[Bug 211219] NIC status does not pass into a state of "no carrier" after disconnecting the cable.

[bugzilla-noreply at freebsd.org]

https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=211219

--- Comment #16 from Franco Fichtner <franco at opnsense.org> ---
I'm not so sure.  All branches received a modified patch that came up
inconclusive in testing, overshadowed by the iflib conversion and multiple
build and subsystem failures due to it.  We circled around a minimal patch
(note the attached patch reverts to the full Intel driver change), but the
weirdness that is "adapter->ims" for just this one chipset code path still
makes no sense in the grand scheme of things.

I asked for a full revert to the working Intel state via e-mail, but that
didn't happen.

I don't think this is sensible for the simple reason that this was introduced
in stable/10, went into 10.3 unnoticed, breaking *a lot* of embedded hardware
with regard to e.g. CARP, then went into 11.0.  It took a long time to identify
the commit that was responsible and unwind it (it was a batch MFC after all).

Settling for anything other than what is attached to this ticket is risky,
especially if there is nobody who can test and verify this in over one and a
half years.


Cheers,
Franco

-- 
You are receiving this mail because:
You are the assignee for the bug.