In response to "Andrew Dunstan" <andrew(at)dunslane(dot)net>:

> Bill Moran wrote:
> > Andrew Dunstan <andrew(at)dunslane(dot)net> wrote:
> >>
> >> Bill Moran wrote:
> >> > + 		if (trace_temp_files != -1)
> >> >
> >>
> >> Might be more robust to say
> >>
> >>     if (trace_temp_files >= 0)
> >
> > Because it would allow for the easy addition of more negative numbers
> > with magic value?
> 
> because ISTM any negative number here should mean no action is to be
> taken. Otherwise how else is it different from 0?

??

I specified in the GUC config that minimum allowable value is -1.

/usr/local/etc/rc.d/postgresql start
FATAL:  -5 is outside the valid range for parameter "trace_temp_files" (-1 .. 2147483647)

set trace_temp_files to -8;
ERROR:  -8 is outside the valid range for parameter "trace_temp_files" (-1 .. 2147483647)

Perhaps there's another reason to use the >= 0 check, but handling invalid
values with POLA doesn't seem to be a good one.

-- 
Bill Moran
Collaborative Fusion Inc.

In response to

• Re: [HACKERS] Patch to log usage of temporary files at 2007-01-04 01:05:27 from Andrew Dunstan

Responses

• Re: [HACKERS] Patch to log usage of temporary files at 2007-01-04 14:34:42 from Andrew Dunstan
• Re: [HACKERS] Patch to log usage of temporary files at 2007-01-04 14:56:18 from Tom Lane

pgsql-hackers by date

Next:	From: Simon Riggs	Date: 2007-01-04 14:27:56
	Subject: wal_checksum = on (default) | off
Previous:	From: Andrew Dunstan	Date: 2007-01-04 14:09:18
	Subject: Re: 8.3 pending patch queue

pgsql-patches by date

Next:	From: Simon Riggs	Date: 2007-01-04 14:27:56
	Subject: wal_checksum = on (default) | off
Previous:	From: Guillaume Smet	Date: 2007-01-04 12:58:12
	Subject: Re: Assorted typos