Re: More robust pg_hba.conf parsing/error logging - Mailing list pgsql-hackers

From Alvaro Herrera
Subject Re: More robust pg_hba.conf parsing/error logging
Date
Msg-id 20090909141817.GC4132@alvh.no-ip.org
Whole thread Raw
In response to More robust pg_hba.conf parsing/error logging  (Rafael Martinez <r.m.guerrero@usit.uio.no>)
List pgsql-hackers
Rafael Martinez wrote:

> Problem:
> - --------
> If you define in pg_hba.conf a database or a user value with 'ALL'
> instead of 'all', you will lose access to *all* databases involved. The
> reload process will not report anything about 'ALL' been an invalid
> value and the new pg_hba.conf will be reloaded.
> 
> This is the only thing in the log file:
> "LOG:  received SIGHUP, reloading configuration files"

Aye, that's surprising.  I think the correct fix here is to change the
strcmp comparisons to pg_strcasecmp() in several places in hba.c.

(BTW the business about appending newlines to special tokens in
next_token() seems ugly and underdocumented.)

-- 
Alvaro Herrera                                http://www.CommandPrompt.com/
PostgreSQL Replication, Consulting, Custom Development, 24x7 support


pgsql-hackers by date:

Previous
From: Pavel Stehule
Date:
Subject: ToDo: array aggregates
Next
From: Alvaro Herrera
Date:
Subject: Re: More robust pg_hba.conf parsing/error logging