On 2018-05-24 13:30:54 -0700, Andres Freund wrote:
> On 2018-05-24 13:08:53 -0400, Alvaro Herrera wrote:
> > Hmm .. surely
> >
> > diff --git a/src/backend/access/heap/heapam.c b/src/backend/access/heap/heapam.c
> > index 5016181fd7..5d7fa1fb45 100644
> > --- a/src/backend/access/heap/heapam.c
> > +++ b/src/backend/access/heap/heapam.c
> > @@ -6690,7 +6690,7 @@ heap_prepare_freeze_tuple(HeapTupleHeader tuple,
> > xid = HeapTupleHeaderGetXmin(tuple);
> > xmin_frozen = ((xid == FrozenTransactionId) ||
> > HeapTupleHeaderXminFrozen(tuple));
> > - if (TransactionIdIsNormal(xid))
> > + if (!xmin_frozen && TransactionIdIsNormal(xid))
> > {
> > if (TransactionIdPrecedes(xid, relfrozenxid))
> > ereport(ERROR,
> >
> >
> > ??
>
> I don't think that's necesary - HeapTupleHeaderGetXmin() returns
> FrozenTransactionId if the tuple is frozen (note the
> HeapTupleHeaderXminFrozen() within).
FWIW, even if that weren't the case: a) there'd be a lot more wrong with
this routine imo. b) some of the tuples affected clearly weren't
frozen...
Greetings,
Andres Freund
