Lost sound in Luci528

Message

PaulBx1 · #41 Post by **PaulBx1** » Thu 29 Sep 2011, 00:44

I have not seen indications in problem reports about which button was clicked during interaction with the alsawizard.

I never do anything but take the defaults, all the way through. I use alsa wizard as a sort of sound reboot.

when you report that sound is not working, I assume you mean that you cannot hear anything from puppy, not that retrovol/speaker-icon has disappeared. Those are separate issues.

I am using "aplay /2barks.au" as my test of whether sound is working (I copied it up from /usr/share/audio for convenience). I have seen these conditions:
1) Speaker icon present and sound works.
2) Speaker icon gone and sound doesn't work.
3) Speaker icon gone but sound still works.

I suspect the two are actually connected in some way since they both went bad in the same time frame (5.2.7 or 5.2.8 ). Fix the speaker icon, you've probably fixed the sound. The icon disappearing also has the advantage that you can see immediately when it goes missing, without the necessity of having to run a test.

PaulBx1, your "module in use" situation is interesting, although I hope it is because you ran 10alsa from a console. It may be that the conflicts are avoided by actions taken before invocation of 10alsa.

Not sure I catch your drift here. Isn't 10alsa supposed to be invokable from a console? I started concentrating on that because 1) I assumed that alsa wizard invokes it (let me know if that is wrong) and 2) it's a heck of a lot faster than running alsa wizard, and 3) it's unlikely two separate things, alsa wizard and 10alsa cause the same symptoms. In other words I imagine I have already narrowed the problem down for you, not only to 10alsa but to the rmmod loop of it. At least it looks that way to me. Maybe there is a problem in the driver when rmmods happen. Maybe rmmod now has a bug in it.

I am attaching the pmodemdiag file. I booted pfix=ram and checked dmesg. Then I ran "./10alsa stop" to make the speaker icon go away (and I got those error messages in the console again). These messages appeared at the end of dmesg as a result:

Code: Select all

HDA Intel 0000:00:1b.0: PCI INT A disabled
HDA Intel 0000:00:1b.0: PCI INT A -> GSI 22 (level, low) -> IRQ 22
HDA Intel 0000:00:1b.0: irq 28 for MSI/MSI-X
HDA Intel 0000:00:1b.0: setting latency timer to 64

I got another set of those same 4 messages if I then ran the alsa wizard.

BTW I tried commenting out the rmmod line in 10alsa, and added 'echo "blah" ' to that loop. The speaker muted but did not disappear when I ran 10alsa stop.

There were 15 "blah"s. Doesn't it matter what order those drivers are removed? Maybe the ordering is wrong, so that we are getting complaints on the removals. There were not 15 error messages, but only about 5 of them.

It might not hurt to have the others here comment out the rmmod in 10alsa, see what happens.

PaulBx1 · #42 Post by **PaulBx1** » Thu 29 Sep 2011, 01:49

I modified that echo statement to say what is being removed, and it does it in the same order lsmod reports it, which is the correct order I assume. There is one question about "snd_seq_dummy" but I'm guessing that is nothing important.

I then uncommented the rmmod and ran it again. Here is what I got:

Code: Select all

sh-4.1# ./10alsa stop 
snd_hda_codec_realtek
ERROR: Module snd_hda_codec_realtek is in use
snd_hda_intel
snd_hda_codec
ERROR: Module snd_hda_codec is in use by snd_hda_codec_realtek
snd_pcm_oss
snd_pcm
ERROR: Module snd_pcm is in use by snd_hda_codec
snd_seq_dummy
snd_mixer_oss
snd_seq_oss
snd_seq_midi
snd_seq_midi_event
snd_rawmidi
snd_seq
snd_seq_device
snd_timer
ERROR: Module snd_timer is in use by snd_pcm
snd
ERROR: Module snd is in use by snd_hda_intel,snd_hda_codec_realtek,snd_hda_codec,snd_pcm,snd_timer
sh-4.1#

Looks like it all depends on the Realtek ethernet controller (wrong conclusion, see later post). I wonder if the others have this RTL8101E controller? I wonder what this controller has to do with sound? BTW I had nothing connected to the ethernet at the time.

I tried rebooting "pfix=ram" to get my speaker icon back, then doing an "rmmod snd_hda_codec_realtek". It still gave me the "in use" error. No extra messages were seen in dmesg. The speaker did not go away.

chard62 · #43 Post by **chard62** » Thu 29 Sep 2011, 02:34

Is it possible that the problem lies with a certain type of sound card?
Mine is built into the Asus motherboard, AC97 type: Realtek alc655

Also, I don't know if this is useful or not. After sound dies, if I try to play a sound file with VLC, I get this error message from VLC:

"Potential ALSA version problem:
VLC failed to initialize your sound output device (if any).
Please update alsa-lib to version 1.0.23-2-g8d80d5f or higher to try to fix this issue"

I don't know how to fix this anyway

PaulBx1 · #44 Post by **PaulBx1** » Thu 29 Sep 2011, 03:07

My sound is the "Intel Corporation N10/ICH 7 Family High Definition Audio Controller (rev 02)"

But yeah, I noticed Realtek was doing audio controllers too. But I don't have one.

Try doing what I'm doing. Boot pfix=ram, then do this:

Code: Select all

cd /etc/init.d
./10alsa stop

See if you get the same error messages.

I wonder if we got a new Realtek driver in 5.2.7 or 5.2.8?

<later>
Ah, now I get it. (doh!) Realtek as a company is involved with multimedia (which means codecs). That doesn't mean this problem has anything to do with my Realtek ethernet controller. It means maybe there is something wrong with that multimedia driver, or with alsa. So the question is, did that driver or alsa get changed in 5.2.7 or 5.2.8? Maybe your message about alsa-lib is the key here.

rerwin · #45 Post by **rerwin** » Thu 29 Sep 2011, 03:35

PaulBx1 and others,
PaulBx1's diagnostic data shows that there was a problem with the loading of the HDA driver, one that I fixed a few days ago for another reason. It helps explain the intermittent nature of the sound disappearance. Although it may not solve the issues completely, at least the possibility of that disruption will be removed.

I have posted my cumulative patch-fix package in the lupu 5.2.8 thread, because it is important to users reporting in several threads.
http://www.murga-linux.com/puppy/viewto ... 604#568604

So, for testing of sound issues, please install lupu528-IU001_rerwin_patch-5.pet. It also contains the "experimental" lupu528-rerwin-test-alsa_patch-1, the alsa-info.sh tool and an updated pmodemdiag that now includes the output from alsa-info.sh. Thanks.
Richard

8-bit · #46 Post by **8-bit** » Thu 29 Sep 2011, 04:28

It appears that the ISO is still available called "pup-431-scsi-intel_modems.iso.
I found it actually from one of your older posts with a link that did not work. But it steered me to where to look.
The direct link to the file is ftp://distro.ibiblio.org/puppylinux/pup ... modems.iso.
So for now, until the next cleanup of old files, that happens. it is available.

Darn it! Replied in the wrong thread. Sorry bout that.

PaulBx1 · #47 Post by **PaulBx1** » Thu 29 Sep 2011, 06:02

Booted pfix=ram, installed Instant Update and the b43 PIO driver and your patch here, then rebooted from the new pupsave. Sorry to say there was no change in the issue I reported.

playdayz · #48 Post by **playdayz** » Fri 30 Sep 2011, 15:29

Here's something to test for possible inclusion in the next Instant Update.

http://diddywahdiddy.net/Puppy500/alsa_ ... 1.0.24.pet

Especially please test this if you can if you are having any problems with sound. We also need to test to make sure it doesn't cause problems for people who are not having problems I am running it and haven't had a problem. Install and reboot. Thanks.

It would be most convenient if any results of this test could be reported in the Lucid 5.2.8 thread. Thanks.
http://www.murga-linux.com/puppy/viewtopic.php?t=70855

playdayz · #49 Post by **playdayz** » Fri 30 Sep 2011, 15:30

Duplicate

PaulBx1 · #50 Post by **PaulBx1** » Sat 01 Oct 2011, 23:16

Booted pfix=ram, installed Instant Update and the b43 PIO driver and alsa_driver-1.0.24.pet (but NOT Rerwin's previous patch - it was not clear if you wanted that or not). Rebooted, creating a new pupsave in the process which I used on the reboot.

My problem is the same, with one exception. I still have sound, alsa wizard still kills the speaker icon. But, alsa wizard now cannot find my PCI Intel sound interface. It stops and asks if it should probe my (nonexistent) ISA bus.

My problem is clearly caused by snd_hda_codec_realtek, which seems broken. Once loaded, it cannot be removed, and the attempt causes the failure of other modules to be removed, which in turn makes my speaker icon go away. Seems to me, I either need snd_hda_codec_realtek not to be loaded in the first place (what application requires it?) or to replace snd_hda_codec_realtek with a newer version, if such is available.

Meanwhile, if someone can tell me where snd_hda_codec_realtek is loaded, so I can prevent it, I could see what effect that has on the system.

PaulBx1 · #51 Post by **PaulBx1** » Sat 01 Oct 2011, 23:44

I found this interesting bit in dmesg. It was after booting, hadn't done anything else yet:

Code: Select all

i801_smbus 0000:00:1f.3: PCI INT B -> GSI 19 (level, low) -> IRQ 19
ACPI: I/O resource 0000:00:1f.3 [0x18a0-0x18bf] conflicts with ACPI region SMBI [0x18a0-0x18af]
ACPI: If an ACPI driver is available for this device, you should use it instead of the native driver
leds_ss4200: no LED devices found
input: PC Speaker as /devices/platform/pcspkr/input/input8
jmb38x_ms 0000:02:00.3: PCI INT A -> GSI 16 (level, low) -> IRQ 16
jmb38x_ms 0000:02:00.3: setting latency timer to 64
HDA Intel 0000:00:1b.0: PCI INT A -> GSI 22 (level, low) -> IRQ 22
HDA Intel 0000:00:1b.0: irq 27 for MSI/MSI-X
HDA Intel 0000:00:1b.0: setting latency timer to 64
input: HDA Digital PCBeep as /devices/pci0000:00/0000:00:1b.0/input/input9
input: HDA Intel Mic as /devices/pci0000:00/0000:00:1b.0/sound/card0/input10
input: HDA Intel Headphone as /devices/pci0000:00/0000:00:1b.0/sound/card0/input11
Error: Driver 'pcspkr' is already registered, aborting...

Don't know if it bears on anything here.

chard62 · #52 Post by **chard62** » Sun 02 Oct 2011, 00:26

I finally am able to get some sound now!!! (without going back to a fresh save file)

I'm still wondering why the fix doesn't survive a reboot, but there are MANY variables in my setup. I want to end up using a newish version of VLC as my player; and there are some definite issues about which version to use for VLC. Someone mentioned the speaker icon in the bottom right, sometimes I have it, and sometimes I don't.

The ALSA driver update has changed my list of possible drivers, and the only one out of 20 or so that works is called "Via" even though my chip is a Realtek...

Thanks for the work guys, I'm seeing some hope for my installation of 5.2.8 now

rerwin · #53 Post by **rerwin** » Mon 03 Oct 2011, 15:24

chard62 wrote:Someone mentioned the speaker icon in the bottom right, sometimes I have it, and sometimes I don't.

The ALSA driver update has changed my list of possible drivers, and the only one out of 20 or so that works is called "Via" even though my chip is a Realtek...

Pupscan reports that all of your hardware reported in the last diag file you got is "VIA".

DESCRIPTION: Multimedia audio controller: VIA Technologies, Inc. VT8233/A/8235/8237 AC97 Audio Controller
VENDOR: 1106 DEVICE: 3059
KERNEL DRIVER: VIA 82xx Audio
KERNEL MODULE: snd-via82xx

The VIA package could use any of the possible chips, but VIA "owns" the whole thing. Therefore, I conclude that the via82xxx driver is appropriate. Or are you talking about a different PC?

Just to be sure you have all of my latest fixes to sound issues, please install the following packages onto 528:

Instant Update 001
the "compiled" ALSA driver
lupu528_patch-6

The last 2 are available here:
http://www.murga-linux.com/puppy/viewto ... 823#569823
Those fix all the problems I am aware of, so maybe they will solve your problem. Or it will help isolate the problem. Then if you still lose sound, please immediately get another pmodemdiag file for me "to chew on." Thanks.
Richard

karm2865 · #54 Post by **karm2865** » Wed 05 Oct 2011, 18:51

HP DC5750
Lucid Puppy 5.2.8 full install to the hard drive duel boot with windows XP-pro
3GB ram / Radeon HD 5430 video / Creative Labs CA0106 Soundblaster

the built in audio on the mother board was not very good and not properly sported by Linux.
install a PCI Creative Labs CA0106 Soundblaster card.

run the sound wizard find the card plays the test sound all good
reboot and no sound check Alsa mixer piddle with the settings checked to make sure the proper driver is in use, it is. scratch my head Google.....
Reboot to windows xp participation card works sounds good .....
reboot to Puppy no sound.

open console
run the sound wizard
alsactl store [enter]
reboot and no sound. swear at the sound card, & driver
(dogs hides under the couch. house mate goes in the other room.)
run the sound wizard. card plays the test sound
open console
alsactl store [enter]
reboot and no sound check Alsa mixer piddle with the settings checked to make sure the proper driver is in use, it is. scratch my head Google...
.

after i use the sound wizard i have nice sound. but the speaker icon is showing muted no big deal almost every thing with sound has a volume control or i can open the Alsa mixer and adjust it there.

so now a week latter of running the sound wizard every puppy boot-up and Google has failed me i am asking is there any way to keep the sound drivers in place?

karm2865 · #55 Post by **karm2865** » Wed 05 Oct 2011, 19:47

tried the updates and fixes now i have no sound or icon and alasmixer refuses to work

Karl Godt · #56 Post by **Karl Godt** » Wed 05 Oct 2011, 19:51

alsactl store [enter]

I dont actually know [ anymore ] if this is important :

Puppy uses the file /etc/asound.state !!!

It should be

Code: Select all

alsactl -f /etc/asound.state store

Puppy-431 seems to

-f,--file # configuration file (default /etc/asound.state)

but Lupu-ubuntu-Puppy-5xx alsactl may have a default configuration file somewhere in /usr/share/alsa* or /usr/lib/alsa* .
[edit_2]/var/lib/alsa/asound.state[/edit_2]

tried the updates and fixes now i have no sound or icon and alasmixer refuses to work

[strike-through]The alsa-1.0.24 dotPET had a buggy or badly compiled snd_intel8x0 driver , that refused to load .[/strike-through]
[edit]
After rebooting the snd_intel8x0 driver now is accepted . Without reboot it had been the [ only ] one of the snd-drivers ,
that had shown "Disagreeing about module layout" while modprobing it manually . Apparently there were some version information of the old alsa-1.0.20 libs somewhere in RAM . I had thought the module was not accepted by the kernel itself .
I must unfortunately confirm , that the driver was not inserted at boot though , but almost all other snd-drivers .
I could only modprobe'ing it after rmmod'ing all loaded snd-drivers , including 'snd' and 'soundcore' .
The scripts of /sbin/pup_event_backend_modrobe* and the files in /etc/modprobe.d are changing quite often ,
so it will take me some time to figure out about it .

Cheers !
[/edit]
[edit_2]
Now I have tested the alsa_driver-1.0.24.pet on a dpup-006 by iguleder : frugal install with no pupsave-file (PUPMODE=5) , installed the alsa_driver-1.0.24.pet , rebooted into a 64MB test-save-file : the modules did not load ; booted again : Still did not load with errs like
snd_ac97_codec: disagrees about version of symbol snd_info_register
snd_ac97_codec: Unknown symbol snd_info_register

I had thought , it may be because of the alsa-lib-version

Code: Select all

strings /usr/lib/libasound.so.2.0.0 |  grep '1\.0\.2'

1.0.23

but decided to run the old ^ depmod-FULL ^ binary instead of ^ depmod ^

I restored /sbin/depmod to the full version. A script can still execute the Busybox depmod by running "busybox depmod"
12 Jan 2011.

Sound detection fixes

::: the next reboot sound was OK , despite the older libs , no errs in /tmp/bootsysinit.log an bootkernel.log .
[/edit_2]

karm2865 · #57 Post by **karm2865** » Wed 05 Oct 2011, 21:10

HP DC5750
Lucid Puppy 5.2.8 full install to the hard drive duel boot with windows XP-pro
3GB ram / Radeon HD 5430 video / Creative Labs CA0106 Soundblaster

well after the updates first i lost sound than video went so i am on a brand new install of puppy using
PupRescue2.3 based on Lucid Puppy 5.2.8 by DPUP5520
I have no clue as to what happened but the new install seems to have fixed the sound problem,

osgradel · #58 Post by **osgradel** » Fri 07 Oct 2011, 09:34

Well, another one with the same problem. At this moment, I had to come back with lupu 5.2.5.

card: Intel 82801DB-ICH4 chip: VIA 1612A

I've tried rerwin's last patch and the new alsa driver, but it doesn't work for me. Anyway, thanks rerwin for your effort. If I can help testing something and giving feedback, count on me.

jafadmin · #59 Post by **jafadmin** » Fri 07 Oct 2011, 17:54

Just to chime in here; Having the same issues as all above with the HDA Intel Realtek ALC662 Rev 1, as well as others.

About half the machines I boot with the USB thumbdrive exhibit this behavior. I have been testing this for the past several days to mixed results, but am unable to isolate any constants other than the HDA Intel aspect.

I'll watch this thread.

playdayz · #60 Post by **playdayz** » Fri 07 Oct 2011, 18:37

I am in a bit of a quandary here. I think I made a mistake by upgrading to alsa-1.0.24 in Lucid 5.2.8. The alsa-driver 1.0.24 pet was compiled in Lucid 5.2.8 and I have no clue why the snd-hda-intel.ko is giving problems. One possibility is to stay with older driver as in 5.2.8 release. Another possibility is to revert to alsa-1.0.21 as in 5.2.5, though I worry that might cause problems for people for whom alsa-1.0.24 is working.

Ah, does this make sense--to post a pet for people who are having sound problems in 5.2.8, that consists of the alsa-utils and libs from 1.0.21--in effect, reverting them to the status of 5.2.5. I will make one up and post it for testing.

Another good possibility for people having sound problems with 5.2.8 is to use Three-Headed Dog, with the 2.6.37.6 kernel from Slacko. If you are having sound problems, please test 3HD if you can to see if they are solved there. 3HD V2T2 is basically identical to 5.2.8 except the newer kernel and newer multimedia components. It is what I use most of the time now.

I am open to suggestions. Thanks.

Here is a pet with alsa-base, libs, and utils from Lucid 525. This is an experimental attempt to solve a problem with sound in 5.2.8.

1. Uninstall the alsa-driver 1.0.24 if it is installed
2. install this pet and reboot

I have tested it with 5.2.8 that had alsa-driver-1.0.24 installed (which I uninstalled), When I rebooted I (still) had sound on both my sound cards--btw, one of my sound cards uses snd-hda-intel.ko but I have not had problems in 528.

This is an experiment--use a backup save file or a fresh pfix=ram boot.

(old)Puppy Linux Discussion Forum

(old)Puppy Linux Discussion Forum