[vdr] background vdr thread eating cpu

Klaus Schmidinger Klaus.Schmidinger at cadsoft.de
Wed Jan 3 18:34:47 CET 2007


Halim Sahin wrote:
> Hi,
> On Mi, Jan 03, 2007 at 04:25:57 +0100, Klaus Schmidinger wrote:
>> Can you identify a particular VDR thread that consumes those
>> 20% more CPU time?
>>
> 
> Thats very dificult.
> After observing this a time here is the output of
> ps -T u -C vdr
> 
> USER       PID  SPID %CPU %MEM    VSZ   RSS TTY      STAT START   TIME COMMAND
> root      5252  5252  0.0  1.0  31536 10628 tty1     Rl+  17:59   0:00 ./vdr -E /dev/shm/epg.data -w 40 -s /sbin/go_down -l 0 -c /etc/vdr
> root      5252  5262  0.0  1.0  31536 10628 tty1     Sl+  17:59   0:00 ./vdr -E /dev/shm/epg.data -w 40 -s /sbin/go_down -l 0 -c /etc/vdr
> root      5252  5263  0.4  1.0  31536 10628 tty1     SNl+ 17:59   0:01 ./vdr -E /dev/shm/epg.data -w 40 -s /sbin/go_down -l 0 -c /etc/vdr
> root      5252  5265  0.0  1.0  31536 10628 tty1     Sl+  17:59   0:00 ./vdr -E /dev/shm/epg.data -w 40 -s /sbin/go_down -l 0 -c /etc/vdr
> root      5252  5266  0.3  1.0  31536 10628 tty1     RNl+ 17:59   0:01 ./vdr -E /dev/shm/epg.data -w 40 -s /sbin/go_down -l 0 -c /etc/vdr
> root      5252  5267  0.0  1.0  31536 10628 tty1     Sl+  17:59   0:00 ./vdr -E /dev/shm/epg.data -w 40 -s /sbin/go_down -l 0 -c /etc/vdr
> root      5252  5268  0.0  1.0  31536 10628 tty1     Sl+  17:59   0:00 ./vdr -E /dev/shm/epg.data -w 40 -s /sbin/go_down -l 0 -c /etc/vdr
> root      5252  5283  0.0  1.0  31536 10628 tty1     Sl+  17:59   0:00 ./vdr -E /dev/shm/epg.data -w 40 -s /sbin/go_down -l 0 -c /etc/vdr
> root      5252  5284  0.0  1.0  31536 10628 tty1     Sl+  17:59   0:00 ./vdr -E /dev/shm/epg.data -w 40 -s /sbin/go_down -l 0 -c /etc/vdr
> root      5252  5285  0.3  1.0  31536 10628 tty1     Sl+  17:59   0:01 ./vdr -E /dev/shm/epg.data -w 40 -s /sbin/go_down -l 0 -c /etc/vdr
> 
> I am not shure that you can read someting out of this.
> I can only tell you, that the cpuload without the dvb-t stick pluged in 
> is lower.

Maybe the CPU load is simply higher for the USB driver,
as suggested by Torgeir Veimo in his posting a few hours ago.

Klaus



More information about the vdr mailing list