Snowball : Device hangs when running shell scripts on Android builds

Bug #1014650 reported by Abhishek Paliwal on 2012-06-18
12
This bug affects 2 people
Affects Status Importance Assigned to Milestone
IglooCommunity
High
Kejun ZHOU
LAVA Dispatcher
Fix Released
High
Yongqin Liu
Linaro Android
Fix Released
High
Kejun ZHOU
Linaro Image Tools
Fix Released
High
Mathieu Poirier

Bug Description

Description:
-------------
Device hangs completely when I'm trying to run shell scripts.
No information about the reason for hang could be found in logcat messages.

Steps to reproduce:
----------------------
1. Bootup snowball with sd card image for build:
https://android-build.linaro.org/builds/~linaro-android/snowball-ics-gcc46-igloo-stable-blob/#build=327

2. Copy the attached shell script(n.sh) to /data location on Android filesystem(use command from HOST machine
 "adb push n.sh /data")
This script is to go to HOME screen after a delay of every 1 second, in a while 1 loop.

3. On shell (serial console over USB interface for snowball) type following commands:
#logcat -v time &
#cd /data
#chmod 755 n.sh
#./n.sh &
Issue: After 59/60 iterations hard hang is observed. This is consistent behavior. Need to reboot device to use it again.

Also, same issues are also seen if I try to run few other shell scripts.

Reproducible:
------------
Yes, just bootup the board and run script, crash is seen everytime.

Logs:
------
Logcat messages attached.

description: updated
vishal (vishalbhoj) wrote :

Affects the system profiling on snowball.

Changed in linaro-android:
importance: Undecided → Critical

Anmar: it is not similar to bug 993168 , its quite abrupt hang, I do not see any crash messages on serial console when board hangs.

description: updated
description: updated
vishal (vishalbhoj) wrote :

I have seen that system hangs when I run Linpack[1] under certain environments. Here is my observation :

1. System doesn't hang when I run Linpack Single/multi thread configuration as a normal application by launching and running it (using monkeyrunner in my case)
2. System hangs when I run the same application when I run the application in conjunction with androidscreencast .

It could be that system hangs only at some high loads .

[1] : http://android.git.linaro.org/gitweb?p=linaro/benchmarks.git;a=blob;f=linpack/com.greenecomputing.linpack-1.apk;h=d0e38866e8a2725836e9c7f228e7ffdece1d1242;hb=ff1690af89b0b26ef5b7d956c54c985bee468ea0

Hang also observed when trying to do bluetooth file transfer. (on same build#327)
Sending a image file from gallery app (Snowball) to another paired phone.

Anmar Oueja (anmar) on 2012-06-26
Changed in igloocommunity:
status: New → In Progress
importance: Undecided → High
assignee: nobody → Kejun ZHOU (kejun-zhou)
milestone: none → 2012.07
Kejun ZHOU (kejun-zhou) wrote :

Hi All,
The bug is hardly reproduced when vmalloc was increased to 380M.

I increased the vmalloc size to 380M by modifing the linaro-image-tools/linaro_image_tools/media_create/android_boards.py
as:

class AndroidSnowballSdConfig(AndroidBoardConfig, SnowballSdConfig):
    boot_script = 'boot.scr'
    initrd_addr = '0x05000000'
    extra_boot_args_options = (
        'earlyprintk mem=128M@0 mali.mali_mem=64M@128M hwmem=168M@192M mem=22M@360M mem_issw=1M@383M mem=640M@384M vmalloc=380M')
    _extra_serial_opts = 'console=ttyAMA2,115200n8'
    android_specific_args = 'init=/init androidboot.console=ttyAMA2'
    dtb_name = None

Zach Pfeffer (pfefferz) on 2012-07-02
Changed in linaro-android:
status: New → In Progress
importance: Critical → High
assignee: nobody → Kejun ZHOU (kejun-zhou)
milestone: none → 12.07
Fathi Boudra (fboudra) on 2012-07-05
Changed in linaro-image-tools:
status: New → Fix Committed
importance: Undecided → High
assignee: nobody → Mathieu Poirier (mathieu.poirier)
milestone: none → 2012.07
Fathi Boudra (fboudra) on 2012-07-05
Changed in lava-dispatcher:
status: New → Fix Committed
importance: Undecided → High
assignee: nobody → Yongqin Liu (liuyq0307)
milestone: none → 2012.07
Kaushik (c-a-subramaniam) wrote :

Can you please check if this fixes the vmalloc issue? We saw something similar in OMAP too with 3.4 kernel

http://www.arm.linux.org.uk/developer/patches/viewpatch.php?id=7438/1

After discussing the issue with Vishal it was decided to increase (once more) vmalloc to 500M. At this limit the system is stable and monkeyrunner isn't failing anymore.

Milo Casagrande (milo) on 2012-07-23
Changed in linaro-image-tools:
status: Fix Committed → Fix Released
Zach Pfeffer (pfefferz) on 2012-07-23
Changed in linaro-android:
status: In Progress → Fix Committed
milestone: 12.07 → 12.08
Zach Pfeffer (pfefferz) on 2012-07-24
Changed in linaro-android:
milestone: 12.08 → 12.07
status: Fix Committed → Fix Released
Kejun ZHOU (kejun-zhou) on 2012-08-01
Changed in igloocommunity:
status: In Progress → Fix Released
Fathi Boudra (fboudra) on 2012-08-03
Changed in lava-dispatcher:
status: Fix Committed → Fix Released
To post a comment you must log in.
This report contains Public information  Edit
Everyone can see this information.

Other bug subscribers