Bug 180760

Summary: Layout Test fast/loader/redirect-to-invalid-url-using-meta-refresh-calls-policy-delegate.html is flaky
Product: WebKit Reporter: Matt Lewis <jlewis3>
Component: Tools / TestsAssignee: Nobody <webkit-unassigned>
Status: NEW ---    
Severity: Normal CC: achristensen, aestes, ap, jacob_uphoff, lforschler, ryanhaddad, tsavell, webkit-bug-importer
Priority: P2 Keywords: InRadar
Version: WebKit Nightly Build   
Hardware: Unspecified   
OS: Unspecified   

Description Matt Lewis 2017-12-13 11:20:33 PST
The following layout test is flaky on [insert platform]

fast/loader/redirect-to-invalid-url-using-meta-refresh-calls-policy-delegate.html

Probable cause:

I think that this revision is a good place to start as it is dealing with the Policy Delegate a few revisions before the first failure of the test. https://trac.webkit.org/changeset/224758/webkit

Flakiness Dashboard:

https://webkit-test-results.webkit.org/dashboards/flakiness_dashboard.html#showAllRuns=true&tests=fast%2Floader%2Fredirect-to-invalid-url-using-meta-refresh-calls-policy-delegate.html

Results:
https://build.webkit.org/results/Apple%20Sierra%20Release%20WK2%20(Tests)/r225841%20(6357)/results.html
https://build.webkit.org/builders/Apple%20Sierra%20Release%20WK2%20(Tests)/builds/6357
Comment 1 Matt Lewis 2017-12-13 11:21:47 PST
(In reply to Matt Lewis from comment #0)
> The following layout test is flaky on [insert platform]
> 
> fast/loader/redirect-to-invalid-url-using-meta-refresh-calls-policy-delegate.
> html
> 
> Probable cause:
> 
> I think that this revision is a good place to start as it is dealing with
> the Policy Delegate a few revisions before the first failure of the test.
> https://trac.webkit.org/changeset/224758/webkit
> 
> Flakiness Dashboard:
> 
> https://webkit-test-results.webkit.org/dashboards/flakiness_dashboard.
> html#showAllRuns=true&tests=fast%2Floader%2Fredirect-to-invalid-url-using-
> meta-refresh-calls-policy-delegate.html
> 
> Results:
> https://build.webkit.org/results/Apple%20Sierra%20Release%20WK2%20(Tests)/
> r225841%20(6357)/results.html
> https://build.webkit.org/builders/Apple%20Sierra%20Release%20WK2%20(Tests)/
> builds/6357
platforms would be macOS WK1, High Sierra Leaks and most recently High Sierra Release WK2
Comment 2 Ryan Haddad 2018-03-27 17:48:59 PDT
--- /Volumes/Data/slave/sierra-debug-tests-wk1/build/layout-test-results/fast/loader/redirect-to-invalid-url-using-meta-refresh-calls-policy-delegate-expected.txt
+++ /Volumes/Data/slave/sierra-debug-tests-wk1/build/layout-test-results/fast/loader/redirect-to-invalid-url-using-meta-refresh-calls-policy-delegate-actual.txt
@@ -1,4 +1,3 @@
-Policy delegate: attempt to load http://a=a&b=b/ with navigation type 'other'
 Tests that we do not redirect to an invalid URL initiated by <meta http-equiv="refresh">. This test PASSED if you see an entry in the dumped frame load callbacks of the form: "willPerformClientRedirectToURL: http://A=a&B=b" followed by "didCancelClientRedirectForFrame".
 
 Note, this test must be run in DumpRenderTree.
Comment 3 Ryan Haddad 2018-03-27 17:49:34 PDT
This test fails alongside fast/loader/redirect-to-invalid-url-using-meta-refresh-disallowed-async-delegates.html
Comment 4 Ryan Haddad 2018-03-27 17:51:19 PDT
(In reply to Ryan Haddad from comment #3)
> This test fails alongside
> fast/loader/redirect-to-invalid-url-using-meta-refresh-disallowed-async-
> delegates.html

https://webkit-test-results.webkit.org/dashboards/flakiness_dashboard.html#showAllRuns=true&tests=fast%2Floader%2Fredirect-to-invalid-url-using-meta-refresh

All recent failures on macOS are seen with debug WK1 bots.
Comment 6 Truitt Savell 2018-08-17 14:37:27 PDT
Text Diff of fast/loader/redirect-to-invalid-url-using-meta-refresh-calls-policy-delegate.html:

--- /Volumes/Data/slave/sierra-debug-tests-wk1/build/layout-test-results/fast/loader/redirect-to-invalid-url-using-meta-refresh-calls-policy-delegate-expected.txt
+++ /Volumes/Data/slave/sierra-debug-tests-wk1/build/layout-test-results/fast/loader/redirect-to-invalid-url-using-meta-refresh-calls-policy-delegate-actual.txt
@@ -1,4 +1,3 @@
-Policy delegate: attempt to load http://a=a&b=b/ with navigation type 'other'
 Tests that we do not redirect to an invalid URL initiated by <meta http-equiv="refresh">. This test PASSED if you see an entry in the dumped frame load callbacks of the form: "willPerformClientRedirectToURL: http://A=a&B=b" followed by "didCancelClientRedirectForFrame".
 
 Note, this test must be run in DumpRenderTree.

text Diff of fast/loader/redirect-to-invalid-url-using-meta-refresh-disallowed-async-delegates.html:

--- /Volumes/Data/slave/sierra-debug-tests-wk1/build/layout-test-results/fast/loader/redirect-to-invalid-url-using-meta-refresh-disallowed-async-delegates-expected.txt
+++ /Volumes/Data/slave/sierra-debug-tests-wk1/build/layout-test-results/fast/loader/redirect-to-invalid-url-using-meta-refresh-disallowed-async-delegates-actual.txt
@@ -1,3 +1,4 @@
+Policy delegate: attempt to load http://a=a&b=b/ with navigation type 'other'
 main frame - didFinishDocumentLoadForFrame
 main frame - didHandleOnloadEventsForFrame
 main frame - willPerformClientRedirectToURL: http://A=a%25B=b 


Issue causing these two tests to fail in tandem is the first line of output missing from the first test ran by test runner and that line appearing in the second tests output.
Comment 7 Alexey Proskuryakov 2020-01-24 23:02:50 PST
*** Bug 206668 has been marked as a duplicate of this bug. ***
Comment 8 Alexey Proskuryakov 2020-01-24 23:03:11 PST
<rdar://problem/58837324>
Comment 9 Ryan Haddad 2020-01-28 09:23:00 PST
*** Bug 206668 has been marked as a duplicate of this bug. ***