Heikki Linnakangas <heikki.linnakangas@enterprisedb.com> writes:
> Hmm, actually, if we're willing to believe PD_ALL_VISIBLE in the page
> header over the xmin/xmax on the tuples, we could simply not bother
> doing anti-wraparound vacuums for pages that have the flag set. I'm not
> sure what changes that would require outside heapam.c, as we'd have to
> be careful to not trust the xmin/xmax if the flag was set.
That seems pretty ugly/dangerous. If we're going to try to do something
here, I much prefer Robert's approach of marking each tuple in the tuple
header.
regards, tom lane