Horton

Re: Avoiding roundoff error in pg_sleep()

Robert Haas <robertmhaas@gmail.com>

View thread View on postgr.es
From: Robert Haas <robertmhaas@gmail.com>
To: Tom Lane <tgl@sss.pgh.pa.us>
Cc: pgsql-hackers@lists.postgresql.org
Date: 2025-09-25T19:41:25Z
Lists: pgsql-hackers 
On Thu, Sep 25, 2025 at 2:42 PM Tom Lane <tgl@sss.pgh.pa.us> wrote:
> I chanced to notice that if you ask pg_sleep for 1ms delay,
> what you actually get is 2ms, for example

Oh, wow. I tested and I get the same behavior.

> Anyway, I propose trying to get rid of this misbehavior by avoiding
> floating point in the delay computation, as attached.

Score one for my blind yet fiery hatred of floating-point arithmetic.
The patch looks good except that (places tongue firmly in cheek) it
will cause problems for users who want to sleep for more than 150,000
years.

-- 
Robert Haas
EDB: http://www.enterprisedb.com

Commits

  1. Try to avoid floating-point roundoff error in pg_sleep().