Horton

Re: Memory Accounting

Peter Geoghegan <pg@bowt.ie>

View thread View on postgr.es
From: Peter Geoghegan <pg@bowt.ie>
To: Jeff Davis <pgsql@j-davis.com>
Cc: Tomas Vondra <tomas.vondra@2ndquadrant.com>, Tom Lane <tgl@sss.pgh.pa.us>, Melanie Plageman <melanieplageman@gmail.com>, Soumyadeep Chakraborty <sochakraborty@pivotal.io>, Robert Haas <robertmhaas@gmail.com>, Pg Hackers <pgsql-hackers@postgresql.org>, soumyadeep2007@gmail.com
Date: 2019-10-04T18:43:06Z
Lists: pgsql-hackers 
On Fri, Oct 4, 2019 at 7:32 AM Jeff Davis <pgsql@j-davis.com> wrote:
> The patch has been floating around for a very long time, so I don't
> remember exactly why I chose a signed value. Sorry.

I am reminded of the fact that int64 is used to size buffers within
tuplesort.c, because it needs to support negative availMem sizes --
when huge allocations were first supported, tuplesort.c briefly used
"Size", which didn't work. Perhaps it had something to do with that.

--
Peter Geoghegan

Commits

  1. Change MemoryContextMemAllocated to return Size

  2. Use Size instead of int64 to track allocated memory

  3. Add transparent block-level memory accounting

  4. Change the way pre-reading in external sort's merge phase works.

  5. Improve memory management for external sorts.

  6. In array_agg(), don't create a new context for every group.