Bug 69257 – [Qt] inspector/styles/styles-disable-inherited.html make the next test fail after r96257

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

[Fehér Zsolt]

Diff:
--- /ramdisk/qt-linux-32-release-qt470/build/layout-test-results/inspector/styles/styles-disable-then-change-expected.txt	2011-10-03 06:06:45.774659416 -0700
+++ /ramdisk/qt-linux-32-release-qt470/build/layout-test-results/inspector/styles/styles-disable-then-change-actual.txt	2011-10-03 06:06:45.774659416 -0700
@@ -1,18 +1,4 @@
+FAIL: Timed out waiting for notifyDone to be called
 Tests that changing a disabled property enables it as well.
 
-Before disable
-[expanded] element.style { ()
-font-weight: bold;
-
-
-After disable
-[expanded] element.style { ()
-/-- overloaded --/ /-- disabled --/ font-weight: bold;
-
-
-After change
-[expanded] element.style { ()
-color: green;
-
-

[Csaba Osztrogonác]

This is a flakey test. It fails when you run all tests, but pass if you run it on itself.

[Csaba Osztrogonác]

Maybe this bug is related to https://bugs.webkit.org/show_bug.cgi?id=66223

[Csaba Osztrogonác]

(In reply to comment #2)
> Maybe this bug is related to https://bugs.webkit.org/show_bug.cgi?id=66223
I checked it, it seems it is unrelated.

[Csaba Osztrogonác]

You can easily reproduce it on Qt:
$ Tools/Scripts/old-run-webkit-tests inspector/styles/styles-disable-inherited.html inspector/styles/styles-disable-then-change.html

[Csaba Osztrogonác]

I don't know how, but http://trac.webkit.org/changeset/96257 broke this test.

[Ryosuke Niwa]

(In reply to comment #5)
> I don't know how, but http://trac.webkit.org/changeset/96257 broke this test.

It seems very unlikely. Given that this test had been flaky, I don't intend to investigate the failure.

[Csaba Osztrogonác]

(In reply to comment #6)
> (In reply to comment #5)
> > I don't know how, but http://trac.webkit.org/changeset/96257 broke this test.
> 
> It seems very unlikely. Given that this test had been flaky, I don't intend to investigate the failure.

But it's true. I can reproduce the bug on r96257, but I can't on r96256.

[Ryosuke Niwa]

(In reply to comment #7)
> (In reply to comment #6)
> > (In reply to comment #5)
> > > I don't know how, but http://trac.webkit.org/changeset/96257 broke this test.
> > 
> > It seems very unlikely. Given that this test had been flaky, I don't intend to investigate the failure.
> 
> But it's true. I can reproduce the bug on r96257, but I can't on r96256.

But it had been flaky anyway so I don't think it's caused by r96257. r96257 probably changed the condition under which the failure occurs. On the bright side, you can now debug it if it reliably fails.

[Csaba Osztrogonác]

inspector/styles/styles-disable-inherited.html(In reply to comment #4)
> You can easily reproduce it on Qt:
> $ Tools/Scripts/old-run-webkit-tests inspector/styles/styles-disable-inherited.html inspector/styles/styles-disable-then-change.html

After skipping inspector/styles/styles-disable-then-change.html,
the next test fails: inspector/styles/styles-disable-then-delete.html

You can reprocude it easily:
$ Tools/Scripts/old-run-webkit-tests inspector/styles/styles-disable-inherited.html inspector/styles/styles-disable-then-delete.html

[Csaba Osztrogonác]

I painted the bots green: http://trac.webkit.org/changeset/96801

But we still need a proper fix for it.

[Csaba Osztrogonác]

They works now, so I unskipped them - http://trac.webkit.org/changeset/107560(In reply to comment #10)
> I painted the bots green: http://trac.webkit.org/changeset/96801
> 
> But we still need a proper fix for it.

Everything works now, so I unskipped the tests - http://trac.webkit.org/changeset/107560