[V8][Chromium] Reenable dedicated worker layout tests

• Component: WebCore Misc.
• Status: RESOLVED FIXED

[Dmitry Lomov]

Now that dedicated workers run in-process, DumpRenderTree supports them.

[Dmitry Lomov]

Created attachment 119100 [details]
Fix

[David Levin]

Comment on attachment 119100 [details]
Fix

View in context: https://bugs.webkit.org/attachment.cgi?id=119100&action=review

> LayoutTests/ChangeLog:11
> +        * platform/chromium/fast/workers/worker-script-error-expected.txt: Added.

It would be great to explain why we have Chromium specific results (in the ChangeLog).

> LayoutTests/platform/chromium/test_expectations.txt:157
> +WONTFIX SKIP : fast/workers/worker-crash-with-invalid-location.html = TEXT

I don't understand this last line.

> LayoutTests/platform/chromium/test_expectations.txt:165
> +WONTFIX SKIP : fast/workers/worker-multi-port.html = CRASH

Odd that we crash when we don't support it. That doesn't seem great.

> LayoutTests/platform/chromium/test_expectations.txt:169
> +

extra blank line.

[Dmitry Lomov]

Comment on attachment 119100 [details]
Fix

View in context: https://bugs.webkit.org/attachment.cgi?id=119100&action=review

>> LayoutTests/platform/chromium/test_expectations.txt:157
>> +WONTFIX SKIP : fast/workers/worker-crash-with-invalid-location.html = TEXT
> 
> I don't understand this last line.

This test tests shared workers.

>> LayoutTests/platform/chromium/test_expectations.txt:165
>> +WONTFIX SKIP : fast/workers/worker-multi-port.html = CRASH
> 
> Odd that we crash when we don't support it. That doesn't seem great.

I'll have a bug opened to fix this (and replace WONTFIX with bug ref.)

[Dmitry Lomov]

Created attachment 119101 [details]
CR feedback

[WebKit Review Bot]

Comment on attachment 119101 [details]
CR feedback

Attachment 119101 [details] did not pass chromium-ews (chromium-xvfb):
Output: http://queues.webkit.org/results/10871027

New failing tests:
fast/workers/worker-script-error.html

[Dmitry Lomov]

Created attachment 119115 [details]
worker-script-error ignored

Filed https://bugs.webkit.org/show_bug.cgi?id=74466

[Dmitry Lomov]

Created attachment 119117 [details]
Rebased

[WebKit Review Bot]

Comment on attachment 119117 [details]
Rebased

Rejecting attachment 119117 [details] from commit-queue.

Failed to run "['/mnt/git/webkit-commit-queue/Tools/Scripts/webkit-patch', '--status-host=queues.webkit.org', '-..." exit_code: 1

ERROR: /mnt/git/webkit-commit-queue/LayoutTests/ChangeLog neither lists a valid reviewer nor contains the string "Unreviewed" or "Rubber stamp" (case insensitive).

Full output: http://queues.webkit.org/results/10874040

[WebKit Review Bot]

Comment on attachment 119117 [details]
Rebased

Clearing flags on attachment: 119117

Committed r102729: <http://trac.webkit.org/changeset/102729>

[WebKit Review Bot]

All reviewed patches have been landed.  Closing bug.

[Li Yin]

I found all the shared workers tests were skipped on chromium.
In the file “LayoutTests/platform/chromium/test_exceptions.txt”, there is a note “// test_shell does not support shared workers”.

So, I have some questions about that:
1.How does chromium test the shared worker feature?
2.If test shell does not support it, why not fix it? Or maybe somebody be doing that?

Thanks very much in advance.

[David Levin]

(In reply to comment #11)
> I found all the shared workers tests were skipped on chromium.
> In the file “LayoutTests/platform/chromium/test_exceptions.txt”, there is a note “// test_shell does not support shared workers”.
> 
> So, I have some questions about that:
> 1.How does chromium test the shared worker feature?

I don't understand the question. Perhaps you are asking how chromium test shared workers. If so, search the chromium code for the name of one of of the shared worker test files to find the answer. (Basically it runs the test in the browser and simulates the necessary bits of the layout test controller.)

> 2.If test shell does not support it, why not fix it? 

Shared workers are separate processes in chromium. They need to be accessible across processes at the very least due to the fact that they are shared and a single domain may occur in different renderer processes.

Now, the question remains should they always be in a separate process. So far the answer has been yes due to various reason with [e-mail] being the strongest advocate of this.


> Or maybe somebody be doing that?

Nope.