Re: qt multi- and broadcast (was Re: dhcpcd not working in simh-vax with xq0:nat networking)

[Johnny Billquist <bqt%softjar.se@localhost>]

On 2021-12-12 13:54, Rhialto wrote:
> On Sat 11 Dec 2021 at 23:42:38 +0100, Johnny Billquist wrote:
> > Uh... If broadcast wasn't working, you'd not have working ARP, more or less.
> > I would expect that people had noticed that a long time ago by now... (There
> > are other protocols also using broadcasts, but ARP is very fundamental, and
> > going on all the time, on any machine using ethernet.)
>
> I should have added "receiving" since the sending of broadcasts works fine.
> But ARP replies aren't broadcast, they are unicast to the requestor:
>
> 13:49:28.641231 a8:5e:45:yy:yy:yy (oui Unknown) > Broadcast, ethertype ARP (0x0806), length 42: Request who-has fritz.box tell zzz.fritz.box, length 28
>
> 13:49:28.642124 08:96:d7:xx:xx:xx (oui Unknown) > a8:5e:45:yy:yy:yy (oui Unknown), ethertype ARP (0x0806), length 60: Reply fritz.box is-at 08:96:d7:xx:xx:xx (oui Unknown), length 46
>
> (output from tcpdump -e -i wm0 arp on another box)
>
> So non-working broadcast reception can go unnoticed for a while.
> Multicast even longer.

Well. Obviously anyone else who wants to talk to that machine will be 
failing, since that requires an ARP resolution by the remote machine 
before any traffic can be sent.

So still - I have a hard time believing that this would have been 
non-functional for all this time on real hardware.

But looking at the code now, including also the 2.11BSD code, where it 
originated... Maybe this actually have been flying under the radar for a 
long time... I'll get back when I've understood what's in there.

  Johnny

--
Johnny Billquist                  || "I'm on a bus
                                  ||  on a psychedelic trip
email: bqt%softjar.se@localhost             ||  Reading murder books
pdp is alive!                     ||  tryin' to stay hip" - B. Idol