alternative solution for ipv6 loopback - Mailing list pgsql-hackers

From Andrew Dunstan
Subject alternative solution for ipv6 loopback
Date
Msg-id 3F579F88.1060202@dunslane.net
Whole thread Raw
List pgsql-hackers
I just thought of something:

For a *real* special case, with no new keywords or other stuff, we could 
just add something like this to hba.c at the start of the 'host*' 
section of parse_hba():

#ifndef HAVE_IPV6       if (strcmp(token,"::1") == 0 || strcmp(token,"::1/128") == 0)           return;
#endif

A bit fragile, but it would allow a default line for the ip6 loopback 
address that wouldn't break ip4 only postmasters.

Any other type of ip6 address would break, as I believe it should. And 
we wouldn't have to write special routines to handle ip6 addresses where 
we otherwise don't know about them.

cheers

andrew




pgsql-hackers by date:

Previous
From: Jon Jensen
Date:
Subject: Re: Win32 native port
Next
From: "Marc G. Fournier"
Date:
Subject: Re: ANONCVS? Is it being updated correctly?