The problem was that kernel source tree from which I was building module was newer than running kernel, and struct ifnet size has changed recently... On Thu, Aug 12, 2004 at 11:59:28AM +0400, Gleb Smirnoff wrote: T> Hmm, some more information: actually if_index is broken T> not only in in mbuf packet headers. T> T> This code with different addresses: T> T> bzero((caddr_t)&ro, sizeof(ro)); T> sin = (struct sockaddr_in *)&ro.ro_dst; T> sin->sin_len = sizeof(*sin); T> sin->sin_family = AF_INET; T> sin->sin_addr = \\some address\\; T> rtalloc_ign(&ro, RTF_CLONING); T> if (ro.ro_rt != NULL) { T> struct rtentry *rt = ro.ro_rt; T> T> printf("rt idx %d, name %s\n", T> rt->rt_ifp->if_index, T> rt->rt_ifp->if_xname); T> T> produces T> T> rt idx 0, name fxp0 T> rt idx 0, name lo0 T> T> On Thu, Aug 12, 2004 at 11:50:53AM +0400, Gleb Smirnoff wrote: T> T> I can see this at least in fxp driver: if I capture packets using T> T> ng_ether(4), and check interface name like this T> T> T> T> printf("idx %d, name %s\n", T> T> (*m)->m_pkthdr.rcvif->if_index, T> T> (*m)->m_pkthdr.rcvif->if_xname); T> T> I get on console: T> T> T> T> idx 0, name fxp0 T> T> T> T> Really fxp's index is 1 (AFAIK if_index must be > 0), I have checked this T> T> using if_nametoindex() from userland. T> T> -- T> Totus tuus, Glebius. T> GLEBIUS-RIPN GLEB-RIPE T> _______________________________________________ T> freebsd-current_at_freebsd.org mailing list T> http://lists.freebsd.org/mailman/listinfo/freebsd-current T> To unsubscribe, send any mail to "freebsd-current-unsubscribe_at_freebsd.org" -- Totus tuus, Glebius. GLEBIUS-RIPN GLEB-RIPEReceived on Thu Aug 12 2004 - 11:03:25 UTC
This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:38:05 UTC