Date: Thu, 11 Jun 2020 17:54:20 +0000 From: bugzilla-noreply@freebsd.org To: bugs@FreeBSD.org Subject: [Bug 247129] [PATCH] Fix some compiler warnings in netinet alias module Message-ID: <bug-247129-227-taWKp28AG0@https.bugs.freebsd.org/bugzilla/> In-Reply-To: <bug-247129-227@https.bugs.freebsd.org/bugzilla/> References: <bug-247129-227@https.bugs.freebsd.org/bugzilla/>
next in thread | previous in thread | raw e-mail | index | archive | help
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=3D247129 --- Comment #4 from Mark Johnston <markj@FreeBSD.org> --- (In reply to nikethmurali from comment #3) A couple more comments: Changing dlen to be unsigned seems dangerous. For example, we have: 450 dlen =3D ntohs(pip->ip_len) - (pip->ip_hl << 2);=20=20=20=20= =20=20=20=20=20=20=20=20=20=20=20=20=20=20=20=20=20 451 if (dlen < ICMP_MINLEN)=20=20=20=20=20=20=20=20=20=20=20=20=20= =20=20=20=20=20=20=20=20=20=20=20=20=20=20=20=20=20=20=20=20=20=20=20=20=20= =20=20=20=20=20=20 452 return (PKT_ALIAS_IGNORED); What happens if a malicious packet defines a header length longer than ip_l= en?=20 If dlen is unsigned, it will end up being a large number and will pass the subsequent check. Regarding the alignment issue, wouldn't it be simpler to modify each of ProtoAliasIn/Out and FragmentIn/Out to take a struct ip * as input, and have them update fields directly? Then those functions know that the ip address fields are not necessarily self-aligned and the compiler can handle it. --=20 You are receiving this mail because: You are the assignee for the bug.=
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?bug-247129-227-taWKp28AG0>