[Webkit-unassigned] [Bug 105330] REGRESSION(r137607): resource load client callbacks are not called for the main resource when loading HTML string

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=105330





--- Comment #9 from Brady Eidson <beidson at apple.com>  2012-12-19 12:04:37 PST ---
(In reply to comment #8)
> (In reply to comment #7)
> > (In reply to comment #6)
> > > (From update of attachment 180188 [details] [details] [details])
> > > View in context: https://bugs.webkit.org/attachment.cgi?id=180188&action=review
> > > 
> > > > Source/WebCore/loader/MainResourceLoader.cpp:697
> > > >  unsigned long MainResourceLoader::identifier() const
> > > >  {
> > > > +    if (m_identifier)
> > > > +        return m_identifier;
> > > >      if (ResourceLoader* resourceLoader = loader())
> > > >          return resourceLoader->identifier();
> > > 
> > > This breaks the "setIdentifier" API on ResourceLoader that WebKit2 relies on.
> > 
> > Now that we've made MainResourceLoader no longer be a ResourceLoader, we're going to start getting in to trouble as we diverge the two.  Having two copies of the identifier seems like a great first step towards breaking things.
> 
> MainResourceLoader::m_identifier should only be used when it is a SubstituteData load. In that case, there should not be a ResourceLoader. So there should still only be one identifier per load.
> 
> Should I consider an assert to that effect here?

Yes please.

On that note, I might also feel better about it having a different name, like in cgarcia's patch.

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