On 7/28/24 10:00, Peter Tribble wrote: > On Sun, Jul 28, 2024 at 5:34 PM Ignacio Soriano Hernandez via > illumos-discuss <discuss@lists.illumos.org> wrote: > >> No worries, it is not working in Virtualbox VM's either after today's >> latest pkg update .. e1000g0 registered/unregistered nwamd_door_req_state: >> ncu interface:e1000g0 not found .. >> > > So the basic question is: has the device not been recognised by the driver, > or is it some > higher level in the stack (like nwam) that's gone wrong? (The fact that the > interface > name appears in the error messages seems to indicate that the device > exists.) > > I'm not exactly sure how to identify this. But: > > Does anything appear in /dev/net? > > Can 'dladm show-phys' see it? > > Does 'prtconf -D | grep e1000' show it? > > The thing is, the last update to the driver was over 4 months ago, so a > broken > driver seems unlikely.
Hi Peter, I'm investigating a regression that I likely introduced with "14237 Want support for pthread_cond_clockwait() and friends" which has put nwam in a bad spot and likely is responsible for this. I think I've root caused this and am building a fix to prove out. My suspicion is that these systems are all using nwam by default in which case it's likely not progressing in getting networking up. Sorry for the trouble. Robert ------------------------------------------ illumos: illumos-discuss Permalink: https://illumos.topicbox.com/groups/discuss/T2b5a4eb486f58cac-Mb7f80343a4bf8e7e88cc9048 Delivery options: https://illumos.topicbox.com/groups/discuss/subscription
