Policy: Re-enabled plugin still disabled

chrome/browser/plugin_service.cc

Index: chrome/browser/plugin_service.cc
diff --git a/chrome/browser/plugin_service.cc b/chrome/browser/plugin_service.cc
index e136f97c9e353d5a42c196914dd7392702e90ccf..ad6f8a19f2e975e2c6a6ff2778bac8e91cfd1f7d 100644
--- a/chrome/browser/plugin_service.cc
+++ b/chrome/browser/plugin_service.cc
@@ -303,7 +303,7 @@ void PluginService::GetAllowedPluginForOpenChannelToPlugin(
   webkit::npapi::WebPluginInfo info;
   bool found = GetFirstAllowedPluginInfo(url, mime_type, &info, NULL);
   FilePath plugin_path;
-  if (found && info.enabled)
+  if (found && webkit::npapi::IsPluginEnabled(info))
     plugin_path = FilePath(info.path);
 
   // Now we jump back to the IO thread to finish opening the channel.
@@ -431,6 +431,7 @@ void PluginService::Observe(NotificationType type,
 #endif
 
     case NotificationType::PLUGIN_ENABLE_STATUS_CHANGED: {
+      webkit::npapi::PluginList::Singleton()->RefreshPlugins();

[jam, 2011/01/19 20:22:09]

I think we want to avoid reloading the plugins from disk when the user
enables/disables plugins, since it's an expensive operation.

[pastarmovj, 2011/01/19 23:39:17]

If we don't do that here policy changes during the runtime won't take effect
until next restart. I don't think that is so bad especially because we only mark
plugins for refresh here we don't enforce it.
And the UI anyway will refresh the plugins always after the user actually
interacts with it.

[jam, 2011/01/20 00:24:25]

I see, I didn't know that the UI always marks it for refresh.  It would be nice
to avoid that, since no need to do all that unnecessary disk access.  But since
it should be rare, ok for now although I'd love to see that changed in the
future :)

       PurgePluginListCache(false);
       break;
     }