bhyve: Clear lid to 0 for internal device errors for NVMe AENs.

[jhb]

Summary

Reported by: GCC

[chuck]

These changes are good but not really necessary from a functionality point of view. In both cases, the emulation is returning an error, making the LID undefined from the guest's perspective. But I'm all for good hygiene.

[imp]

+1 on chuck's results. This can be anything.... but unassigned variables are allowed to have any behavior when accessing them now (well, always, but usually it was just some random value)... So I'm good with the change. It might be simpler to assign lid=0 before the case statement, though.

[jhb]

I think the real issue here is avoiding leaking host hypervisor stack memory to a guest, not if NVMe clients care about the value of lid for these specific cases.