[Webkit-unassigned] [Bug 51686] [GTK] Closing a window during an onload event can trigger serious GLib warnings

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Tue Dec 28 15:04:50 PST 2010


https://bugs.webkit.org/show_bug.cgi?id=51686





--- Comment #3 from Martin Robinson <mrobinson at webkit.org>  2010-12-28 15:04:50 PST ---
(From update of attachment 77573)
View in context: https://bugs.webkit.org/attachment.cgi?id=77573&action=review

Thanks for the review!

>> WebKit/gtk/ChangeLog:13
>> +
> 
> If this were the problem just reffing the view before emitting the signal would be enough; from our talk I understand that it must survive for a while after this function is finished, which is different. Am I wrong?

I think my entry doesn't explain the problem fully. close-web-view essentially kills the view (because the client may remove the last reference), but the caller (FrameLoader) depends on the view still existing.

>> WebKit/gtk/WebCoreSupport/ChromeClientGtk.cpp:255
>> +    gboolean isHandled = FALSE;
> 
> No need to initialize this, and perhaps you could call it 'dummy'.

I'm fine with removing the initialization (I prefer to initialize things unconditionally), but I think isHandled better explains what it is and why I'm ignoring it.

>> WebKit/gtk/WebCoreSupport/ChromeClientGtk.cpp:267
>>  
> 
> Aren't you checking for m_closeSoonTimer right before?

Oh right. I separated the check out after adding it to line 263 and forgot to remove the old one. Thanks!

>> WebKit/gtk/WebCoreSupport/ChromeClientGtk.cpp:268
>> +    m_webView->priv->corePage->setGroupName("");
> 
> And this?

It's necessary so that JavaScript can't find the page. It originates from the other ports. I'll leave a comment.

>> WebKit/gtk/WebCoreSupport/ChromeClientGtk.cpp:272
>> +    // signal, but our caller my need to send more signals web view. For
> 
> 'more signals web view'? Missing 'to the'?

Oh yeah. Should be "more signals to the web view"

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.



More information about the webkit-unassigned mailing list