[Webkit-unassigned] [Bug 27655] [v8] cache v8 strings when converting from webcore string to v8 string

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=27655





--- Comment #16 from anton muhin <antonm at chromium.org>  2009-07-30 07:02:43 PDT ---
David, thanks a lot!

(In reply to comment #14)
> (From update of attachment 33769 [details])
> Last thing to fix....(the paths for the files).
> 
> > Index: third_party/WebKit/WebCore/ChangeLog
> You seem to be creating these patches from outside of your WebKit enlistment.
> 
> These lines should start with WebCore (or else committing them is going to be
> annoying).
> 
> If you create the patch when you are in the third_party/WebKit directory this
> should all work. Alternately, you could just carefully edit the patch by hand.

svn-create-patch doesn't quite work for me (or I just misuse it).  Tried to
manually fix it, sorry if missed anything.

> 
> > ===================================================================
> > --- third_party/WebKit/WebCore/ChangeLog	(revision 46575)
> > +++ third_party/WebKit/WebCore/ChangeLog	(working copy)
> > @@ -1,3 +1,24 @@
> > +2009-07-30  Anton Muhin  <antonm at chromium.org>
> > +
> > +        Reviewed by NOBODY (OOPS!).
> > +
> > +        Cache v8 strings when converting from WebCore::String to v8 string.
> > +        https://bugs.webkit.org/show_bug.cgi?id=27655
> > +
> > +        * bindings/v8/V8Binding.cpp:
> > +        (WebCore::v8String): now just immediately calls v8ExternalString
> > +        (WebCore::enableStringImplCache): enables caching of conversions from WebCore::StringImpl to
> > +        v8::String
> > +        (WebCore::makeExternalString): utilty function to create external v8::String out of
> > +        WebCore::Strin
> typo: Strin
> 
> I can fix this on landing, but if you upload a new patch, it would be nice to
> fix it.

Fixed.

> 
> 
> > I'm lacking knowledge of our build process, but are you sure we have to
> > separate binaries for worker and render?  I was under impression that the
> > binary is the same and it flags which turn it into a worker or renderer
> > process.
> 
> Yes. Note: Workers in chromium only run one thread, but we can leave this as a
> runtime flag as it all appears to work.
> 
> The api feels a bit narrow in scope "enableStringImplCache" vs
> "setSingleThreadedMode", but this is fine. It can be broadened if there is ever
> any other code that needs this info.

So as was discussed over IM, let's leave it for now.

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