kern/55090: block with spin-mutex held assertion when bringing iwn(4) up

[venture37%geeklan.co.uk@localhost]

>Number:         55090
>Category:       kern
>Synopsis:       block with spin-mutex held assertion when bringing iwn(4) up
>Confidential:   no
>Severity:       serious
>Priority:       medium
>Responsible:    kern-bug-people
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Fri Mar 20 03:40:00 +0000 2020
>Originator:     Sevan Janiyan
>Release:        NetBSD 9.99.50
>Organization:
>Environment:
NetBSD 9.99.50 amd64 x86_64

>Description:
Brought iwn interface down, stopped dhcpcd, let some time pass and brought the interface back up. System panicked with 
kernel diagnostic assertion "ci->ci_mtx_count == -1" failed: file "/usr/src/sys/kern/kern_synch.c", line 740 mi_switch: cpu2: ci_mtx_count (-2) != -1 (block with spin-mutex held)

bt
vpanic()
kern_assert()
mi_switch()
sleepq_block()
iwn5000_load_firmware()
iwn_hw_init()
iwn_init()
ifmedia_change()
ifmedia_ioctl()
ieee80211_ioctl()
iwn_ioctl()
doifioctl()
sys_ioctl()
syscall()
--- syscall (number 54) ---

Photo of panic:
https://pbs.twimg.com/media/EThRIqFUMAAtRtp?format=jpg&name=orig

>How-To-Repeat:
flap iwn(4) with a delay in between up and down states or some minutes?
>Fix: