same for both devices so the the Linux kernel loads 
the serial driver, preventing the associated software from working.

     My current work around for this is to blacklist usb_serial_simple. 
Given that this driver change was recent and still under testing, could 
we find a solution for this that doesn't involve such hacks at my end.

     I  was able to root cause this issue only because the USB device 
and  software was working correctly before. A new user would just assume 
that the device is not Linux compatible.

     I've filed a bugzilla report here : 
https://bugzilla.kernel.org/show_bug.cgi?id=220193

     This is the software : https://github.com/florentbr/OWON-VDS1022

regards,

Amit

Re: [BUG] usb-serial-simple: device conflict with Owon VDS1022I

[Greg KH]

On Fri, Jun 06, 2025 at 05:00:29PM +0530, Amit wrote:
> Hi,
> 
>     PATCH 6.1 082/167 adds serial support for OWON HDS200 through the simple
> usb serial driver (). This causes issues with Owon VDS1022. The VID and PID
> is the same for both devices so the the Linux kernel loads the serial
> driver, preventing the associated software from working.

As vendors are not supposed to use the same device id for different
devices, that's against the USB.org rules, so that's on them, not us :)

>     My current work around for this is to blacklist usb_serial_simple. Given
> that this driver change was recent and still under testing, could we find a
> solution for this that doesn't involve such hacks at my end.
> 
>     I  was able to root cause this issue only because the USB device and 
> software was working correctly before. A new user would just assume that the
> device is not Linux compatible.
> 
>     I've filed a bugzilla report here :
> https://bugzilla.kernel.org/show_bug.cgi?id=220193
> 
>     This is the software : https://github.com/florentbr/OWON-VDS1022

As that's just using libusb, it can disconnect the device from the
usb-serial driver if it needs to from the program itself.  Why not
suggest this change to the authors of the software, OR just change it to
use the serial connection that the kernel now presents instead of
attempting to talk to it directly from userspace?

thanks,

greg k-h

Re: [BUG] usb-serial-simple: device conflict with Owon VDS1022I

[Amit]

Hi Greg,

     You've made valid points on both. Let me check with the authors of 
the software, there is an ongoing discussion on this.

regards,

Amit

On 06/06/25 17:47, Greg KH wrote:
> On Fri, Jun 06, 2025 at 05:00:29PM +0530, Amit wrote:
>> Hi,
>>
>>      PATCH 6.1 082/167 adds serial support for OWON HDS200 through the simple
>> usb serial driver (). This causes issues with Owon VDS1022. The VID and PID
>> is the same for both devices so the the Linux kernel loads the serial
>> driver, preventing the associated software from working.
> As vendors are not supposed to use the same device id for different
> devices, that's against the USB.org rules, so that's on them, not us :)
>
>>      My current work around for this is to blacklist usb_serial_simple. Given
>> that this driver change was recent and still under testing, could we find a
>> solution for this that doesn't involve such hacks at my end.
>>
>>      I  was able to root cause this issue only because the USB device and
>> software was working correctly before. A new user would just assume that the
>> device is not Linux compatible.
>>
>>      I've filed a bugzilla report here :
>> https://bugzilla.kernel.org/show_bug.cgi?id=220193
>>
>>      This is the software : https://github.com/florentbr/OWON-VDS1022
> As that's just using libusb, it can disconnect the device from the
> usb-serial driver if it needs to from the program itself.  Why not
> suggest this change to the authors of the software, OR just change it to
> use the serial connection that the kernel now presents instead of
> attempting to talk to it directly from userspace?
>
> thanks,
>
> greg k-h