New revision of 0c45:8419 Microdia/Wivitouch USB Infrared Touchscreen no longer working under 14.04

[Cesar Araujo]

Filing here per request at:

  http://irclogs.ubuntu.com/2015/11/24/%23ubuntu-devel.html#t14:35

The issue is regarding infra-red overlay touch screen panels from Wivitouch in China:

  http://www.wivitouch.com/sdp/1079694/4/pd-5199630/7116924-2132434/Multitouch_screen_infrared_touch_screen_19.html
  http://www.diytrade.com/china/pd/7116927/Multitouch_screen_infrared_touch_screen_18_5.html
  http://img.diytrade.com/smimg/1079694/17764824-1256906-0/nn/58f1.jpg

I have approximately 300 USB touch panels, from two different batches, but all report the same USB ID 0c45:8419. The manufacturer told me they had changed the firmware without changing the USB ID/version because they did not consider the change important. Both batches work correctly on Ubuntu 10.04, but only the first batch works on Ubuntu 14.04. The remaining batch does not work under Ubuntu 14.04 so this is a regression.

After plugging in, both batches appear equivalently under:

  /dev/input/eventNN

and for both 'evtest' reports:

  Input device ID: bus 0x3 vendor 0xc45 product 0x8419 version 0x100
  Input device name: "SONiX USB Device"
  …

Presumably something is occuring that causes a version version of 'libev' to handle one batch differently. Attached is the output of running:

  $ sudo lsusb -vvvv -c -d 0c45:8419 > ….txt

for one device from each of the two batches, which confirms that they appear identically. I have tested 250/250 (working) from the first batch, and 10/50 (not-working) from the second batch.

The only way to tell the batches apart is by the length of the USB cable, the first (working) batch have USB cables that are 15-20 centimetres long, and the second (non-working) batch have USB cables that are ~1 metre long.

[Cesar Araujo (ca-b)]

This works on ubuntu 10.04 the new batch and the old batch, out of the box.

[Cesar Araujo (ca-b)]

$ udevadm info --name=/dev/hidraw1
P: /devices/pci0000:00/0000:00:14.0/usb2/2-4/2-4.4/2-4.4:1.0/0003:0C45:8419.000E/hidraw/hidraw1
N: hidraw1
E: DEVNAME=/dev/hidraw1
E: DEVPATH=/devices/pci0000:00/0000:00:14.0/usb2/2-4/2-4.4/2-4.4:1.0/0003:0C45:8419.000E/hidraw/hidraw1
E: MAJOR=251
E: MINOR=1
E: SUBSYSTEM=hidraw

$ udevadm info --name=/dev/hidraw0
P: /devices/pci0000:00/0000:00:14.0/usb2/2-4/2-4.1/2-4.1:1.0/0003:0C45:8419.000D/hidraw/hidraw0
N: hidraw0
E: DEVNAME=/dev/hidraw0
E: DEVPATH=/devices/pci0000:00/0000:00:14.0/usb2/2-4/2-4.1/2-4.1:1.0/0003:0C45:8419.000D/hidraw/hidraw0
E: MAJOR=251
E: MINOR=0
E: SUBSYSTEM=hidraw

[Paul Sladen (sladen)]

Cesar, there is only one copy of the output of "sudo lsusb -vvvv -d 0c45:8419", please can you run if for *both* the old and new firmware-revision devices, so that it's possible to compare and see if there is a way to tell them apart.

Please can you include the information from the manufacturer that confirms that they changed the firmware, without changing the USB ID.

[Cesar Araujo (ca-b)]

Not working dmesg:

[30258.104003] usb 2-4.4: new low-speed USB device number 13 using xhci_hcd
[30258.124635] usb 2-4.4: New USB device found, idVendor=0c45, idProduct=8419
[30258.124643] usb 2-4.4: New USB device strings: Mfr=1, Product=2, SerialNumber=0
[30258.124649] usb 2-4.4: Product: USB Device
[30258.124655] usb 2-4.4: Manufacturer: SONiX
[30258.124967] usb 2-4.4: ep 0x81 - rounding interval to 32 microframes, ep desc says 40 microframes
[30258.132064] input: SONiX USB Device as /devices/pci0000:00/0000:00:14.0/usb2/2-4/2-4.4/2-4.4:1.0/input/input45
[30258.132567] hid-multitouch 0003:0C45:8419.0011: input,hiddev0,hidraw0: USB HID v1.00 Mouse [SONiX USB Device] on usb-0000:00:14.0-4.4/input0

Working dmesg

[30298.065870] usb 2-4.1: new low-speed USB device number 14 using xhci_hcd
[30298.088751] usb 2-4.1: New USB device found, idVendor=0c45, idProduct=8419
[30298.088762] usb 2-4.1: New USB device strings: Mfr=1, Product=2, SerialNumber=0
[30298.088768] usb 2-4.1: Product: USB Device
[30298.088773] usb 2-4.1: Manufacturer: SONiX
[30298.088979] usb 2-4.1: ep 0x81 - rounding interval to 32 microframes, ep desc says 40 microframes
[30298.097192] input: SONiX USB Device as /devices/pci0000:00/0000:00:14.0/usb2/2-4/2-4.1/2-4.1:1.0/input/input47
[30298.097575] hid-multitouch 0003:0C45:8419.0012: input,hiddev0,hidraw0: USB HID v1.00 Mouse [SONiX USB Device] on usb-0000:00:14.0-4.1/input0

[Paul Sladen (sladen)]

(Extracted from tarball and re-attached).

[Paul Sladen (sladen)]

(Extracted from tarball and re-attached).

[Cesar Araujo (ca-b)]

python replay.py

[Cesar Araujo (ca-b)]

python replay.py

[Cesar Araujo (ca-b)]

evtestt of working

[Cesar Araujo (ca-b)]

evtestt of not working

[Cesar Araujo (ca-b)]

evtestt of not working

[Cesar Araujo (ca-b)]

The new batch has an usb cable longer than the original like a meter and the old batch the working one has a usb cable of 15/20 cms.

[Stephen Kitt (steve-sk2)]

When you say the second batch doesn't work, is it only in evtest that they don't work, or more generally? I'd expect this to be a kernel bug rather than an evtest or libevdev bug...

[Cesar Araujo (ca-b)]

Hello Stephen.

Thanks in advance for the response.

When I say second batch it means the second purchase we made of 100 touch screens. We don’t buy them every day but we do use them every day hundreds of times a day in kiosks, so when we buy, we buy in in batches of 50 or 150. About now we have 300 touch screens working and the "second batch" which was purchased November did not work. If I can help making giving more information or running some commands or even some code I will be glad.

Once again thanks in advance.