Bug 206474 – Regression r254652: fast/xmlhttprequest/xmlhttprequest-multiple-sync-xhr-during-unload.html fails on windows

• Component: Tools / Tests
• Status: RESOLVED FIXED

[Aakash Jain]

r254652 added test fast/xmlhttprequest/xmlhttprequest-multiple-sync-xhr-during-unload.html which is consistently failing on windows.

Results database: https://results.webkit.org/?suite=layout-tests&test=fast%2Fxmlhttprequest%2Fxmlhttprequest-multiple-sync-xhr-during-unload.html

[Aakash Jain]

e.g.: https://build.webkit.org/results/Apple%20Win%2010%20Debug%20(Tests)/r254663%20(3323)/results.html

--- /home/buildbot/worker/win10-debug-tests/build/layout-test-results/fast/xmlhttprequest/xmlhttprequest-multiple-sync-xhr-during-unload-expected.txt
+++ /home/buildbot/worker/win10-debug-tests/build/layout-test-results/fast/xmlhttprequest/xmlhttprequest-multiple-sync-xhr-during-unload-actual.txt
@@ -1,9 +1,9 @@
 frame "<!--frame1-->" - has 1 onunload handler(s)
-CONSOLE MESSAGE: line 9: XMLHttpRequest cannot load xmlhttprequest-responsetype-json.json?0.
-CONSOLE MESSAGE: line 9: XMLHttpRequest cannot load xmlhttprequest-responsetype-json.json?1.
-CONSOLE MESSAGE: line 9: XMLHttpRequest cannot load xmlhttprequest-responsetype-json.json?2.
-CONSOLE MESSAGE: line 9: XMLHttpRequest cannot load xmlhttprequest-responsetype-json.json?3.
-CONSOLE MESSAGE: line 9: XMLHttpRequest cannot load xmlhttprequest-responsetype-json.json?4.
+CONSOLE MESSAGE: line 9: XMLHttpRequest cannot load xmlhttprequest-responsetype-json.json
+CONSOLE MESSAGE: line 9: XMLHttpRequest cannot load xmlhttprequest-responsetype-json.json
+CONSOLE MESSAGE: line 9: XMLHttpRequest cannot load xmlhttprequest-responsetype-json.json
+CONSOLE MESSAGE: line 9: XMLHttpRequest cannot load xmlhttprequest-responsetype-json.json
+CONSOLE MESSAGE: line 9: XMLHttpRequest cannot load xmlhttprequest-responsetype-json.json

[Aakash Jain]

r254652 might have also caused http/tests/websocket/tests/hybi/simple-wss.html to fail, not 100% sure at the moment.

[David Kilzer (:ddkilzer)]

*** Bug 206515 has been marked as a duplicate of this bug. ***

[Chris Dumez]

Looks to me that we need a windows-specific baseline.

[Ryan Haddad]

The test was marked as failing in https://trac.webkit.org/changeset/254851/webkit

[Chris Dumez]

(In reply to Ryan Haddad from comment #5)
> The test was marked as failing in
> https://trac.webkit.org/changeset/254851/webkit

This is not a failure, it just needs a windows baseline:
--- /home/buildbot/worker/win10-debug-tests/build/layout-test-results/fast/xmlhttprequest/xmlhttprequest-multiple-sync-xhr-during-unload-expected.txt
+++ /home/buildbot/worker/win10-debug-tests/build/layout-test-results/fast/xmlhttprequest/xmlhttprequest-multiple-sync-xhr-during-unload-actual.txt
@@ -1,9 +1,9 @@
 frame "<!--frame1-->" - has 1 onunload handler(s)
-CONSOLE MESSAGE: line 9: XMLHttpRequest cannot load xmlhttprequest-responsetype-json.json?0.
-CONSOLE MESSAGE: line 9: XMLHttpRequest cannot load xmlhttprequest-responsetype-json.json?1.
-CONSOLE MESSAGE: line 9: XMLHttpRequest cannot load xmlhttprequest-responsetype-json.json?2.
-CONSOLE MESSAGE: line 9: XMLHttpRequest cannot load xmlhttprequest-responsetype-json.json?3.
-CONSOLE MESSAGE: line 9: XMLHttpRequest cannot load xmlhttprequest-responsetype-json.json?4.
+CONSOLE MESSAGE: line 9: XMLHttpRequest cannot load xmlhttprequest-responsetype-json.json
+CONSOLE MESSAGE: line 9: XMLHttpRequest cannot load xmlhttprequest-responsetype-json.json
+CONSOLE MESSAGE: line 9: XMLHttpRequest cannot load xmlhttprequest-responsetype-json.json
+CONSOLE MESSAGE: line 9: XMLHttpRequest cannot load xmlhttprequest-responsetype-json.json
+CONSOLE MESSAGE: line 9: XMLHttpRequest cannot load xmlhttprequest-responsetype-json.json
 Makes sure that we throw when doing sync XHRs during unload.
 
 On success, you will see a series of "PASS" messages, followed by "TEST COMPLETE".

[Ryan Haddad]

(In reply to Chris Dumez from comment #6)
> (In reply to Ryan Haddad from comment #5)
> > The test was marked as failing in
> > https://trac.webkit.org/changeset/254851/webkit
> 
> This is not a failure, it just needs a windows baseline:
> ---
> /home/buildbot/worker/win10-debug-tests/build/layout-test-results/fast/
> xmlhttprequest/xmlhttprequest-multiple-sync-xhr-during-unload-expected.txt
> +++
> /home/buildbot/worker/win10-debug-tests/build/layout-test-results/fast/
> xmlhttprequest/xmlhttprequest-multiple-sync-xhr-during-unload-actual.txt
> @@ -1,9 +1,9 @@
>  frame "<!--frame1-->" - has 1 onunload handler(s)
> -CONSOLE MESSAGE: line 9: XMLHttpRequest cannot load
> xmlhttprequest-responsetype-json.json?0.
> -CONSOLE MESSAGE: line 9: XMLHttpRequest cannot load
> xmlhttprequest-responsetype-json.json?1.
> -CONSOLE MESSAGE: line 9: XMLHttpRequest cannot load
> xmlhttprequest-responsetype-json.json?2.
> -CONSOLE MESSAGE: line 9: XMLHttpRequest cannot load
> xmlhttprequest-responsetype-json.json?3.
> -CONSOLE MESSAGE: line 9: XMLHttpRequest cannot load
> xmlhttprequest-responsetype-json.json?4.
> +CONSOLE MESSAGE: line 9: XMLHttpRequest cannot load
> xmlhttprequest-responsetype-json.json
> +CONSOLE MESSAGE: line 9: XMLHttpRequest cannot load
> xmlhttprequest-responsetype-json.json
> +CONSOLE MESSAGE: line 9: XMLHttpRequest cannot load
> xmlhttprequest-responsetype-json.json
> +CONSOLE MESSAGE: line 9: XMLHttpRequest cannot load
> xmlhttprequest-responsetype-json.json
> +CONSOLE MESSAGE: line 9: XMLHttpRequest cannot load
> xmlhttprequest-responsetype-json.json
>  Makes sure that we throw when doing sync XHRs during unload.
>  
>  On success, you will see a series of "PASS" messages, followed by "TEST
> COMPLETE".
Understood, just wanted to call out the fact that it was marked as failing via the dupe.

[Alexey Proskuryakov]

Out of curiosity, why is it OK for Windows to have this difference? The diff makes it look like the test is not doing the expected work.

[Chris Dumez]

(In reply to Alexey Proskuryakov from comment #8)
> Out of curiosity, why is it OK for Windows to have this difference? The diff
> makes it look like the test is not doing the expected work.

It looks to me that the search query is getting stripped on windows before the URL gets printed.

The test still says:
PASS xhrExceptionCount is 5

So to me it does look like the test is passing. The only difference is in console logging.

[Chris Dumez]

(In reply to Chris Dumez from comment #9)
> (In reply to Alexey Proskuryakov from comment #8)
> > Out of curiosity, why is it OK for Windows to have this difference? The diff
> > makes it look like the test is not doing the expected work.
> 
> It looks to me that the search query is getting stripped on windows before
> the URL gets printed.
> 
> The test still says:
> PASS xhrExceptionCount is 5
> 
> So to me it does look like the test is passing. The only difference is in
> console logging.

Code looks like:
    String messageEnd = error.isAccessControl() ? " due to access control checks."_s : "."_s;
    context.addConsoleMessage(MessageSource::JS, MessageLevel::Error, makeString(messageStart, error.failingURL().string(), messageEnd));

For some reason, it seems error.failingURL() is missing the query parameter.

[Chris Dumez]

Committed r254860: <https://trac.webkit.org/changeset/254860>

[Radar WebKit Bug Importer]

<rdar://problem/58763098>

[Alexey Proskuryakov]

That seems like something to fix (at least eventually), so that we don't have to maintain separate expectations.