Add USB HID tests #10096

fkjagodzinski · 2019-03-14T11:26:37Z

Description

Add tests-usb_device-hid test suite for USBHID, USBMouse & USBKeyboard.

Pull request type

[ ] Fix
[ ] Refactor
[ ] Target update
[ ] Functionality change
[ ] Docs update
[x] Test update
[ ] Breaking change

Reviewers

@c1728p9, @jamesbeyond, @maciejbocianski

Release Notes

fkjagodzinski · 2019-03-14T11:32:15Z

requirements.txt

@@ -23,4 +23,5 @@ git+https://github.com/armmbed/[email protected]
 icetea>=1.2.1,<1.3
 pycryptodome>=3.7.2,<=3.7.3
 pyusb>=1.0.0,<2.0.0
+hidapi>=0.7.99,<0.8.0


@c1728p9, @jamesbeyond I'd like to clarify the licensing of this module. Based on its LICENSE file it can be used under one of 3 licenses. Should I specify, which one we want to use? If so, how do I do that?

@0xc0170 Might be a question for you as well

I need more context here, one of 3 licenses? Does this module hidapi have dual license or?

Yes 3

cython-hidapi may be used by one of three licenses as outlined in LICENSE.txt

Got it . This is missing in our license file (capturing python dependencies there!) , can you fix that ? IT would be added there as well - hidapi under BSD license

Created #10123

ciarmcom · 2019-03-14T12:00:39Z

@fkjagodzinski, thank you for your changes.
@maciejbocianski @jamesbeyond @c1728p9 @ARMmbed/mbed-os-maintainers please review.

fkjagodzinski · 2019-03-14T13:39:02Z

I can see Travis CI failed when installing the hidapi Python module added to requirements.txt in this PR.

@0xc0170 @cmonr How can I add the missing prerequisites to the travis config? Or should we avoid introducing such modification just to be able to run one test case? (Currently, all USB tests are skipped on the CI anyway.) I can update the python code to handle the import failure and skip this test case.

fkjagodzinski · 2019-03-14T16:33:34Z

OK, with a temporary fix, tests also work on Windows hosts. I'll prepare a proper fix shortly.

cmonr · 2019-03-15T03:13:12Z

How can I add the missing prerequisites to the travis config?

Oh jeez, this just got really nasty really quick.

if I understand correctly, the new python module also requires additional system packages?
Not sure I want to force users to install something else before even doing a pip install.

fkjagodzinski · 2019-03-15T11:47:56Z

if I understand correctly, the new python module also requires additional system packages?
Not sure I want to force users to install something else before even doing a pip install.

Yeah, this is not ideal, but applies only to Linux users. I can see two possible solutions here:

make that single test case optional (run only if the host machine has hidapi already installed), and possibly add the TESTS/usb_device/hid/{requirements.txt,README} files,
rewrite the test case to use pyusb -- the transmitted data is treated as raw anyway, no HID specific parsing here.

The second option seems better, but has a drawback -- the host side HID driver will be excluded, making the test less valuable. USB transfers through pyusb are already tested with basic USB tests.

@c1728p9 what do you think?

c1728p9 · 2019-03-15T15:32:11Z

TESTS/usb_device/hid/main.cpp

+    usb_hid.connect();
+    greentea_send_kv(MSG_KEY_TEST_RAW_IO, REPORT_SIZE);
+    usb_hid.wait_ready();
+    wait_ms(500);


Why the delay?

I added that temporarily to handle host side driver setup. It turned out to take much longer on Windows than Linux. It is replaced with a greentea_parse_kv() now.

c1728p9 · 2019-03-15T16:26:07Z

Hi @fkjagodzinski I would prefer to go with option 1 - make the test optional on linux. As you mentioned, I don't think pyusb will add much testing value as it bypasses the host's driver. @donatieng what do you think?

cmonr · 2019-03-15T16:56:42Z

CC @bulislaw

fkjagodzinski · 2019-03-26T15:24:02Z

The PR is ready for review again. Latest updates provide:

a proper fix for variable setup time of the host HID driver -- this was exposed on Windows; now the device simply waits for a greentea message from the host,
update for the requirements.txt file -- hidapi is skipped on Linux because of its dependencies; the module can still be easily installed manually; if the module is missing on the host machine, the one test case that requires it is skipped,
update for LICENSE.md providing info about hidapi and pyusb licenses.

adbridge · 2019-03-27T16:35:49Z

@c1728p9, @jamesbeyond, @maciejbocianski Can we get this re-reviewed please?

The first 4 bytes received were lost due to a wrong address. Read the output report into HID_REPORT.data.

To successfully use pyusb on Windows hosts, a Zadig configuration has to be performed. Since config for basic tests has already been provided, use it again.

Wait for the host driver to finish setup before sending any HID reports from the device. USBHID::wait_ready() blocks until the device reaches 'configured' state, but the state of the host HID driver remains unknown to the device.

Every test case waits at greentea_parse_kv() anyway.

This test case uses `hidapi` -- a cross-platform Python module. To keep the initial Mbed setup as simple as possible, the `hidapi` module is skipped on Linux hosts because of its external dependancies for this platform. The module can be easily installed following instructions from the README file. The test case is skipped if the host machine lacks `hidapi` module.

fkjagodzinski · 2019-04-09T15:14:15Z

I've rebased the commits onto current master and fixed the LICENSE.md conflict.

cmonr · 2019-04-09T18:40:20Z

CI started

mbed-ci · 2019-04-09T22:03:49Z

Test run: FAILED

Summary: 2 of 11 test jobs failed
Build number : 1
Build artifacts

Failed test jobs:

jenkins-ci/mbed-os-ci_cloud-client-test
jenkins-ci/mbed-os-ci_exporter

fkjagodzinski · 2019-04-10T10:13:36Z

I've checked the errors; they're unrelated to this PR.

jenkins-ci/mbed-os-ci_cloud-client-test

ExampleAppCloudTest01_Rel1_2.K66F

Dut Initialization failed, reason(s): DUT index 1 - RaaS DUT flashing failed

jenkins-ci/mbed-os-ci_exporter

# Failed build combinations
################################################################################
# mbed-os-example-blinky NUCLEO_F429ZI make_armc6

armclang: error: Failed to check out a license.

0xc0170 · 2019-04-10T10:26:43Z

Tests restarted

cmonr · 2019-04-12T01:20:54Z

CI job restarted: jenkins-ci/cloud-client-test

mbed-ci · 2019-04-17T04:25:23Z

Test run: SUCCESS

Summary: 11 of 11 test jobs passed
Build number : 2
Build artifacts

fkjagodzinski commented Mar 14, 2019

View reviewed changes

ciarmcom requested review from c1728p9, jamesbeyond, maciejbocianski and a team March 14, 2019 12:00

ciarmcom added the needs: review label Mar 14, 2019

fkjagodzinski force-pushed the test-usb-hid branch from 2f97c22 to a845863 Compare March 14, 2019 16:29

c1728p9 reviewed Mar 15, 2019

View reviewed changes

fkjagodzinski force-pushed the test-usb-hid branch from a845863 to 8c6f63e Compare March 22, 2019 16:13

cmonr added needs: work and removed needs: review labels Mar 25, 2019

maciejbocianski approved these changes Mar 28, 2019

View reviewed changes

c1728p9 approved these changes Mar 28, 2019

View reviewed changes

Filip Jagodzinski added 7 commits April 9, 2019 17:03

Tests: USBHID: Add tests

572beb6

USBHID: Fix the initial HID report read operation

565ceb0

The first 4 bytes received were lost due to a wrong address. Read the output report into HID_REPORT.data.

Tests: USBHID: Reuse VID & PID from basic test

4f88069

To successfully use pyusb on Windows hosts, a Zadig configuration has to be performed. Since config for basic tests has already been provided, use it again.

Tests: USBHID: Handle variable driver setup time

36eb700

Wait for the host driver to finish setup before sending any HID reports from the device. USBHID::wait_ready() blocks until the device reaches 'configured' state, but the state of the host HID driver remains unknown to the device.

Tests: USBHID: Remove unnecessary wait_ready call

111f485

Every test case waits at greentea_parse_kv() anyway.

Tests: USBHID: Specify hidapi license

491ea55

fkjagodzinski force-pushed the test-usb-hid branch from e7021b2 to 491ea55 Compare April 9, 2019 15:08

0xc0170 approved these changes Apr 17, 2019

View reviewed changes

0xc0170 added ready for merge release-version: 5.12.2 and removed needs: work labels Apr 17, 2019

0xc0170 merged commit 6b9d157 into ARMmbed:master Apr 17, 2019

0xc0170 removed the ready for merge label Apr 17, 2019

fkjagodzinski deleted the test-usb-hid branch April 17, 2019 14:01

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Add USB HID tests #10096

Add USB HID tests #10096

fkjagodzinski commented Mar 14, 2019

fkjagodzinski Mar 14, 2019

cmonr Mar 15, 2019

0xc0170 Mar 15, 2019

fkjagodzinski Mar 15, 2019

0xc0170 Mar 15, 2019

0xc0170 Mar 15, 2019

ciarmcom commented Mar 14, 2019

fkjagodzinski commented Mar 14, 2019

fkjagodzinski commented Mar 14, 2019

cmonr commented Mar 15, 2019 •

edited

Loading

fkjagodzinski commented Mar 15, 2019

c1728p9 Mar 15, 2019

fkjagodzinski Mar 22, 2019

c1728p9 commented Mar 15, 2019

cmonr commented Mar 15, 2019

fkjagodzinski commented Mar 26, 2019

adbridge commented Mar 27, 2019

fkjagodzinski commented Apr 9, 2019

cmonr commented Apr 9, 2019

mbed-ci commented Apr 9, 2019

fkjagodzinski commented Apr 10, 2019

0xc0170 commented Apr 10, 2019

cmonr commented Apr 12, 2019

mbed-ci commented Apr 17, 2019

Add USB HID tests #10096

Add USB HID tests #10096

Conversation

fkjagodzinski commented Mar 14, 2019

Description

Pull request type

Reviewers

Release Notes

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

ciarmcom commented Mar 14, 2019

fkjagodzinski commented Mar 14, 2019

fkjagodzinski commented Mar 14, 2019

cmonr commented Mar 15, 2019 • edited Loading

fkjagodzinski commented Mar 15, 2019

Choose a reason for hiding this comment

Choose a reason for hiding this comment

c1728p9 commented Mar 15, 2019

cmonr commented Mar 15, 2019

fkjagodzinski commented Mar 26, 2019

adbridge commented Mar 27, 2019

fkjagodzinski commented Apr 9, 2019

cmonr commented Apr 9, 2019

mbed-ci commented Apr 9, 2019

fkjagodzinski commented Apr 10, 2019

0xc0170 commented Apr 10, 2019

cmonr commented Apr 12, 2019

mbed-ci commented Apr 17, 2019

cmonr commented Mar 15, 2019 •

edited

Loading