Android lollipop 5.0.0 crashes in every 1 hour interval

[damienhuang]

Hi
I am using the lollipop5.0.0 image from the impatient section within boundary device website on my customized imx6 sabrelite board. But the android interface always crash & restart in about 1 hour interval except the first crash which was at 1 hour 53 minutes. Each crash always started with “binder: 180:301 transaction failed 29189, size 68-0″

for example below is a section of the crash message captured by dmesg (please see attachment for the full dmesg log).

”
[     89.894034] nf_conntrack: automatic helper assignment is deprecated and it will be removed soon. Use the iptables CT target to attach helpers instead.
[ 6803.895395] binder: 180:301 transaction failed 29189, size 68-0
[ 6803.907601] alarm_release: clear alarm, pending 0
[ 6803.911029] alarm_release: clear pending alarms 4
[ 6804.230239] request_suspend_state: wakeup (0->0) at 6804224170477 (2016-06-03 09:28:30.450868469 UTC)
[ 6804.370109] init: untracked pid 748 killed by signal 9
[ 6804.396867] init: untracked pid 783 killed by signal 9
[ 6804.403220] init: untracked pid 833 killed by signal 9
[ 6804.509155] init: untracked pid 893 killed by signal 9
[ 6804.523157] init: untracked pid 547 killed by signal 9
[ 6804.527250] init: untracked pid 667 killed by signal 9
[ 6804.531660] mtp_release
[ 6804.533276] init: untracked pid 571 killed by signal 9
[ 6804.659739] init: untracked pid 1690 killed by signal 9
[ 6804.673300] init: untracked pid 1181 killed by signal 9
[ 6804.710552] init: untracked pid 955 killed by signal 9
[ 6804.768529] init: untracked pid 1098 killed by signal 9
[ 6804.774956] init: untracked pid 1119 killed by signal 9
[ 6804.782339] init: untracked pid 1053 killed by signal 9
[ 6804.788160] init: untracked pid 1077 killed by signal 9
[ 6804.792283] init: untracked pid 1148 killed by signal 9
[ 6814.066928] request_suspend_state: wakeup (0->0) at 6814060858811 (2016-06-03 09:28:40.287557804 UTC)
[ 6820.452401] write_reg: set register 0x00 to 0x00
[ 6820.452410] set_mode: changed mode to 0x00
[ 6820.452901] write_reg: set register 0x00 to 0x00
[ 6820.452908] set_mode: changed mode to 0x00
[ 6820.547417] write_reg: set register 0x00 to 0x00
[ 6820.547427] set_mode: changed mode to 0x00
[ 6820.548308] write_reg: set register 0x00 to 0x00
[ 6820.548316] set_mode: changed mode to 0x00
[ 6820.626837] lowmemorykiller: lowmem_shrink: convert oom_adj to oom_score_adj:
[ 6820.632720] lowmemorykiller: oom_adj 0 => oom_score_adj 0
[ 6820.636891] lowmemorykiller: oom_adj 1 => oom_score_adj 58
[ 6820.641117] lowmemorykiller: oom_adj 2 => oom_score_adj 117
[ 6820.641121] lowmemorykiller: oom_adj 3 => oom_score_adj 176
[ 6820.641124] lowmemorykiller: oom_adj 9 => oom_score_adj 529
[ 6820.641127] lowmemorykiller: oom_adj 15 => oom_score_adj 1000
[ 6822.979227] acc_open
[ 6822.980133] acc_release
[ 6825.287630] mtp_open
[ 6826.932613] fec 2188000.ethernet eth0: Freescale FEC PHY driver [Micrel KSZ9031 Gigabit PHY] (mii_bus:phy_addr=2188000.ethernet:06, irq=188)
[ 7864.490347] init: untracked pid 1633 exited with status 0
[10440.220770] alarm_release: clear pending alarms c
[10440.222857] binder: 180:301 transaction failed 29189, size 68-0
[10440.589058] request_suspend_state: wakeup (0->0) at 10440582989910 (2016-06-03 10:29:06.809688903 UTC)
[10441.062400] init: untracked pid 29866 killed by signal 9
[10441.066761] init: untracked pid 29568 killed by signal 9
[10441.071253] init: untracked pid 29546 killed by signal 9
[10441.075586] init: untracked pid 29591 killed by signal 9
[10441.079824] init: untracked pid 29838 killed by signal 9
[10441.084013] init: untracked pid 29889 killed by signal 9
[10441.088285] init: untracked pid 29643 killed by signal 9
[10441.091224] mtp_release
[10441.093822] init: untracked pid 29307 killed by signal 9
[10441.098091] init: untracked pid 29286 killed by signal 9
[10441.102343] init: untracked pid 29464 killed by signal 9
[10441.106959] init: untracked pid 29716 killed by signal 9
[10441.111669] init: untracked pid 29913 killed by signal 9
[10441.115873] init: untracked pid 29943 killed by signal 9
[10441.120362] init: untracked pid 29981 killed by signal 9
[10441.124629] init: untracked pid 30218 killed by signal 9
[10450.363138] request_suspend_state: wakeup (0->0) at 10450357068245 (2016-06-03 10:29:16.583766904 UTC)
[10456.829500] write_reg: set register 0x00 to 0x00
[10456.829510] set_mode: changed mode to 0x00
[10456.829998] write_reg: set register 0x00 to 0x00
[10456.830004] set_mode: changed mode to 0x00
[10456.936864] write_reg: set register 0x00 to 0x00
[10456.936873] set_mode: changed mode to 0x00
[10456.937359] write_reg: set register 0x00 to 0x00
[10456.937366] set_mode: changed mode to 0x00
[10457.010034] lowmemorykiller: lowmem_shrink: convert oom_adj to oom_score_adj:
[10457.015890] lowmemorykiller: oom_adj 0 => oom_score_adj 0
[10457.020048] lowmemorykiller: oom_adj 1 => oom_score_adj 58
[10457.024247] lowmemorykiller: oom_adj 2 => oom_score_adj 117
[10457.028573] lowmemorykiller: oom_adj 3 => oom_score_adj 176
[10457.032862] lowmemorykiller: oom_adj 9 => oom_score_adj 529
[10457.037231] lowmemorykiller: oom_adj 15 => oom_score_adj 1000
[10458.938584] acc_open
[10458.939490] acc_release
[10461.011971] mtp_open
[10462.423779] fec 2188000.ethernet eth0: Freescale FEC PHY driver [Micrel KSZ9031 Gigabit PHY] (mii_bus:phy_addr=2188000.ethernet:06, irq=188)
[10462.439043] IPv6: ADDRCONF(NETDEV_UP): eth0: link is not ready
”

Can someone help to shed some light on this?

Cheers,

Original Attachment has been moved to: dmesg.log.zip

Original Attachment has been moved to: ps.log.zip

Original Attachment has been moved to: logcat.log.zip

[damienhuang]

Is there something wrong with the partition table? it seems the kernel have some errors in this area.

[vivekkaushik]

Hello Damien,

Board has booted completely and will remains working unless I go and start navigating the settings menu, as soon as I enter in settings menu and start navigating any menu option system resets. I don't think any thing wrong with partition table other wise system would not have booted completely. but still can you point where do you see is the problem related to partition table. 

Kind Regards

Vivek

[vivekkaushik]

Hello All,

We have ported android lollipop-5.0.0 for our 7” display custom product. We used the below part in out design

Processor: MCIMX6D5EYM12AD

PMIC: MMPF0100F0AEP

Android is booting completely on the system, but we are observing that it is resetting the system as soon as we start navigating through settings menu. Attached complete android logs for your reference.

Any Idea what is going wrong in the system.

Kind Regards

Vivek

[damienhuang · Accepted solution]

After I removed bluetooth & wifi from file "required_hardware.xml",  "../boundary/nitrogen6x/init.rc" and "../fsl/imx6/etc/init.rc", plus removed uim service from "../boundary/nitrogen6x/init.rc", logcat did not showing excessive error messages and the crash disappeared.

[gary_bisson]

Hi Damien,

Please copy/paste your logcat to pastebin and reply with the link.

Also, please specify what application is running at the time of the crash. It looks like the low memory killer kicks in. As a FYI, Android 5.0.0 has been known to have some memory leaks which have been fixed in Android 5.1.1, therefore I recommend to switch to the 5.1.1 version.

Regards,

Gary

[damienhuang]

Thanks Gary for the message.

please see attachment for

     1) ps result before the crash happen

     2) logcat message after the crash.

[gary_bisson]

Hi Damien,

Nothing obvious is in the logcat output except that the platform tries to use Bluetooth and WiFi whereas it's not there. I suggest you remove the Bluetooth and WiFi features from /system/etc/permissions/required_hardware.xml.

Also, please specify what modifications you've made to the original SabreLite design since this issue has not been reproduced on ours. If you have modified the RAM chips or layout, please perform some DDR test as Igor suggested.

Regards,

Gary

[damienhuang]

Thanks Gary for the suggestion, I will remove Bluetooth and wifi from the xml file and do a DDR test asap.

Look at the logcat file, I see the below pair of messages is keep popping up every second or two, do you know what this message pair regards to? And how can I adjust to remove it? Will this message relate to the crash since it may running out of memory eventually?

“

E/uim-sysfs( 1254): uim: Inside main

E/uim-sysfs( 1254): unexpected - found 0 kim devices

“

Cheers,

[damienhuang]

Regarding the above uim message, can I comment out the uim section inside "nitrogen6x/init.rc" as below?

"

#shared transport user space mgr service for Bluetooth, FM and GPS

#service uim /system/bin/uim-sysfs

#    class core

#    user bluetooth

#    group bluetooth net_bt_admin

#    seclabel u:r:hci_attach:s0

"

Cheers,

[igorpadykov]

Hi Damien

is working with just linux, had you run ddr test ?

https://community.freescale.com/docs/DOC-105652

Best regards

igor

-----------------------------------------------------------------------------------------------------------------------

Note: If this post answers your question, please click the Correct Answer button. Thank you!

-----------------------------------------------------------------------------------------------------------------------