Horton

Re: Should we add xid_current() or a int8->xid cast?

Tom Lane <tgl@sss.pgh.pa.us>

View thread View on postgr.es
From: Tom Lane <tgl@sss.pgh.pa.us>
To: Andres Freund <andres@anarazel.de>
Cc: Alvaro Herrera <alvherre@2ndquadrant.com>, Thomas Munro <thomas.munro@gmail.com>, Fujii Masao <masao.fujii@oss.nttdata.com>, Mark Dilger <hornschnorter@gmail.com>, pgsql-hackers <pgsql-hackers@postgresql.org>
Date: 2020-04-02T19:20:30Z
Lists: pgsql-hackers 
Andres Freund <andres@anarazel.de> writes:
> On 2020-04-02 14:33:18 -0300, Alvaro Herrera wrote:
>> Hmm, for some reason I had it in my head that we would make these use an
>> "epoch/val" output format rather than raw uint64 values.

> Why would we do that? IMO the goal should be to reduce awareness of the
> 32bitness of normal xids from as many places as possible, and treat them
> as an internal space optimization.

If they're just int64s then you don't need special functions to do
things like finding the min or max in a column of them.

			regards, tom lane

Commits

  1. Introduce xid8-based functions to replace txid_XXX.

  2. Add SQL type xid8 to expose FullTransactionId to users.

  3. Replace the former method of determining snapshot xmax --- to wit, calling