Heavy stutter without system load

[EricDHH]

Binary package hint: pulseaudio

Description: Ubuntu lucid (development branch)
Release: 10.04 AMD64

With turn from karmic to lucid beta the sound is poping and stuttering while playing trough pulseaudio. When this happens i got tons of these lines in messages.
Apr 19 12:13:29 agamemnon pulseaudio[5718]: ratelimit.c: 5 events suppressed
Apr 19 12:18:05 agamemnon pulseaudio[5718]: ratelimit.c: 1 events suppressed
Sometimes pulse goes to sleep_state and the sound stops completely, audio applications grey out in this case.

The chip is a rt888 HDA. I tryed to set higher priority at /etc/pulse/daemon.conf but -11 and forbidden realtime scheduling seems to be the reality. I tryed different resample-method without an effect. In karmic pulse runs with -20 and everything was fine.

pulseaudio:
  Installed: 1:0.9.22~0.9.21+stable-queue-32-g8478-0ubuntu14

There is 2x 2.8GHz CPU power, 4gig of 1333 mem but load isn't needed to let pulse stutter und popple. Other distros enable pulse-rt to workaround this, ubuntu did not not have this group.

ProblemType: Bug
DistroRelease: Ubuntu 10.04
Package: pulseaudio 1:0.9.22~0.9.21+stable-queue-32-g8478-0ubuntu14
ProcVersionSignature: Ubuntu 2.6.32-21.32-genusername 2.6.32.11+drm33.2
Uname: Linux 2.6.32-21-generic x86_64
NonfreeKernelModules: nvidia
AlsaVersion: Advanced Linux Sound Architecture Driver Version 1.0.21.
Architecture: amd64
AudioDevicesInUse:
 USER PID ACCESS COMMAND
 /dev/snd/controlC0: eric 5718 F.... pulseaudio
 /dev/snd/pcmC0D0p: eric 5718 F...m pulseaudio
Card0.Amixer.info:
 Card hw:0 'SB'/'HDA ATI SB at 0xfe024000 irq 16'
   Mixer name : 'Realtek ALC888'
   Components : 'HDA:10ec0888,1458e601,00100001'
   Controls : 39
   Simple ctrls : 21
CheckboxSubmission: 72ffaca291c05635776d4a13f9e644e4
CheckboxSystem: 32980ee028b29311fef8e195a759b031
Date: Mon Apr 19 12:17:10 2010
ProcEnviron:
 LANGUAGE=de_DE:de:en_GB:en
 LANG=de_DE.utf8
 SHELL=/bin/bash
SourcePackage: pulseaudio
dmi.bios.date: 12/09/2009
dmi.bios.vendor: Award Software International, Inc.
dmi.bios.version: F2
dmi.board.name: GA-MA770T-UD3
dmi.board.vendor: Gigabyte Technology Co., Ltd.
dmi.board.version: x.x
dmi.chassis.type: 3
dmi.chassis.vendor: Gigabyte Technology Co., Ltd.
dmi.modalias: dmi:bvnAwardSoftwareInternational,Inc.:bvrF2:bd12/09/2009:svnGigabyteTechnologyCo.,Ltd.:pnGA-MA770T-UD3:pvr:rvnGigabyteTechnologyCo.,Ltd.:rnGA-MA770T-UD3:rvrx.x:cvnGigabyteTechnologyCo.,Ltd.:ct3:cvr:
dmi.product.name: GA-MA770T-UD3
dmi.sys.vendor: Gigabyte Technology Co., Ltd.
mtime.conffile..etc.pulse.daemon.conf: 2010-04-18T16:17:13.703498

[Daniel T Chen (crimsun)]

Can you reproduce this symptom using linux-alsa-driver-modules-$(uname -r) from ppa:ubuntu-audio-dev?

[EricDHH (ericdhh)]

Checking.....

Everything fine when idle (pulseaudio prio -11)

and with quad burnK7 on heavy load, the mouse reacts slow but the sound comes perfect. The only messages lines are these.

Apr 19 19:25:01 agamemnon kernel: [ 70.926118] hda-intel: IRQ timing workaround is activated for card #0. Suggest a bigger bdl_pos_adj.
Apr 19 19:25:25 agamemnon kernel: [ 94.711410] CE: hpet increasing min_delta_ns to 15000 nsec

These ppa parts should go to lucid asap;-)

[Daniel T Chen (crimsun)]

They won't go into Lucid, which releases in less than two weeks;
nothing remotely resembling a kitten-killer is in this bug.

That said, I'm a bit perplexed. Your SSID is in the LPIB quirk table,
which was necessary for a matching board. Do you have anything in
/etc/modprobe.d/alsa-base.conf that you added manually?

[EricDHH (ericdhh)]

My alsa-base.conf has an old date, i did not change things there cause years ago everything works fine. But the system was updated from 8.10 till now.

-rw-r--r-- 1 root root 2386 Jan 29 01:01 alsa-base.conf

I append the file to give a chance to review.

[Daniel T Chen (crimsun)]

(setting status because of fix being in linux-alsa-driver-modules)

[EricDHH (ericdhh)]

Did further investigations, with ppa the sound is mostly clear but sometimes the stutter occurs with ratelimit again. So i searched for the vendor id 0x10ec0888 and found that the chip seems to be the alc1200. Trying some module options in alsa-base.conf did not help to fix this completely.
I found pointings at suse and fedora to use the driver from realtek.com.tw. Got the 5.14 audiopack from their server and compile the drivers well in lucid beta. Now there is no stutter and crackle anymore, their driver are from 04/2010.

[EricDHH (ericdhh)]

Today i got kernel 2.6.32-24 for amd64. The alc888 does already the same stutter as before, there is no machine load required for that stutter in audio, but there are no more loglines that count the stutter.
Jul 8 21:52:53 agamemnon kernel: [ 78.996943] hda-intel: IRQ timing workaround is activated for card #0. Suggest a bigger bdl_pos_adj.
And like before the realtek linux driverpack 5.14 can fix this. There must be differences between the kernel alsa and their drivers.

[EricDHH (ericdhh)]

Got kernel package 2.6.32-25 today, same hiccup and stutter without any machine load. THis is somewhat ugly and none alsa tip fix that.

[ 27.747264] hda-intel: IRQ timing workaround is activated for card #0. Suggest a bigger bdl_pos_adj.
[ 260.859848] CE: hpet increasing min_delta_ns to 15000 nsec

With realteks alsa drivers anything is fine like before: realtek-linux-audiopack-5.15. Why will noone build these into the next kernel package and everything is fine?

[EricDHH (ericdhh)]

Description: Ubuntu 10.10
Release: 10.10

New kernel, new ubuntu but stutter and blobs again in the sound. This happens without system load and give again following errors.

[ 36.066639] hda-intel: IRQ timing workaround is activated for card #0. Suggest a bigger bdl_pos_adj.

Please maverick users check this on your system. I can fix this with the realtek alsa source, so there must be a difference to the maverick alsa package.

[EricDHH (ericdhh)]

Opening a new bug against the alsa driver in kernel now, this can be closed.

[Launchpad Janitor (janitor)]

[Expired for alsa-driver (Ubuntu) because there has been no activity for 60 days.]