I wrote: > I still think that we really need to add a check in ExecProcNode().
Actually ... to what extent could a check in ExecInitNode() substitute for that? Or do we need both? How about ExecEndNode() and ExecReScan()? You could argue that the latter two tree traversals are unlikely either to consume more stack than ExecInitNode() or to be called from a more deeply nested point. So maybe they're okay. But I'm not sure I believe that initialization stack needs always exceed execution stack needs, though sometimes they might. regards, tom lane -- Sent via pgsql-hackers mailing list (email@example.com) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-hackers