The code that this test relates to is gone. It was added by http://trac.webkit.org/changeset/20694, which fixed a crash in WebKit/Plugins/WebNetscapePluginStream.mm. The crash was caused by the lines removed below: loader->documentLoader()->addPlugInStreamLoader(_loader); - if (!_loader->load(request)) - _loader->documentLoader()->removePlugInStreamLoader(_loader); + _loader->load(request); The implementation is now shared, or at least WebBaseNetscapePluginStream.mm maintains a copy of PluginStream.cpp which has always had: void PluginStream::start() { ASSERT(!m_loadManually); m_loader = NetscapePlugInStreamLoader::create(m_frame, this); m_loader->setShouldBufferData(false); m_loader->documentLoader()->addPlugInStreamLoader(m_loader.get()); m_loader->load(m_resourceRequest); } Also, the crash situation doesn't arise for ports that do not implement a ResourceLoadDelegate. So should such ports implement the layoutTestController method the test requires and some drt specific code into the frameloaders to pass this test or should it be removed/moved to platform-specific tests?
Created attachment 58677 [details] Patch
(In reply to comment #1) > Created an attachment (id=58677) [details] > Patch LGTM, but you only missed unskipping plugins/get-url-that-the-resource-load-delegate-will-disallow.html
Created attachment 59287 [details] Patch
(In reply to comment #2) > (In reply to comment #1) > > Created an attachment (id=58677) [details] [details] > > Patch > > LGTM, but you only missed unskipping plugins/get-url-that-the-resource-load-delegate-will-disallow.html whoops!
Committed r61580: <http://trac.webkit.org/changeset/61580>
http://trac.webkit.org/changeset/61580 might have broken SnowLeopard Intel Release (Tests) The following changes are on the blame list: http://trac.webkit.org/changeset/61579 http://trac.webkit.org/changeset/61580