From alsa-user-owner@alsa.jcu.cz  Thu Jan  7 03:29:45 1999
Received: from executioner.lis.net.au (mail@executioner.lis.net.au [203.35.83.3])
	by marvin.jcu.cz (8.9.1a/8.9.1) with ESMTP id DAA02118
	for <alsa-user@alsa.jcu.cz>; Thu, 7 Jan 1999 03:27:19 +0100
Received: from uucp by executioner.lis.net.au with local-rmail (Exim 2.05 #4)
	id 0zy5AK-0002iX-00 (Debian); Thu, 7 Jan 1999 13:27:16 +1100
Received: from kerry by gotss1.gotss.ml.org with local (Exim 1.92 #1)
	id 0zy4xE-0000Ls-00 (Debian); Thu, 7 Jan 1999 13:13:44 +1100
Message-ID: <19990107131344.B1317@gotss.ml.org>
Date: Thu, 7 Jan 1999 13:13:44 +1100
From: Kerry Hoath <kerry@gotss.ml.org>
To: alsa-user@alsa.jcu.cz
Subject: Re: ALSA and GUS MAX, on kernel 2.0.34 / Debian 2.0
References: <19990105141426.A9981@gotss.ml.org> <XFMail.990106162811.scode_ufp@usa.net>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
In-Reply-To: <XFMail.990106162811.scode_ufp@usa.net>; from Peter Schuller on Wed, Jan 06, 1999 at 04:28:11PM +0100
Reply-To: alsa-user@alsa.jcu.cz
Sender: alsa-user-owner@alsa.jcu.cz
Precedence: list

I don't understand low-level sound programming but here is what basically 
happens:
An application dumps a buffer of digital audio to be played. The driver says,
"Start a DMA transferr and tell me when the transferr is finished or to a
certain point".
Guess how that happens? The soundcard triggers an interrupt that is picked
up by the pc. Basically the soundcard says,
"I've finished... eyes this way... attention to me" <smile>

Check your /proc/pci and make sure nothing is using IRQ5. If a PCI device
grabs it, the ISA cards will appear never to generate interrupts. Symptoms
of this are looping sound fragments, and messages to kernel log.

Regards, Kerry.
On Wed, Jan 06, 1999 at 04:28:11PM +0100, Peter Schuller wrote:
> > This is an interesting situation, I am running Debian 2.0 and my 2 gusmax
> > cards operate flawlessly. Was the /proc/interrupts snapshot taken after you
> > had played sounds?
> 
> Yes it was. After at least a few sessions of Q2 :)
> 
> > It's a longshot but I noticed that the interrupt count
> > on IRQ5 was zero. Does this count increase when you play sound?
> 
> Apperantly not. Just to be sure, I just entered Q2 again, made some noice and
> checked it again. Still 0. And it didn't help to try one of those applications
> in which I have problems. Still 0.
> 
> But is it really supposed to increase? I don't know much about hardware, but
> isn't the interrupts there to make it possible for the soundcard so speak to
> the comp., so to speak? Isn't the DMA channel used when playing sounds, i.e.
> when passing data on TO the soundcard?
> 
> Ah well, that is probably 100% incorrect...
> 
> > My first
> > hunch is if the sound loops it is an interrupt conflict. I know I had to
> > go into the CMOS setup, PCI/Plug 'n' play, set resources assigned by manual
> > and set IRQ5 (the one that has the gusmax on it) to Legacy ISA.
> 
> Well first of all, as I said, it worked fine before, using the same HW, ICQ and
> DMA. And secondly, my comp. is an old P133, so the BIOS doesn't have that kind
> of settings yet, so I can't try chaning that. It does have PnP though, but the
> MAX doesn't support it, so I guess that can't be the problem?
> 
> And btw, I am not setting the DMA/IRQ manually. It's autodetected. I just do
> "modprobe snd-gusmax".
> 
> Anyway, thanks a lot!
> 
> / Peter Schuller
> ----------------
> E-Mail: scode_ufp@usa.net
> Web: http://hem.passagen.se/petersch
> 
> ------
> To unsubscribe from <alsa-user@alsa.jcu.cz> mailing list send message
> 'unsubscribe' in the body of message to <alsa-user-request@alsa.jcu.cz>.
> 

-- 
--
Kerry Hoath:
kerry@gotss.ml.org or khoath@lis.net.au
ICQ UIN: 8226547


