Bug 565883

Summary: CPU consumption goes over 100%, duplicate entries in sound preferences, new mail sound cut off, unable to downgrade
Product: [Fedora] Fedora Reporter: udo <udovdh>
Component: pulseaudioAssignee: udo <udovdh>
Status: CLOSED DUPLICATE QA Contact: Fedora Extras Quality Assurance <extras-qa>
Severity: high Docs Contact:
Priority: high    
Version: 12CC: lkundrak, lpoetter
Target Milestone: ---Keywords: Reopened
Target Release: ---   
Hardware: All   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2010-02-18 11:58:10 UTC Type: ---
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:
Attachments:
Description Flags
The output you asked for. none

Description udo 2010-02-16 16:08:53 UTC
Description of problem: CPU consumption goes over 100%


Version-Release number of selected component (if applicable):
Version     : 0.9.21                            Vendor: Fedora Project
Release     : 4.fc12                        Build Date: Mon 18 Jan 2010 02:15:00 AM CET


How reproducible:
Run Fedora 12, use Firefox, Thunderbird. Watch youtube videos. See CPU consumption go to 100+% after a while.
  PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND            
  392 udo        9 -11  497m 6396 4688 R 100.1  0.2  33:24.89 pulseaudio        
 1027 udo       20   0  423m 122m  20m R 83.8  3.1  17:14.18 npviewer.bin       
  979 udo       20   0  829m 180m  25m S  3.7  4.6   2:49.18 firefox   

This is on AMD Phenom(tm) 9550 Quad-Core Processor so one full core is wasted for playing cd-quality audio from youtube vids. (!!)

Steps to Reproduce:
1. See above.
2.
3.
  
Actual results:
See above.

Expected results:
Negligible CPU consumption for this type of usage pattern. 

Additional info:
$ rpm -qa|grep puls
pulseaudio-0.9.21-4.fc12.x86_64
xmms-pulse-0.9.4-8.fc12.x86_64
alsa-plugins-pulseaudio-1.0.22-1.fc12.x86_64
pulseaudio-libs-glib2-0.9.21-4.fc12.x86_64
pulseaudio-utils-0.9.21-4.fc12.x86_64
pulseaudio-module-zeroconf-0.9.21-4.fc12.x86_64
pulseaudio-module-x11-0.9.21-4.fc12.x86_64
pulseaudio-module-bluetooth-0.9.21-4.fc12.x86_64
pulseaudio-gdm-hooks-0.9.21-4.fc12.x86_64
pulseaudio-libs-0.9.21-4.fc12.i686
pulseaudio-esound-compat-0.9.21-4.fc12.x86_64
alsa-plugins-pulseaudio-1.0.22-1.fc12.i686
pulseaudio-libs-0.9.21-4.fc12.x86_64
pulseaudio-module-gconf-0.9.21-4.fc12.x86_64

a `killall pulseaudio` fixes the issue. pulseaudio is restarted and consumes less CPU for a while.

Comment 1 udo 2010-02-16 16:11:22 UTC
I sometimes notice the effects mentioned at https://bugzilla.redhat.com/show_bug.cgi?id=564390 when the CPU is at 100% for pulseaudio.

Comment 2 udo 2010-02-16 16:13:13 UTC
I also noticed an incomplete sound for `new mail` as described at https://bugzilla.redhat.com/show_bug.cgi?id=559467.

Comment 3 udo 2010-02-16 16:24:13 UTC
Due to dependency-hell, downgrading is not as intuitive as `yum downgrade pulseaudio`.
This, together with the combination of the three observed issues this does not imply that the bug should be lower in priority or be labelled 'duplicate'.

Comment 4 udo 2010-02-16 16:24:35 UTC
(to be clear: downgrading hasn't happened/worked for me)

Comment 5 udo 2010-02-16 18:49:17 UTC
I see multiple lines in /var/log/messages saying stuff like:
protocol-esound.c: pulsecore/protocol-esound.c: Failed to create sink input.

or

ratelimit.c: 14 events suppressed

or

sink-input.c: Failed to create sink input: too many inputs per sink.

Comment 6 Lennart Poettering 2010-02-18 02:47:47 UTC
Please provide the output of "pacmd ls" when this happens.

Comment 7 Lennart Poettering 2010-02-18 03:04:26 UTC

*** This bug has been marked as a duplicate of bug 559467 ***

Comment 8 udo 2010-02-18 06:31:08 UTC
Created attachment 394861 [details]
The output you asked for.

The output you asked for.

Comment 9 Lennart Poettering 2010-02-18 11:58:10 UTC

*** This bug has been marked as a duplicate of bug 559467 ***