Per experimentation with a recent example, in which unreasonable amounts
of time could elapse before the backend would respond to a query-cancel.
This might be something to back-patch, but the patch doesn't apply cleanly
because this code was rewritten for 9.1. Given the lack of field
complaints I won't bother for now.
Cédric Villemain
#include "executor/hashjoin.h"
#include "executor/nodeHash.h"
#include "executor/nodeHashjoin.h"
+#include "miscadmin.h"
#include "utils/memutils.h"
case HJ_SCAN_BUCKET:
+ /*
+ * We check for interrupts here because this corresponds to
+ * where we'd fetch a row from a child plan node in other
+ * join types.
+ */
+ CHECK_FOR_INTERRUPTS();
+
/*
* Scan the selected hash bucket for matches to current outer
*/