search

luochuang1 / enh

Automatically exported from code.google.com/p/enh
0 stars 0 forks source link

DalvikExplorer reports one core when hardware has two #33

Closed GoogleCodeExporter closed 9 years ago

GoogleCodeExporter commented 9 years ago 
What steps will reproduce the problem?
1.Run the application "dalvik explorer" on a dual core phone
2.observe the result. 
3.note that twin dd if=/dev/zero of=/dev/null 
  in a shell changes(corrects) the result.

What is the expected output? 
    Manufacturer: motorola
    Model: MB865
    CPU ABI: armeabi-v7a
    Cores: 2

What do you see instead?
    Manufacturer: motorola
    Model: MB865
    CPU ABI: armeabi-v7a
    Cores: 1

What version of the product are you using? On what operating system?
Same result on Samsung/SGH-I897/SGH-I897 and this Moto Atrix.
+++++++++++++++++++++++++++++
android.vm.dexfile=true
file.encoding=UTF-8
file.separator=/
http.agent=Dalvik/1.4.0 (Linux; U; Android 2.3.6; MB865 
Build/5.5.1-175_EDMR1.25)
java.boot.class.path=/system/framework/core.jar:/system/framework/bouncycastle.j
ar:/system/framework/ext.jar:/system/framework/framework.jar:/system/framework/a
ndroid.policy.jar:/system/framework/services.jar:/system/framework/core-junit.ja
r:/system/framework/com.motorola.android.frameworks.jar:/system/framework/com.mo
torola.android.widget.jar
java.class.path=.
java.class.version=46.0
java.compiler=

++++++++++++++++++++++++++++++++++++
Manufacturer: motorola
Model: MB865

CPU ABI: armeabi-v7a
Cores: 1

Brand: MOTO
Board: p3
Device: edison

Build Fingerprint: 
MOTO/edison_att_us/edison:2.3.6/5.5.1-175_EDMR1.25/5.51.175.25:user/release-keys

DalvikVM Heap Size: 42 MiB

Screen Density: 240dpi (1.5x DIP)
Screen Size: 540 x 960 pixels
Exact DPI: 254.0 x 256.67368
Approximate Dimensions: 2.1" x 3.7" (4.3" diagonal)

Features:
  android.hardware.wifi
  android.hardware.location.network
  android.hardware.telephony
  android.hardware.location

Please provide any additional information below.
Running two coppies of:
   dd if=/dev/zero of=/dev/null 
changes the result to the expected two cores.

A better call than "Runtime.getRuntime().availableProcessors()"
might help or perhaps the base libraries under this call 
could be improved.

Original issue reported on code.google.com by nifty.mi...@gmail.com on 30 May 2012 at 8:11

GoogleCodeExporter commented 9 years ago 
sysconf _SC_PROCESSORS_ONLN was broken in bionic until some time in 2010, which 
probably means Gingerbread was broken. I'll reimplement it in Java in Dalvik 
Explorer so we can support old builds better.

Original comment by elliott....@gmail.com on 31 May 2012 at 5:04

GoogleCodeExporter commented 9 years ago 
fixed by revision 562 (version 2.6).

Original comment by elliott....@gmail.com on 31 May 2012 at 5:07

GoogleCodeExporter commented 9 years ago 
(let me know if that fixes things for you.)

Original comment by elliott....@gmail.com on 31 May 2012 at 6:26

GoogleCodeExporter commented 9 years ago 
Not quite.
I will grab /proc/cpuinfo and /proc/stat for you.

On my two phones what I now see is:
++++++++++++++++++
My Samsung.
http.agent=Dalvik/1.4.0 (Linux; U; Android 2.3.5; SAMSUNG-SGH-I897
Build/GINGERBREAD)
shows.....
Manufacturer: SAMSUNG
Model: SAMSUNG-SGH-I897

CPU ABI: armeabi-v7a
Cores: 0 (enabled: 1)
+++++++++++++++++++++++++++++++++++
My Motorola phone is unchanged:
http.agent=Dalvik/1.4.0 (Linux; U; Android 2.3.6; MB865
Build/5.5.1-175_EDMR1.25)
Manufacturer: motorola
Model: MB865
CPU ABI: armeabi-v7a
Cores: 1
Brand: MOTO
Board: p3
Device: edison
Build Fingerprint:
MOTO/edison_att_us/edison:2.3.6/5.5.1-175_EDMR1.25/5.51.175.25:user/release-keys
++++++++++++++++++++++
When this is done:
It might also be valuable to put a header on the report with a
date:time and DalvikExplorer version
and perhaps something unique about the phone so in a week or a month
the data could be sorted out.
The verson has double value in that it lets you know what version of
DE is running as well as it permits
someone to parse the file should the format change in the future.
Right now I believe that DE did up-date
on the Moto phone but since it acts exactly the same (dd trick changes
Cores count) I will have to
double check.

Thanks,
mitch

Original comment by nifty.mi...@gmail.com on 31 May 2012 at 8:25

GoogleCodeExporter commented 9 years ago 
For the Samsung phone. /proc/stat and /proc/cpuinfo
$ cat  /proc/stat
cpu  593950 37287 330868 2264692 129664 98 23759 0 0 0
cpu0 593950 37287 330868 2264692 129664 98 23759 0 0 0
intr 21969188 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1083 0 0 0 1083 0 0 0
3412 0 7 0 2574 0 0 0 0 0 0 0 1 5 0 0 0 42 0 0 0 3 0 0 0 0 988011 0 0 0 0 0
159 0 0 6585414 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 148779 0 0 0 0 1
0 9962191 3240635 94150 0 0 0 0 384400 0 0 0 0 7389 0 0 0 147529 0 0 0 7565
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 377418 3 39 2 0 28 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 15797 0 0 0 0 0 0 0 0 0 0 0 0 1083 0 0 0 0 0 0 0 0 0
0 0 0 385 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ctxt 46800737
btime 1338463260
processes 37335
procs_running 2
procs_blocked 0
softirq 16530820 0 2824421 1 574294 0 0 12495175 0 6918 630011
$ cat /proc/cpuinfo
Processor       : ARMv7 Processor rev 2 (v7l)
BogoMIPS        : 996.00
Features        : swp half thumb fastmult vfp edsp thumbee neon vfpv3
CPU implementer : 0x41
CPU architecture: 7
CPU variant     : 0x2
CPU part        : 0xc08
CPU revision    : 2

Hardware        : herring
Revision        : 0030
Serial          : 3332cfba413300ec

Original comment by nifty.mi...@gmail.com on 31 May 2012 at 9:05

GoogleCodeExporter commented 9 years ago 
the internet implies that part 0xc08 is an OMAP3430, which is single-core, so i 
think Dalvik Explorer and the Linux kernel are in agreement here, and correct. 
(i wish i had a canonical list of ARM cpuid part numbers. why isn't there a 
wikipedia page for that?)

so i think this is a Nexus S (or similar), which was single core. oh, but 
you're saying i report 0? oops. i'll have to test on a single-core phone. i 
only tested on >= 2, stupidly. yeah, judging by your output it looks like the 
SMP kernel's output is different; the uniprocessor kernel doesn't output the 
"processor" lines; just a single "Processor" line. which is also there in the 
SMP output, but needs to be ignored. ugh.

i'll look at fixing this tonight if i get time. i like your version number idea 
too.

Original comment by elliott....@gmail.com on 31 May 2012 at 9:49

GoogleCodeExporter commented 9 years ago 
The below update is info from my Atrix2 which has two cores.

Update:
DalvikExp 2.6 thinks this Atrix2 has 1 core.which switches to 2 cores
with a dd if=/dev/zero of=/dev/null running.

The root of the problem may be that  /proc/stat also changes so it
may be that the OS shuts down the second core and there is no easy
way to know short of a look up table by cpu type which will not tell you if the
OS will ever light it up again.  Perhaps report "Currently Active Cores"

ATRIX....MB865  processor is TI OMAP4430 and Moto has a big old "Dual
core" splash screen at startup
$ cat /proc/stat
cpu  448442 40045 198354 6469762 107462 25 3847 0 0 0
cpu0 405103 34002 183907 2987157 100389 25 3632 0 0 0
intr 5537088 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 466482 119311 0 0 0 0 0 0 0 0 544776 0 0 0
136916 78721 0 0 0 0 0 0 0 0 0 0 404025 0 7 0 0 0 0 0 0 0 0 0 0 0 0 0
412321 0 0 901 747754 397 0 0 1 452569 0 0 0 0 5193 0 0 7661 167946 5
234 0 0 0 0 0 0 0 0 0 21158 0 0 1283241 0 0 0 0 0 0 0 285738 0 0 0 0
127681 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 6 0 0 0 0 0 0 0 0 0 0 0
0 0 0 26249 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 191184 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
1 0 17670 0 0 0 0 0 38939 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0
ctxt 27963118
btime 1338330310
processes 46635
procs_running 2
procs_blocked 0
softirq 3147469 175426 1509764 2043 247347 1848 1848 90087 96816 2716 1019574

$cat /proc/cpuinfo
Processor       : ARMv7 Processor rev 3 (v7l)
processor       : 0
BogoMIPS: 1996.68

Features: swp half thumb fastmult vfp edsp thumbee neon vfpv3
CPU implementer : 0x41
CPU architecture: 7
CPU variant     : 0x1
CPU part: 0xc09
CPU revision    : 3

Hardware: mapphone_UMTS
Revision: 0000
Serial  : 0000000000000000
$ Atrix

Original comment by nifty.mi...@gmail.com on 31 May 2012 at 10:53

GoogleCodeExporter commented 9 years ago 
odd. my expectation was that you'd get "Cores: 2 (enabled: 1)" in cases like 
this.

i'd be curious to see "ls -l /sys/devices/system/cpu" in a case like this. 
(also "ls -l /sys/devices/system/cpu/cpu*".)

(i don't actually have any hardware that shuts cores down myself.)

Original comment by elliott....@gmail.com on 31 May 2012 at 11:07

GoogleCodeExporter commented 9 years ago 
Hmm... I am not sure they do shut down...
This is on the Motorola Atrix2...
$ ls -l /sys/devices/system/cpu
-r--r--r-- root     root                 4096 2012-05-31 16:39 online
-r--r--r-- root     root                 4096 2012-05-31 16:39 possible
-r--r--r-- root     root                 4096 2012-05-31 16:39 present
-r--r--r-- root     root                 4096 2012-05-31 16:39 kernel_max
-r--r--r-- root     root                 4096 2012-05-31 16:39 offline
drwxr-xr-x root     root                        2012-05-31 16:39 cpufreq
drwxr-xr-x root     root                        2012-05-31 16:39 cpuidle
drwxr-xr-x root     root                        2012-05-31 15:57 cpu0
drwxr-xr-x root     root                        2012-05-31 16:39 cpu1
app_152@edison:/$

...............DE version 2.6............
Build/Device Details

Manufacturer: motorola
Model: MB865

CPU ABI: armeabi-v7a
Cores: 1

Brand: MOTO
Board: p3
Device: edison

Build Fingerprint:
MOTO/edison_att_us/edison:2.3.6/5.5.1-175_EDMR1.25/5.51.175.25:user/release-keys

DalvikVM Heap Size: 42 MiB

Screen Density: 240dpi (1.5x DIP)
Screen Size: 540 x 960 pixels
Exact DPI: 254.0 x 256.67368
Approximate Dimensions: 2.1" x 3.7" (4.3" diagonal)

Features:
  android.hardware.wifi
  android.hardware.location.network
  android.hardware.telephony
  android.hardware.location
  android.hardware.touchscreen.multitouch
  android.hardware.sensor.compass
  android.hardware.camera
  android.hardware.bluetooth
  android.hardware.touchscreen.multitouch.distinct
  android.hardware.sensor.proximity
  android.hardware.microphone
  android.hardware.sensor.light
  android.hardware.location.gps
  com.motorola.vpn.cisco
  android.hardware.camera.autofocus
  android.hardware.telephony.gsm
  android.hardware.camera.front
  android.hardware.sensor.accelerometer
  android.software.live_wallpaper
  android.hardware.touchscreen
  android.hardware.camera.flash

OpenGL ES version: 2.0

DalvikVM features:
  method-trace-profiling
  method-trace-profiling-streaming
  hprof-heap-dump
  hprof-heap-dump-streaming

Shared Java libraries:
  com.motorola.android.telephony
  android.test.runner
  com.motorola.blur.library.service
  com.motorola.blur.library.messaging
  com.motorola.android.storage
  com.cequint.platform.1.4.75
  com.motorola.vpn.ext
  com.motorola.blur.library.image
  com.motorola.android.datamanager.base
  com.motorola.android.customization
  com.scalado.caps
  jcifs-krb5-1.3.12
  com.motorola.blur.library.utilities
  com.android.location.provider
  com.motorola.android.imirrorservice
  com.motorola.android.camera
  javax.obex
  com.motorola.android.iextdispservice
  com.motorola.android.mobad
..................
=======================

$ ls -l /sys/devices/system/cpu/cpu*
-rw-r--r-- root     root                 4096 2012-05-31 16:47 online
drwxr-xr-x root     root                        2012-05-31 16:47 topology
drwxr-xr-x root     root                        2012-05-31 16:47 cpuidle
drwxr-xr-x root     root                        2012-05-31 15:57 cpufreq
-rw-r--r-- root     root                 4096 2012-05-31 16:47 online
drwxr-xr-x root     root                        2012-05-31 16:47 cpuidle
drwxr-xr-x root     root                        2012-05-31 16:47 mot_hotplug
-r--r--r-- root     root                 4096 2012-05-31 16:47 current_driver
-r--r--r-- root     root                 4096 2012-05-31 16:47
current_governor_ro
app_152@edison:/$

Original comment by nifty.mi...@gmail.com on 31 May 2012 at 11:51

GoogleCodeExporter commented 9 years ago 
why don't you think they shut down? i think that's exactly what's going on. (i 
think that's what the mot_hotplug stuff is all about, and why /proc/stat and 
/proc/cpuinfo don't match /sys/devices/system/cpu.)

could you show the content of the cpu0 and cpu1 directories separately? i 
forgot about the cpufreq and cpuidle directories. could you also cat the 
"online" files in those directories? i'm assuming they say "1" and "0" when one 
core is asleep, and "1" and "1" when they're both up.

Original comment by elliott....@gmail.com on 1 Jun 2012 at 12:09

GoogleCodeExporter commented 9 years ago 
You could be correct...  it can be easy to explore

sure... and one better I hope.
I can run modest shell scripts like this for you:
#!/system/bin/sh
set -x
cat /proc/stat
echo ===========================================
cat  /proc/cpuinfo
echo ===========================================
ls -lR  /sys/devices/system/cpu
echo ===========================================
ls -l /sys/devices/system/cpu/cpu0
echo ===========================================
ls -l /sys/devices/system/cpu/cpu1

sleep 2
dd if=/dev/zero of=/dev/null &

cat /proc/stat
echo ===========================================
cat  /proc/cpuinfo
echo ===========================================
ls -lR  /sys/devices/system/cpu
echo ===========================================
ls -l /sys/devices/system/cpu/cpu0
echo ===========================================
ls -l /sys/devices/system/cpu/cpu1

sleep 2
kill %1
+++++++++++++++++++++++++++++++++++++++++++++++
The above results in:
$ sh < tinker

+ cat /proc/stat
cpu  29470 7289 16968 363032 19300 3 169 0 0 0
cpu0 23678 5075 14313 170267 15318 3 137 0 0 0
intr 560568 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 25788 928 0 0 0 0 0 0 0 0 65551 0 0 0
11468 5 0 0 0 0 0 0 0 0 0 0 28526 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 34654
0 0 248 95688 397 0 0 1 7624 0 0 0 0 400 0 0 4785 21927 77 0 0 0 0 0 0
0 0 0 0 18754 0 0 211764 0 0 0 0 0 2721 3308 6672 0 0 0 0 3396 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 12 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2522 1
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 6408 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 3148 0 0 0 0
0 3793 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ctxt 2320618
btime 1338504921
processes 6108
procs_running 1
procs_blocked 0
softirq 265746 22490 110674 141 9637 129 129 7428 11633 172 103313

+ echo ===========================================
===========================================

+ cat /proc/cpuinfo
Processor       : ARMv7 Processor rev 3 (v7l)
processor       : 0
BogoMIPS        : 1996.68

Features        : swp half thumb fastmult vfp edsp thumbee neon vfpv3
CPU implementer : 0x41
CPU architecture: 7
CPU variant     : 0x1
CPU part        : 0xc09
CPU revision    : 3

Hardware        : mapphone_UMTS
Revision        : 0000
Serial          : 0000000000000000

+ echo ===========================================
===========================================

+ ls -lR /sys/devices/system/cpu
-lR: No such file or directory
online
possible
present
kernel_max
offline
cpufreq
cpuidle
cpu0
cpu1

+ echo ===========================================
===========================================

+ ls -l /sys/devices/system/cpu/cpu0
-rw-r--r-- root     root                 4096 2012-05-31 17:40 online
drwxr-xr-x root     root                        2012-05-31 17:40 topology
drwxr-xr-x root     root                        2012-05-31 17:40 cpuidle
drwxr-xr-x root     root                        2012-05-31 17:38 cpufreq

+ echo ===========================================
===========================================

+ ls -l /sys/devices/system/cpu/cpu1
-rw-r--r-- root     root                 4096 2012-05-31 17:40 online
drwxr-xr-x root     root                        2012-05-31 17:40 cpuidle

+ sleep 2

+ dd if=/dev/zero of=/dev/null

+ cat /proc/stat
cpu  29475 7289 16977 363226 19300 3 169 0 0 0
cpu0 23682 5075 14321 170461 15318 3 137 0 0 0
intr 560847 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 25798 928 0 0 0 0 0 0 0 0 65586 0 0 0
11478 5 0 0 0 0 0 0 0 0 0 0 28559 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 34682
0 0 248 95804 397 0 0 1 7624 0 0 0 0 400 0 0 4785 21927 77 0 0 0 0 0 0
0 0 0 0 18754 0 0 211788 0 0 0 0 0 2726 3313 6674 0 0 0 0 3396 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 12 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2525 1
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 6410 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 3154 0 0 0 0
0 3793 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ctxt 2321432
btime 1338504921
processes 6115
procs_running 3
procs_blocked 0
softirq 265827 22490 110713 141 9637 129 129 7428 11633 172 103355

+ echo ===========================================
===========================================

+ cat /proc/cpuinfo
Processor       : ARMv7 Processor rev 3 (v7l)
processor       : 0
BogoMIPS        : 598.90

Features        : swp half thumb fastmult vfp edsp thumbee neon vfpv3
CPU implementer : 0x41
CPU architecture: 7
CPU variant     : 0x1
CPU part        : 0xc09
CPU revision    : 3

Hardware        : mapphone_UMTS
Revision        : 0000
Serial          : 0000000000000000

+ echo ===========================================
===========================================

+ ls -lR /sys/devices/system/cpu
-lR: No such file or directory
online
possible
present
kernel_max
offline
cpufreq
cpuidle
cpu0
cpu1

+ echo ===========================================
===========================================

+ ls -l /sys/devices/system/cpu/cpu0
-rw-r--r-- root     root                 4096 2012-05-31 17:40 online
drwxr-xr-x root     root                        2012-05-31 17:40 topology
drwxr-xr-x root     root                        2012-05-31 17:40 cpuidle
drwxr-xr-x root     root                        2012-05-31 17:38 cpufreq

+ echo ===========================================
===========================================

+ ls -l /sys/devices/system/cpu/cpu1
-rw-r--r-- root     root                 4096 2012-05-31 17:40 online
drwxr-xr-x root     root                        2012-05-31 17:40 cpuidle

+ sleep 2

+ kill %1
[1]   Terminated              sh <tinker
app_152@edison:/mnt/sdcard$

Original comment by nifty.mi...@gmail.com on 1 Jun 2012 at 12:49

GoogleCodeExporter commented 9 years ago 
manually taking processors offline, i see that x86 and ARM kernels differ. on 
ARM, offline processors go missing from both /proc files, so we can't use /proc 
for _SC_NPROCESSORS_CONF. (this means bionic's still broken. i'll have to fix 
that.)

revision 565 switches to using /sys/devices/system/cpu/ for the total (not 
online) CPUs. glibc does that. interestingly, it uses both 
/sys/devices/system/cpu/online _and_ /proc/stat for the online processors. i'm 
not sure why, but since both seem to work fine, i'll stick with /proc/stat for 
now since it's easier.

Original comment by elliott....@gmail.com on 1 Jun 2012 at 5:05

GoogleCodeExporter commented 9 years ago 
(for the record, 0xc08 just means Cortex-A8, not specifically an OMAP.)

Original comment by elliott....@gmail.com on 1 Jun 2012 at 6:28

GoogleCodeExporter commented 9 years ago 
okay, should be fixed in version 2.7, which also decodes some of the 
/proc/cpuinfo stuff.

Original comment by elliott....@gmail.com on 1 Jun 2012 at 6:48

GoogleCodeExporter commented 9 years ago 
Thank you!

Original comment by nifty.mi...@gmail.com on 1 Jun 2012 at 6:43

GoogleCodeExporter commented 9 years ago 
also fixed in bionic: https://android-review.googlesource.com/37568

Original comment by elliott....@gmail.com on 6 Jun 2012 at 7:06