hashmap iterator can iterate twice the same item if deleting items via delcurrent_next

When iterating through the hashtable that is downsizing (or upsizing), the item could move from first internal table to the second one and could be iterated twice. This usually is not a problem because we delete all or none, but in the hashmap_test, we delete every second item and the non-delete item could be iterated first in the first table and then second when walking through the second table.

The other possibility of double walking the same item is when the items in the table get reordered due to psl value change.