Closed Bug 601069 Opened 15 years ago Closed 15 years ago

Some mochitest-browser-chrome tests in suite/ time out, after bug 543800

Categories

(SeaMonkey :: Testing Infrastructure, defect)

defect
Not set
normal

Tracking

(Not tracked)

VERIFIED FIXED
seamonkey2.1b2

People

(Reporter: bugzilla, Assigned: bugzilla)

References

(Blocks 1 open bug)

Details

(Keywords: regression)

Attachments

(1 file, 2 obsolete files)

The first one is "TEST-UNEXPECTED-FAIL | chrome://mochitests/content/browser/suite/browser/test/browser_page_style_menu.js | Test timed out" but others follow. See it so far on Linux and OSX, Windows hasn't run it so far. Example log (OSX): http://tinderbox.mozilla.org/showlog.cgi?log=SeaMonkey/1285903268.1285905799.29670.gz Regression range: http://hg.mozilla.org/mozilla-central/pushloghtml?fromchange=8dcc5e960d5c&tochange=669eb2d837d7 Could be the results from the changes for mochitest, perhaps we have to port some of the changes that happened since our inital port (which was done 04/2009): http://hg.mozilla.org/mozilla-central/log/6eb42326798b/browser/base/content/test/browser_page_style_menu.js The main bug to look at here could be bug 592859.
Summary: mochitest-browser-chrome tests in suite/ time out → mochitest-browser-chrome tests in suite/ intermittently times out
(In reply to comment #0) > http://tinderbox.mozilla.org/showlog.cgi?log=SeaMonkey/1285903268.1285905799.29670.gz OS X 10.5 comm-central-trunk debug test mochitest-other on 2010/09/30 20:21:08 { TEST-UNEXPECTED-FAIL | chrome://mochitests/content/browser/suite/browser/test/browser_page_style_menu.js | Test timed out TEST-UNEXPECTED-FAIL | chrome://mochitests/content/browser/suite/browser/test/browser_page_style_menu.js | Found a tab after previous test timed out: jar:file:///var/folders/nt/ntruOvnCGaqeHuQCa+uQrU+++Tw/-Tmp-/tmp3dmKdq/extensions/mochikit@mozilla.org/chrome/mochikit.jar!/content/browser/suite/browser/test/page_style_sample.html } The second error makes obvious what the cause is ;->
No longer blocks: SmTestFail
Status: NEW → RESOLVED
Closed: 15 years ago
Resolution: --- → DUPLICATE
Whiteboard: [orange]
Keywords: regression
Summary: mochitest-browser-chrome tests in suite/ intermittently times out → Some mochitest-browser-chrome tests in suite/ time out
This won't be fixed by the Makefile.in changes in bug 601113. We need to do the same as bug 592859.
Assignee: nobody → aqualon
Status: RESOLVED → REOPENED
Attachment #480109 - Flags: review?(bugspam.Callek)
Resolution: DUPLICATE → ---
Comment on attachment 480109 [details] [diff] [review] remove usage of chrome://mochikit from browser-chrome tests (In reply to comment #2) > This won't be fixed by the Makefile.in changes in bug 601113. Correct, but that change was just the start of that bug... At least, you should have commented there: no need we do double work :-/ > We need to do the same as bug 592859. Damn, some m-c tests were fixed in that bug too :-< Yet, one bug should be enough on our side. >- content.location = >- "chrome://mochikit/content/browser/suite/browser/test/page_style_sample.html"; >+ let rootDir = getRootDirectory(gTestPath); >+ content.location = rootDir + "page_style_sample.html"; Do you know why they used that getRootDirectory() instead of "usual" http://mochi.test:8888 ?
Hardware: x86 → All
Sorry serge, but I thought your bug should be about our make package-tests issue, which has nothing to do with the mochitest-browser-chrome test failures. And I saw your dupe first when I uploaded the patch here. > Do you know why they used that getRootDirectory() instead of "usual" > http://mochi.test:8888 ? Probably to be prepared for further changes of the url. You don't need to change the tests then.
(In reply to comment #4) > > Do you know why they used that getRootDirectory() instead of "usual" > > http://mochi.test:8888 ? > Probably to be prepared for further changes of the url. You don't need to > change the tests then. http://elvis314.wordpress.com/2010/10/01/mochikit-jar-changes-are-in-mozilla-central/ also contains some information about getRootDirectory().
(In reply to comment #4) So I got double-confused: 1) They started to update tests in bug 592859 then completed that in bug 543800. As I saw test updates in the latter, I didn't looked for the former :-/ And, as often, they didn't let us know about that work beforehand :-( 2) I noticed that the unjarred files are still present in the package. But I missed to check that http://mochi.test:8888 was working only because it was actually accessing them and not the jarred files :-/ *** A mochitest-chrome test to fix too: { 7729 ERROR TEST-UNEXPECTED-FAIL | chrome://mochitests/content/chrome/suite/common/tests/chrome/test_idcheck.xul | [SimpleTest/SimpleTest.js, window.onerror] An error occurred - loadLocalMailAccount is not defined at chrome://mochitests/content/chrome/suite/common/tests/chrome/test_idcheck.xul:214 } I assume it fails to load mailTestUtils.js .
Blocks: 543800
Status: REOPENED → ASSIGNED
Depends on: 592859
This also fixes test_idcheck.xul, thanks for spotting this serge!
Attachment #480109 - Attachment is obsolete: true
Attachment #480250 - Flags: review?(bugspam.Callek)
Attachment #480109 - Flags: review?(bugspam.Callek)
Comment on attachment 480250 [details] [diff] [review] remove usage of chrome://mochikit from browser-chrome tests v2 I didn't test this patch, but the list of tests is correct and the fixes look fine... >diff --git a/suite/browser/test/browser/browser_pluginnotification.js b/suite/browser/test/browser/browser_pluginnotification.js >@@ -1,9 +1,10 @@ >-const gTestRoot = "chrome://mochikit/content/browser/suite/browser/test/"; >+//const gTestRoot = "chrome://mochikit/content/browser/suite/browser/test/"; >+const gTestRoot = getRootDirectory(gTestPath); Nit: drop the commented out line. >diff --git a/suite/common/tests/chrome/test_idcheck.xul b/suite/common/tests/chrome/test_idcheck.xul >@@ -43,24 +43,31 @@ >+ >+ let scriptLoader = Cc["@mozilla.org/moz/jssubscript-loader;1"]. >+ getService(Ci.mozIJSSubScriptLoader); >+ >+ let rootDir = getRootDirectory(window.location.href); >+ scriptLoader.loadSubScript(rootDir + "/mailTestUtils.js", this); >+ Remove the "/", as is bug 543800.
Attachment #480250 - Flags: feedback+
Comment on attachment 480250 [details] [diff] [review] remove usage of chrome://mochikit from browser-chrome tests v2 >--- a/suite/browser/test/browser/browser_pluginnotification.js >+//const gTestRoot = "chrome://mochikit/content/browser/suite/browser/test/"; Nit, don't use this comment. Everything else looks good to me, I am asking serge to give this an "official" review [though he's not a peer]. Serge what I am looking for from you is: * Anything obvious in this diff that I missed. * Any other test files you have noticed that might be missing here. Thank You!
Attachment #480250 - Flags: review?(sgautherie.bz)
Attachment #480250 - Flags: review?(bugspam.Callek)
Attachment #480250 - Flags: review+
I already included the review/feedback comments into my local queue, will upload a new patch after Serge had another look at it. A grep on all test files in suite/ hasn't found other occurrences of chrome://mochikit except those needed by the test harness itself.
(In reply to comment #10) > Serge what I am looking for from you is: Are you able to review this by tomorrow evening?, if not I will probably just land it, since I did give a review, just felt it was worth a second set of eyes.
This is the updated version of the patch, I won't have time to update the patch until Sunday evening if Serge finds any other issues with the patch. Feel free to do the update yourself to get it checked-in.
Attachment #480250 - Attachment is obsolete: true
Attachment #481801 - Flags: review?
Attachment #480250 - Flags: review?(sgautherie.bz)
Attachment #481801 - Flags: review? → review?(sgautherie.bz)
Comment on attachment 481801 [details] [diff] [review] remove usage of chrome://mochikit from browser-chrome tests v3 [Checked in: See comment 15] (In reply to comment #10) > * Anything obvious in this diff that I missed. > * Any other test files you have noticed that might be missing here. My comment 8 feedback already covers (most of) these ;-) (In reply to comment #13) > Feel free to do the update yourself to get it checked-in. I found two (non-code) nits which I'll fix, then I'll push the patch. NB: This time, I took time to run the tests (as Callek asked for my "review") and noticed issues with 2 of them, but I believe they are unrelated and will file separate bugs.
Attachment #481801 - Flags: review?(sgautherie.bz) → review+
Comment on attachment 481801 [details] [diff] [review] remove usage of chrome://mochikit from browser-chrome tests v3 [Checked in: See comment 15] http://hg.mozilla.org/comm-central/rev/2994d32dff9c
Attachment #481801 - Attachment description: remove usage of chrome://mochikit from browser-chrome tests v3 → remove usage of chrome://mochikit from browser-chrome tests v3 [Checked in: See comment 15]
Status: ASSIGNED → RESOLVED
Closed: 15 years ago15 years ago
Flags: in-testsuite+
Resolution: --- → FIXED
Target Milestone: --- → seamonkey2.1b2
(In reply to comment #14) > NB: ... I ... noticed issues with 2 of them, but I believe they are unrelated > and will file separate bugs. I filed bug 603028 and bug 603031. Bug 602633 fixed DOMi hang in test_idcheck.xul, while I was reviewing ;-)
Summary: Some mochitest-browser-chrome tests in suite/ time out → Some mochitest-browser-chrome tests in suite/ time out, after bug 543800
(In reply to comment #14) > (In reply to comment #13) > > Feel free to do the update yourself to get it checked-in. > > I found two (non-code) nits which I'll fix, then I'll push the patch. > I'm curious what these are (without having to compare checked in cset to the diff here) Thanks though!
(In reply to comment #16) > Bug 602633 fixed DOMi hang in test_idcheck.xul, while I was reviewing ;-) Yes, been a bit slower than usual with updating this while I was at Mozilla...
(In reply to comment #17) > I'm curious what these are (without having to compare checked in cset to the > diff here) 2-3 whitespaces, and const before var. ***** V.Fixed per http://tinderbox.mozilla.org/showlog.cgi?log=SeaMonkey/1286608666.1286610806.16791.gz Linux comm-central-trunk debug test mochitest-other on 2010/10/09 00:17:46
Status: RESOLVED → VERIFIED
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Created:
Updated:
Size: