On Saturday 25 June 2016 07:41:19 Paul Goyette wrote: > On Fri, 24 Jun 2016, Michael van Elst wrote: > > [email protected] ("John D. Baker") writes: > >> Jun 18 12:56:53 hostname dhcpcd[PID]: wm0: invalid UDP packet from > >> 19.100.192.168 Jun 18 12:56:53 hostname dhcpcd[PID]: wm0: invalid UDP > >> packet from 20.12.192.168 Jun 18 14:15:22 hostname dhcpcd[PID]: wm0: > >> invalid UDP packet from 119.16.192.168 Jun 23 21:48:35 hostname > >> dhcpcd[PID]: wm0: invalid UDP packet from 150.129.192.168 Jun 23 > >> 18:57:32 hostname dhcpcd[PID]: wm0: invalid UDP packet from > >> 163.85.192.168 Jun 24 02:56:29 hostname dhcpcd[PID]: wm0: invalid UDP > >> packet from 76.24.192.168 > >> > >> Needless to say, the purported source IPs are not on my network. > > > > Obviously these are not IP addresses. Each ends with 192.168, so there > > is a off-by-2 error when accessing the address field. > > Or some strange byte/word swap error...
Or probably a bug with the BPF reader being re-worked so it was interruptible. I think it's due to BPF queue having >1 packet. I'll look into fixing it, it happens very occasionally on my dev machine, but my network is small so it's hard to reproduce. Roy
