LAVA Validation Lab

LAVA Android test "skia" failed on android build "origen-ics-gcc47-samsunglt-stable-blob/#build=40"

Bug #1006745 reported by Abhishek Paliwal
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
LAVA Validation Lab
Invalid
Low
Yongqin Liu

Bug Description

Description:
-----------------

On Android build:
https://android-build.linaro.org/builds/~linaro-android/origen-ics-gcc47-samsunglt-stable-blob/#build=40

skia test failed.

device is not found when executing this test even though device is connected.

From Logs:
(http://validation.linaro.org/lava-server/scheduler/job/20742/log_file#entry112)
---------
<LAVA_DISPATCHER>2012-05-28 10:03:59 AM INFO: [ACTION-B] lava_android_test_run is started with {u'test_name': u'skia'}
/srv/lava/instances/production/bin/lava-android-test
[ 1892.250000] binder: 2409:2409 transaction failed 29189, size 56-0

root@linaro: <LAVA_DISPATCHER>2012-05-28 10:04:01 AM INFO: adb connect over default network interface
LC_ALL=C ping -W4 -c1 192.168.1.10
PING 192.168.1.10 (192.168.1.10) 56(84) bytes of data.
64 bytes from 192.168.1.10: icmp_seq=1 ttl=64 time=0.266 ms

--- 192.168.1.10 ping statistics ---
1 packets transmitted, 1 received, 0% packet loss, time 0ms
rtt min/avg/max/mdev = 0.266/0.266/0.266/0.000 ms
root@linaro: ifconfig eth0
eth0: ip 192.168.1.125 mask 255.255.255.0 flags [up broadcast running multicast]
root@linaro: <LAVA_DISPATCHER>2012-05-28 10:04:07 AM INFO: Execute adb command on host: adb disconnect 192.168.1.125
No such device 192.168.1.125:5555
<LAVA_DISPATCHER>2012-05-28 10:04:07 AM INFO: Execute adb command on host: adb connect 192.168.1.125
connected to 192.168.1.125:5555
getprop init.svc.bootanim
stopped
root@linaro: <LAVA_DISPATCHER>2012-05-28 10:04:11 AM INFO: Execute command on host: lava-android-test run skia -s 192.168.1.125:5555 -o /tmp/tmpIeIVnH/skia.2012-05-28T10:04:11Z.bundle
ERROR: Test execution error: Failed to execute command(skia_bench -repeat 6 -timers w -config 565 -match repeat):ret_code=255
LAVA: --Start Operation: Run test (skia) on device(192.168.1.125:5555)
LAVA: Begin to execute command: adb -s 192.168.1.125:5555 shell logcat -c
LAVA: Begin to execute command: adb -s 192.168.1.125:5555 shell skia_bench -repeat 6 -timers w -config 565 -match bitmap
LAVA: Begin to execute command: adb -s 192.168.1.125:5555 shell skia_bench -repeat 6 -timers w -config 565 -match rects
LAVA: Begin to execute command: adb -s 192.168.1.125:5555 shell skia_bench -repeat 6 -timers w -config 565 -match repeat
LAVA: (stdout) error: device not found
<LAVA_DISPATCHER>2012-05-29 04:13:26 AM INFO: Execute adb command on host: adb disconnect 192.168.1.125
No such device 192.168.1.125:5555
<LAVA_DISPATCHER>2012-05-29 04:13:26 AM WARNING: [ACTION-E] lava_android_test_run is finished with error (Failed to run test case(skia) on device(192.168.1.125:5555) with return value: 1).

Tags: lava-lab
Revision history for this message
Andy Doan (doanac) wrote :

I think the "device not found" might be a red herring. skia runs many more tests than what is shown there, so I suspect the ski_bench binary crashed and we don't get that information back properly from adb or something

Revision history for this message
Paul Larson (pwlars) wrote :

I'm not sure, it seems like it might be related to another bug Abhishek was showing me yesterday where many of the tests failed to run because even though the ping check succeeded, adb failed to get a connection to the board. This *could* be a network hiccup or something, or maybe even adb being flaky... Abhishek, have you ever tried doing a loop of adb attach, run some quick shell command, adb detach? Similar to what lava is doing here, to try to say whether adb is really stable? It seems like it would make a good test.

Andy Doan (doanac)
Changed in lava-lab:
milestone: none → 2012.06
Yongqin Liu (liuyq0307)
Changed in lava-lab:
assignee: nobody → Yongqin Liu (liuyq0307)
Revision history for this message
Yongqin Liu (liuyq0307) wrote :

There is a similar error in job#21395
but there seems be a manually killing job operation in the job#21395.
because there is "KILLING JOB FOR EXCEEDING TIMEOUT" displayed in the log file
"http://validation.linaro.org/lava-server/scheduler/job/21395/log_file"

And for this job #20742, there is also "KILLING JOB FOR EXCEEDING TIMEOUT" displayed.
But it was displayed during the CTS test.

And for the jobs after #20742, all the skia tests are successful if the skia test was run.
So I intend to think there were some manual operations that was not showed in the lava log.
And I suggest that if the skia test does not occur the same error in the following five jobs from now on,
we can make this bug invalid.

Changed in lava-lab:
status: New → Confirmed
importance: Undecided → Low
Revision history for this message
Michael Hudson-Doyle (mwhudson) wrote : Re: [Bug 1006745] Re: LAVA Android test "skia" failed on android build "origen-ics-gcc47-samsunglt-stable-blob/#build=40"

Yongqin Liu <email address hidden> writes:

> There is a similar error in job#21395
> but there seems be a manually killing job operation in the job#21395.
> because there is "KILLING JOB FOR EXCEEDING TIMEOUT" displayed in the log file
> "http://validation.linaro.org/lava-server/scheduler/job/21395/log_file"
>
> And for this job #20742, there is also "KILLING JOB FOR EXCEEDING
> TIMEOUT" displayed. But it was displayed during the CTS test.
>
> And for the jobs after #20742, all the skia tests are successful if
> the skia test was run. So I intend to think there were some manual
> operations that was not showed in the lava log. And I suggest that if
> the skia test does not occur the same error in the following five jobs
> from now on, we can make this bug invalid.

The KILLING JOB FOR EXCEEDING TIMEOUT messages are automatic, not
manual. But the timeout is 24 hours, so if we're hitting that,
something has gone pretty badly wrong.

Cheers,
mwh

Revision history for this message
Yongqin Liu (liuyq0307) wrote :

About one month from then. so make it incomplete.
There is also the possibility that the adb connection lost as in other bugs.
Then let it fixed in other bugs.

Here set it incomplete first.

Changed in lava-lab:
status: Confirmed → Incomplete
Fathi Boudra (fboudra)
Changed in lava-lab:
milestone: 2012.06 → none
milestone: none → 2012.07
Andy Doan (doanac)
Changed in lava-lab:
milestone: 2012.07 → 2012.08
Fathi Boudra (fboudra)
Changed in lava-lab:
status: Incomplete → Invalid
milestone: 2012.08 → none
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.
Subscribing...

Other bug subscribers

Related blueprints

Remote bug watches

Bug watches keep track of this bug in other bug trackers.