Castor Fu wrote:
> I got a panic in suspendsched() (lost the dump and the panic strings)
> on NetBSD-i386 using the 1.5 branch as of about last sunday.  I haven't
> been able to reproduce this yet.  
> 
> This code is all signficantly newer than what we run internally.  Has
> anyone seen anything similar?

Yes, I've seen this once on reboot on i386, NetBSD 1.5E (-current) kernel.
IIRC it was classic trap()-type panic. It happened when
the final sync started, before any info about synced buffers was
printed. I'm not using MFS or any other "special" filesystem,
just FFS & NTFS.

Does the proclist walking in suspendsched() need to be guarded by
proclist_(un)lock_write() ?

Jaromir
-- 
Jaromir Dolecek <jdolecek@NetBSD.org>      http://www.ics.muni.cz/~dolecek/
@@@@  Wanna a real operating system ? Go and get NetBSD, damn!  @@@@