Hello,
the machine trips assert in bpf_mtap_ether():
#if NVLAN > 0
KASSERT(m->m_len >= ETHER_HDR_LEN);
memcpy(&evh, mtod(m, char *), ETHER_HDR_LEN);
evh.evl_proto = evh.evl_encap_proto;
evh.evl_encap_proto = htons(ETHERTYPE_VLAN);
evh.evl_tag = htons(m->m_pkthdr.ether_vtag);
can you also share output of dmesg command? perhaps also
some information about USB-ethernet device. looking more
closely at bpf_mtap_ether() function it looks like the
mbuf which trips the assert has M_VLANTAG set.
the callstack is:
bpf_mtap_ether()
ifiq_input()
ure_rxeof()
usb_transfer_complete()
ehci_softintr()
softintr_dispatch()
Is downstream box connected via ethernet cable? or is
there some kind of switch involved between the boxes?
also how ifconfig(8) output on OpenBSD laptop looks like?
thanks and
regards
sashan
On Tue, Jan 21, 2025 at 04:22:56PM -0500, T Harpster wrote:
> Bug team.
>
> I'm running 7.6 on a Lenovo laptop.
>
> I was testing a USB Ethernet adapter and had set it up to work as a router.
>
> The built in Ethernet port is configured as the WAN interface.
> When I connected another computer downstream (to the internal interface in
> PF) which was the USB Ethernet converter and tried to do a speed test, the
> kernel panics.
>
> Attached are photos of the messages.
>
> I'm just sending you the photos in case it's not a known issue.
>
> If it isn't a known issue, I can replicate the error again and send you
> proper documentation.
>
> I apologize if this isn't the best documentation for you, but it's the
> first bug I've come across in OpenBSD in the 6+ years of running it.
>
> Thank you for all you do.
>
> Todd Harpster
>
>
> Please forgive any typos.
> Sent from my phone.
