Like these WebKit1 Windows ones: IWebFrameLoadDelegatePrivate2::didDisplayInsecureContent IWebFrameLoadDelegatePrivate2::didRunInsecureContent
<rdar://problem/8392724>
Created attachment 71940 [details] proposed patch
Attachment 71940 [details] did not pass style-queue: Failed to run "['WebKitTools/Scripts/check-webkit-style']" exit_code: 1 WebKit2/UIProcess/API/C/WKPage.h:91: Extra space between WKPageDidFirstLayoutForFrameCallback and didFirstLayoutForFrame [whitespace/declaration] [3] WebKit2/UIProcess/API/C/WKPage.h:92: Extra space between WKPageDidFirstVisuallyNonEmptyLayoutForFrameCallback and didFirstVisuallyNonEmptyLayoutForFrame [whitespace/declaration] [3] WebKit2/UIProcess/API/C/WKPage.h:94: Extra space between WKPageDidDisplayInsecureContentForFrameCallback and didDisplayInsecureContentForFrame [whitespace/declaration] [3] WebKit2/UIProcess/API/C/WKPage.h:95: Extra space between WKPageDidRunInsecureContentForFrameCallback and didRunInsecureContentForFrame [whitespace/declaration] [3] Total errors found: 4 in 9 files If any of these errors are false positives, please file a bug against check-webkit-style.
Attachment 71940 [details] did not build on qt: Build output: http://queues.webkit.org/results/4828036
Comment on attachment 71940 [details] proposed patch View in context: https://bugs.webkit.org/attachment.cgi?id=71940&action=review > WebKit2/ChangeLog:11 > + Added the delegate. Just like the bundle version, it misses WebOrigin parameter that Mac > + delegate call used to have. It doesn't seem necessary for clients. > + I believe the origin is necessary to implement origin tainting (eg. if a page in one origin runs insecure content, we need to taint all other pages in that origin). This tracking may be something we should do in WebCore in WebKit instead of in the app though. > WebKit2/WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:785 > + > + RefPtr<APIObject> userData; > + WebProcess::shared().connection()->send(Messages::WebPageProxy::DidDisplayInsecureContentForFrame(m_frame->frameID(), InjectedBundleUserMessageEncoder(userData.get())), webPage->pageID()); > } This is just passing a null userData to the UIProcess. We either need to pass the userData to the bundle callback or not pass it to the UIProcess. > WebKit2/WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:796 > webPage->injectedBundleLoaderClient().didRunInsecureContentForFrame(webPage, m_frame); > + > + RefPtr<APIObject> userData; > + WebProcess::shared().connection()->send(Messages::WebPageProxy::DidRunInsecureContentForFrame(m_frame->frameID(), InjectedBundleUserMessageEncoder(userData.get())), webPage->pageID()); Here too.
Created attachment 71954 [details] pass userData
Attachment 71954 [details] did not pass style-queue: Failed to run "['WebKitTools/Scripts/check-webkit-style']" exit_code: 1 WebKit2/UIProcess/API/C/WKPage.h:91: Extra space between WKPageDidFirstLayoutForFrameCallback and didFirstLayoutForFrame [whitespace/declaration] [3] WebKit2/UIProcess/API/C/WKPage.h:92: Extra space between WKPageDidFirstVisuallyNonEmptyLayoutForFrameCallback and didFirstVisuallyNonEmptyLayoutForFrame [whitespace/declaration] [3] WebKit2/UIProcess/API/C/WKPage.h:94: Extra space between WKPageDidDisplayInsecureContentForFrameCallback and didDisplayInsecureContentForFrame [whitespace/declaration] [3] WebKit2/UIProcess/API/C/WKPage.h:95: Extra space between WKPageDidRunInsecureContentForFrameCallback and didRunInsecureContentForFrame [whitespace/declaration] [3] Total errors found: 4 in 17 files If any of these errors are false positives, please file a bug against check-webkit-style.
Committed <http://trac.webkit.org/changeset/70584>.
Attachment 71954 [details] did not build on qt: Build output: http://queues.webkit.org/results/4857026
Landed build fix in <http://trac.webkit.org/changeset/70587>.