Closed Bug 1205197 Opened 10 years ago Closed 10 years ago

Intermittent app_test.js | Homescreen app hashchange should scroll to the top of the page - timeout of 10000ms exceeded | Homescreen app hashchange should cancel dialogs - timeout of 10000ms exceeded

Categories

(Firefox OS Graveyard :: Gaia::Homescreen, defect)

ARM
Gonk (Firefox OS)
defect
Not set
normal

Tracking

(Not tracked)

RESOLVED FIXED

People

(Reporter: nigelb, Assigned: cwiiis)

Details

(Keywords: intermittent-failure, qawanted)

Attachments

(1 file)

No description provided.
Is this a regression? I've not seen this fail before... (not to say it wasn't, but still)
Adding qawanted to check if this is a regression, and if so, what from.
Keywords: qawanted
QAwanted will need steps to reproduce first. What kind of test is this? Do you have a theory on how to reproduce this manually?
Flags: needinfo?(chrislord.net)
(In reply to Pi Wei Cheng [:piwei] from comment #70) > QAwanted will need steps to reproduce first. What kind of test is this? Do > you have a theory on how to reproduce this manually? This tests that pressing the home button scrolls to the top of the page. I don't know why it's failing, the unit test works fine and consistently locally, I don't have any theories as to why it would fail. I don't think there are any random or time-based elements at play... However, I don't know how our build machines running unit tests differs to running them locally and that may yield some clues. I was hoping that QA would have the info of whether this has always happened, or only started happening at a certain point, and would be able to do the bisection if that didn't highlight a specific commit as causing this.
Flags: needinfo?(chrislord.net)
I believe this is marionette testing correct? We don't have the setup for this. No-jun or John do you know more that could help here?
Flags: needinfo?(npark)
Flags: needinfo?(jdorlus)
Asking martijn in case he know who to talk to. This looks like the marionette-js test, which we don't handle though.
Flags: needinfo?(npark) → needinfo?(martijn.martijn)
I saw this message at dev-gaia mailing list: https://groups.google.com/forum/#!msg/mozilla.dev.gaia/7OLMu487Dno/7XMPNj3HFwAJ Aren't Gij tests on taskcluster not just started running since 2 weeks only? There are also bug 1187717 and bug 1189286, which seems to indicate that there are still some stability issues running these tests on Mulet. Alexandre would perhaps know more about this. I'm currently not involved with Gij tests.
Flags: needinfo?(martijn.martijn) → needinfo?(lissyx+mozillians)
(In reply to Martijn Wargers [:mwargers] (QA) from comment #89) > I saw this message at dev-gaia mailing list: > https://groups.google.com/forum/#!msg/mozilla.dev.gaia/7OLMu487Dno/ > 7XMPNj3HFwAJ > Aren't Gij tests on taskcluster not just started running since 2 weeks only? > There are also bug 1187717 and bug 1189286, which seems to indicate that > there are still some stability issues running these tests on Mulet. > Alexandre would perhaps know more about this. I'm currently not involved > with Gij tests. I am not either. There was some unit tests failing with lockscreen. Turns out this was because the test lacked specifying some window size. Maybe there is something similar and this is only getting exposed by mulet?
Flags: needinfo?(lissyx+mozillians)
I am digging into this. Off the top of my head. I am not sure.
Flags: needinfo?(jdorlus)
Contrary to some comments, this is a unit test failing intermittently, not an integration test.
(In reply to Chris Lord [:cwiiis] from comment #92) > Contrary to some comments, this is a unit test failing intermittently, not > an integration test. Chris, I am referring to Gaia Unit chunk 9, where Greg Weng fixed some unit tests :)
Namely this was bug 1198140 which was perma fail. I cannot tell but maybe some similar reason can explain intermittent failure ? I think I have spotted some possible intermittent resize of the window at boot time in Mulet, maybe it is related?
Flags: needinfo?(chrislord.net)
Comment on attachment 8668856 [details] [review] [gaia] Cwiiis:bug1205197-new-homescreen-intermittent-hashchange-test > mozilla-b2g:master It could be this, possibly? document.hidden is undefined locally, but if it was defined on some test machine (somehow?) and returned true, these tests would fail. Pre-emptively flagging for review, it might be sensible to do this anyway (and also fixes asserting before calling done).
Flags: needinfo?(chrislord.net)
Attachment #8668856 - Flags: review?(gmarty)
Assignee: nobody → chrislord.net
Status: NEW → ASSIGNED
Comment on attachment 8668856 [details] [review] [gaia] Cwiiis:bug1205197-new-homescreen-intermittent-hashchange-test > mozilla-b2g:master LGTM.
Attachment #8668856 - Flags: review?(gmarty) → review+
Well, 10 green runs, I'm going to call that fixed. We can reopen/refile if it isn't. Merged: https://github.com/mozilla-b2g/gaia/commit/22b991afdb1e748bb8b6a1c5b0af5352c8c910f8
Status: ASSIGNED → RESOLVED
Closed: 10 years ago
Resolution: --- → FIXED
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Created:
Updated:
Size: