svn commit: r193664 - in head: sbin/ifconfig share/man/man4
sys/net sys/netinet sys/netinet6 sys/sys
Bjoern A. Zeeb
bz at FreeBSD.org
Mon Jun 8 21:00:08 UTC 2009
On Sun, 7 Jun 2009, Hiroki Sato wrote:
Hi,
> Author: hrs
> Date: Sun Jun 7 23:00:40 2009
> New Revision: 193664
> URL: http://svn.freebsd.org/changeset/base/193664
>
> Log:
> Fix and add a workaround on an issue of EtherIP packet with reversed
> version field sent via gif(4)+if_bridge(4). The EtherIP
> implementation found on FreeBSD 6.1, 6.2, 6.3, 7.0, 7.1, and 7.2 had
> an interoperability issue because it sent the incorrect EtherIP
> packets and discarded the correct ones.
>
> This change introduces the following two flags to gif(4):
>
> accept_rev_ethip_ver: accepts both correct EtherIP packets and ones
> with reversed version field, if enabled. If disabled, the gif
> accepts the correct packets only. This flag is enabled by
> default.
>
> send_rev_ethip_ver: sends EtherIP packets with reversed version field
> intentionally, if enabled. If disabled, the gif sends the correct
> packets only. This flag is disabled by default.
>
> These flags are stored in struct gif_softc and can be set by
> ifconfig(8) on per-interface basis.
>
> Note that this is an incompatible change of EtherIP with the older
> FreeBSD releases. If you need to interoperate older FreeBSD boxes and
> new versions after this commit, setting "send_rev_ethip_ver" is
> needed.
>
> Reviewed by: thompsa and rwatson
> Spotted by: Shunsuke SHINOMIYA
> PR: kern/125003
> MFC after: 2 weeks
>
> Added:
> head/sbin/ifconfig/ifgif.c (contents, props changed)
> Modified:
> head/sbin/ifconfig/Makefile
> head/share/man/man4/gif.4
> head/share/man/man4/if_bridge.4
> head/sys/net/if_gif.c
> head/sys/net/if_gif.h
> head/sys/netinet/in_gif.c
> head/sys/netinet6/in6_gif.c
> head/sys/sys/priv.h
>
....
> Modified: head/sys/net/if_gif.c
> ==============================================================================
> --- head/sys/net/if_gif.c Sun Jun 7 22:55:48 2009 (r193663)
> +++ head/sys/net/if_gif.c Sun Jun 7 23:00:40 2009 (r193664)
> @@ -653,6 +670,7 @@ gif_ioctl(ifp, cmd, data)
> struct gif_softc *sc = ifp->if_softc;
> struct ifreq *ifr = (struct ifreq*)data;
> int error = 0, size;
> + u_int options;
> struct sockaddr *dst, *src;
> #ifdef SIOCSIFMTU /* xxx */
> u_long mtu;
> @@ -887,6 +905,24 @@ gif_ioctl(ifp, cmd, data)
> /* if_ioctl() takes care of it */
> break;
>
> + case GIFGOPTS:
> + options = sc->gif_options;
> + error = copyout(&options, ifr->ifr_data,
> + sizeof(options));
> + break;
> +
> + case GIFSOPTS:
> + if ((error = priv_check(curthread, PRIV_NET_GIF)) != 0)
> + break;
> + if ((error = copyin(&options, &sc->gif_options,
> + sizeof(sc->gif_options)))) {
> + if ((options | GIF_FULLOPTS) == GIF_FULLOPTS)
> + ifr->ifr_data = (caddr_t)options;
> + else
> + error = EINVAL;
> + }
> + break;
> +
> default:
> error = EINVAL;
> break;
As noticed by ps there is a build problem here.
It seems it is the result of some reversed logic. Does this patch look ok?
http://people.freebsd.org/~bz/20090608-03-if_gif.c.diff
--
Bjoern A. Zeeb The greatest risk is not taking one.
More information about the svn-src-all
mailing list