Horton

Re: Making aggregate deserialization (and WAL receive) functions slightly faster

David Rowley <dgrowleyml@gmail.com>

View thread View on postgr.es
From: David Rowley <dgrowleyml@gmail.com>
To: Tom Lane <tgl@sss.pgh.pa.us>
Cc: Amit Kapila <amit.kapila16@gmail.com>, Michael Paquier <michael@paquier.xyz>, PostgreSQL Developers <pgsql-hackers@lists.postgresql.org>
Date: 2023-10-26T04:00:29Z
Lists: pgsql-hackers

Attachments

On Thu, 26 Oct 2023 at 08:43, Tom Lane <tgl@sss.pgh.pa.us> wrote:
> I think that we can make that assumption starting with v17.
> Back-patching it would be hazardous perhaps; but if there's some
> function out there that depends on NUL termination, testing should
> expose it before too long.  Wouldn't hurt to mention this explicitly
> as a possible incompatibility in the commit message.
>
> Looking over the v5 patch, I have some nits:

Thanks for looking at this again. I fixed up each of those and pushed
the result, mentioning the incompatibility in the commit message.

Now that that's done, I've attached a patch which makes use of the new
initReadOnlyStringInfo initializer function for the original case
mentioned when I opened this thread. I don't think there are any
remaining objections to this, but I'll let it sit for a bit to see.

David

Commits

  1. Make use of initReadOnlyStringInfo() in more places

  2. Optimize various aggregate deserialization functions, take 2

  3. Introduce the concept of read-only StringInfos

  4. Revert "Optimize various aggregate deserialization functions"

  5. Optimize various aggregate deserialization functions