822514 - Possible android perma-orange w/ crash @ libdvm.so + 0x48022

Status: RESOLVED WORKSFORME

(Core :: General, defect)

Description

[Daniel Holbert [:dholbert]]

Looks like we've got an android crash affecting all unit tests -- not sure if it's a code issue or a releng issue. Closing the tree, filing the 

The affected push is:
  https://tbpl.mozilla.org/?tree=Mozilla-Inbound&rev=53ab929e9386
and every single "Android 2.2 Armv6 opt" test run on that job is orange w/ this crash.

(That's the latest push w/ android results at the moment)

Comment 1

[Daniel Holbert [:dholbert]]

(In reply to Daniel Holbert [:dholbert] from comment #0)
> Closing the tree, filing the 

*Closing the tree, filing this as a Releng bug for the moment.

Comment 2

[Daniel Holbert [:dholbert]]

The crash looks like:
{
rgs: ['../hostutils/bin/xpcshell', '-g', '/builds/tegra-163/test/build/hostutils/xre', '-v', '170', '-f', '/builds/tegra-163/test/build/tests/reftest/reftest/components/httpd.js', '-e', "const _PROFILE_PATH = '/tmp/tmp9aGIO6';const _SERVER_PORT = '30163'; const _SERVER_ADDR ='10.250.48.218';", '-f', '/builds/tegra-163/test/build/tests/reftest/server.js']
INFO | remotereftests.py | Server pid: 46574
{'uptime': ['0 days 0 hours 4 minutes 38 seconds 299 ms'], 'power': ['Power status:', ' AC power ONLINE', ' Battery charge NO BATTERY', ' Remaining charge: 0%', ' Battery Temperature: 0.0 (c)'], 'process': [['10031', '1510', 'com.mozilla.SUTAgentAndroid'], ['10018', '1233', 'com.android.launcher'], ['10013', '1481', 'com.cooliris.media'], ['10009', '1464', 'com.android.quicksearchbox'], ['10002', '1473', 'com.android.music'], ['10007', '1220', 'com.android.inputmethod.latin'], ['1000', '1020', 'system'], ['1001', '1228', 'com.android.phone'], ['10004', '1332', 'android.process.media'], ['10032', '1442', 'com.mozilla.watcher'], ['10006', '1422', 'com.android.mms'], ['10010', '1407', 'com.android.providers.calendar'], ['10014', '1391', 'com.android.email'], ['10017', '1367', 'com.android.bluetooth'], ['10029', '1348', 'com.android.deskclock'], ['10015', '1271', 'android.process.acore'], ['1000', '1242', 'com.android.settings']], 'screen': ['X:1600 Y:1200'], 'memory': ['PA:839487488, FREE: 760778752'], 'systime': ['2012/12/17 04:46:44:789'], 'rotation': ['ROTATION:0'], 'disk': [], 'os': ['harmony-eng 2.2 FRF91 20110202.102810 test-keys'], 'id': ['00:26:e8:d4:a6:82'], 'uptimemillis': ['278321']}
pushing directory: /tmp/tmpxsY2nk to /mnt/sdcard/tests/reftest/profile
pushing directory: /tmp/tmpxsY2nk to /mnt/sdcard/tests/reftest/profile
REFTEST INFO | runreftest.py | Running tests: start.

FIRE PROC: '"MOZ_CRASHREPORTER=1,XPCOM_DEBUG_BREAK=stack,MOZ_CRASHREPORTER_NO_REPORT=1,NO_EM_RESTART=1,MOZ_PROCESS_LOG=/tmp/tmp9MEho3pidlog,XPCOM_MEM_BLOAT_LOG=/tmp/tmpxsY2nk/runreftest_leaks.log" org.mozilla.fennec -no-remote -profile /mnt/sdcard/tests/reftest/profile/'
INFO | automation.py | Application pid: 1613

INFO | automation.py | Application ran for: 0:00:12.444083
INFO | automation.py | Reading PID log: /tmp/tmp9MEho3pidlog
getting files in '/mnt/sdcard/tests/reftest/profile/minidumps/'
Downloading symbols from: http://ftp-scl3.mozilla.com/pub/mozilla.org/mobile/tinderbox-builds/mozilla-inbound-android-armv6/1355788179/fennec-20.0a1.en-US.android-arm-armv6.crashreporter-symbols.zip
PROCESS-CRASH | remoteautomation.py | application crashed [@ libdvm.so + 0x48022]
Crash dump filename: /tmp/tmp2y_QOS/6ce16244-3d1d-dc1c-09fe465a-6ec0da9d.dmp
Operating system: Android
                  0.0.0 Linux 2.6.32.9-00002-gd8084dc-dirty #1 SMP PREEMPT Wed Feb 2 11:32:06 PST 2011 armv7l nvidia/harmony/harmony/harmony:2.2/FRF91/20110202.102810:eng/test-keys
CPU: arm
     0 CPUs

Crash reason:  SIGSEGV
Crash address: 0x0

Thread 21 (crashed)
 0  libdvm.so + 0x48022
     r4 = 0x0032b808    r5 = 0x00000000    r6 = 0x433126c4    r7 = 0x562ffae4
     r8 = 0x562ffcbc    r9 = 0x00000000   r10 = 0x00000001    fp = 0x562ffc80
     sp = 0x562ffa78    lr = 0x8084e4bf    pc = 0x80848022
    Found by: given as instruction pointer in context
 1  libdvm.so + 0x35b3b
     sp = 0x562ffa80    pc = 0x80835b3d
    Found by: stack scanning
 2  libdvm.so + 0x85c02
     sp = 0x562ffa90    pc = 0x80885c04
    Found by: stack scanning
 3  dalvik-LinearAlloc (deleted) + 0x2466c2
     sp = 0x562ffa98    pc = 0x433126c4
    Found by: stack scanning
 4  libdvm.so + 0x3bb73
     sp = 0x562ffaa0    pc = 0x8083bb75
    Found by: stack scanning
 5  libdvm.so + 0x85c02
     sp = 0x562ffaa4    pc = 0x80885c04
    Found by: stack scanning
 6  libxul.so!_JNIEnv::CallVoidMethod(_jobject*, _jmethodID*, ...) [jni.h : 631 + 0xe]
     sp = 0x562ffad0    pc = 0x53aaf7dc
    Found by: stack scanning
 7  libxul.so!AndroidGLController::SetGLVersion(int) [AndroidLayerViewWrapper.cpp : 54 + 0x16]
     sp = 0x562ffae8    pc = 0x53e243a8
    Found by: call frame info
 8  libxul.so!mozilla::AndroidBridge::RegisterCompositor(_JNIEnv*, bool) [AndroidBridge.cpp : 1201 + 0xa]
     r4 = 0x0023df70    r5 = 0x54c84cec    r6 = 0x00000000    sp = 0x562ffb08
     pc = 0x53e21f60
    Found by: call frame info
 9  libxul.so!mozilla::gl::GLContextProviderEGL::CreateForWindow(nsIWidget*) [GLContextProviderEGL.cpp : 2193 + 0x16]
     r4 = 0x54c84cd4    r5 = 0x00000010    r6 = 0x4e46ed80    r7 = 0x562ffcb8
     sp = 0x562ffb30    pc = 0x541d3e40
    Found by: call frame info
10  libxul.so!mozilla::layers::LayerManagerOGL::CreateContext() [LayerManagerOGL.cpp : 466 + 0x6]
     r4 = 0x560f7600    r5 = 0x00000000    r6 = 0x562ffcc0    r7 = 0x562ffcb8
     sp = 0x562ffbc8    pc = 0x541a80d0
    Found by: call frame info
11  libxul.so!mozilla::layers::LayerManagerOGL::Initialize(bool) [LayerManagerOGL.cpp : 59 + 0x2]
}
https://tbpl.mozilla.org/php/getParsedLog.php?id=18032958&tree=Mozilla-Inbound

Comment 3

[Daniel Holbert [:dholbert]]

This was broken on "Android 2.2 opt", too, and the next push mysteriously went green on that platform (no results yet for Android 2.2 Armv6 opt on that next push, but I assume they'll be green, too).

So: not sure what happened, but this seems to have fixed itself.

Comment 4

[Daniel Holbert [:dholbert]]

By pushlog-blame, this seems to have been caused by bug 820167's push,
 https://hg.mozilla.org/integration/mozilla-inbound/rev/e14c7409ede7

(That was the push before the fully-orange-on-android TBPL cycle that I linked in comment 0, and it incidentally didn't get any android builds, but it was backed out for other reasons *after* the TBPL cycle from comment 0, and that's the one that went green.)

Moving to Core|General and tentatively marking dependency.