Peter Eisentraut <peter(dot)eisentraut(at)2ndquadrant(dot)com> writes:
> It seems to me that it would make sense if InsertPgAttributeTuple() were
> to set attcacheoff to -1 instead of taking it from the caller.
Looked this over, no objections.
I wonder whether we should set that field to -1 when we *read*
pg_attribute rows from disk, and be less fussed about what gets written
out. The only real advantage is that this'd protect us from foolish
manual changes to pg_attribute.attcacheoff entries, but that doesn't
seem negligible.
regards, tom lane