1212811 - TEST-UNEXPECTED-FAIL | test_windows.py TestBaseWindow.test_switch_to_and_focus | AssertionError: False is not true

Status: RESOLVED FIXED

(Testing :: Firefox UI Tests, defect)

Description

[Henrik Skupin [:whimboo][⌚️UTC+2]]

Traceback (most recent call last):
File "/home/mozauto/jenkins/workspace/mozilla-central_functional/venv/local/lib/python2.7/site-packages/marionette/marionette_test.py", line 296, in run
testMethod()
File "/home/mozauto/jenkins/workspace/mozilla-central_functional/venv/lib/python2.7/site-packages/firefox_puppeteer/tests/test_windows.py", line 153, in test_switch_to_and_focus
self.assertTrue(win2.focused)
AssertionError: False is not true

Comment 1

[Intermittent Failures Robot]

11 automation job failures were associated with this bug in the last 7 days.

Repository breakdown:
* mozilla-beta: 7
* mozilla-aurora: 3
* mozilla-release: 1

Platform breakdown:
* linux32: 6
* linux64: 5

For more details, see:
https://brasstacks.mozilla.com/orangefactor/?display=Bug&bugid=1212811&startday=2015-10-12&endday=2015-10-18&tree=all

Comment 2

[Intermittent Failures Robot]

8 automation job failures were associated with this bug in the last 7 days.

Repository breakdown:
* mozilla-aurora: 4
* mozilla-central: 2
* mozilla-beta: 2

Platform breakdown:
* linux64: 4
* linux32: 4

For more details, see:
https://brasstacks.mozilla.com/orangefactor/?display=Bug&bugid=1212811&startday=2015-10-19&endday=2015-10-25&tree=all

Comment 3

[Henrik Skupin [:whimboo][⌚️UTC+2]]

We had a couple of those reports for the latest beta tests. I have checked the affected Linux machines and each of those had a crash report dialog open for the Keyboard input methods application: /usr/bin/ibus-daemon.

I retriggered the builds but looks like nearly all of them are still failing. So that dialog is not the reason.

Comment 4

[Intermittent Failures Robot]

11 automation job failures were associated with this bug in the last 7 days.

Repository breakdown:
* mozilla-beta: 11

Platform breakdown:
* linux64: 6
* linux32: 5

For more details, see:
https://brasstacks.mozilla.com/orangefactor/?display=Bug&bugid=1212811&startday=2015-11-23&endday=2015-11-29&tree=all

Comment 5

[Intermittent Failures Robot]

15 automation job failures were associated with this bug in the last 7 days.

Repository breakdown:
* mozilla-beta: 15

Platform breakdown:
* linux32: 9
* linux64: 6

For more details, see:
https://brasstacks.mozilla.com/orangefactor/?display=Bug&bugid=1212811&startday=2015-11-30&endday=2015-12-06&tree=all

Comment 6

[Intermittent Failures Robot]

10 automation job failures were associated with this bug in the last 7 days.

Repository breakdown:
* mozilla-release: 5
* mozilla-beta: 5

Platform breakdown:
* linux32: 6
* linux64: 4

For more details, see:
https://brasstacks.mozilla.com/orangefactor/?display=Bug&bugid=1212811&startday=2016-02-08&endday=2016-02-14&tree=all

Comment 7

[Intermittent Failures Robot]

7 automation job failures were associated with this bug in the last 7 days.

Repository breakdown:
* mozilla-beta: 7

Platform breakdown:
* linux64: 6
* linux32: 1

For more details, see:
https://brasstacks.mozilla.com/orangefactor/?display=Bug&bugid=1212811&startday=2016-02-22&endday=2016-02-28&tree=all

Comment 8

[Henrik Skupin [:whimboo][⌚️UTC+2]]

Ok, so the problem here could be in the `focused_chrome_window_handle()` method:

https://dxr.mozilla.org/mozilla-central/source/testing/puppeteer/firefox/firefox_puppeteer/ui/windows.py#42

In some cases - I assume on slower or highly utilized machines - the focus is not set immediately to the new window. Instead it takes a moment and exactly when performing the focus change no window has the focus. The above mentioned method would return `null` then. As result the failure in the BaseWindow test would happen.

Maybe the best option here would be to safe-guard `focused_chrome_window_handle()` and only return if a valid handle has been found or a timeout happens.

Comment 9

[Henrik Skupin [:whimboo][⌚️UTC+2]]

Ok, I was able to reproduce this problem on an affected Linux machine:

** Focusing of handle: 3
** focused_chrome_window_handle: None
** focused_chrome_window_handle: 15


We should indeed safe-guard the `focused_chrome_window_handle()` method.

Comment 10

[Henrik Skupin [:whimboo][⌚️UTC+2]]

Attachment: MozReview Request: Bug 1212811 - focused_chrome_window_handle() should wait until focus transition has been finished. r=maja_zf

Review commit: https://reviewboard.mozilla.org/r/38455/diff/#index_header
See other reviews: https://reviewboard.mozilla.org/r/38455/

Comment 11

[Henrik Skupin [:whimboo][⌚️UTC+2]]

Comment on attachment 8727357 [details]
MozReview Request: Bug 1212811 - focused_chrome_window_handle() should wait until focus transition has been finished. r=maja_zf

Review request updated; see interdiff: https://reviewboard.mozilla.org/r/38455/diff/1-2/

Comment 12

[Henrik Skupin [:whimboo][⌚️UTC+2]]

Comment on attachment 8727357 [details]
MozReview Request: Bug 1212811 - focused_chrome_window_handle() should wait until focus transition has been finished. r=maja_zf

Review request updated; see interdiff: https://reviewboard.mozilla.org/r/38455/diff/2-3/

Comment 13

[Maja Frydrychowicz :impossibus (was :maja_zf) (needinfo me)]

Comment on attachment 8727357 [details]
MozReview Request: Bug 1212811 - focused_chrome_window_handle() should wait until focus transition has been finished. r=maja_zf

https://reviewboard.mozilla.org/r/38455/#review35071

r+wc -- I have a small doubt, but otherwise this looks good.

::: testing/puppeteer/firefox/firefox_puppeteer/ui/windows.py:64
(Diff revision 3)
> +        return Wait(self.marionette).until(get_active_handle,

Can `get_active_handle` ever legitimately return 0? In that case, wouldn't `until` fail due to 0 being falsey?

Comment 14

[Henrik Skupin [:whimboo][⌚️UTC+2]]

(In reply to Maja Frydrychowicz (:maja_zf) from comment #13)
> Can `get_active_handle` ever legitimately return 0? In that case, wouldn't
> `until` fail due to 0 being falsey?

No, we do not have outer window ids of 0. The first visible chrome window which is the browser window has the id 3.

Comment 15

[Pulsebot]

https://hg.mozilla.org/integration/mozilla-inbound/rev/560000c81653

Comment 16

[Wes Kocher (:KWierso) (Not reading bugmail; email directly if needed)]

https://hg.mozilla.org/mozilla-central/rev/560000c81653

Comment 17

[Henrik Skupin [:whimboo][⌚️UTC+2]]

No test failure for this particular issue anymore the last days. Please get this patch landed on aurora and beta. It's a test-only change. Thanks.

Comment 18

[Carsten Book [:Tomcat]]

https://hg.mozilla.org/releases/mozilla-aurora/rev/8006cfdb9bff

Comment 19

[Carsten Book [:Tomcat]]

https://hg.mozilla.org/releases/mozilla-beta/rev/99f70ae29229