Clarified JSGlobalData (JavaScript VM) lifetime
Created attachment 139371 [details] Patch
Comment on attachment 139371 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=139371&action=review > Source/JavaScriptCore/ChangeLog:37 > + notwithstaning, C++ does in fact know how to execute destructors in Speling error, notwithstaning. > Source/JavaScriptCore/runtime/JSGlobalData.h:155 > + Heap heap; Could you add a comment here about the destruction order so nobody will accidentally move this?
Comment on attachment 139371 [details] Patch Attachment 139371 [details] did not pass qt-wk2-ews (qt): Output: http://queues.webkit.org/results/12552666
Comment on attachment 139371 [details] Patch Attachment 139371 [details] did not pass qt-ews (qt): Output: http://queues.webkit.org/results/12553679
Comment on attachment 139371 [details] Patch Attachment 139371 [details] did not pass win-ews (win): Output: http://queues.webkit.org/results/12558684
Committed r115579: <http://trac.webkit.org/changeset/115579>
Comment on attachment 139371 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=139371&action=review > Source/JavaScriptCore/ChangeLog:42 > + (JSGlobalContextCreate): Removed compatibility shim for older > + operating systems because it's no longer used. This was not a shim for older operating systems, this was for applications linked against older versions of WebKit. I think that this removal is a mistake.
Reopening, because Alexey is right.
Created attachment 140839 [details] Patch
Geoff, please file a Radar for the follow-up fix.
<rdar://problem/11437914>
Committed r116809: <http://trac.webkit.org/changeset/116809>