> * You had consistently changed the simple_heap_update calls to do the
> wrong thing. (I'm surprised it didn't blow up on you in your testing.)
> In a sequence like
>
> newtuple = heap_modifytuple(tup, rel, repl_val, repl_null, repl_repl);
>
> simple_heap_update(rel, &newtuple->t_self, newtuple);
> CatalogUpdateIndexes(rel, newtuple);
>
> the second parameter to simple_heap_update *must* be newtuple->t_self
> not tup->t_self. The reason is that simple_heap_update stores the new
> physical location of the updated tuple back into that parameter, and
> then the CatalogUpdateIndexes call relies on newtuple->t_self to
> generate new index entries. The way you had it coded, it was generating
> new index entries pointing at the old version of the tuple ...
Strange. I guess I must have been testing with a database that had
short enough system catalogs that the indexes were never used?
Chris