Re: specfs/spec_vnops.c diagnostic assertion panic

[Robert Elz <kre%munnari.OZ.AU@localhost>]

    Date:        Thu, 11 Aug 2022 13:01:34 +0000
    From:        Taylor R Campbell <campbell+netbsd-current-users%mumble.net@localhost>
    Message-ID:  <20220811130139.B310860843%jupiter.mumble.net@localhost>

  | If you cvs up spec_vnops.c to 1.211, you may get a different assertion
  | firing, or not; which assertion fires now might potentially help to
  | diagnose the problem.

I shall do that - but don't hold your breath waiting on results, as I
indicated, this does not happen often, and if there's a trigger, I have
no idea what it is.

  | If that still doesn't help, you could try the attached patch (which
  | I'm not committing at the moment because it changes the kernel ABI).

I can try that too (from the subsequent msg) (the ABI change won't affect me).

  | If so, the `sd_closing=0xffffbc0013315500' part gives you an lwp
  | address, which you can feed to `bt/a' to get a stack trace:
  |
  | db> bt/a 0xffffbc0013315500

So I guess I better include DDB in the kernel as well...  (normally I
don't, as in general the system is always running X, and I've never
had any luck accessing ddb when the system crashes while X is running.
Not an issue for this one - this crash happens, when it happens, way
too early for X to be anywhere near running).

kre