Date: Fri, 11 Nov 2011 09:23:19 -0500 From: Ben Kaduk <minimarmot@gmail.com> To: Gleb Smirnoff <glebius@freebsd.org> Cc: svn-src-stable@freebsd.org, svn-src-all@freebsd.org, "Bjoern A. Zeeb" <bz@freebsd.org>, src-committers@freebsd.org, svn-src-stable-9@freebsd.org Subject: Re: svn commit: r226572 - stable/9/sys/netinet Message-ID: <CAK2BMK7E=0-Gb0hP4m4zXyEP=mN0fW8g_cpFFZgE7guhM%2BctCg@mail.gmail.com> In-Reply-To: <9A674915-EF68-4CEA-93CB-CDE349082019@FreeBSD.org> References: <201110201558.p9KFw550081440@svn.freebsd.org> <9A674915-EF68-4CEA-93CB-CDE349082019@FreeBSD.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Thu, Oct 20, 2011 at 12:40 PM, Bjoern A. Zeeb <bz@freebsd.org> wrote: > On 20. Oct 2011, at 15:58 , Gleb Smirnoff wrote: > >> Author: glebius >> Date: Thu Oct 20 15:58:05 2011 >> New Revision: 226572 >> URL: http://svn.freebsd.org/changeset/base/226572 >> >> Log: >> =A0MFhead 226401,226402: >> >> =A0 =A0Remove last remnants of classful addressing: >> >> =A0 =A0- Remove ia_net, ia_netmask, ia_netbroadcast from struct in_ifadd= r. >> =A0 =A0- Remove net.inet.ip.subnetsarelocal, I bet no one need it in 201= 1. >> =A0 =A0- fix bug when we were not forwarding to a host which matches cla= ssful >> =A0 =A0 =A0net address. For example router having 192.168.x.y/16 network= attached, >> =A0 =A0 =A0would not forward traffic to 192.168.*.0, which are legal IPs= in >> =A0 =A0 =A0CIDR world. >> =A0 =A0- For compatibility, leave autoguessing of mask based on class. >> >> =A0 =A0Reviewed by: =A0andre, bz, rwatson >> >> =A0 =A0Add support for IPv4 /31 prefixes, as described in RFC3021. >> >> =A0 =A0To run a /31 network, participating hosts MUST drop support >> =A0 =A0for directed broadcasts, and treat the first and last addresses >> =A0 =A0on subnet as unicast. The broadcast address for the prefix >> =A0 =A0should be the link local broadcast address, INADDR_BROADCAST. > ... >> Modified: stable/9/sys/netinet/in_var.h >> =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D >> --- stable/9/sys/netinet/in_var.h =A0 =A0 Thu Oct 20 15:55:01 2011 =A0 = =A0 =A0 =A0(r226571) >> +++ stable/9/sys/netinet/in_var.h =A0 =A0 Thu Oct 20 15:58:05 2011 =A0 = =A0 =A0 =A0(r226572) >> @@ -60,12 +60,9 @@ struct in_ifaddr { >> =A0 =A0 =A0 struct =A0ifaddr ia_ifa; =A0 =A0 =A0 =A0 =A0/* protocol-inde= pendent info */ >> #define =A0 =A0 =A0 ia_ifp =A0 =A0 =A0 =A0 =A0ia_ifa.ifa_ifp >> #define ia_flags =A0 =A0 =A0ia_ifa.ifa_flags >> - =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 = =A0 /* ia_{,sub}net{,mask} in host order */ >> - =A0 =A0 u_long =A0ia_net; =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 /* network n= umber of interface */ >> - =A0 =A0 u_long =A0ia_netmask; =A0 =A0 =A0 =A0 =A0 =A0 /* mask of net p= art */ >> - =A0 =A0 u_long =A0ia_subnet; =A0 =A0 =A0 =A0 =A0 =A0 =A0/* subnet numb= er, including net */ >> - =A0 =A0 u_long =A0ia_subnetmask; =A0 =A0 =A0 =A0 =A0/* mask of subnet = part */ >> - =A0 =A0 struct =A0in_addr ia_netbroadcast; /* to recognize net broadca= sts */ >> + =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 = =A0 /* ia_subnet{,mask} in host order */ >> + =A0 =A0 u_long =A0ia_subnet; =A0 =A0 =A0 =A0 =A0 =A0 =A0/* subnet addr= ess */ >> + =A0 =A0 u_long =A0ia_subnetmask; =A0 =A0 =A0 =A0 =A0/* mask of subnet = */ >> =A0 =A0 =A0 LIST_ENTRY(in_ifaddr) ia_hash; =A0/* entry in bucket of inet= addresses */ >> =A0 =A0 =A0 TAILQ_ENTRY(in_ifaddr) ia_link; /* list of internet addresse= s */ >> =A0 =A0 =A0 struct =A0sockaddr_in ia_addr; =A0 =A0/* reserve space for i= nterface name */ > > This should at least break parts of netstat and ifmcstat as I had told yo= u in > private email before. Just so we're clear, it changes the API (at the RC stage!), causing at least the net/openafs port to fail to compile. I presume, given the commit message, that I am just not supposed to use ia_net and ia_netmask any more, and that ia_submet and ia_subnetmask are presumed to be sufficient? -Ben Kaduk
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CAK2BMK7E=0-Gb0hP4m4zXyEP=mN0fW8g_cpFFZgE7guhM%2BctCg>