Re: Current test failures

[Taylor R Campbell <campbell+netbsd-current-users%mumble.net@localhost>]

> Date: Sat, 7 Dec 2019 11:19:49 +0200
> From: Andreas Gustafsson <gson%NetBSD.org@localhost>
> 
> Martin Husemann wrote:
> > Here is a simple recipe to reproduce the massive test lossage in -current:
> > 
> > 	cd /usr/tests/dev/raidframe && atf-run
> 
> I have now bisected it down to the following commits:
> 
>   2019.12.05.03.21.08 riastradh src/sys/kern/subr_percpu.c 1.20

Unlikely to be relevant -- this only makes an assertion fire in fewer
circumstances, so it can't reasonably cause _more_ crashes.

>   2019.12.05.03.21.17 riastradh src/sys/kern/subr_pserialize.c 1.16

OOPS -- rmind removed pserialize_init from rump_init, so the mutex
never got initialized.  Fixed in rump.c 1.337!

>   2019.12.05.03.21.29 riastradh src/sys/kern/subr_pserialize.c 1.17

Unlikely to be relevant -- this only makes an evcnt attach statically
rather than in pserialize_init.

>   2019.12.05.03.21.42 riastradh src/external/cddl/osnet/sys/sys/opentypes.h 1.5

Can't imagine how this could be relevant -- only affects the tools
build!