#12538 NORM 4-softw: XO-4 CPU usage at idle is high

Zarro Boogs per Child bugtracker at laptop.org
Thu Feb 14 12:16:32 EST 2013

#12538: XO-4 CPU usage at idle is high
           Reporter:  dsd           |       Owner:  cjb          
               Type:  defect        |      Status:  new          
           Priority:  normal        |   Milestone:  4-software   
          Component:  not assigned  |     Version:  not specified
         Resolution:                |    Keywords:               
        Next_action:  never set     |    Verified:  0            
Deployment_affected:                |   Blockedby:               
           Blocking:                |  

Comment(by dsd):

 Testing 13.1.0 build 29 on an XO-4 B1 (1GHz, 8686) and an XO-4 C1 (1.2GHz,
 8787), trying to measure the effect of external SD card polling. Fresh
 reflash, automatic power management disabled in Sugar control panel.

 Having left the system a few minutes after boot idle on the sugar home
 screen, I measured load average with:
 sleep 1m && uptime
 sleep 1m && uptime
 sleep 1m && uptime

 And averaged the results in the left column, to get the load average over
 3 minutes.

 Then I measured CPU idle percentage in top, just by observation, reporting
 the number where the system seems to settle the most.

 I did these tests on 3 configurations on each laptop: shipped kernel,
 shipped kernel with external SD card inserted, and a custom kernel with
 these lines removed:

 pr_info("%s: FIXME: setting broken card detection quirk

 Results are:

 1. try on normal kernel, no SD card
 load 0.35
 idle 95%

 2. add SD card
 load 0.27
 idle 96%

 3. try on kernel with no polling
 load 0.35
 idle 97%

 1. try on normal kernel, no SD card
 load 0.42
 idle 80%

 2. add SD card
 idle 82%
 load 0.25

 3. try on kernel with no polling
 idle 85%
 load 0.36

 ***Quick analysis***

 It is obvious that the polling of the external SD card slot has some
 impact on system resources but it seems relatively small in my testing.

 The most interesting observation is that the B1 idle percentage looks just
 fine, whereas the C1 looks bad.

Ticket URL: <http://dev.laptop.org/ticket/12538#comment:1>
One Laptop Per Child <http://laptop.org/>
OLPC bug tracking system

More information about the Bugs mailing list