[Openswan Users] kernel 2.4.26 + Openswan 2.1.4 + net-snmp-5.1 = Kernel Oops

[Robert Story (Users)]

On Thu, 23 Sep 2004 15:12:34 +0200 (MET DST) Paul wrote:
PW> > On Sun, 2004-08-01 at 06:27, Martin Bene wrote:
PW> >> Martin Bene wrote:
PW> >> <agent/mibgroup/mibII/interfaces.c>
PW> >> ...
PW> >> if (ioctl(fd, 0x8947, &ifr) >= 0) {
PW> >>      new_ioctl_nums = 1;
PW> >> } else if (ioctl(fd, SIOCDEVPRIVATE, &ifr) >= 0) {
PW> >>      new_ioctl_nums = 0;
PW> >> ...
PW> >>
PW> >> So, it first tries SIOCGMIIREG ioctl; if that doesn't work, it falls
PW> >> back to SIOCDEVPRIVATE which seems to have been previously used for this
PW> >> purpose.
PW> >>
PW> >> ipsec however uses SIOCDEVPRIVATE completely differently:
PW> >>
PW> >> /* attach a virtual ipsec? device to a physical device */
PW> >> #define IPSEC_SET_DEV (SIOCDEVPRIVATE)
PW> >>
PW> >> ==> Result: ipsec bombs out trying to handle SIOCDEVPRIVATE ioctl for
PW> >> (already configured) ipsec0.

If someone can cook up a configure test that we could use to detect if there
will be a conflict, we can make the fallback ioctl conditional.

-- 
Robert Story; NET-SNMP Junkie <http://www.net-snmp.org/>
<irc://irc.freenode.net/#net-snmp>  
Archive: <http://sourceforge.net/mailarchive/forum.php?forum=net-snmp-users>

You are lost in a twisty maze of little standards, all different.