WebWidgetClient::screenInfo() no longer does a synchronous IPC.

content/browser/renderer_host/render_view_host.cc

Index: content/browser/renderer_host/render_view_host.cc
diff --git a/content/browser/renderer_host/render_view_host.cc b/content/browser/renderer_host/render_view_host.cc
index 22d9b89b03f31bccf2ae1d54f0538890f105cb60..e02d3d73ce1f186ac093801d0f3a254ce2d38c63 100644
--- a/content/browser/renderer_host/render_view_host.cc
+++ b/content/browser/renderer_host/render_view_host.cc
@@ -197,6 +197,17 @@ bool RenderViewHost::CreateRenderView(const string16& frame_name,
   params.session_storage_namespace_id = session_storage_namespace_->id();
   params.frame_name = frame_name;
   params.next_page_id = next_page_id;
+#if defined(OS_POSIX) || defined(USE_AURA)
+  if (view())
+    view()->GetScreenInfo(&params.screen_info);
+  else
+    RenderWidgetHostView::GetDefaultScreenInfo(&params.screen_info);
+#else
+  params.screen_info =
+      params.WebKit::WebScreenInfoFactory::screenInfo(

[lain Merrick, 2012/01/31 12:42:09]

params.WebKit looks wrong.

And this may still call screenInfo() with NULL, right? Maybe just add a NULL
check here?

[Fady Samuel, 2012/01/31 19:54:12]

It probably is. I'm trying to update and test on Windows now (I mostly working
on Linux so there's a bit of latency to get a fresh sync'ed version of chromium
tested with this patch on Windows).

+          gfx::NativeViewFromId(GetNativeViewId()));
+#endif
+
   Send(new ViewMsg_New(params));
 
   // If it's enabled, tell the renderer to set up the Javascript bindings for