Subject: [wrstuden@netbsd.org: Re: Recent macppc kernels hang under load]
To: None <port-macppc@netbsd.org>
From: Bill Studenmund <wrstuden@netbsd.org>
List: port-macppc
Date: 10/01/2003 15:14:06
--WfZ7S8PLGjBY9Voh
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

Sorry for the mis-reply. I'm still learning mutt.

Take care,

Bill
----- Forwarded message from Bill Studenmund <wrstuden@netbsd.org> -----

Date: Wed, 1 Oct 2003 15:10:39 -0700
From: Bill Studenmund <wrstuden@netbsd.org>
To: David Young <dyoung@pobox.com>
Cc: tech-kern@netbsd.org
Subject: Re: Recent macppc kernels hang under load

On Tue, Sep 30, 2003 at 12:27:42AM -0500, David Young wrote:
> On Mon, Sep 29, 2003 at 11:52:56PM -0500, Dave Huang wrote:
> > I have a G4 dual 500MHz, IDE disk with a single UFS partition for
> > NetBSD mounted on / (and an HFS+ one for OS X too). /usr/src mounted
> > over UDP NFS, 100BaseTX, gem ethernet driver. /usr/src.local union
> > mounted on top of /usr/src.
>=20
> The union mount is probably the problem: unionfs is buggy and it
> causes hangs for me all the time.

1) Then why didn't you say anything when I asked about unionfs issues
on -current a while back? :-)

2) unionfs should not cause problems that will hang the kernel such that
the programmer switch doesn't work. unionfs should not mess w/ interrupts.
You can deadlock processes, and if you start the race for root, you can
wedge the whole userland, but the programmer switch (which reacts to
interrupts as I understand it) should still get you to ddb.

Take care,

Bill



----- End forwarded message -----

--WfZ7S8PLGjBY9Voh
Content-Type: application/pgp-signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (NetBSD)

iD8DBQE/e1GtWz+3JHUci9cRArgGAKCD2wrA6RfGi8KDbM5sjvC1kwKm7QCfaM5w
7T7sErsiLDjDNtyHO472pAU=
=AxHb
-----END PGP SIGNATURE-----

--WfZ7S8PLGjBY9Voh--