JSC should use BindingState to determine the activeDOMWindow
Created attachment 156787 [details] Patch
Created attachment 156789 [details] work in progress
Comment on attachment 156789 [details] work in progress Attachment 156789 [details] did not pass gtk-ews (gtk): Output: http://queues.webkit.org/results/13436956
Created attachment 156796 [details] Patch for landing
Created attachment 156797 [details] Patch for landing
Created attachment 156803 [details] Patch
Comment on attachment 156803 [details] Patch LGTM.
Created attachment 156808 [details] resolve project.pbxproj merge conflict
Committed r124835: <http://trac.webkit.org/changeset/124835>
Comment on attachment 156808 [details] resolve project.pbxproj merge conflict View in context: https://bugs.webkit.org/attachment.cgi?id=156808&action=review > Source/WebCore/bindings/js/JSDOMBinding.h:-414 > - // FIXME: Remove these functions in favor of activeContext and > - // firstContext, which return ScriptExecutionContext*. We prefer to use > - // ScriptExecutionContext* as the context object in the bindings. Adam: Is this comment obsolete? Did you intentionally delete it rather than moving it to BindingState.h?
(In reply to comment #10) > (From update of attachment 156808 [details]) > View in context: https://bugs.webkit.org/attachment.cgi?id=156808&action=review > > > Source/WebCore/bindings/js/JSDOMBinding.h:-414 > > - // FIXME: Remove these functions in favor of activeContext and > > - // firstContext, which return ScriptExecutionContext*. We prefer to use > > - // ScriptExecutionContext* as the context object in the bindings. > > Adam: Is this comment obsolete? Did you intentionally delete it rather than moving it to BindingState.h? I added these comments the last time I tried to fix this problem (Bug 75793). I'm not 100% sure which approach will work. Replacing these functions with activeContext and firstContext is one promising approach. If I don't succeed this time around, I'll add comments explaining where we ended up.