lava-android-test install failed for no adb connection

[Yongqin Liu]

lava-android-test failed at install tests.
<LAVA_DISPATCHER>2012-01-13 03:20:12 AM INFO: Execute command on host: lava-android-test install 0xbench -s 192.168.1.159:5555
ERROR: Test installation error: Failed to create directory(/data/lava-android-test/share/installed-tests/0xbench) for test(0xbench)
LAVA: --Start Operation: Install test (0xbench) on device(192.168.1.159:5555)
<LAVA_DISPATCHER>2012-01-13 03:20:12 AM INFO: Execute adb command on host: adb disconnect 192.168.1.159
No such device 192.168.1.159:5555
<LAVA_DISPATCHER>2012-01-13 03:20:13 AM INFO: Action lava_android_test_install finished.

the information "No such device 192.168.1.159:5555" means that before execute "adb disconnect", the adb connection has been disconnected.

this is because the the android image becomes suspended and cause the network can be pinged:(

http://validation.linaro.org/lava-server/scheduler/job/8583

[Yongqin Liu (liuyq0307)]

lp:~liuyq0307/lava-dispatcher/disable-suspend-new branch can decrease the rate of this problem,
but it seem not works 100%,

 IMO, we need to check it more deeply.

and below is the output information from conmux-console on my local machine about the suspend and ip disconnect
---------------------------------------------------------------------------

[ 4707.146118] Freezing remaining freezable tasks ... (elapsed 0.01 seconds) done.
[ 4707.519592] PM: suspend of devices complete after 296.844 msecs
[ 4707.565307] PM: late suspend of devices complete after 16.571 msecs
[ 4707.594299] Disabling non-boot CPUs ...
[ 4707.622161] CPU1: shutdown
[ 4813.349182] Suspended for 105.700 seconds
[ 4813.375793] wakeup wake lock: alarm
[ 4813.401855] Enabling non-boot CPUs ...
[ 4813.470458] CPU1: Booted secondary processor
[ 4813.470458] CPU1: Unknown IPI message 0x1
[ 4813.470550] Calibrating delay loop (skipped) already calibrated this CPU
[ 4813.484558] Switched to NOHz mode on CPU #1
[ 4813.581451] thermal_cooling_dev_register:Found cpu cpu
[ 4813.581451] thermal_init_thermal_state:Not all components registered for cpu domain sensor (null), gov=ef9089a0, cooling=c08bcfcc
[ 4813.581451] CPU1 is up
[ 4813.964141] ------------[ cut here ]------------
[ 4813.975830] WARNING: at /mnt/jenkins/workspace/linaro-android_staging-panda/build/kernel/drivers/video/omap2/dss/dss.c:716 dss_runtime_get+0x34/0x40()
[ 4813.975891] Modules linked in: wl12xx wl1273_core gator btwilink rfcomm hci_uart bnep bluetooth st_drv
[ 4814.014007] [<c0019874>] (unwind_backtrace+0x0/0xf8) from [<c004b850>] (warn_slowpath_common+0x4c/0x64)
[ 4814.031036] [<c004b850>] (warn_slowpath_common+0x4c/0x64) from [<c004b884>] (warn_slowpath_null+0x1c/0x24)
[ 4814.031036] [<c004b884>] (warn_slowpath_null+0x1c/0x24) from [<c031759c>] (dss_runtime_get+0x34/0x40)
[ 4814.065185] [<c031759c>] (dss_runtime_get+0x34/0x40) from [<c0317e24>] (dispc_runtime_resume+0x8/0x1340)
[ 4814.082427] [<c0317e24>] (dispc_runtime_resume+0x8/0x1340) from [<c038c2e8>] (pm_generic_runtime_resume+0x2c/0x40)
[ 4814.100555] [<c038c2e8>] (pm_generic_runtime_resume+0x2c/0x40) from [<c00353e0>] (_od_resume_noirq+0x40/0x4c)
[ 4814.118286] [<c00353e0>] (_od_resume_noirq+0x40/0x4c) from [<c038d054>] (pm_noirq_op+0x98/0x160)
[ 4814.134857] [<c038d054>] (pm_noirq_op+0x98/0x160) from [<c038d164>] (dpm_resume_noirq+0x48/0x144)
[ 4814.151519] [<c038d164>] (dpm_resume_noirq+0x48/0x144) from [<c00844f8>] (suspend_devices_and_enter+0x114/0x354)
[ 4814.169616] [<c00844f8>] (suspend_devices_and_enter+0x114/0x354) from [<c0084824>] (enter_state+0xec/0x128)
[ 4814.187255] [<c0084824>] (enter_state+0xec/0x128) from [<c00857c4>] (suspend+0x5c/0x194)
[ 4814.187255] [<c00857c4>] (suspend+0x5c/0x194) from [<c0063ec0>] (process_one_work+0x12c/0x48c)
[ 4814.219512] [<c0063ec0>] (process_one_work+0x12c/0x48c) from [<c0064598>] (worker_thread+0x17c/0x37c)
[ 4814.236511] [<c0064598>] (worker_thread+0x17c/0x37c) from [<c00689c0>] (kthread+0x80/0x88)
[ 4814.236511] [<c00689c0>] (kthread+0x80/0x88) from [<c0013e50>] (kernel_thread_exit+0x0/0x8)
[ 4814.236511] ---[ end trace 46af755aa476bada ]---
[ 4814.280609] ------------[ cut here ]------------
...

[Yongqin Liu (liuyq0307)]

another possibility is that the there is an expired adb connection that remained by last job or last action,
and this cause the creation of new adb connection failed.

so it's better to clear the expired adb connection before create a new one.
this has been added in branch lp:~liuyq0307/lava-dispatcher/mount-data-partition.
and we can see after this modification whether this problem still occurs.