Re: 2.6.12-rc1-mm3, sound card lost id

Re: 2.6.12-rc1-mm3, sound card lost id

[Jean Delvare]

> ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.12-rc1/2.6.12-rc1-mm3/
> (...)
>  bk-alsa.patch

This one made /proc/asound/card0/id change from "Live" to "Unknown" on
one of my systems, preventing alsatcl from properly restoring my mixer
settings.

I guess this wasn't exactly expected?

00:0d.0 Multimedia audio controller: Creative Labs SB Live! EMU10k1 (rev 06)
        Subsystem: Creative Labs CT4832 SBLive! Value
        Flags: bus master, medium devsel, latency 48, IRQ 5
        I/O ports at 8800 [size=32]

Class:     0401
Device:    1102:0002
Subsystem: 1102:8027

Thanks,
-- 
Jean Delvare


-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click

Re: Re: 2.6.12-rc1-mm3, sound card lost id

[Takashi Iwai]

At Sat, 26 Mar 2005 11:19:45 +0100,
Jean Delvare wrote:
> 
> > ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.12-rc1/2.6.12-rc1-mm3/
> > (...)
> >  bk-alsa.patch
> 
> This one made /proc/asound/card0/id change from "Live" to "Unknown" on
> one of my systems, preventing alsatcl from properly restoring my mixer
> settings.
> 
> I guess this wasn't exactly expected?

Hmm, perhaps it's a side effect of chip detection patch by James.
But "Unknown" is bad, of course.

How does /proc/asound/cards look?


Takashi


-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click

Re: Re: 2.6.12-rc1-mm3, sound card lost id

[Jean Delvare]

Hi Takashi,

> > This one made /proc/asound/card0/id change from "Live" to "Unknown"
> > on one of my systems, preventing alsatcl from properly restoring my
> > mixer settings.
> 
> Hmm, perhaps it's a side effect of chip detection patch by James.
> But "Unknown" is bad, of course.
> 
> How does /proc/asound/cards look?

0 [Unknown        ]: EMU10K1 - SB Live [Unknown]
                     SB Live [Unknown] (rev.6, serial:0x80271102) at 0x8800, irq 5

With the bk-alsa patch reverted, it looks like:

0 [Live           ]: EMU10K1 - Sound Blaster Live!
                     Sound Blaster Live! (rev.6, serial:0x80271102) at 0x8800, irq 5

Hope that helps. If you need any additional information, just ask.

Thanks,
-- 
Jean Delvare


-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click

Re: Re: 2.6.12-rc1-mm3, sound card lost id

[Lee Revell]

On Tue, 2005-03-29 at 19:57 +0200, Jean Delvare wrote:
> Hi Takashi,
> 
> > > This one made /proc/asound/card0/id change from "Live" to "Unknown"
> > > on one of my systems, preventing alsatcl from properly restoring my
> > > mixer settings.
> > 
> > Hmm, perhaps it's a side effect of chip detection patch by James.
> > But "Unknown" is bad, of course.
> > 
> > How does /proc/asound/cards look?
> 
> 0 [Unknown        ]: EMU10K1 - SB Live [Unknown]
>                      SB Live [Unknown] (rev.6, serial:0x80271102) at 0x8800, irq 5
> 
> With the bk-alsa patch reverted, it looks like:
> 
> 0 [Live           ]: EMU10K1 - Sound Blaster Live!
>                      Sound Blaster Live! (rev.6, serial:0x80271102) at 0x8800, irq 5
> 
> Hope that helps. If you need any additional information, just ask.

I think we just have to add this PCI id to the table.  I got the same
result before James added the SBLive! platinum detection.

What is the output of 'lspci -v | grep -1 EMU10k1'?

Lee



-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click

Re: Re: 2.6.12-rc1-mm3, sound card lost id

[Jean Delvare]

Hi Lee,

> I think we just have to add this PCI id to the table.  I got the same
> result before James added the SBLive! platinum detection.
> 
> What is the output of 'lspci -v | grep -1 EMU10k1'?

00:0d.0 Multimedia audio controller: Creative Labs SB Live! EMU10k1 (rev 06)
        Subsystem: Creative Labs CT4832 SBLive! Value

and the one you didn't ask for:

00:0d.0 Class 0401: 1102:0002 (rev 06)
        Subsystem: 1102:8027

This made me realize that I could still try to hack it myself. The
following patch somehow helped:

--- linux-2.6.12-rc1-mm3/sound/pci/emu10k1/emu10k1_main.c.orig	2005-03-29 20:38:12.000000000 +0200
+++ linux-2.6.12-rc1-mm3/sound/pci/emu10k1/emu10k1_main.c	2005-03-29 22:32:23.000000000 +0200
@@ -680,6 +680,10 @@
 	 .driver = "EMU10K1", .name = "E-mu APS [4001]", 
 	 .emu10k1_chip = 1,
 	 .ecard = 1} ,
+	{.vendor = 0x1102, .device = 0x0002, .subsystem = 0x80271102,
+	 .driver = "EMU10K1", .name = "SB Live Player 1024", 
+	 .emu10k1_chip = 1,
+	 .ac97_chip = 1} ,
 	{.vendor = 0x1102, .device = 0x0002, .subsystem = 0x80641102,
 	 .driver = "EMU10K1", .name = "SB Live 5.1", 
 	 .emu10k1_chip = 1,


Now the card will be listed as "S1024" instead of "Unknown" so that's a
change. Looks like the short name is auto-generated? Unfortunately
that's still not "Live" as before so my mixer settings are not back yet.
And I believe that "Live" was a much better name than "S1024" too.

Thanks,
-- 
Jean Delvare


-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click

Re: Re: 2.6.12-rc1-mm3, sound card lost id

[Lee Revell]

On Tue, 2005-03-29 at 22:46 +0200, Jean Delvare wrote:
> Hi Lee,
> 
> > I think we just have to add this PCI id to the table.  I got the same
> > result before James added the SBLive! platinum detection.
> > 
> > What is the output of 'lspci -v | grep -1 EMU10k1'?
> 
> 00:0d.0 Multimedia audio controller: Creative Labs SB Live! EMU10k1 (rev 06)
>         Subsystem: Creative Labs CT4832 SBLive! Value
> 
> and the one you didn't ask for:
> 
> 00:0d.0 Class 0401: 1102:0002 (rev 06)
>         Subsystem: 1102:8027
> 
> This made me realize that I could still try to hack it myself. The
> following patch somehow helped:
> 
> --- linux-2.6.12-rc1-mm3/sound/pci/emu10k1/emu10k1_main.c.orig	2005-03-29 20:38:12.000000000 +0200
> +++ linux-2.6.12-rc1-mm3/sound/pci/emu10k1/emu10k1_main.c	2005-03-29 22:32:23.000000000 +0200
> @@ -680,6 +680,10 @@
>  	 .driver = "EMU10K1", .name = "E-mu APS [4001]", 
>  	 .emu10k1_chip = 1,
>  	 .ecard = 1} ,
> +	{.vendor = 0x1102, .device = 0x0002, .subsystem = 0x80271102,
> +	 .driver = "EMU10K1", .name = "SB Live Player 1024", 
> +	 .emu10k1_chip = 1,
> +	 .ac97_chip = 1} ,
>  	{.vendor = 0x1102, .device = 0x0002, .subsystem = 0x80641102,
>  	 .driver = "EMU10K1", .name = "SB Live 5.1", 
>  	 .emu10k1_chip = 1,
> 
> 
> Now the card will be listed as "S1024" instead of "Unknown" so that's a
> change. Looks like the short name is auto-generated? Unfortunately
> that's still not "Live" as before so my mixer settings are not back yet.
> And I believe that "Live" was a much better name than "S1024" too.
> 

Here is the patch (against ALSA CVS) in its preferred format.  You will
probably have to apply it by hand.  If the mixer settings can't be
restored you'll have to do it manually or edit asound.state by hand.

Lee

Index: alsa/alsa-kernel/pci/emu10k1/emu10k1_main.c
===================================================================
RCS file: /cvsroot/alsa/alsa-kernel/pci/emu10k1/emu10k1_main.c,v
retrieving revision 1.49
diff -u -r1.49 emu10k1_main.c
--- alsa/alsa-kernel/pci/emu10k1/emu10k1_main.c	27 Mar 2005 14:00:54 -0000	1.49
+++ alsa/alsa-kernel/pci/emu10k1/emu10k1_main.c	29 Mar 2005 20:51:44 -0000
@@ -693,6 +693,10 @@
 	 .driver = "EMU10K1", .name = "SBLive! Platinum [CT4760P]", 
 	 .emu10k1_chip = 1,
 	 .ac97_chip = 1} ,
+	{.vendor = 0x1102, .device = 0x0002, .subsystem = 0x80271102,
+	 .driver = "EMU10K1", .name = "SBLive! Value [CT4832]", 
+	 .emu10k1_chip = 1,
+	 .ac97_chip = 1} ,
 	{.vendor = 0x1102, .device = 0x0002,
 	 .driver = "EMU10K1", .name = "SB Live [Unknown]", 
 	 .emu10k1_chip = 1,

Lee



-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click

Re: Re: 2.6.12-rc1-mm3, sound card lost id

[James Courtier-Dutton]

Lee Revell wrote:
> On Tue, 2005-03-29 at 23:13 +0200, Jean Delvare wrote:
> 
>>>+	{.vendor = 0x1102, .device = 0x0002, .subsystem = 0x80271102,
>>>+	 .driver = "EMU10K1", .name = "SBLive! Value [CT4832]", 
>>>+	 .emu10k1_chip = 1,
>>>+	 .ac97_chip = 1} ,
>>
>>Unsurprisingly, my card is now named CT4832. I had to edit
>>/etc/asound.state manually to get my mixer settings back (with some
>>warnings, but I get some sound).
>>
>>Not sure I quite see the idea of renaming from "Live", which the user
>>will understand, to (I suppose) the exact chip name on the card, while
>>the user has certainly no idea what it is. But heh I'm not an ALSA
>>developer, there must be a good reason.
> 
> 
> Blame Creative.  They have released so many different cards under the
> Live! name that it's become meaningless.  There are now two separate
> classes of cards marketed as Live! (or Audigy) that don't even use the
> same driver (ca0106 and emu10k1x), the hardware isn't even similar.
> Unsurprisingly, the user confusion is massive.
> 
> The model number is the simplest unique identifier.  This is also what
> the "good" Windows drivers for these devices, http://www.kxproject.com,
> use.
> 
> Lee
> 

It looks to me that alsactl is to blame here.
If a user has 2 sound cards of exactly the same model, the state will
not be saved, as the state.xxxx will be the same for both cards.
I think we need to change alsactl to make the saved state unique.
Maybe a combination of driver name and card number, and not use short
name at all.
Any comments?


-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click

Re: Re: 2.6.12-rc1-mm3, sound card lost id

[Jean Delvare]

Hi Lee,

> Here is the patch (against ALSA CVS) in its preferred format.  You
> will probably have to apply it by hand.  If the mixer settings can't
> be restored you'll have to do it manually or edit asound.state by
> hand.
> 
> Lee
> 
> Index: alsa/alsa-kernel/pci/emu10k1/emu10k1_main.c
> ===================================================================
> RCS file: /cvsroot/alsa/alsa-kernel/pci/emu10k1/emu10k1_main.c,v
> retrieving revision 1.49
> diff -u -r1.49 emu10k1_main.c
> --- alsa/alsa-kernel/pci/emu10k1/emu10k1_main.c	27 Mar 2005 14:00:54 -0000	1.49
> +++ alsa/alsa-kernel/pci/emu10k1/emu10k1_main.c	29 Mar 2005 20:51:44 -0000
> @@ -693,6 +693,10 @@
>  	 .driver = "EMU10K1", .name = "SBLive! Platinum [CT4760P]", 
>  	 .emu10k1_chip = 1,
>  	 .ac97_chip = 1} ,
> +	{.vendor = 0x1102, .device = 0x0002, .subsystem = 0x80271102,
> +	 .driver = "EMU10K1", .name = "SBLive! Value [CT4832]", 
> +	 .emu10k1_chip = 1,
> +	 .ac97_chip = 1} ,
>  	{.vendor = 0x1102, .device = 0x0002,
>  	 .driver = "EMU10K1", .name = "SB Live [Unknown]", 
>  	 .emu10k1_chip = 1,

Unsurprisingly, my card is now named CT4832. I had to edit
/etc/asound.state manually to get my mixer settings back (with some
warnings, but I get some sound).

Not sure I quite see the idea of renaming from "Live", which the user
will understand, to (I suppose) the exact chip name on the card, while
the user has certainly no idea what it is. But heh I'm not an ALSA
developer, there must be a good reason.

Thanks,
-- 
Jean Delvare


-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click

Re: Re: 2.6.12-rc1-mm3, sound card lost id

[Lee Revell]

On Tue, 2005-03-29 at 23:13 +0200, Jean Delvare wrote:
> > +	{.vendor = 0x1102, .device = 0x0002, .subsystem = 0x80271102,
> > +	 .driver = "EMU10K1", .name = "SBLive! Value [CT4832]", 
> > +	 .emu10k1_chip = 1,
> > +	 .ac97_chip = 1} ,
> Unsurprisingly, my card is now named CT4832. I had to edit
> /etc/asound.state manually to get my mixer settings back (with some
> warnings, but I get some sound).
> 
> Not sure I quite see the idea of renaming from "Live", which the user
> will understand, to (I suppose) the exact chip name on the card, while
> the user has certainly no idea what it is. But heh I'm not an ALSA
> developer, there must be a good reason.

Blame Creative.  They have released so many different cards under the
Live! name that it's become meaningless.  There are now two separate
classes of cards marketed as Live! (or Audigy) that don't even use the
same driver (ca0106 and emu10k1x), the hardware isn't even similar.
Unsurprisingly, the user confusion is massive.

The model number is the simplest unique identifier.  This is also what
the "good" Windows drivers for these devices, http://www.kxproject.com,
use.

Lee



-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click

Re: Re: 2.6.12-rc1-mm3, sound card lost id

[Jaroslav Kysela]

On Tue, 29 Mar 2005, James Courtier-Dutton wrote:

> > The model number is the simplest unique identifier.  This is also what
> > the "good" Windows drivers for these devices, http://www.kxproject.com,
> > use.

> It looks to me that alsactl is to blame here.
> If a user has 2 sound cards of exactly the same model, the state will
> not be saved, as the state.xxxx will be the same for both cards.

Nope. The card id is autogenerated from the long name (identification)  
by default. An index is appended for second and later cards to create
unique ids. Also, you may override the card identification using the "id"
kernel module parameter, if you don't like autogenerated id. Then you can 
assign name like "SBLive" or so to avoid confusion...

						Jaroslav

-----
Jaroslav Kysela <perex@suse.cz>
Linux Kernel Sound Maintainer
ALSA Project, SUSE Labs


-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click

Re: Re: 2.6.12-rc1-mm3, sound card lost id

[Takashi Iwai]

At Tue, 29 Mar 2005 23:13:45 +0200,
Jean Delvare wrote:
> 
> Hi Lee,
> 
> > Here is the patch (against ALSA CVS) in its preferred format.  You
> > will probably have to apply it by hand.  If the mixer settings can't
> > be restored you'll have to do it manually or edit asound.state by
> > hand.
> > 
> > Lee
> > 
> > Index: alsa/alsa-kernel/pci/emu10k1/emu10k1_main.c
> > ===================================================================
> > RCS file: /cvsroot/alsa/alsa-kernel/pci/emu10k1/emu10k1_main.c,v
> > retrieving revision 1.49
> > diff -u -r1.49 emu10k1_main.c
> > --- alsa/alsa-kernel/pci/emu10k1/emu10k1_main.c	27 Mar 2005 14:00:54 -0000	1.49
> > +++ alsa/alsa-kernel/pci/emu10k1/emu10k1_main.c	29 Mar 2005 20:51:44 -0000
> > @@ -693,6 +693,10 @@
> >  	 .driver = "EMU10K1", .name = "SBLive! Platinum [CT4760P]", 
> >  	 .emu10k1_chip = 1,
> >  	 .ac97_chip = 1} ,
> > +	{.vendor = 0x1102, .device = 0x0002, .subsystem = 0x80271102,
> > +	 .driver = "EMU10K1", .name = "SBLive! Value [CT4832]", 
> > +	 .emu10k1_chip = 1,
> > +	 .ac97_chip = 1} ,
> >  	{.vendor = 0x1102, .device = 0x0002,
> >  	 .driver = "EMU10K1", .name = "SB Live [Unknown]", 
> >  	 .emu10k1_chip = 1,
> 
> Unsurprisingly, my card is now named CT4832. I had to edit
> /etc/asound.state manually to get my mixer settings back (with some
> warnings, but I get some sound).
> 
> Not sure I quite see the idea of renaming from "Live", which the user
> will understand, to (I suppose) the exact chip name on the card, while
> the user has certainly no idea what it is. But heh I'm not an ALSA
> developer, there must be a good reason.

Changing the name to more descriptive one is a good thing, IMO.
But, of course, it's a regression if the system gets (somehow) broken
after the update.

We can add a hack to set id explicitly to be compatible with older
version...


Takashi


-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click