netdev - ICMP host relookup support doubts

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [day] [month] [year] [list]

Message-ID: <Pine.LNX.4.58.0803281002210.3562@u.domain.uli>
Date:	Fri, 28 Mar 2008 10:51:32 +0200 (EET)
From:	Julian Anastasov <ja@....bg>
To:	Herbert Xu <herbert@...dor.apana.org.au>
cc:	netdev@...r.kernel.org
Subject: ICMP host relookup support doubts


	Hello,

	Am I missing something but it seems recent code (12 Dec 2007)
in icmp_send uses fl.fl4_src for output and input routes while it can
be initialized with saddr = 0 for non-local address. While this was
suited for the initial output route to avoid error for non-local
saddr, I suspect the relookup support uses it for additional output
and input route. Isn't the end result that no ICMP error is sent
after ip_route_input error for saddr=0? I assume no ICMP is generated
for forwarded traffic, expect if xfrm_lookup changes fl somewhere.
May be someone should fix it by replacing fl.fl4_src usage in the
new code with original iph->daddr? Eg. fl2.fl4_dst and the saddr arg
for ip_route_input? I don't know what is right to be done here.
saddr=0 should be used only as saddr arg to output route. OTOH, 
sysctl_icmp_errors_use_inbound_ifaddr can change the saddr too.

Regards

--
Julian Anastasov <ja@....bg>
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html