Since node()->shadowAncestorNode() does not return NULL, this if statement will never fail. This should be shadowHost().
Created attachment 166158 [details] Patch
I'm confused. How is this covered by existing tests? Is it a behavior change?
(In reply to comment #2) > I'm confused. How is this covered by existing tests? Is it a behavior change? It's not, it's a fix in the logic. shadowAncestorNode is a weird cousin of shadowRoot. We're trying to get rid of this weird cousin and in this particular case shadowRoot and shadowAncestorNode can be used interchangeably. So, the ChangeLog should say: No change in logic, refactoring.
Comment on attachment 166158 [details] Patch Please fix the ChangeLog before landing.
Created attachment 167687 [details] Patch for landing
Comment on attachment 167687 [details] Patch for landing Rejecting attachment 167687 [details] from commit-queue. Failed to run "['/mnt/git/webkit-commit-queue/Tools/Scripts/webkit-patch', '--status-host=queues.webkit.org', '-..." exit_code: 1 ERROR: /mnt/git/webkit-commit-queue/Source/WebCore/ChangeLog neither lists a valid reviewer nor contains the string "Unreviewed" or "Rubber stamp" (case insensitive). Full output: http://queues.webkit.org/results/14214605
Created attachment 181472 [details] Patch for landing
I didn't notice this patch was not committed yet. rebase on ToT, and cq+.
Comment on attachment 181472 [details] Patch for landing Clearing flags on attachment: 181472 Committed r138922: <http://trac.webkit.org/changeset/138922>
All reviewed patches have been landed. Closing bug.