Closed Bug 544601 Opened 16 years ago Closed 15 years ago

Intermittent failure in browser_480148.js | Timed out, followed by several "Found a browser window after previous test timed out"

Categories

(Firefox :: Session Restore, defect)

Product:
Firefox
Component:
Session Restore
Platform:
x86
Windows Server 2003
Type:
defect
Priority:
Not set
Severity:
normal

Tracking

()

Status:
RESOLVED FIXED

People

(Reporter: philor, Unassigned)

References

Details

(Keywords: intermittent-failure)

Attachments

(1 file)

request doubled timeout [checked in]
865 bytes, patch
zpao
: review+
Details | Diff | Splinter Review
http://tinderbox.mozilla.org/showlog.cgi?log=Firefox/1265403566.1265406289.13981.gz Linux mozilla-central debug test mochitest-other on 2010/02/05 12:59:26 s: moz2-linux-slave07 TEST-PASS | chrome://mochikit/content/browser/browser/components/sessionstore/test/browser/browser_480148.js | Only one browser window should be open eventually ++DOMWINDOW == 213 (0xc4d30e8) [serial = 1841] [outer = 0xbfaf2f0] --DOCSHELL 0xbc22760 == 70 --DOCSHELL 0xbbaa0d8 == 69 ++DOMWINDOW == 214 (0xc31eef0) [serial = 1842] [outer = 0xc0f40e0] WARNING: NS_ENSURE_TRUE(doc) failed: file /builds/moz2_slave/mozilla-central-linux-debug/build/dom/base/nsDOMWindowUtils.cpp, line 786 --DOCSHELL 0xb3303d8 == 68 --DOCSHELL 0xc0850a8 == 67 TEST-UNEXPECTED-FAIL | chrome://mochikit/content/browser/browser/components/sessionstore/test/browser/browser_480148.js | Timed out
OS X 10.5.2 mozilla-central debug test mochitest-other [testfailed] Started 14:28, finished 15:05 http://tinderbox.mozilla.org/showlog.cgi?log=Firefox/1266618530.1266620661.8397.gz
http://tinderbox.mozilla.org/showlog.cgi?log=Firefox/1268376723.1268378734.26356.gz OS X 10.5.2 mozilla-central debug test mochitest-other on 2010/03/11 22:52:03 s: moz2-darwin9-slave23
http://tinderbox.mozilla.org/showlog.cgi?log=Firefox/1270587679.1270590012.14827.gz Linux mozilla-central debug test mochitest-other on 2010/04/06 14:01:19 s: moz2-linux-slave02
Summary: Intermittent failure in browser_480148.js | Timed out → Intermittent failure in browser_480148.js | Timed out, followed by many many more errors in subsequent tests due to windows remaining open
philringnalda%gmail.com http://tinderbox.mozilla.org/showlog.cgi?log=Firefox/1281460894.1281464239.19889.gz WINNT 5.2 mozilla-central debug test mochitest-other on 2010/08/10 10:21:34 s: win32-slave42 7612 ERROR TEST-UNEXPECTED-FAIL | chrome://mochikit/content/chrome/toolkit/content/tests/chrome/test_bug558403.xul | [SimpleTest/SimpleTest.js, window.onerror] An error occurred - s1 is not defined at chrome://mochikit/content/chrome/toolkit/content/tests/chrome/test_bug558403.xul:80 PROCESS-CRASH | Main app process exited normally | application crashed (minidump found) Thread 0 (crashed) PROCESS-CRASH | Main app process exited normally | application crashed (minidump found) Thread 0 (crashed) TEST-UNEXPECTED-FAIL | chrome://mochikit/content/browser/browser/components/sessionstore/test/browser/browser_480148.js | Timed out TEST-UNEXPECTED-FAIL | chrome://mochikit/content/browser/toolkit/mozapps/extensions/test/browser/browser_uninstalling.js | Timed out TEST-UNEXPECTED-FAIL | chrome://mochikit/content/browser/toolkit/mozapps/extensions/test/browser/browser_uninstalling.js | Found a tab after previous test timed out: about:addons
Not that anybody is actually looking at this, but it seems like something is going wrong somewhere else TEST-INFO | chrome://mochikit/content/browser/browser/components/sessionstore/test/browser/browser_480148.js | Console message: CoCreateInstance() FAILED. which is coming from http://mxr.mozilla.org/mozilla-central/source/gfx/layers/d3d9/Nv3DVUtils.cpp#90 At some point the tests were actually running and passing, but that's not in the recent logs.
OS: Linux → Windows Server 2003
There have been two occurrences so far today that seem to be a bit different in terms of the errors afterwards: http://tinderbox.mozilla.org/showlog.cgi?log=Firefox/1284146703.1284151472.18405.gz WINNT 5.2 mozilla-central debug test mochitest-other on 2010/09/10 12:25:03 http://tinderbox.mozilla.org/showlog.cgi?log=Firefox/1284156981.1284161347.810.gz WINNT 5.2 mozilla-central debug test mochitest-other on 2010/09/10 15:16:21 In both cases, they have a few "Found a tab after previous test timed out: about:addons" starting with browser_bug577990.js (for just a few tests), and then go into a long string of "Should be no active installs at the end of the test" starting with browser_auth.js.
Summary: Intermittent failure in browser_480148.js | Timed out, followed by many many more errors in subsequent tests due to windows remaining open → Intermittent failure in browser_480148.js | Timed out, followed by several "Found a browser window after previous test timed out"
Comment 364 through comment 367 weren't this, they were a regression since backed out.
While making sure I was right that the CoCreateInstance spew wasn't related, by looking in a successful log and noting that not only do several of them appear in this test when it passes, too, but also several appear in a great many other tests, I happened to notice the real diagnostic: In this bug's web page, cmd+f and search for -ix- I forget the details, but the slaves with -ix- in the name are the faster new ones. In that passing log on an -ix- slave, this test took 15 seconds. In the first passing log I found on an old slow slave, "Test took 29.435s to complete." That's pushing the 30 second timeout pretty hard, finishing with just over half a second to spare.
Attachment #482108 - Flags: review?(paul)
Comment on attachment 482108 [details] [diff] [review] request doubled timeout [checked in] Thanks for looking. I had another change I wanted to make to this test which might help - it currently just runs all the test cases at once as opposed to sequentially. Might not actually help the test, but would give use consistent results.
Attachment #482108 - Flags: review?(paul) → review+
Comment on attachment 482108 [details] [diff] [review] request doubled timeout [checked in] http://hg.mozilla.org/mozilla-central/rev/4f238d3d3e6a
Attachment #482108 - Attachment description: request doubled timeout → request doubled timeout [checked in]
Looks like this has quieted down and is only happening on branches that haven't merged m-c into them since the timeout extension. I'll leave it open for a little while longer, but it seems like we can close this.
Yup, I think I only left it open because I wasn't sure whether comment 539 meant "another change I want to make, using this bug, after which I'll try backing out the longer timeout" or "another change that's in some other bug."
Whatever fix Paul was thinking about, should most likely go in a new bug since looks like this timeout was only due to test taking too much (1 month without oranges is a good value). Making other changes to the same test here risks to show a new orange caused by those changes as this old one. Thus I'm resolving.
Status: NEW → RESOLVED
Closed: 15 years ago
Resolution: --- → FIXED
Whiteboard: [orange]
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Creator:
Created:
Updated:
Size: