Bug 90594 – REGRESSION(r121891): It made two tests fail on Qt and on Lion

• Status: RESOLVED
• Resolution: FIXED
• Component: Tools / Tests

[Kristóf Kosztyó]

After the r121891 these two tests start to fail.

http/tests/inspector/console-cd-completions.html:

--- /ramdisk/qt-linux-64-release/build/layout-test-results/http/tests/inspector/console-cd-completions-expected.txt 
+++ /ramdisk/qt-linux-64-release/build/layout-test-results/http/tests/inspector/console-cd-completions-actual.txt 
@@ -1,6 +1,6 @@
 Test that completions in the context of an iframe with a different origin will result in names of its global variables. Test passes if all global variables are found among completions AND there are NO console messages. Bug 65457.
 
 
-myGlobalVar
-myGlobalFunction
+NOT FOUND: myGlobalVar
+NOT FOUND: myGlobalFunction

http/tests/inspector/console-cd.html:

--- /ramdisk/qt-linux-64-release/build/layout-test-results/http/tests/inspector/console-cd-expected.txt 
+++ /ramdisk/qt-linux-64-release/build/layout-test-results/http/tests/inspector/console-cd-actual.txt 
@@ -1,5 +1,5 @@
 Tests that console evaluation can be performed in an iframe context.Bug 19872.
 
 
-foo = 2011
+foo = ReferenceError: Can't find variable: foo

[Csaba Osztrogonác]

No, it isn't Qt specific bug, it fails on Apple Lion too. (http://build.webkit.org/builders/Apple%20Lion%20Release%20WK1%20%28Tests%29/builds/844 - It isn't finished, but you can see these failing tests in the stdio) It's strange for me that layout tests takes more  than an hour on the Apple bots with 8 paralel threads. It takes for Qt port 18 minutes on 1 thread.

[Csaba Osztrogonác]

It is skipped on Qt by https://trac.webkit.org/changeset/121894 . Please unskip it with the proper fix. Thanks.

[Peter Wang]

It's weird. I can manually do the test well. I need to investigate the DRT test.

[Csaba Osztrogonác]

They pass on Qt now, I unskipped them - r130733