Subject: Re: Recording with wss(4) hardware seems to be broken in
To: Brian Buhrow <buhrow@lothlorien.nfbcal.org>
From: Jonathan A. Kollasch <jakllsch@kollasch.net>
List: port-i386
Date: 08/09/2007 22:29:47
--2VXyA7JGja7B50zs
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Thu, Aug 09, 2007 at 10:32:47AM -0700, Brian Buhrow wrote:
> 	Hello.  I have an old Dell Inspiron 3500 which has been running NetBSD
> for years.  For many years I used it as my digital audio recording machin=
e,
> for making Cds from vinyl records, recording meetings and the like.  About
> a year ago, I upgraded it to NetBSD-3, but didn't have occasion to use its
> recording capabilities.  Recently, I began work on a project which will
> require its trusty service again.  When I pulled it out to test it, I fou=
nd
> that recording didn't work with its 3.x kernel.  All recording processes
> would get stuck in "aud_rd" and no bytes would be written to disk.
> Playing, however, seems to work fine.  Thinking that perhaps because the
> machine had been sitting idle for a long time, perhaps it has just died, I
> downgraded it to NetBSD-2.0, where recording works as expected.
>=20
> 	Here's what the attachment looks like under both NetBSD-2.x and 3.x
>=20
>=20
> wss1 at pnpbios0 index 17 (CSC0000, WSS/SB)
> wss1: io 530-537 388-38b 220-22f, irq 5, DMA 1 0
> wss1: CS4237B
> audio0 at wss1: full duplex, mmap
> opl0 at wss1: model OPL3
> midi0 at opl0: WSS Yamaha OPL3
>=20
> Has anyone else experienced this same problem when going from NetBSD-2.x =
to
> 3.x?  Recording with other audio(4) drivers seems to work OK.  Does anyone
> have any ideas where I might look to solve this issue?  I can live with
> NetBSD-2.x on this box, but it would be nice to know that the problem is
> fixed going forward after 2.x.  I have not tried a 4.x kernel, perhaps I
> should.

Well, wss(4) recording doesn't work here in 4.99.26.  audiorecord
ends up getting a EINVAL in the setup phase and bails out.
cat(1)ing /dev/sound acts about like a blocking /dev/null.
vmstat -i shows the device interrupting while trying
to record.

There have been some seemingly significant modifications
between 2.x and 3.x in the CVS logs.  It's hard
(well, harder than I want to work on this now)
to track as there are like 3-7 files that wss(4) comprises of.

> Any thoughts would be greatly appreciated.

Good luck.  Also, ISA hardware isn't fun anymore.

	Jonathan Kollasch

--2VXyA7JGja7B50zs
Content-Type: application/pgp-signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.7 (NetBSD)

iD8DBQFGu9urOjx1ye3hmokRAiphAKCUhEVzesDy/OXVycNGFkitPLyWVACfTJPB
S6mmontnV7qU4cWeiza0kNk=
=zSrT
-----END PGP SIGNATURE-----

--2VXyA7JGja7B50zs--