TLDR: Workaround: Bypass the name resolution attempt and provide the
IPv4 address directly. I cannot upload a fixed package to Debian at the
moment -- I'd be very grateful if somebody prepared an NMU.
The crash happens in sock_proxychain_new()
[sock_proxy.c:nsock_proxy.c:106] when a proxy entry has not been
properly initialized. In the reported scenario, there seem to be two
connected causes:
(1) proxy_socks4_node_new() exits with an error if the resulting address
as returned by proxy_resolve() is not an AF_INET address. Running
nmap -v --proxies socks4://ipv6.google.com:8909
reproducably leads to a segfault at the same place.
(2) As the *_node_new() function has failed, proxy_noew_new()
[nsock_proxy.c:371] returns a NULL pointer which is dereferenced in
nsock_proxychain_new().
It does not matter why proxy_socks4_node_new() fails, either: Running
nmap -v --proxies socks4://foo:8909
also segfaults if "foo" cannot be resolved to an IPv4 address.
Cheers,
-Hilko
--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
