ac97, alc101+kt8235 sound

ac97, alc101+kt8235 sound

[Benson Chow]

I guess this ac97 stuff is pretty confusing.

Anyway, my motherboard has a kt8235 southbridge and an ALC101 AC97
decoder.  I read in some posting to hack the ALC101 into the
drivers/sound/ac97_codec.c with something like this:

        {0x414C4730, "ALC101",             &null_ops},

and then use the via82cxxx_audio driver.  It didn't do anything.  So, I
tried hacking the PCI device number in via82cxxx_audio.c by changing

static struct pci_device_id via_pci_tbl[] __initdata = {
        { PCI_VENDOR_ID_VIA, PCI_DEVICE_ID_VIA_82C686_5,
          PCI_ANY_ID, PCI_ANY_ID, },
        { 0, }
};

to

static struct pci_device_id via_pci_tbl[] __initdata = {
        { PCI_VENDOR_ID_VIA, PCI_DEVICE_ID_VIA_8233_5,
          PCI_ANY_ID, PCI_ANY_ID, },
        { 0, }
};

hoping that these via chips were pretty close.  Unfortunately no, it
still doesn't work.  It did, however, find the AC97 codec fine (I added
some printk's), but no sound is produced.  Any ideas on how to get this
vt8235-based motherboard sound working?  (and ALSA-0.9.2 seems to do
nothing but segfault it seems.)

Running the 2.4.20 kernel on an ecs p4vxasd2+ board (yeah, I know...)

Thanks,

-bc

WARNING: All HTML emails get deleted.  DO NOT SEND HTML MAIL.

Re: ac97, alc101+kt8235 sound

[Alan Cox]

On Maw, 2003-04-15 at 12:49, Benson Chow wrote:
> hoping that these via chips were pretty close.  Unfortunately no, it
> still doesn't work.  It did, however, find the AC97 codec fine (I added
> some printk's), but no sound is produced.  Any ideas on how to get this
> vt8235-based motherboard sound working?  (and ALSA-0.9.2 seems to do
> nothing but segfault it seems.)

See 2.4.21pre - that has the driver for VIA8233/5

Re: ac97, alc101+kt8235 sound

[Benson Chow]

Cool, my ecs p4vxasd2+ (p4x333/kt8235/alc101) works with onboard sound
now.

What's the normal flow to get this added into ac97_codecs.c?

+        {0x414C4730, "ALC101",             &null_ops},

Adding this line into the table in ac97_codecs.c (with a few missing
#defines fixed... then I noticed they're already in -ac1 :) in
2.4.21-pre7 made sound work fine.

Now hopefully those of us who bought those Fry's "free" p4vxasd2+ 5.0's
can have sound.  I just mplayer'ed an avi with the new table entry and it
appears to work fine.  Volume controls look functional.  Will do some more
testing (record, etc.) tonight.

Thanks,

-bc

On 15 Apr 2003, Alan Cox wrote:

> Date: 15 Apr 2003 12:36:51 +0100
> From: Alan Cox <alan@lxorguk.ukuu.org.uk>
> To: Benson Chow <blc+lkml@q.dyndns.org>
> Cc: Linux Kernel Mailing List <linux-kernel@vger.kernel.org>
> Subject: Re: ac97, alc101+kt8235 sound
>
> On Maw, 2003-04-15 at 12:49, Benson Chow wrote:
> > hoping that these via chips were pretty close.  Unfortunately no, it
> > still doesn't work.  It did, however, find the AC97 codec fine (I added
> > some printk's), but no sound is produced.  Any ideas on how to get this
> > vt8235-based motherboard sound working?  (and ALSA-0.9.2 seems to do
> > nothing but segfault it seems.)
>
> See 2.4.21pre - that has the driver for VIA8233/5
>

WARNING: All HTML emails get deleted.  DO NOT SEND HTML MAIL.

Re: ac97, alc101+kt8235 sound

[Alan Cox]

On Maw, 2003-04-15 at 18:09, Benson Chow wrote:
> What's the normal flow to get this added into ac97_codecs.c?
> 
> +        {0x414C4730, "ALC101",             &null_ops},
> 
> Adding this line into the table in ac97_codecs.c (with a few missing
> #defines fixed... then I noticed they're already in -ac1 :) in
> 2.4.21-pre7 made sound work fine.

This really shouldnt make any difference. Does it work without that
patch as well ?

Re: ac97, alc101+kt8235 sound

[Benson Chow]

You're right, it's not necessary.

It just reports as 'Unknown' but works same as with the patch.

After some more testing, it plays fine, mixer works mostly, minus the
issue with recording.  Attempting to record seems to inject 60 cycle hum
into the system and recorded output either bombs out or is just the hum -
no line-in signal is recorded.  Looks like the realtek alc101 has some
quirks...

-bc

On 15 Apr 2003, Alan Cox wrote:

> Date: 15 Apr 2003 17:28:29 +0100
> From: Alan Cox <alan@lxorguk.ukuu.org.uk>
> To: Benson Chow <blc@q.dyndns.org>
> Cc: Linux Kernel Mailing List <linux-kernel@vger.kernel.org>
> Subject: Re: ac97, alc101+kt8235 sound
>
> On Maw, 2003-04-15 at 18:09, Benson Chow wrote:
> > What's the normal flow to get this added into ac97_codecs.c?
> >
> > +        {0x414C4730, "ALC101",             &null_ops},
> >
> > Adding this line into the table in ac97_codecs.c (with a few missing
> > #defines fixed... then I noticed they're already in -ac1 :) in
> > 2.4.21-pre7 made sound work fine.
>
> This really shouldnt make any difference. Does it work without that
> patch as well ?
>
>
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/
>

WARNING: All HTML emails get deleted.  DO NOT SEND HTML MAIL.

Re: ac97, alc101+kt8235 sound

[Mark Rutherford]

Since we are on this subject...

I tried this kernel about 2 days ago, and had unexpected results at times.
1. the kernel would lock up, and so dead that not even the sysrq key did
anything (I usually turn that debugging stuff on)
2. it would work, but more than 1 application accessing the sound would cause
the kernel or give it up and once again, lock up.
3. it would just lock up the second app trying to gain access to the sound.
I do not have any output from the kernel, it crashed that hard (why..?)
first thing I did was apply the 2.4.21 pre patch, then the -ac patch to a
2.4.20 tree, this right?
anyone has any ideas on getting this to work right, and have at least more
than 1 application access the card, let me know.
im trying to get quake + teamspeak to co-exist.


Alan Cox wrote:

> On Maw, 2003-04-15 at 12:49, Benson Chow wrote:
> > hoping that these via chips were pretty close.  Unfortunately no, it
> > still doesn't work.  It did, however, find the AC97 codec fine (I added
> > some printk's), but no sound is produced.  Any ideas on how to get this
> > vt8235-based motherboard sound working?  (and ALSA-0.9.2 seems to do
> > nothing but segfault it seems.)
>
> See 2.4.21pre - that has the driver for VIA8233/5
>
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/

--
Regards,
Mark Rutherford
mark@justirc.net

PGP key: http://www.justirc.net/~mark/markrutherford.asc
fingerprint: 1CF2 6229 306D A2C8 2C89  46BE FFD6 D910 5170 4FA9

Re: ac97, alc101+kt8235 sound

[Alan Cox]

On Iau, 2003-04-17 at 02:08, Mark Rutherford wrote:
> 2. it would work, but more than 1 application accessing the sound would cause
> the kernel or give it up and once again, lock up.

The lockup reports are not something I've seen before. As to multiple
applications - use ALSA, the OSS driver doesn't currently support the
secondary directsound channels.

I'll try and duplicate the multiple app hangs but I've not seen those
reported so far.

Re: ac97, alc101+kt8235 sound (2.4.21-pre7-ac1)

[Benson Chow]

Seems that the VT8235 has some issues with ac97 recording.  Using aumix,
and setting the input to line-in, cat /dev/dsp returns the same byte over
and over again, as if there is no signal.  Dumping this file and dumping
back to /dev/dsp confirms the silence - no sound got record.  Messing
around with the mixer doesn't appear to do much.  The mixer subsystem, I
can control the volume of line-in and line-out, and hear the volume going
up and down.  I also tried Windows, seems it works fine in Windows.

I tried the same command sequence on my SiS735 (i810-compatible) AC97
machine and the results differ (it works :-)  So it seems that there's
something that's wrong with the kt8235 side of things?

Playback works fine.  Just cannot record.  Is the kt8235 driver not quite
primetime yet?

Thanks,

-bc

On 15 Apr 2003, Alan Cox wrote:

> Date: 15 Apr 2003 12:36:51 +0100
> From: Alan Cox <alan@lxorguk.ukuu.org.uk>
> To: Benson Chow <blc+lkml@q.dyndns.org>
> Cc: Linux Kernel Mailing List <linux-kernel@vger.kernel.org>
> Subject: Re: ac97, alc101+kt8235 sound
>
> On Maw, 2003-04-15 at 12:49, Benson Chow wrote:
> > hoping that these via chips were pretty close.  Unfortunately no, it
> > still doesn't work.  It did, however, find the AC97 codec fine (I added
> > some printk's), but no sound is produced.  Any ideas on how to get this
> > vt8235-based motherboard sound working?  (and ALSA-0.9.2 seems to do
> > nothing but segfault it seems.)
>
> See 2.4.21pre - that has the driver for VIA8233/5
>

WARNING: All HTML emails get deleted.  DO NOT SEND HTML MAIL.

Re: ac97, alc101+kt8235 sound (2.4.21-pre7-ac1)

[Alan Cox]

On Gwe, 2003-04-18 at 04:37, Benson Chow wrote:
> I tried the same command sequence on my SiS735 (i810-compatible) AC97
> machine and the results differ (it works :-)  So it seems that there's
> something that's wrong with the kt8235 side of things?
> 
> Playback works fine.  Just cannot record.  Is the kt8235 driver not quite
> primetime yet?

Not owning a microphone I've never tested that side, I assumed that
since it worked on the 686 it worked on the 8233/5 8)