> We'd still want XID keys for the locks that are used to wait for a
> particular transaction to complete (eg when waiting to update
> a tuple). I think that's OK since VACUUM doesn't need to hold any
> such locks, but it'd probably mean making separate lmgr API entry
> points for those locks as opposed to normal table-level locks.
In this case XID is used as key in LOCKTAG, ie in lock identifier,
but we are going to change XIDTAG, ie just holder identifier.
No new entry will be required.
Vadim