Re: softints stuck for 16 seconds: stuck in swapping?

[Taylor R Campbell <riastradh%NetBSD.org@localhost>]

> From: Thomas Klausner <wiz%netbsd.org@localhost>
> Date: Fri, 26 Jun 2026 19:21:26 +0200
> 
> In a system with very high disk load (swapping a lot), could it be
> possible that
> 
> panic: cpu0: softints stuck for 16 seconds
> 
> happens even though it would recover?

Sure, it's possible the system would recover.  But if there is any
softing waiting for swap I/O, that in itself is a bug.  And if there
is any softint doing that much computation on a CPU before yielding to
lower-priority softints, there's probably something else wrong.

If you drop into ddb, can you get stack traces from:

1. all the softint lwps (ps to find them, then bt/a 0x...), and
2. all the CPUs (mach cpu 0; bt; mach cpu 1; bt; ...)?