Re: kern/54908: kernel semi-lock-up

[Andrew Doran <ad%netbsd.org@localhost>]

The following reply was made to PR kern/54908; it has been noted by GNATS.

From: Andrew Doran <ad%netbsd.org@localhost>
To: gnats-bugs%netbsd.org@localhost
Cc: 
Subject: Re: kern/54908: kernel semi-lock-up
Date: Thu, 30 Jan 2020 19:07:26 +0000

 Other things that would be useful:
 
 - Straight "bt" to trace back through the interrupt into the CPU hog would
   be good.
 
 - If arm supports "mach cpu 0" / "mach cpu 1" to teleport onto the other CPU
   and do a "bt" there that would be enlightening.  Trace by LWP generally
   doesn't work on running LWPs because the values saved in the PCB are out
   of date.
 
 - If any of that produces decent results, the routine / line of code where
   the hang in pgdaemon / softclk is.
 
 Andrew