Closed Bug 963048 Opened 10 years ago Closed 10 years ago

Intermittent TEST-UNEXPECTED-FAIL | chrome://mochitests/content/browser/browser/base/content/test/newtab/browser_newtab_perwindow_private_browsing.js | Test timed out

Categories

(Firefox :: Private Browsing, defect)

Product:
Firefox
Component:
Private Browsing
Platform:
x86
Linux
Type:
defect
Priority:
Not set
Severity:
normal

Tracking

()

Status:
RESOLVED WORKSFORME

People

(Reporter: cbook, Unassigned)

References

(
URL
)

Details

(Keywords: intermittent-failure) 

Rev3 Fedora 12x64 mozilla-b2g26_v1_2 debug test mochitest-browser-chrome on 2014-01-22 15:45:04 PST for push 4494940fb3ea

slave: talos-r3-fed64-040

https://tbpl.mozilla.org/php/getParsedLog.php?id=33428503&tree=Mozilla-B2g26-v1.2

TEST-UNEXPECTED-FAIL | chrome://mochitests/content/browser/browser/base/content/test/newtab/browser_newtab_perwindow_private_browsing.js | Test timed out
This burst appears to be a regression from bug 911307.
while working on bug 992911 to run each directory standalone, I found that this test (browesr_newtab_perwindow_private_browsing.js) was causing the newtab/ directory to timeout during shutdown almost consistently on windows 8 opt:
https://tbpl.mozilla.org/?tree=Try&rev=ce66258f33e5

here is a link to the log:
https://tbpl.mozilla.org/php/getParsedLog.php?id=39914353&tree=Try

For additional data, I see a lot of privatebrowsing tests causing timeouts during shutdown, so there might be a single fix that we can do to clean up a lot of our problems.

I have cc'd the folks who have contributed to this specific test case.

:ehsan, any thoughts on this?
Flags: needinfo?(ehsan)
I would be inclined to comment out parts of the runTests method and see if there's anything in particular that contributes to the timeouts.
(In reply to Joel Maher (:jmaher) from comment #10)
> while working on bug 992911 to run each directory standalone, I found that
> this test (browesr_newtab_perwindow_private_browsing.js) was causing the
> newtab/ directory to timeout during shutdown almost consistently on windows
> 8 opt:
> https://tbpl.mozilla.org/?tree=Try&rev=ce66258f33e5
> 
> here is a link to the log:
> https://tbpl.mozilla.org/php/getParsedLog.php?id=39914353&tree=Try
> 
> For additional data, I see a lot of privatebrowsing tests causing timeouts
> during shutdown, so there might be a single fix that we can do to clean up a
> lot of our problems.
> 
> I have cc'd the folks who have contributed to this specific test case.
> 
> :ehsan, any thoughts on this?

There is nothing in the log which either explains what the browser is doing here, or anything that ties the timeout to PB tests.  Also, I don't really understand what the patches in your try push do.  I'm not sure if I can provide any useful information here, we need to know more about what's going on.
Flags: needinfo?(ehsan)
Thanks for replying!

I am running mochitest with --test-path browser/base/content/test/newtab, so just those tests are running.  The logs never have anything useful for these timeout on shutdown issues, but that is the most common issue I see.

To determine the PB test as the cause, I did a bunch of pushes with various tests commented out, this is the one that is causing it to timeout while shutting down.  If there is a better way to help narrow down these failures to a test case, let me know :)
(In reply to comment #13)
> Thanks for replying!
> 
> I am running mochitest with --test-path browser/base/content/test/newtab, so
> just those tests are running.  The logs never have anything useful for these
> timeout on shutdown issues, but that is the most common issue I see.
> 
> To determine the PB test as the cause, I did a bunch of pushes with various
> tests commented out, this is the one that is causing it to timeout while
> shutting down.  If there is a better way to help narrow down these failures to
> a test case, let me know :)

So do you get the shutdown timeout when running only this test through mach?

Have you attached a debugger to see what Firefox is doing when it's sitting around?
The reason I'm asking the first question is that the log you linked to is definitely running other tests.  :-)
got it.  The patch is uploaded to bug 992911, and it builds a list of leaf directories and runs each directory with a fresh profile/browser.  It does this by setting --test-path=<leaf_dir>

The result is we run a series of mochitest runs.  When doing this on mochitest-plain, we have 1 failure in all the tests, for browser-chrome we have 70 failures and I haven't touched the devtools yet :(
I am not seeing this failure in the recent try runs.
Status: NEW → RESOLVED
Closed: 10 years ago
Resolution: --- → WORKSFORME
You need to log in before you can comment on or make changes to this bug.