kern/55154: Running sysctl test on macppc in qemu triggers panic

[gson%gson.org@localhost (Andreas Gustafsson)]

>Number:         55154
>Category:       kern
>Synopsis:       Running sysctl test on macppc in qemu triggers panic
>Confidential:   no
>Severity:       serious
>Priority:       medium
>Responsible:    kern-bug-people
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Wed Apr 08 07:30:01 +0000 2020
>Originator:     Andreas Gustafsson
>Release:        NetBSD-current, source date >= 2020.03.16.21.20.13
>Organization:

>Environment:
System: NetBSD
Architecture: powerpc
Machine: macppc
>Description:

>How-To-Repeat:

Running the ATF tests on -current/macppc under qemu, the guest system
panics running the sbin/sysctl/t_sysctl/basic test case:

  sbin/sysctl/t_sysctl (544/843): 1 test cases
      basic: [ 25617.0219410] trap: kernel read DSI trap @ 0xfff6c528 by 0x125f40 (DSISR 0x40000000, err=14), lr 0x3cd078
  [ 25617.0219410] panic: trap
  [ 25617.0219410] cpu0: Begin traceback...
  [ 25617.0219410] 0x034a2b60: at 0x6bf0e4
  [ 25617.0219410] 0x034a2b90: at 0x6bf1e8
  [ 25617.0219410] 0x034a2bd0: at 0x130fdc
  [ 25617.0219410] 0x034a2c80: kernel DSI read trap @ 0xfff6c528 by 0x125f40: srr1=0x9032
  [ 25617.0219410]	     r1=0x34a2d50 cr=0x24008844 xer=0 ctr=0 dsisr=0x40000000
  [ 25617.0219410] 0x034a2d50: at 0x694f38
  [ 25617.0219410] 0x034a2e10: at 0x696fbc
  [ 25617.0219410] 0x034a2e50: at 0x697208
  [ 25617.0219410] 0x034a2ec0: at 0x127e80
  [ 25617.0219410] 0x034a2f20: user SC trap #202 by 0xfde215e0: srr1=0xf032
  [ 25617.0219410]	     r1=0xffffd470 cr=0x44004844 xer=0 ctr=0xfde215d8
  [ 25617.0219410] cpu0: End traceback...
  [ 25617.0219410] dumpsys: TBD
  [ 25617.0219410] rebooting

A bisection points to the commit by pgoyette at 2020.03.16.21.20.09
through 2020.03.16.21.20.09 with the commit message

  Use the module subsystem's ability to process SYSCTL_SETUP() entries to
  automate installation of sysctl nodes.

  Note that there are still a number of device and pseudo-device modules
  that create entries tied to individual device units, rather than to the
  module itself.  These are not changed.

Logs are at:

  http://www.gson.org/netbsd/bugs/build/macppc/commits-2020.03.html#2020.03.16.21.20.13

Martin's macppc tests running on real hardware are not affected:

   http://www.netbsd.org/~martin/macppc-atf/

>Fix: