System and CPU hanging on RPi

[John Klos <john%ziaspace.com@localhost>]

I often forget what the status of support for each model is in NetBSD. I 
have a number of RPi2 and RPi3 systems which are happy and run fine with 
NetBSD 8.

I recently tried aarch64-current on an RPi3B+ (too many qualifiers - the 
latest, gigabit RPi), but the whole system hung while CVSing pkgsrc. This 
was from, I believe, May 5th, 2019, or the 6th.

I then tried earmv7hf-current on the same system from the same date, and 
it, too, hung the whole system. I couldn't get in to the debugger in 
either case (janky keyboard?), and I don't have a serial console attached.

So then I tried NetBSD-8 from May 9th, 2019. I then realized that mue* 
isn't working in NetBSD-8, so I connected an axe* gigabit ethernet to USB 
adapter. It CVSed just fine, then proceeded to compile a number of 
packages without issues.

However, it did hang in a different way after a while:

load averages:  0.55,  0.24,  0.11;               up 0+06:48:34    23:50:2
39 processes: 36 sleeping, 3 on CPU
CPU0: 53.1% user,  0.0% nice, 29.1% system,  0.0% interrupt, 17.8% idle
CPU1:  0.0% user,  0.0% nice,  0.0% system,  0.0% interrupt,  0.0% idle
CPU2:  1.8% user,  0.0% nice,  1.6% system,  0.0% interrupt, 96.6% idle
CPU3: 10.0% user,  0.0% nice,  4.2% system,  0.0% interrupt, 85.8% idle
Memory: 499M Act, 119M Inact, 11M Wired, 11M Exec, 569M File, 34M Free
Swap: 2048M Total, 2048M Free

Notice that CPU1 is 0.0% across the board. A sleep command from a 
configure was sleeping on CPU/1, and killing it allowed the running 
compile to continue.

But what's going on with CPU1?

John