Subject: Re: Consistent vnlock dead locks in low memory conditions
To: Stephen Jones <smj@cirr.com>
From: Bill Studenmund <wrstuden@netbsd.org>
List: tech-net
Date: 06/22/2006 19:25:25
--r5Pyd7+fXNt84Ff3
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Thu, Jun 22, 2006 at 09:59:30AM -0700, Stephen Jones wrote:
> I have a file server that used to operate fairly reliably with 1GB of =20
> RAM.  A few weeks ago the memory tested bad and the system would not =20
> boot, however the only known spares available were two 128MB modules =20
> giving me 256MB of ram.  Since bringing the server up with this =20
> amount of ram, I am able to achieve multiple vnlock-deadlocks

Please note that you need to do a bit more digging to really say it's a
vnlock-deadlock. I'm certainly not trying to say you aren't having a hang,
just that "deadlock" usually implies a loop in locking (thread A has lock=
=20
P and is sleeping for Q, while thread B has Q and is sleeping for P).

The reason I mention this is it could well be that something's holding a=20
vnode lock but is in fact sleeping for something else which is blocked.=20
Thus the process that is really the problem is NOT in vnlock. :-|

Take care,

Bill

--r5Pyd7+fXNt84Ff3
Content-Type: application/pgp-signature
Content-Disposition: inline

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

iD8DBQFEm1EUWz+3JHUci9cRArz7AJwLSz/+WJB93wzUUHA7H6HTLCrJoACdFb1q
i+8hf7A5KCQ0x1PR8Liv9ro=
=okVj
-----END PGP SIGNATURE-----

--r5Pyd7+fXNt84Ff3--