How To Have Sound On A Toshiba P105/100 Series (oss 10.3)

This is a very simple and fast "howto"

For this HowTo, I've installed the GNOME CD release v10.3 and Mike Petersen's Addon CD previously, though this might not be necessary.

After installation has completely finished, on your first boot up add the next command in the kernell options:

acpi_osi=!Linux[/b]

Test your sound and...
That's al folks!!!

Luis

IMPORTANT This method works with NO custom DSDT.aml, and it's completely HARMLESS.
NOTE. Most likely you will want to modiy your /boot/grub/menu.lst to boot permanently with this option in case this hack works for you.

Confirmed this worked for me, thanks for the tip

Is there any chance you could help provide more information as to what your sound configuration is, such that we can characterize what PCs this approach might apply to?

If so, could you download and run the following script:
http://bulletproof.servebeer.com/als...s/alsa-info.sh

To run it, just open a konsole, navigate to the directory where it is located, and type "bash alsa-info.sh"

This is a script that user wishie from IRC channel #alsa wrote, to gather ALSA information from the machine its run on. When run, it will automatically post the output on pastebin.ca and give you the URL where its posted.

Here is an example of the sort of thing to expect.. http://pastebin.ca/724516

Please post the URL here and from that we can see what the details are of your sound chipset, and have a flavour as to what soundchips this interesting fix will apply to (as it may work on more than just a Toshiba P105/100 series).

Thanks

I ran the script like you asked and it has posted the output here: http://pastebin.ca/748941

I hope this provided the information you wanted and will help others down the road. Also if you were interested I have a Toshiba Satellite p105-S9312.

I ran the script like you asked and it has posted the output here: http://pastebin.ca/748941I hope this provided the information you wanted and will help others down the road. Also if you were interested I have a Toshiba Satellite p105-S9312.[/b]

Thanks, ... It appears it has a Conexant CX20551 (Waikiki) chipset/codec, that has both an analog and digital playback device.
The alsa drive that it uses is the snd_hda_intel. The CX20551 was previously refered to as the CXT5047.

It was initially added in ala-1.0.14rc1 (as CXT5047) and then later some issues were fixed, including fixes to:
* silent ouptput,
* automute, (I think supposed to fix when headphones are plugged , laptop speakers don't mute)
* recording issues
* headphone squeal and noise
in alsa-1.0.14rc3/4 (together with renaming to CX20551 version)

I am wondering if this is specific to the 2.6.22 kernel?
http://www.ussg.iu.edu/hypermail/linux/ker...707.0/0408.html


I can't say I know what it does. ... One site (on ACPI patches for 2.6.22-rc3) noted:

The boot option "acpi_osi=" has always disabled Linux _OSI support, thus disabling all OS Interface strings which are advertised by Linux to the BIOS.

acpi_osi="string" adds the interface string, and
acpi_osi="!string" invalidates the pre-defined interface string[/b]

http://www.mail-archive.com/linux-acpi@vge...g/msg07070.html

with the example:

eg. acpi_osi="!Windows 2006" would disable Linux's claim of Vista compatibility. [/b]

Purportedly the HP DV1000, DV2000 and DV3000 also have Conexant CXT5047. But I could not find reference to other CXT5047 users applying the acpi_osi=!Linux fix. I did find reference to older SUSE-10.0 CXT5047 users struggling with their sound.

Anyway, the bottom line is it is good to have a suggested fix available (that has been proven to work).

Is there any chance you could help provide more information as to what your sound configuration is, such that we can characterize what PCs this approach might apply to?

If so, could you download and run the following script:
http://bulletproof.servebeer.com/als...s/alsa-info.sh

To run it, just open a konsole, navigate to the directory where it is located, and type "bash alsa-info.sh"

This is a script that user wishie from IRC channel #alsa wrote, to gather ALSA information from the machine its run on. When run, it will automatically post the output on pastebin.ca and give you the URL where its posted.

Here is an example of the sort of thing to expect.. http://pastebin.ca/724516

Please post the URL here and from that we can see what the details are of your sound chipset, and have a flavour as to what soundchips this interesting fix will apply to (as it may work on more than just a Toshiba P105/100 series).

Thanks
[/b]

Done....
http://pastebin.ca/750025

Do u also need my dmidecode? (probabily there will be much more info)

Cheers

Luis

PS: This hack also worked in Mandriva 2008.0 (GNOME and KDE)
PS2. I was one of the beta testers from ALSA's team (GrueMaster) with 1.0.14rcx
PS3. This is a Toshiba Satellite P105-SP921 (Latin America Model) PSPA6U-006005

Confirmed this worked for me, thanks for the tip
[/b]

Tyriel.
Cuold you, please do the following as root

#dmidecode > dmidecode.txt

and email the dmidecode.txt file to:linux-acpi@vger.kernel.org providng information about the problem and solution.
TIA
Luis

It was initially added in ala-1.0.14rc1 (as CXT5047) and then later some issues were fixed, including fixes to:
* silent ouptput,
* automute, (I think supposed to fix when headphones are plugged , laptop speakers don't mute)
* recording issues
* headphone squeal and noise
in alsa-1.0.14rc3/4 (together with renaming to CX20551 version)[/b]

You are just right!

I am wondering if this is specific to the 2.6.22 kernel?
http://www.ussg.iu.edu/hypermail/linux/ker...707.0/0408.html
I can't say I know what it does. ... One site (on ACPI patches for 2.6.22-rc3) noted:

http://www.mail-archive.com/linux-acpi@vge...g/msg07070.html

with the example:
Purportedly the HP DV1000, DV2000 and DV3000 also have Conexant CXT5047. But I could not find reference to other CXT5047 users applying the acpi_osi=!Linux fix. I did find reference to older SUSE-10.0 CXT5047 users struggling with their sound.[/b]

That's because this fix only applies to kernel 2.6.22 and before we had to apply a custom DSDT in initram. However, I've seen very many people with HPs DVXseries applying a custom DSDT.aml/hex to kernel.

Anyway, the bottom line is it is good to have a suggested fix available (that has been proven to work).
[/b]

Hopefully this will be fixed in a near future if everyone sends the correct info to acpi.org.

Kind Regards

Luis

Do u also need my dmidecode? (probabily there will be much more info)[/b]

Interesting stuff, but I would not know what to do with the dmidecode information. ... Maybe I could learn?

About the only time I've used dmidecode (in the past) was to figure out exactly what memory my mother's SUSE-PC has via ssh in a remote konsole session (as when I travel across the Atlantic next year for a brief visit, I want to upgrade the memory on her PC, and I don't have time to get things wrong ...) and hence
dmidecode | less +/bank > memorystatus.txt
worked for that. A user on IRC#suse passed that code to me.

I'm just a regular SuSE user who likes playing with multimedia, and hence I have tried to focus my support, to other users, more on the multimedia side of things.

IMaybe I could learn?[/b]

LOL.... I think you already know too much

...) and hence
dmidecode | less +/bank > memorystatus.txt
worked for that. A user on IRC#suse passed that code to me.[/b]

I do not see the difference between that "fancy"command and dmidecode > dmidecode.txt ("diffing" them gives me the same results )

I'm just a regular SuSE user who likes playing with multimedia, and hence I have tried to focus my support, to other users, more on the multimedia side of things.
[/b]

And that's much appreciated..

Kind Regards

Luis