Re: should check interrupts in BuildRelationExtStatistics ?
Tom Lane <tgl@sss.pgh.pa.us>
From: Tom Lane <tgl@sss.pgh.pa.us>
To: Thomas Munro <thomas.munro@gmail.com>
Cc: Justin Pryzby <pryzby@telsasoft.com>,
Michael Paquier <michael@paquier.xyz>,
Robert Haas <robertmhaas@gmail.com>,
Tomas Vondra <tomas.vondra@enterprisedb.com>,
pgsql-hackers <pgsql-hackers@postgresql.org>
Date: 2022-07-07T19:50:51Z
Lists: pgsql-hackers
Attachments
- check-for-interrupts-in-statistics-qsorts.patch (text/x-diff) patch
Thomas Munro <thomas.munro@gmail.com> writes: > On Wed, Jul 6, 2022 at 11:37 AM Tom Lane <tgl@sss.pgh.pa.us> wrote: >> qsort_interruptible > +1 So here's a patch that does it that way. I first meant to put the new file into src/port/, but after remembering that that directory has no backend-only functions, I went with src/backend/utils/sort/ instead. For the moment I contented myself with changing qsort[_arg] calls that occur during statistics collection. We have a lot more of course, but many of them can be expected to not be dealing with much data, and in some cases we might want some closer analysis to be sure there's no performance hit. So I'm inclined to not worry too much about the remaining qsort calls until somebody complains. This could be back-patched to v14 without much worry, I should think. regards, tom lane
Commits
-
Invent qsort_interruptible().
- e64cdab00302 16.0 landed
- af72b0889441 14.5 landed
- 12c99c88442d 15.0 landed