Bug 88780 – Web Inspector: Reenabled tests in r119962 still fails

• Component: Web Inspector (Deprecated)
• Status: RESOLVED INVALID

[Csaba Osztrogonác]

r119962 reenabled tests after implementing breaking from native callback, but tests are still failing.

[Konrad Piascik]

Is there a concept of V8 and JSC specific tests?

[Csaba Osztrogonác]

(In reply to comment #1)
> Is there a concept of V8 and JSC specific tests?

As far as I know the defcato rule is that platform independent expected
files(next to the test) mean JSC specific and expected files in
platform/chromium mean V8 specific results.

[Csaba Osztrogonác]

Skipped on Qt again until proper fix to paint the bot green - http://trac.webkit.org/changeset/119976

[Peter Wang]

Working on it.

[Peter Wang]

Created attachment 147061 [details]
Patch

[Peter Wang]

(In reply to comment #5)
> Created an attachment (id=147061) [details]
> Patch

There is a bug in xhr-breakpoints.html. The first parameter of "_setBreakpoint(url, bool)" should be assigned as "null" rather than an empty string, since in this function has a conditional statement "if (!url) ".
Other test cases is ok.

[Pavel Feldman]

(In reply to comment #6)
> (In reply to comment #5)
> > Created an attachment (id=147061) [details] [details]
> > Patch
> 
> There is a bug in xhr-breakpoints.html. The first parameter of "_setBreakpoint(url, bool)" should be assigned as "null" rather than an empty string, since in this function has a conditional statement "if (!url) ".
> Other test cases is ok.

!"" == true

[Pavel Feldman]

Comment on attachment 147061 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=147061&action=review

> LayoutTests/inspector/debugger/xhr-breakpoints.html:44
> +            pane._setBreakpoint(null, true);

This should be "". "" stands for any XHR. Other than that, lgtm.

[Peter Wang]

(In reply to comment #8)
> (From update of attachment 147061 [details])
> View in context: https://bugs.webkit.org/attachment.cgi?id=147061&action=review
> 
> > LayoutTests/inspector/debugger/xhr-breakpoints.html:44
> > +            pane._setBreakpoint(null, true);
> 
> This should be "". "" stands for any XHR. Other than that, lgtm.

Yes, I've investigated the related code and noticed it. But it's wired that Qt browser works well if I manually set breakpoint for "Any XHR", but failed to DRT test.:(
The only different is that "_setBreakpoint()" is invoked by a click event in my manual operation and in DRT case it is directly called with parameters.

[Peter Wang]

Created attachment 147262 [details]
test result after enlarge the value of m_timeout

I assigned "m_timeout" of "Tools/DumpRenderTree/qt/LayoutTestControllerQt.cpp" to 90000, the test result seems more close to expected. I have no idea should I just increase its value or write this case to "Layouttests/platform/qt/Skipped"?

[Peter Wang]

And if I split the case into three according to test suites. Every cases can pass the test.  
Pavel, do you think it's a option to split the case to let Qt Browser walk around timeout problem?

[Pavel Feldman]

(In reply to comment #11)
> And if I split the case into three according to test suites. Every cases can pass the test.  
> Pavel, do you think it's a option to split the case to let Qt Browser walk around timeout problem?

I think splitting the test will be masking the issue. I'd leave it skipped and file appropriate bug.

[Peter Wang]

Created attachment 147474 [details]
Patch

[Peter Wang]

I've filed https://bugs.webkit.org/show_bug.cgi?id=89056 to record the problem of "xhr-breakpoints.html". Now only this case is left in Skipped of Qt Browser.
And I'm sure this case can pass the test if we walk around the timeout problem.
Csaba and Pavel, do you think the result is acceptable?

[WebKit Review Bot]

Comment on attachment 147474 [details]
Patch

Attachment 147474 [details] did not pass chromium-ews (chromium-xvfb):
Output: http://queues.webkit.org/results/12966009

New failing tests:
inspector/debugger/step-through-event-listeners.html
inspector/debugger/event-listener-breakpoints.html

[WebKit Review Bot]

Created attachment 147562 [details]
Archive of layout-test-results from ec2-cr-linux-01

The attached test failures were seen while running run-webkit-tests on the chromium-ews.
Bot: ec2-cr-linux-01  Port: <class 'webkitpy.common.config.ports.ChromiumXVFBPort'>  Platform: Linux-2.6.35-28-virtual-x86_64-with-Ubuntu-10.10-maverick

[Peter Wang]

(In reply to comment #14)
> I've filed https://bugs.webkit.org/show_bug.cgi?id=89056 to record the problem of "xhr-breakpoints.html". Now only this case is left in Skipped of Qt Browser.
> And I'm sure this case can pass the test if we walk around the timeout problem.
> Csaba and Pavel, do you think the result is acceptable?

(In reply to comment #16)
> Created an attachment (id=147562) [details]
> Archive of layout-test-results from ec2-cr-linux-01
> 
> The attached test failures were seen while running run-webkit-tests on the chromium-ews.
> Bot: ec2-cr-linux-01  Port: <class 'webkitpy.common.config.ports.ChromiumXVFBPort'>  Platform: Linux-2.6.35-28-virtual-x86_64-with-Ubuntu-10.10-maverick

Sorry, I cannot find anything related with my test cases? :(