I can not find any reason to initialize an EventPath lazily. We can calculate an EventPath in EventDispatcher's constructor and remove EventDispatcher::ensureEventPath() after bug 109898 is fixed.
Created attachment 188496 [details] Calculates an EventPath in the constructor
The uploaded patch depends on the patch in bug 109898.
Created attachment 188798 [details] Patch for landing
Comment on attachment 188798 [details] Patch for landing Rejecting attachment 188798 [details] from commit-queue. New failing tests: fast/events/touch/gesture/pad-gesture-cancel.html Full output: http://queues.webkit.org/results/16610104
Hmm. fast/events/touch/gesture/pad-gesture-cancel.html has been crashed without this patch locally. Let me retry cq+. (In reply to comment #4) > (From update of attachment 188798 [details]) > Rejecting attachment 188798 [details] from commit-queue. > > New failing tests: > fast/events/touch/gesture/pad-gesture-cancel.html > Full output: http://queues.webkit.org/results/16610104
Comment on attachment 188798 [details] Patch for landing Rejecting attachment 188798 [details] from commit-queue. New failing tests: fast/events/touch/gesture/pad-gesture-cancel.html Full output: http://queues.webkit.org/results/16545079
Comment on attachment 188798 [details] Patch for landing Attachment 188798 [details] did not pass chromium-ews (chromium-xvfb): Output: http://queues.webkit.org/results/16613055 New failing tests: fast/events/touch/gesture/pad-gesture-cancel.html
Created attachment 189034 [details] Patch for landing
Let me cq+ tomorrow to watch the result. (In reply to comment #8) > Created an attachment (id=189034) [details] > Patch for landing
Comment on attachment 189034 [details] Patch for landing Clearing flags on attachment: 189034 Committed r143426: <http://trac.webkit.org/changeset/143426>
All reviewed patches have been landed. Closing bug.