Make MutationEvents a RuntimeEnabled feature
I think this might need to be a per document setting.
Created attachment 158397 [details] Patch
Please wait for approval from abarth@webkit.org, dglazkov@chromium.org, fishd@chromium.org, jamesr@chromium.org or tkent@chromium.org before submitting, as this patch contains changes to the Chromium public API. See also https://trac.webkit.org/wiki/ChromiumWebKitAPI.
See http://crbug.com/142648 for the Chromium side of this patch.
Comment on attachment 158397 [details] Patch I haven't verified that you've hooked all the right code points, but the rest of the patch LGTM
Created attachment 158429 [details] Patch
Comment on attachment 158429 [details] Patch You missed some events. At the very least the DOMNodeRemoved* ones.
Created attachment 159226 [details] Patch
(In reply to comment #7) > (From update of attachment 158429 [details]) > You missed some events. At the very least the DOMNodeRemoved* ones. Thanks, fixed. Dmitri, is there any way to test ContextFeatures from DRT? That would have easily caught this.
Comment on attachment 159226 [details] Patch How about hooking Document's hasListenerType method instead? It'd be much less code and easier to be sure you catch everything.
Or, even better, hook addListenerType to not add if the type is a mutation event type. addListenerType is called a lot less often and in much less hot codepaths is why this might be better than hooking hasListenerType.
Waiting for some nearby cleanup to carry out Ojan's suggestion.
Created attachment 159541 [details] Patch
Comment on attachment 159541 [details] Patch API change LGTM
Comment on attachment 159541 [details] Patch Clearing flags on attachment: 159541 Committed r126113: <http://trac.webkit.org/changeset/126113>
All reviewed patches have been landed. Closing bug.