Paste the COMPLETE build information from "Help->About Wireshark", "wireshark -v", or "tshark -v".
As noted at http://mail.ojctech.com/pipermail/radiotap/2007-August/000045.html, Wireshark interprets bit 14 in the Radiotap header as IEEE80211_RADIOTAP_FCS. The correct interpretation according to http://netbsd.gw.com/cgi-bin/man-cgi?ieee80211_radiotap+9+NetBSD-current is IEEE80211_RADIOTAP_RX_FLAGS.
Is it OK to have constructions like this in the dissectors?Or should this be handled by configure?#ifdef __FreeBSD__ IEEE80211_RADIOTAP_XCHANNEL = 14;#elif __NetBSD__ IEEE80211_RADIOTAP_RX_FLAGS = 14;#else IEEE80211_RADIOTAP_FCS = 14,#endif
No, and no. The interpretation of a Radiotap file should *not* be dependent on the OS on which you're running Wireshark!IEEE80211_RADIOTAP_XCHANNEL was moved to 18 when I mentioned this problem to David Young and Sam Leffler. IEEE80211_RADIOTAP_FCS was, as far as I know, never used.I'm working on the one remaining issue, namely collisions between some OpenBSD bits and bits used in other OSes. Hopefully we can come up with a standard that all OSes can adopt and continue to adhere to. If not, we might have to define multiple DLT_ values for different flavors of radiotap; that wouldn't be the ideal, but at least it'd mean that a file capture on one OS could be interpreted by Wiretap (or tcpdump, or...) running on *any* OS.