dhcpcd-discuss

Re: Re: Remotely exploitable crash in dhcpcd

Roy Marples

Tue Aug 01 18:51:14 2017

On 26/06/2017 12:57, Jason A. Donenfeld wrote:
> I saw this via tcpdump. I'm away til Weds but will send more extensive
> debugging when home.

I didn't get any extensive debugging :)

Anyway, I've fixed a few more outliers and committed the patch to trunk.
It works fine for me on NetBSD and Linux
There's a few outstanding todos but they are minor

  *  Ideally kernels need a mechanism not to ARP announce on address
addition. (I can adjust for this NetBSD at least - does Linux have a
/proc knob?)
  *  The BPF code could be adjusted to iterate all active states across
all interfaces and list each xid (and chaddr IF all chaddr fit into the
space).

Roy

Follow-Ups:
Re: Re: Remotely exploitable crash in dhcpcdJason A. Donenfeld
References:
Remotely exploitable crash in dhcpcdJason A. Donenfeld
Re: Remotely exploitable crash in dhcpcdJason A. Donenfeld
Re: Re: Remotely exploitable crash in dhcpcdRoy Marples
Re: Re: Remotely exploitable crash in dhcpcdJason A. Donenfeld
Re: Re: Remotely exploitable crash in dhcpcdRoy Marples
Re: Re: Remotely exploitable crash in dhcpcdRoy Marples
Re: Re: Remotely exploitable crash in dhcpcdJason A. Donenfeld
Re: Re: Remotely exploitable crash in dhcpcdJason A. Donenfeld
Re: Re: Remotely exploitable crash in dhcpcdRoy Marples
Re: Re: Remotely exploitable crash in dhcpcdJason A. Donenfeld
Archive administrator: postmaster@marples.name