Bug#721300: defaults to ipv6 address for [local_ip] and fails to call ipv4 endpoint

[Alexander Zangerl]

Package: sip-tester
Version: 1:3.1-3.1
Severity: normal
Tags: ipv6

i'm using a custom scenario file to call a number via my sip provider,
which uses [local_ip] to generate the Via: and Contact: sip headers.

this scenario file works fine on a ipv4-only box, but not on a dual-stack
system:

there sipp defaults to that box's ipv6 address (even though the remote
host given on the commandline is ipv4 only). sipp subsequently inserts 
the ipv6 address in the scenario file (visible when testing with -trace_msg), 
and then terminates with the slightly unhelpful 
message 'can't send udp packet: operation not permitted'.

workaround: running sipp with -i <your ipv4 address> ... overrides its
automatic address selection.

ideally sipp should either verify that the sip target we're trying to
reach is an ipv6 address/has an aaaa record, or offer the user to select v4/v6
operation with a commandline option.

regards
az