Make ScopeChainNode GC allocated
Created attachment 83911 [details] Patch
Attachment 83911 [details] did not build on qt: Build output: http://queues.webkit.org/results/8043579
Attachment 83911 [details] did not build on win: Build output: http://queues.webkit.org/results/8031574
Attachment 83911 [details] did not build on win: Build output: http://queues.webkit.org/results/8043603
Attachment 83911 [details] did not build on gtk: Build output: http://queues.webkit.org/results/8035544
Attachment 83911 [details] did not build on chromium: Build output: http://queues.webkit.org/results/8019600
Comment on attachment 83911 [details] Patch Per our discussion I suggest making scope chain an object so that it can hold properties in anonymous slots, and do away with custom marking (also, function should also be holding scope chain in an anonymous slot now). As such I'm going to r- for the time being. Hope this sounds right to you, feel free to argue we should land this as a cell for now if you thing this is a useful stepping stone & I'll re-review.
Attachment 83911 [details] did not build on chromium: Build output: http://queues.webkit.org/results/8031878
Created attachment 84092 [details] Patch
Comment on attachment 84092 [details] Patch Okay, having function have different types of reference to its global object & scope is badness, but I understand that we'll revisit this.
Attachment 84092 [details] did not build on win: Build output: http://queues.webkit.org/results/8073418
Attachment 84092 [details] did not build on qt: Build output: http://queues.webkit.org/results/8073420
Committed r79904: <http://trac.webkit.org/changeset/79904>
Attachment 84092 [details] did not build on gtk: Build output: http://queues.webkit.org/results/8077151
(In reply to comment #12) > Attachment 84092 [details] did not build on qt: > Build output: http://queues.webkit.org/results/8073420 Qt buildfix landed in http://trac.webkit.org/changeset/79916 (https://bugs.webkit.org/show_bug.cgi?id=55407)