pointing devices appear to have multiple sys paths creating redundant tests

[Jeff Lane ]

Looking at the new pointing_device tests on this system:

201105-8051

and it appears that the touchpad and USB mouse are showing up as multiple devices:

https://certification.canonical.com/hardware/201105-8051/submission/iJ9VlCKh1zll4z5/result/test:input%252Fpointing_input7:input%252Fpointing
https://certification.canonical.com/hardware/201105-8051/submission/iJ9VlCKh1zll4z5/result/test:input%252Fpointing_input8:input%252Fpointing

both point to the Synaptics touchpad

and

https://certification.canonical.com/hardware/201105-8051/submission/iJ9VlCKh1zll4z5/result/test:input%252Fpointing_input5:input%252Fpointing
https://certification.canonical.com/hardware/201105-8051/submission/iJ9VlCKh1zll4z5/result/test:input%252Fpointing_input10:input%252Fpointing
https://certification.canonical.com/hardware/201105-8051/submission/iJ9VlCKh1zll4z5/result/test:input%252Fpointing_input15:input%252Fpointing

all point to a Logitech USB mouse?

[Brendan Donegan (brendan-donegan)]

Is this still an issue? It's turned one year old yesterday.

[Daniel Manrique (roadmr)]

Is there a way to get a submission from this system? I'd like to run it through the recently-updated udev parser, I'm confident the duplicated devices will not appear anymore. The most recent submission I saw on c3 dates from back in July, way before the new parser was operational, and also (I think) before c3 started storing raw submission.xml files so I have no data to test with :(

[Jeff Lane  (bladernr)]

https://certification.canonical.com/certificates/1305-5240/

Read the notes there... possibly not. It's in Taipei, so you could ask Yung, Sam or Taishiang, but that system may have be dead from other problems...

Yet another reason why I detest getting "engineering samples" for certification... they've usually been quite thoroughly abused and I really can't say if this problem was because of a legitimate bug or because the hardware was messed up. :(

[Daniel Manrique (roadmr)]

Tagging "job" as we may need to fine-tune the job itself, but also some work may need to be done in the parser to better "squash" these devices.

If we can't get the submission from the original offending systems, maybe we can scour recent submissions to see if this is an issue in more systems.

[Daniel Manrique (roadmr)]

I came up with a quick test script that can be copied to SUTs to verify. For instance, output on my system shows 3 pointing devices only, which is accurate (trackpoint, touchpad and USB mouse):

$ bash /tmp/quicktest.sh
 name: input/pointing_Logitech_USB_Receiver_MOUSE_input16
 requires: device.path == "/devices/pci0000:00/0000:00:1a.0/usb1/1-1/1-1.1/1-1.1:1.0/input/input16"

 name: input/pointing_SynPS2_Synaptics_TouchPad_TOUCHPAD_input5
 requires: device.path == "/devices/platform/i8042/serio1/input/input5"

 name: input/pointing_TPPS2_IBM_TrackPoint_MOUSE_input15
 requires: device.path == "/devices/platform/i8042/serio1/serio2/input/input15"

Here's the script:

#!/bin/bash
PATH=$PATH:/usr/share/checkbox/scripts
cat <<'EOF' | run_templates -t -s 'udev_resource | filter_templates -w "category=MOUSE" -w "category=TOUCHPAD" -w "category=TOUCHSCREEN"'
 name: input/pointing_`echo "${product}_${category}" | sed 's/ /_/g;s/[^_a-zA-Z0-9-]//g'`_`basename $path`
 requires: device.path == "$path"
EOF

[Zygmunt Krynicki (zyga)]

Setting to triaged as Daniel did a fair amount of debugging on this one