Re: Regressions

[Jason Thorpe <thorpej%me.com@localhost>]

> On Mar 1, 2020, at 5:26 AM, Andreas Gustafsson <gson%gson.org@localhost> wrote:
> 
> 55018 atf tests for pppoe sometimes leave rump_server processes around

I'm trying to think of a solution for this one, but this seems squarely on rump, and not an actual issue with the core code that's being tested, FWIW.

The issue seems to be that rump really wants to join threads that are created for work queues when the rump server exits.  But in this particular case, there's a global work queue that never goes away because in the real kernel, there's no need to do this before the system reboots / shuts down.  Any change to fix this will be 100% for the appeasement of rump.

-- thorpej