F14: high temperatures after coming back from hibernate

JD jd1008 at gmail.com
Tue Nov 9 03:35:19 UTC 2010


On 11/08/2010 07:22 PM, Ranjan Maitra wrote:
> On Fri, 5 Nov 2010 18:55:19 -0500 Gordon Messmer<yinyang at eburg.com>
> wrote:
>
>> On 11/04/2010 05:40 AM, Ranjan Maitra wrote:
>>> I have noticed a strange problem after I come back from hibernate and
>>> upgrade in F14 on my Thinkpad T61. The temperatures are off the charts,
>>> hitting as high as 95C, and staying there, soon after waking up. The
>>> laptop has become unusable, basically. Here is the output of top: the
>>> topmost processes. Note that the laptop has 4 GB of memory,
>>>
>>>    1751 maitra    20   0  422m 2232  964 R 99.7  0.1  22:05.74 pulseaudio
>> It looks like pulseaudio is in some sort of loop.  I'd attach strace to
>> it and see what it's doing.  The PID is 1751, so you'd do something like:
>>
>> strace -f -s 256 -p 1751
>>
>> and cancel with Ctrl+C.  You'll probably notice that there are sections
>> of output that simply repeat.  Send a bit back to the list.  It's hard
>> to say whether or not the problem will be apparent, but that'll be a
>> place to start.
>> -- 
>> users mailing list
>> users at lists.fedoraproject.org
>> To unsubscribe or change subscription options:
>> https://admin.fedoraproject.org/mailman/listinfo/users
>> Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
>>
> Hi,
>
> Many thanks for your e-mail. It appears that you are right because the
> following loops incessantly, and only when I go into pm-hibernate, not
> when I go into pm-suspend.
>
> Here is the necessary output:
>
> % top
>
>   PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND
>   1589 maitra     9 -11  422m 2176 1048 R 94.9  0.1   8:48.81 pulseaudio
>
> % strace -f -s 256 -p 1589
>
> The following keeps on scrolling:
>
> [pid  1589] ppoll([{fd=5, events=POLLIN}, {fd=21, events=POLLIN}, {fd=9, events=0}, {fd=4, events=POLLIN}, {fd=20, events=POLLIN}, {fd=26, events=POLLIN}, {fd=31, events=POLLIN|POLLERR|POLLHUP}, {fd=31, events=0}, {fd=30, events=POLLIN}, {fd=29, events=POLLIN}, {fd=22, events=POLLIN}, {fd=25, events=POLLIN}, {fd=16, events=POLLIN}, {fd=19, events=POLLIN}, {fd=15, events=POLLIN|POLLERR|POLLHUP}, {fd=15, events=0}, {fd=14, events=POLLIN}, {fd=13, events=POLLIN}, {fd=7, events=POLLIN}], 19, NULL, NULL, 8) = 1 ([{fd=9, revents=POLLHUP}])
>
> Killing pulseaudio (the process) seems to take care of it, and the
> temperature cools.
>
> Once again, I am running the F14-LXDE spin.
>
> Many thanks,
> Ranjan
>
Hello Ranjan,
I wrote about the pulseaudio taking 40% of cpu when nothing was playing.
The cause of this constant looping is a bug in the Adobe flash plugin.
Pulseaudio daemon listens on a socket FD for incoming audio.
Each client (such as the adobe flash plugin) is supposed to close
the socket after the media finishes playing. Well, the adobe flash plugin
does not close the socket, causing pulseaudio to be in a tight loop,
polling the socket and finding nothing to process.

In contradistinction to the Adobe plugin, run
/usr/bin/ffplay -autoexit somefile.mp3 (or any media file).
While the file is being played, you will notice pulseaudio
take a lot of the CPU. Once ffplay exits and closes the audio
socket, pulseaudio will consume a very negligible amount
of cpu, approaching 0%.

Cheers,

JD


More information about the users mailing list