Patch to follow.
Created attachment 172774 [details] Patch
Joshua, Alec could one of you please check sanity of this before I land?
Comment on attachment 172774 [details] Patch One totally minor issue...other than that this looks good. View in context: https://bugs.webkit.org/attachment.cgi?id=172774&action=review > Source/WebCore/inspector/InspectorIndexedDBAgent.cpp:453 > + idbCursor->close(); Can you avoid calling close() and just let the cursor go out of scope (close() is not a method in the spec, so JS just lets it get GC'ed - we'll probably be making it private soon)
(In reply to comment #3) > (From update of attachment 172774 [details]) > One totally minor issue...other than that this looks good. > > > View in context: https://bugs.webkit.org/attachment.cgi?id=172774&action=review > > > Source/WebCore/inspector/InspectorIndexedDBAgent.cpp:453 > > + idbCursor->close(); > > Can you avoid calling close() and just let the cursor go out of scope (close() is not a method in the spec, so JS just lets it get GC'ed - we'll probably be making it private soon) Sure, I actually forgot to remove it when debugging an issue of inspector blocking database. It turned out that I just needed to call database.close() instead.
Committed r133855: <http://trac.webkit.org/changeset/133855>