On Wed, Jul 24, 2024 at 06:00:59AM -0700, Noah Misch wrote:
> I'm still seeing need for s/int/int64/ at:
Nice catches! I've missed these.
> - "pagesegno" variable
> - return value of MultiXactIdToOffsetSegment()
Only used in four places for two elog(DEBUG1) entries with %x.
> - return value of MXOffsetToMemberSegment()
Also used in four places for two elog(DEBUG1) entries with %x, plus
three callers in PerformMembersTruncation(), nothing fancy.
> Only the first should be a live bug, since multixact isn't electing the higher
> pageno ceiling.
Yes, and it makes a switch to long segment names everywhere easier.
There is a patch in the air to do that, without the pg_upgrade
additions required to do the transfer, though.
I am attaching a patch for all these you have spotted, switching the
logs to use %llx. Does that look fine for you?
--
Michael