Closed Bug 1119906 Opened 9 years ago Closed 6 years ago

Intermittent browser/base/content/test/newtab/browser_newtab_bug722273.js | fake site is gone

Categories

(Firefox :: New Tab Page, defect, P3)

Product:
Firefox
Component:
New Tab Page
Platform:
x86_64
Windows 8.1
Type:
defect

Tracking

()

Status:
RESOLVED FIXED
Milestone:
Firefox 60
Tracking Flags:
Tracking Status
firefox41 --- disabled
firefox42 --- disabled
firefox43 --- disabled
firefox-esr38 --- unaffected
firefox60 --- fixed

People

(Reporter: RyanVM, Assigned: tracy)

References

Details

(Keywords: intermittent-failure, Whiteboard: [leave open][stockwell unknown])

Attachments

(1 file, 3 obsolete files)

bug_1119906_skip-if_redo.patch
1.20 KB, patch
jmaher
: review+
Details | Diff | Splinter Review 
02:21:38 INFO - 650 INFO TEST-START | browser/base/content/test/newtab/browser_newtab_bug722273.js
02:21:39 INFO - 651 INFO checking window state
02:21:39 INFO - 652 INFO innerHeight,776,884
02:21:39 INFO - 653 INFO Changing browser innerHeight from 776 to 884
02:21:39 INFO - 654 INFO innerWidth,1186,1030
02:21:39 INFO - 655 INFO TEST-PASS | browser/base/content/test/newtab/browser_newtab_bug722273.js | first site is our fake site
02:21:39 INFO - 656 INFO TEST-UNEXPECTED-FAIL | browser/base/content/test/newtab/browser_newtab_bug722273.js | the fake site is gone -
02:21:39 INFO - Stack trace:
02:21:39 INFO - chrome://mochitests/content/browser/browser/base/content/test/newtab/browser_newtab_bug722273.js:runTests:24
02:21:39 INFO - self-hosted:InterpretGeneratorResume:659
02:21:39 INFO - self-hosted:next:617
02:21:39 INFO - chrome://mochitests/content/browser/browser/base/content/test/newtab/head.js:TestRunner.next:146
02:21:39 INFO - chrome://mochikit/content/browser-test.js:testScope/test_executeSoon/<.run:881
02:21:39 INFO - null:null:0
02:21:39 INFO - 657 INFO MEMORY STAT vsize after test: 691724288
02:21:39 INFO - 658 INFO MEMORY STAT vsizeMaxContiguous after test: 7828021288960
02:21:39 INFO - 659 INFO MEMORY STAT residentFast after test: 252329984
02:21:39 INFO - 660 INFO MEMORY STAT heapAllocated after test: 69173838
02:21:39 INFO - 661 INFO TEST-OK | browser/base/content/test/newtab/browser_newtab_bug722273.js | took 709ms
Attached patch patch (obsolete) — — Splinter Review 
Assignee: nobody → mbrubeck
Status: NEW → ASSIGNED

        Attachment #8638940 -
        Flags: review?(ttaubert)

    
    

    
  
Comment on attachment 8638940 [details] [diff] [review]
patch

Actually, I don't think this is correct. Sorry.

        Attachment #8638940 -
        Attachment is obsolete: true

        Attachment #8638940 -
        Flags: review?(ttaubert)
Assignee: mbrubeck → nobody
Status: ASSIGNED → NEW

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
  
Tim, do you have time to look into this or can you suggest someone who might? This is one of the top oranges right now and needs an owner.
Flags: needinfo?(ttaubert)

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
  
Didn't work :(

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
  
Tim, any news here? This is a top orange and I'm leaning towards skipping it at this point.
Flags: needinfo?(ttaubert)

    
    

    
      
  

    
    

    
      
  

    
    

    
  
(In reply to Ryan VanderMeulen [:RyanVM UTC-4] from comment #441)
> Tim, any news here? This is a top orange and I'm leaning towards skipping it
> at this point.

Sorry, I'm out of ideas. If the content services folks that own the newtab page these days have no idea either then skipping sounds fair to me.
Flags: needinfo?(ttaubert)

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
  
Test disabled on m-c. ni? myself to disable on Aurora/Beta as well.

          status-firefox41:
          --- → affected

          status-firefox42:
          --- → affected

          status-firefox43:
          --- → disabled

          status-firefox-esr38:
          --- → unaffected
Flags: needinfo?(ryanvm)
Whiteboard: [test disabled][leave open]

    
    

    
      
  

    
    

    
      
  

    
  

          status-firefox41:
          affected → disabled

          status-firefox42:
          affected → disabled
Flags: needinfo?(ryanvm)

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
  
why is this bug still open?

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
  
Bulk assigning P3 to all open intermittent bugs without a priority set in Firefox components per bug 1298978.
Priority: -- → P3

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
  
RyanVM re-enabled this on 2016-05-05 with https://hg.mozilla.org/mozilla-central/rev/7b05c8371114. browser_newtab_bug722273.js is still failing intermittently on OS X 10.10 debug. https://brasstacks.mozilla.com/orangefactor/?display=Bug&bugid=1119906&entireHistory=true&tree=mozilla-central

RyanVM: would you like to disable this for (os = Mac && debug)?
Flags: needinfo?(ryanvm)
Whiteboard: [test disabled][leave open] → [leave open]

    
    

    
      
  

    
    

    
  
That sounds like a great idea if we can't find anybody to investigate it.
Flags: needinfo?(ryanvm)

    
    

    
  
Ok, I'm going to dig into the test case and ask around if anyone has any ideas what is going on here.  If I am unable to fix it for Mac debug, I'll resort to disabling for that configuration.

    
    

    
  
I can manually clear history and the about:newtab page does show history cleared.  So that functionality isn't actually broken.  For some reason about:newtab doesn't auto-reload on Mac debug (works in opt) when the history is cleared via the test method.   

Question:  Is the intent of this test to check that about:newtab correctly reflects cleared history and/or to test about:newtab reloading on clear history?  If it's it the former, I have a patch in try that works around the page apparently not auto-reloading by loading up another about:newtab. https://treeherder.mozilla.org/#/jobs?repo=try&revision=7bd0e3e51f20. If it's the latter, then I am stuck and we should probably just disable for Mac debug, unless anyone has any ideas.

    
    

    
  
not sure how that other bug got intertwined into my try run there.  That seems to have caused the Linux runs to be skipped.  trying try again: https://treeherder.mozilla.org/#/jobs?repo=try&revision=1edb625b9753. which appears to be a clean attempt.  Let's see how it goes.  

Neil, is this patch's approach correct, or am I now avoiding what is supposed to be tested? (see previous comment Question)
Flags: needinfo?(enndeakin)

    
    

    
  
why is try skipping a big chunk of tests in bc1 on Linux, including this test?

    
    

    
  
(In reply to Tracy Walker [:tracy] from comment #515)
> why is try skipping a big chunk of tests in bc1 on Linux, including this
> test?

These newtab tests are skipped on linux - 
http://searchfox.org/mozilla-central/source/browser/base/content/test/newtab/browser.ini#2

    
    

    
  
(In reply to Tracy Walker [:tracy] from comment #514)
> Neil, is this patch's approach correct, or am I now avoiding what is
> supposed to be tested? (see previous comment Question)

I think it should be testing that the original tab's contents updated. But switching back to tabA (using BrowserTestUtils.switchTab) before calling performOnCell would probably be sufficient for the test, assuming that works.
Flags: needinfo?(enndeakin)

    
    

    
  

      
      
      
      

        Attached patch
          
          
        bug_1119906.patch (obsolete)
        — 
          — Splinter Review
      

    


  
The head.js function, whenPagesUpdated(), is supposed to be waiting for the page to be updated.  But on Mac debug there is some underlying delay in play that slips through and causes the test to fail.   

Not quite sure why this hack works. But I can't otherwise figure out how to delay the performOnCell check of the data under test until the data has actually been updated.  well, a hard setTimeout before performOnCell is called works.  But isn't doing so worse than this patch?
Assignee: nobody → twalker
Status: NEW → ASSIGNED

        Attachment #8801208 -
        Flags: review?(enndeakin)

    
    

    
      
  

    
    

    
  
Comment on attachment 8801208 [details] [diff] [review]
bug_1119906.patch

I found that the history sanitization isn't happening timely.  Well Sanitizer() is asynchronous. So for some reason it's signaling complete when it is not actually complete.

anyway, going to obsolete this patch and dig into sanitize.js.

        Attachment #8801208 -
        Flags: review?(enndeakin)

    
  

        Attachment #8801208 -
        Attachment is obsolete: true

    
    

    
      
  

    
    

    
  
Tim, best I can tell, the use of function whenPagesUpdated() from browser/base/content/test/newtab/head.js isn't working for this particular test case, on Mac debug.  The check for the cell site being null happens before the page is actually refreshed.

I can force the test case into working by running this after the history is cleared, instead of whenPagesUpdated().  It usually takes just 3 or 4 iterations through the loop to break out successfully.

  let isGone;
  for (let i = 0; i < 100; i++) {
    isGone = yield performOnCell(0, cell => { return cell.site == null; });
    if (isGone == true) {
      break;
    }
  }
  
  ok(isGone, "fake site is gone"); 

But is there something we can listen for here to ensure the page has refreshed before the final check?  Some DOM state or something?
Flags: needinfo?(ttaubert)

    
    

    
  
Haven't look at this code for a loooong while, sorry. I'm not sure what the best way is to fix this test, your best bet is probably to ask the people currently working on it. I'm not sure who that is though...
Flags: needinfo?(ttaubert)

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
  
Tracy, this bug has really increased in frequency over the last week, it looks like November 14th was a trigger day for this bug- mind taking another look at this?
Flags: needinfo?(twalker)

    
    

    
  
Joel,  yes,  I'll try to figure out what may have regressed this to fail even more often.  I'll also check if my hacky test fix still eliminates the raciness in the test.  Which may be a temporary solution, 'til I'm able to sort out a better observer helper than the one that is intermittently failing in the test.

    
    

    
      
  

    
    

    
  
Incidentally, I think this test is flakey in general. I can trivially get it to timeout by running it through rr on Linux. I get a bunch of console spew like this in the setup phase (around: `yield promiseAddFakeVisits()`)

89 INFO Console message: [JavaScript Error: "linkMap is undefined" {file: "resource://gre/modules/NewTabUtils.jsm" line: 1159}]
Links_onLinkChanged@resource://gre/modules/NewTabUtils.jsm:1159:9
PlacesProvider__callObservers@resource://gre/modules/NewTabUtils.jsm:798:11
PlacesProvider_onFrecencyChanged@resource://gre/modules/NewTabUtils.jsm:768:7
89 INFO Console message: [JavaScript Error: "linkMap is undefined" {file: "resource://gre/modules/NewTabUtils.jsm" line: 1159}]
Links_onLinkChanged@resource://gre/modules/NewTabUtils.jsm:1159:9
PlacesProvider__callObservers@resource://gre/modules/NewTabUtils.jsm:798:11
PlacesProvider_onFrecencyChanged@resource://gre/modules/NewTabUtils.jsm:768:7

    
    

    
  

      
      
      
      

        Attached patch
          
          
        bug_1119906_skip-if.patch (obsolete)
        — 
          — Splinter Review
      

    


  
Yes, flakey on debug. Test case works fine in opt.  

Let's disable this test case for the time being; 'til I can figure out a reliable and non-cludgel method of solving the raciness with test method at the page reload.  I will figure this out, at the latest in Hawaii.

Joel, please find simple patch attached for review.
Flags: needinfo?(twalker)

        Attachment #8813734 -
        Flags: review?(jmaher)

    
    

    
  
Comment on attachment 8813734 [details] [diff] [review]
bug_1119906_skip-if.patch

Review of attachment 8813734 [details] [diff] [review]:
-----------------------------------------------------------------

should be easy to fix up and r+

::: browser/base/content/test/newtab/browser.ini
@@ +7,5 @@
>  [browser_newtab_background_captures.js]
>  [browser_newtab_block.js]
>  [browser_newtab_bug721442.js]
>  [browser_newtab_bug722273.js]
> +skip-if (os = "mac" && debug) # temporary sip-if due to increase in intermittent failures on Mac debug - bug 1119906

this should be: os == "mac" <- note the == vs =

nit: small typo, should be: skip-if

        Attachment #8813734 -
        Flags: review?(jmaher) → review-

    
    

    
  
Comment on attachment 8813734 [details] [diff] [review]
bug_1119906_skip-if.patch

Review of attachment 8813734 [details] [diff] [review]:
-----------------------------------------------------------------

::: browser/base/content/test/newtab/browser.ini
@@ +7,5 @@
>  [browser_newtab_background_captures.js]
>  [browser_newtab_block.js]
>  [browser_newtab_bug721442.js]
>  [browser_newtab_bug722273.js]
> +skip-if (os = "mac" && debug) # temporary sip-if due to increase in intermittent failures on Mac debug - bug 1119906

Also needs to be |skip-if = ...|

    
    

    
  


  
yep, got the correct skip-if statement here now and the fixed typo in comment

        Attachment #8813734 -
        Attachment is obsolete: true

        Attachment #8813772 -
        Flags: review?(jmaher)

    
    

    
  
Comment on attachment 8813772 [details] [diff] [review]
bug_1119906_skip-if_redo.patch

Review of attachment 8813772 [details] [diff] [review]:
-----------------------------------------------------------------

thank you!

        Attachment #8813772 -
        Flags: review?(jmaher) → review+

    
    

    
  
(In reply to Joel Maher ( :jmaher) from comment #539)
> Comment on attachment 8813772 [details] [diff] [review]
> bug_1119906_skip-if_redo.patch
> 
> Review of attachment 8813772 [details] [diff] [review]:
> -----------------------------------------------------------------
> 
> thank you!

thank you!  


keeping the [leave open] for actual fix
Keywords: checkin-needed

    
    

    
  
Pushed by ryanvm@gmail.com:
https://hg.mozilla.org/integration/mozilla-inbound/rev/70a5d943ef16
Temporarily skip browser_newtab_bug722273.js on OSX debug until it can be made more reliable. r=jmaher
Keywords: checkin-needed

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  
Summary: Intermittent browser_newtab_bug722273.js | the fake site is gone → Intermittent browser_newtab_bug722273.js | fake site is gone

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
  
Assignee: twalker → nobody
Status: ASSIGNED → NEW

    
    

    
  
this picked up again on April 11th, on windows 7 debug.

this test is already skipped on osx/debug (this bug ~5 months ago):
https://dxr.mozilla.org/mozilla-central/source/browser/base/content/test/newtab/browser.ini#10

I am doing some retriggers on april 11th to determine if there is a root cause:
https://treeherder.mozilla.org/#/jobs?repo=mozilla-inbound&filter-searchStr=win%20debug%20vm%20browser-chrome-5&tochange=c05000863d49dd34483a779725de39a47deedbe1&fromchange=7d6d6e8ec1cc55acae0a01688ad47dd9363c3c36&selectedJob=90240375

ni myself to follow up on retriggers and get more context
Flags: needinfo?(jmaher)
Summary: Intermittent browser_newtab_bug722273.js | fake site is gone → Intermittent browser/base/content/test/newtab/browser_newtab_bug722273.js | fake site is gone
Whiteboard: [leave open] → [leave open][stockwell needswork]

    
    

    
  
to add to this, this failure is only showing up on taskcluster mochitest, not buildbot.  The environment is slightly different (in a good way).

this also fails in non-e10s and e10s.

As this failed 46 times in the last week, we should look into the failures.

a recent log is here:
https://treeherder.mozilla.org/logviewer.html#?repo=autoland&job_id=91992133&lineNumber=4010

and related screenshot:
https://queue.taskcluster.net/v1/task/AyWO0uFJQSypr2Kna4IK8w/runs/0/artifacts/public/test_info/mozilla-test-fail-screenshot_9iaekt.png

and here is related information from the above mentioned log:
00:03:57     INFO -  80 INFO TEST-START | browser/base/content/test/newtab/browser_newtab_bug722273.js
00:03:57     INFO -  GECKO(4264) | [Parent 4264] WARNING: Cannot set transparency mode on top-level windows.: file z:/build/build/src/widget/windows/nsWindow.cpp, line 3057
00:03:57     INFO -  GECKO(4264) | [Parent 4264] WARNING: Cannot set transparency mode on top-level windows.: file z:/build/build/src/widget/windows/nsWindow.cpp, line 3057
00:03:57     INFO -  GECKO(4264) | --DOCSHELL 084CA000 == 2 [pid = 4368] [id = {45359cd6-2c49-479d-bba5-ab5698589403}]
00:03:57     INFO -  GECKO(4264) | --DOCSHELL 053BD000 == 1 [pid = 4368] [id = {f6381e4c-7442-4d9a-9783-aeeaea27c4e5}]
00:03:57     INFO -  GECKO(4264) | [Parent 4264] WARNING: Cannot set transparency mode on top-level windows.: file z:/build/build/src/widget/windows/nsWindow.cpp, line 3057
00:03:57     INFO -  GECKO(4264) | [Parent 4264] WARNING: Cannot set transparency mode on top-level windows.: file z:/build/build/src/widget/windows/nsWindow.cpp, line 3057
00:03:57     INFO -  GECKO(4264) | ++DOCSHELL 0EBE6000 == 22 [pid = 4264] [id = {f370b13e-60e3-45ff-886c-a6639da22ae3}]
00:03:57     INFO -  GECKO(4264) | ++DOMWINDOW == 64 (1388F800) [pid = 4264] [serial = 109] [outer = 00000000]
00:03:57     INFO -  GECKO(4264) | [Parent 4264] WARNING: 'NS_FAILED(rv)', file z:/build/build/src/caps/BasePrincipal.cpp, line 380
00:03:57     INFO -  GECKO(4264) | ++DOMWINDOW == 65 (13C60C00) [pid = 4264] [serial = 110] [outer = 1388F800]
00:03:57     INFO -  GECKO(4264) | [Parent 4264] WARNING: 'NS_FAILED(rv)', file z:/build/build/src/caps/BasePrincipal.cpp, line 380
00:03:57     INFO -  GECKO(4264) | [Parent 4264] WARNING: 'NS_FAILED(rv)', file z:/build/build/src/caps/BasePrincipal.cpp, line 380
00:03:58     INFO -  GECKO(4264) | [Parent 4264] WARNING: Cannot set transparency mode on top-level windows.: file z:/build/build/src/widget/windows/nsWindow.cpp, line 3057
00:03:58     INFO -  GECKO(4264) | ++DOMWINDOW == 66 (0CD1D800) [pid = 4264] [serial = 111] [outer = 1388F800]
00:03:58     INFO -  GECKO(4264) | [Parent 4264] WARNING: 'NS_FAILED(rv)', file z:/build/build/src/dom/xhr/XMLHttpRequestMainThread.cpp, line 2000
00:03:58     INFO -  GECKO(4264) | [Parent 4264] WARNING: 'NS_FAILED(rv)', file z:/build/build/src/dom/xhr/XMLHttpRequestMainThread.cpp, line 2000
00:03:58     INFO -  GECKO(4264) | [Parent 4264] WARNING: 'NS_FAILED(rv)', file z:/build/build/src/dom/xhr/XMLHttpRequestMainThread.cpp, line 2000
00:03:58     INFO -  TEST-INFO | started process screenshot
00:03:58     INFO -  TEST-INFO | screenshot: exit 0
00:03:58     INFO -  Buffered messages logged at 00:03:57
00:03:58     INFO -  81 INFO innerHeight,,884
00:03:58     INFO -  82 INFO innerWidth,,1030
00:03:58     INFO -  83 INFO Entering test bound setup
00:03:58     INFO -  84 INFO Leaving test bound setup
00:03:58     INFO -  85 INFO Entering test bound
00:03:58     INFO -  Buffered messages logged at 00:03:58
00:03:58     INFO -  86 INFO TEST-PASS | browser/base/content/test/newtab/browser_newtab_bug722273.js | first site is our fake site -
00:03:58     INFO -  Buffered messages finished
00:03:58    ERROR -  87 INFO TEST-UNEXPECTED-FAIL | browser/base/content/test/newtab/browser_newtab_bug722273.js | fake site is gone -
00:03:58     INFO -  Stack trace:
00:03:58     INFO -      chrome://mochitests/content/browser/browser/base/content/test/newtab/browser_newtab_bug722273.js:null:27
00:03:58     INFO -      process@resource://gre/modules/Promise.jsm -> resource://gre/modules/Promise-backend.js:922:23
00:03:58     INFO -      walkerLoop@resource://gre/modules/Promise.jsm -> resource://gre/modules/Promise-backend.js:806:7
00:03:58     INFO -      Promise*scheduleWalkerLoop@resource://gre/modules/Promise.jsm -> resource://gre/modules/Promise-backend.js:739:11
00:03:58     INFO -      schedulePromise@resource://gre/modules/Promise.jsm -> resource://gre/modules/Promise-backend.js:770:7
00:03:58     INFO -      completePromise@resource://gre/modules/Promise.jsm -> resource://gre/modules/Promise-backend.js:707:7
00:03:58     INFO -      TaskImpl_run@resource://gre/modules/Task.jsm:324:15
00:03:58     INFO -      promise callback*TaskImpl_handleResultValue@resource://gre/modules/Task.jsm:396:7
00:03:58     INFO -      TaskImpl_run@resource://gre/modules/Task.jsm:327:15
00:03:58     INFO -      promise callback*TaskImpl_handleResultValue@resource://gre/modules/Task.jsm:396:7
00:03:58     INFO -      TaskImpl_run@resource://gre/modules/Task.jsm:327:15
00:03:58     INFO -      TaskImpl@resource://gre/modules/Task.jsm:277:3
00:03:58     INFO -      asyncFunction@resource://gre/modules/Task.jsm:252:14
00:03:58     INFO -      Task_spawn@resource://gre/modules/Task.jsm:166:12
00:03:58     INFO -      setup@chrome://mochitests/content/browser/browser/base/content/test/newtab/head.js:133:22
00:03:58     INFO -      TaskImpl_run@resource://gre/modules/Task.jsm:319:42
00:03:58     INFO -      TaskImpl@resource://gre/modules/Task.jsm:277:3
00:03:58     INFO -      asyncFunction@resource://gre/modules/Task.jsm:252:14
00:03:58     INFO -      Task_spawn@resource://gre/modules/Task.jsm:166:12
00:03:58     INFO -      TaskImpl_handleResultValue@resource://gre/modules/Task.jsm:389:16
00:03:58     INFO -      TaskImpl_run@resource://gre/modules/Task.jsm:327:15
00:03:58     INFO -      TaskImpl@resource://gre/modules/Task.jsm:277:3
00:03:58     INFO -      asyncFunction@resource://gre/modules/Task.jsm:252:14
00:03:58     INFO -      Task_spawn@resource://gre/modules/Task.jsm:166:12
00:03:58     INFO -      Tester_execTest@chrome://mochikit/content/browser-test.js:752:9
00:03:58     INFO -      Tester.prototype.nextTest</<@chrome://mochikit/content/browser-test.js:672:7
00:03:58     INFO -      SimpleTest.waitForFocus/waitForFocusInner/focusedOrLoaded/<@chrome://mochikit/content/tests/SimpleTest/SimpleTest.js:791:59
00:03:58     INFO -  88 INFO Leaving test bound
00:03:58     INFO -  GECKO(4264) | [Parent 4264] WARNING: Cannot set transparency mode on top-level windows.: file z:/build/build/src/widget/windows/nsWindow.cpp, line 3057
00:03:58     INFO -  GECKO(4264) | [Parent 4264] WARNING: Cannot set transparency mode on top-level windows.: file z:/build/build/src/widget/windows/nsWindow.cpp, line 3057
00:03:58     INFO -  GECKO(4264) | [Parent 4264] WARNING: Cannot set transparency mode on top-level windows.: file z:/build/build/src/widget/windows/nsWindow.cpp, line 3057
00:03:59     INFO -  GECKO(4264) | MEMORY STAT | vsize 707MB | vsizeMaxContiguous 822MB | residentFast 281MB | heapAllocated 108MB
00:03:59     INFO -  89 INFO TEST-OK | browser/base/content/test/newtab/browser_newtab_bug722273.js | took 1890ms


:pdehaan, can you help find someone on your team to look at this test and help reduce the failures or fix it?
Blocks: 1351272
Flags: needinfo?(jmaher) → needinfo?(pdehaan)

    
    

    
      
  

    
    

    
  
this bug has reduced in frequency last week and this week- not sure if something is fixed or this test is sensitive to certain timing or memory pressures that might have changed recently.

:pdehaan, as the triage owner, this is a long time for a response, it would be nice to put this bug in the right hands so in case it does spike up again we have a path to action.
Whiteboard: [leave open][stockwell needswork] → [leave open][stockwell unknown]

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
  
it looks like this picked up on the 13th (took a break for the weekend) and had a few failures/day since the 16th - this is all on windows7-debug.

Here is a recent log:
https://treeherder.mozilla.org/logviewer.html#?repo=autoland&job_id=138741919

and related screenshot:
https://public-artifacts.taskcluster.net/Ic6OOrHYS-eJhbAO0ZfKVg/0/public/test_info/mozilla-test-fail-screenshot_0qetyy.png

here is related text to the failure:
12:26:34     INFO -  24 INFO Leaving test bound setup
12:26:34     INFO -  25 INFO Entering test bound
12:26:34     INFO -  Buffered messages logged at 12:26:34
12:26:34     INFO -  26 INFO TEST-PASS | browser/base/content/test/newtab/browser_newtab_bug722273.js | first site is our fake site -
12:26:34     INFO -  Buffered messages finished
12:26:34    ERROR -  27 INFO TEST-UNEXPECTED-FAIL | browser/base/content/test/newtab/browser_newtab_bug722273.js | fake site is gone -
12:26:34     INFO -  Stack trace:
12:26:34     INFO -  chrome://mochitests/content/browser/browser/base/content/test/newtab/browser_newtab_bug722273.js:null:27
12:26:34     INFO -  28 INFO Leaving test bound
12:26:34     INFO -  GECKO(2680) | --DOMWINDOW == 9 (0D7D5000) [pid = 2296] [serial = 16] [outer = 00000000] [url = http://mochi.test:8888/#2]
12:26:34     INFO -  GECKO(2680) | --DOMWINDOW == 8 (0E50E000) [pid = 2296] [serial = 11] [outer = 00000000] [url = about:blank]
12:26:34     INFO -  GECKO(2680) | --DOMWINDOW == 7 (0D7D9000) [pid = 2296] [serial = 13] [outer = 00000000] [url = http://mochi.test:8888/#0]
12:26:34     INFO -  GECKO(2680) | MEMORY STAT | vsize 761MB | vsizeMaxContiguous 570MB | residentFast 268MB | heapAllocated 107MB
12:26:34     INFO -  29 INFO TEST-OK | browser/base/content/test/newtab/browser_newtab_bug722273.js | took 1136ms


:tim can you help get this prioritized based on work load and get in a queue for a fix.
Flags: needinfo?(pdehaan) → needinfo?(tspurway)
Whiteboard: [leave open][stockwell unknown] → [leave open][stockwell needswork]

    
    

    
  
deferring to :mardak
Flags: needinfo?(tspurway) → needinfo?(edilee)

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
  
Looks like the failure rate has dropped quite a bit, and these tests will probably go away when old tiles are removed. Activity Stream has its own tests that cover clearing history.
Flags: needinfo?(edilee)

    
    

    
      
  

    
    

    
      
  

    
    

    
      
  

    
    

    
  
Test was removed in bug 1433133 https://hg.mozilla.org/mozilla-central/rev/fc80dd3e4fae
Status: NEW → RESOLVED
Closed: 6 years ago

          status-firefox60:
          --- → fixed
Depends on: 1433133
Resolution: --- → FIXED
Target Milestone: --- → Firefox 60
Assignee: nobody → twalker

          You need to log in
          before you can comment on or make changes to this bug.
        






  

    
  







  

    

      
Attachment

      

      
      
      
      
    

    

      

        

          

            
General

            

              Creator: 



            

            
Created: 

            
Updated: 

            
Size: