[Webkit-unassigned] [Bug 147824] NetworkProcess: DNS prefetch happens in the Web Process

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=147824

--- Comment #26 from Carlos Garcia Campos <cgarcia at igalia.com> ---
(In reply to comment #25)
> (In reply to comment #23)
> > (In reply to comment #21)
> > > Comment on attachment 262588 [details]
> > > Updated patch
> > > 
> > > View in context:
> > > https://bugs.webkit.org/attachment.cgi?id=262588&action=review
> > > 
> > > > Source/WebCore/loader/FrameLoaderClient.h:33
> > > > +#include "DNS.h"
> > > 
> > > We should keep the implementation of the virtual function out of the header,
> > > that way we don’t need to include DNS.h in this header. Generally speaking
> > > it’s not so valuable to have implementations of virtual functions in the
> > > header rather than a cpp file since it’s not common to be able to expand
> > > them in line at the call site.
> > 
> > I agree, the advantage in this case that we don't have a common cpp file was
> > to avoid having the same implementation in all FrameLoader clients expect
> > the WebKit2 one. I'll make it pure virtual then, to ensure all clients
> > implement it.
> > 
> > > > Source/WebKit2/WebProcess/WebPage/ios/WebPageIOS.mm:632
> > > > +        m_page->mainFrame().loader().client().prefetchDNS(downcast<Element>(*node).absoluteLinkURL().host());
> > > 
> > > Are we guaranteed that m_page is non-null?
> > 
> > Same question ap made in comment #5, which confirms I should add an assert
> > there :-)
> 
> m_page can be null after WebPage::close() is called so I think we should
> null-check.

Yes, but I guess you can't tap on a closed page. This is a private method called from two WebPage::potentialTapAtPosition() and WebPage::tapHighlightAtPosition(). Both methods are message handlers, that aren't called if the page is closed, and both methods use m_page without any null check before calling sendTapHighlightForNodeIfNecessary().

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.webkit.org/pipermail/webkit-unassigned/attachments/20151020/108b4880/attachment.html>