[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Bug#434571: marked as done (Module snd-sb16 set interrupts wrong)

Your message dated Mon, 17 Aug 2009 19:01:37 +0200
with message-id <20090817170137.GA21492@inutil.org>
and subject line Re: Module snd-sb16 set interrupts wrong
has caused the Debian Bug report #434571,
regarding Module snd-sb16 set interrupts wrong
to be marked as done.

This means that you claim that the problem has been dealt with.
If this is not the case it is now your responsibility to reopen the
Bug report if necessary, and/or fix the problem forthwith.

(NB: If you are a system administrator and have no idea what this
message is talking about, this may indicate a serious mail system
misconfiguration somewhere. Please contact owner@bugs.debian.org
immediately.)


-- 
434571: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=434571
Debian Bug Tracking System
Contact owner@bugs.debian.org with problems
--- Begin Message --- 
Package: linux-image
Version: 2.6.21-2

I have the Sound Blaster 16 OEM card CT1790 which I have mainly used
to play CDs thru.  It has jumpers to set IRQs and it's set at irq2.  I
seem to only be able to get it to work OK at irq5 which is the default
for the card.  Except that my Modem is on irq 5.  That's why I set the
jumper to irq2.  Now the card, while not PnP, can have it's interrupts
set by software (by Linux software ?).  Well, yes, because when I load
the snd-sb16 module for it and specify irq=2 I find from
/proc/interrupts that it's been set to irq 5 and the sound card works
OK (at least with .wav files).  So the linux software must have changed
the irq 2 (that I jumpered) to irq 5.  This happens when I haven't
used my modem yet and irq 5 is still available.

But now I can't use my modem.  If I remove the module, irq 5 is
available (per /proc/int*) and the modem will grab irq 5 and try to
dial.  But I get all sorts of errors due the the shorting out of the
irq 5 line by the sound card since irq 5 is still set in the sound
card even though the sound card module has been removed and irq 5 has
been released by the kernel.  My sound card is working right since
it's sending 0 volts (a grounded line --actually low impedance) on the
irq 5 line to indicate that it isn't sending any interrupts.  So the
kernel needs to have a list of resources that exist in hardware but
are not currently supported by any loaded module, or the like.

Once the interrupt has been changed to 5, the only way I know to get
it back to 2 is to power-down (turn off the power) and then boot again.
This is not a power-on reboot.

Now, if when loading the snd-sb16 module after I start the modem so
that irq 5 is in use with irq=2 on the command line, /proc/int*
does show that irq 9 is for SoundBlaster (since irq 2 and 9 are
really the same irq).  OK so far but irq 9 is never sent by the
sound-card when I try to use it, although the module is apparently
listening for it.  Nor is irq 5 sent either as may be checked by
starting the modem and checking the number of interrupts sent per
/proc/int* after trying to play a .wav file.  Playing a .wav file only
plays the first couple of words since apparently no interrupts are
sent.

So it seems to me that the snd-sb16 module may not be able to change
the interrupts except perhaps to reset the card so that the default
irq 5 is set it the card, overriding what may be set by jumpers.
If the driver has such limitations, they need to be either fixed or
documented.

			David Lawyer

--- End Message ---
--- Begin Message --- 
Version: 2.6.26-1

David Lawyer wrote:
> > > So it seems to me that the snd-sb16 module may not be able to change
> > > the interrupts except perhaps to reset the card so that the default
> > > irq 5 is set it the card, overriding what may be set by jumpers.
> > > If the driver has such limitations, they need to be either fixed or
> > > documented.
> > 
> > Do you still own that card?
>
> > There are several changes in git related to the interrupt handling
> > of non-pnp Soundblaster cards. Does this still occur with more
> > recent kernels (2.6.26 from Lenny or later kernels from unstable)?
>
> Yes but I'm no longer using it since I'm using another PC for
> multimedia.  On my old PC which I'm using now it's 2.6.26 but I'm
> using it from a dumb terminal that doesn't have sound so it's not easy
> to check out.  So I may never get around to looking into this.

Ok, since this is most likely fixed upstream, I'm marking the bug fixed
in the Lenny version. If you ever get around to test this on the hardware
again and you find it non-working, please reopen the bug.

Thanks,
        Moritz

--- End Message ---
Reply to: