im pretty sure this was fixed in src/usr.sbin/vmd/vionet.c r1.27. the quickest fix for me on 7.8 was to update the whole src/usr.sbin/vmd directory to HEAD and build and install it. you could probably just apply the specific commit above and get the fix too.
upgrading to a snapshot on the host will work too. On Thu, Dec 25, 2025 at 08:53:49AM +1000, David Gwynne wrote: > i'm guessing this is a VirtIO Network Packet header, as described at > https://wiki.osdev.org/Virtio. i dont know what's gone wrong between the > guest and hypervisor that you're seeing it "on the wire". > > dlg > > > On 24 Dec 2025, at 17:49, David Gwynne <[email protected]> wrote: > > > > It looks like there's an extra 12 bytes of 0s before what I think is the > > real Ethernet header. > > > > On Wed, 24 Dec 2025, 12:56 izzy Meyer, <[email protected]> wrote: > > On Wed, 24 Dec 2025 10:11:39 +1000 > > David Gwynne <[email protected]> wrote: > > > i would try following the packets through the host with tcpdump. in > > > this setup you should see the packets coming from the vm into the > > > host on the tap interface, through veb0, and into vport0. so `tcpdump > > > -ni tap0`, then `tcpdump -ni veb0`, and `tcpdump -ni vport`. > > > > > > if you dont see anything on tap0 then it sounds like a vm or vmm > > > problem. > > > > > > did you ugprade the host at the same time? or just the vm? > > > > I upgraded the Alpine VM well after upgrading to 7.8 on the host. > > > > > while im here, i have a couple of notes on your setup below. > > > > > > if you're only going to be running the one vm then you don't need > > > need veb in your topology, you can talk to the vm directly on the tap > > > interface. without veb, you can think of the tap interface as having > > > a cable between the host and the vm. adding veb plugs that connection > > > into a switch, and vport then plugs the host into that same switch. > > > > Oh, okay. That's much simpler. Hopefully I got this correct: > > > > ``` > > ~ $ cat /etc/hostname.tap0 > > lladdr fe:e1:ba:d1:ca:57 > > inet 172.16.100.1 255.255.255.0 > > ~ $ cat /etc/vm.conf > > vm "alpine" { > > disable > > memory 768M > > cdrom "/home/izder456/Downloads/alpine-virt-3.23.2-x86_64.iso" > > disk "/home/izder456/VMs/Alpine.qcow2" > > interface tap0 { > > lladdr de:ed:ba:d4:99:a8 > > } > > owner izder456 > > } > > ~ $ cat /etc/pf.conf > > set skip on lo > > set block-policy drop > > set optimization normal > > > > # vmm NAT > > match out on egress received-on tap0 nat-to (egress) > > > > # Multicast junk > > pass proto udp to 224.0.0.251 port mdns allow-opts > > pass inet6 proto udp to ff02::fb port mdns allow-opts > > pass proto udp to 239.255.255.250 port ssdp allow-opts > > pass inet6 proto udp to { ff02::c, ff05::c, ff08::c } port ssdp > > allow-opts > > > > # ICMP > > pass inet proto icmp > > pass inet6 proto icmp6 > > ``` > > > > I *did* use the same mac addresses as you did for simplicity, shouldn't > > be an issue though. > > > > Did I get my setup correct based on how you explained it? > > > > I deleted the disk of the old install as I had nothing important on it > > just to start fresh, and attempted a tcpdump after setting up the > > interfaces in the guest's installer environment. > > > > Alpine guest (with above config): > > ``` > > localhost:~# setup-interfaces > > Available interfaces are: eth0. > > Enter '?' for help on bridges, bonding and vlans. > > Which one do you want to initialize? (or '?' or 'done') [eth0] > > Ip address for eth0? (or 'dhcp', 'none', '?') [dhcp] 172.16.100.2 > > Netmask? [255.255.0.0] 255.255.255.0 > > Gateway? (or 'none') [none] 172.16.100.1 > > Configuration for eth0: > > type=static > > address=172.16.100.2 > > netmask=255.255.255.0 > > gateway=172.16.100.1 > > Do you want to do any manual network configuration? (y/n) [n] > > localhost:~# rc-service networking restart > > * Starting networking ... > > * lo ... > > [ ok ] > > * eth0 ... > > [ ok ] > > localhost:~# ping 8.8.8.8 > > PING 8.8.8.8 (8.8.8.8): 56 data bytes > > ^C > > --- 8.8.8.8 ping statistics --- > > 2 packets transmitted, 0 packets received, 100% packet loss > > ``` > > > > Host: > > ``` > > ~ $ doas tcpdump -ni tap0 > > tcpdump: listening on tap0, link-type EN10MB > > 20:44:57.951248 00:00:00:00:00:00 00:00:00:00:00:00 ffff 54: > > ffff ffff deed bad4 99a8 0806 0001 0800 > > 0604 0001 deed bad4 99a8 ac10 6402 0000 > > 0000 0000 ac10 6401 > > 20:44:58.991110 00:00:00:00:00:00 00:00:00:00:00:00 ffff 54: > > ffff ffff deed bad4 99a8 0806 0001 0800 > > 0604 0001 deed bad4 99a8 ac10 6402 0000 > > 0000 0000 ac10 6401 > > ^C > > 2 packets received by filter > > 0 packets dropped by kernel > > ``` > > > > What's curious here is that tcpdump sees packets coming from the mac > > address assigned to the VM's virtual ethernet card, but things aren't > > going out to the internet through my host? What am I doing wrongly here? > > > > -- > > iz (she/her) > > > > > i like to say mundane things, > > > there are too many uninteresting things > > > that go unnoticed. > > > > izder456 (dot) neocities (dot) org >
