1207461 - No usable stack from Robocop crashes

Status: RESOLVED FIXED

(Firefox for Android Graveyard :: General, defect)

Description

[Nigel Babu [:nigelb]]

22:04:13     INFO -  Loading JavaScript test from http://mochi.test:8888/tests/robocop/robocop_javascript.html?slug=1442984636834&path=testBrowserDiscovery.js
22:04:24     INFO -  INFO | automation.py | Application ran for: 0:01:24.110011
22:04:24     INFO -  INFO | zombiecheck | Reading PID log: /tmp/tmpfsQapFpidlog
22:04:25     INFO -  /data/tombstones does not exist; tombstone check skipped
22:04:27     INFO -  mozcrash Saved minidump as /builds/slave/test/build/blobber_upload_dir/3d7115ff-b66e-e9c3-5f5d1368-26d9af17.dmp
22:04:27     INFO -  mozcrash Saved app info as /builds/slave/test/build/blobber_upload_dir/3d7115ff-b66e-e9c3-5f5d1368-26d9af17.extra
22:04:27  WARNING -  PROCESS-CRASH | testBrowserDiscovery | application crashed [None]
22:04:27     INFO -  Crash dump filename: /tmp/tmp64II9i/3d7115ff-b66e-e9c3-5f5d1368-26d9af17.dmp
22:04:27     INFO -  No symbols path given, can't process dump.
22:04:27     INFO -  0 ERROR runApp() exited with code 1
22:04:28  WARNING -  PROCESS-CRASH | Automation Error: Missing end of test marker (process crashed?)
22:04:29     INFO -  09-22 22:03:05.583 I/dalvikvm( 1031): Could not find method android.app.ActivityManager.getLauncherLargeIconSize, referenced from method org.mozilla.gecko.GeckoAppShell.getPreferredIconSize

Comment 1

[Nigel Babu [:nigelb]]

Geoff, do you know what's going on here? I see this across trees. Did we land something recently that broke this?

Comment 2

[Nigel Babu [:nigelb]]

Instances of this happening

https://treeherder.mozilla.org/logviewer.html#?job_id=14555541&repo=mozilla-inbound
https://treeherder.mozilla.org/logviewer.html#?job_id=14536866&repo=mozilla-inbound
https://treeherder.mozilla.org/logviewer.html#?job_id=14545317&repo=mozilla-inbound
https://treeherder.mozilla.org/logviewer.html#?job_id=14552174&repo=mozilla-inbound

Comment 3

[Geoff Brown [:gbrown]]

I don't have insight into the cause of the crashes. I don't see them happening recently -- I think the cause of the crashes has been resolved now.

However, I am concerned about the lack of crash report and particularly the "No symbols path" message.

https://treeherder.mozilla.org/#/jobs?repo=try&revision=d40f567efbf0 introduces an intentional crash. Note that mochitest and reftest crashes are reported correctly with a crash report -- but not on robocop, where the "No symbols path" error is seen consistently.

Comment 4

[Geoff Brown [:gbrown]]

I think I broke robocop crash reporting when I introduced runrobocop.py.

Comment 5

[Geoff Brown [:gbrown]]

Attachment: pass symbolsPath to automation.runApp

options.symbolsPath was not being passed to automation's runApp() (which eventually calls down to mozcrash), from runrobocop.py only. This passes the symbols path correctly.

Fix demonstrated at https://treeherder.mozilla.org/#/jobs?repo=try&revision=40c42ce8891c:

12:32:14  WARNING -  PROCESS-CRASH | testGeckoProfile | application crashed [@ nsWindow::OnGlobalAndroidEvent]
12:32:14     INFO -  Crash dump filename: /tmp/tmpz8M1Bj/69e2e639-b7d1-6b0e-246ddf32-42848959.dmp
12:32:14     INFO -  Operating system: Android
12:32:14     INFO -                    0.0.0 Linux 2.6.29-g4bb8fa0 #1 Fri Feb 20 17:28:01 MST 2015 armv7l generic/sdk/generic:2.3.7/GINGERBREAD/eng.ubuntu.20140123.014351:eng/test-keys
12:32:14     INFO -  CPU: arm
12:32:14     INFO -       ARMv0
12:32:14     INFO -       0 CPUs
12:32:14     INFO -  Crash reason:  SIGSEGV
12:32:14     INFO -  Crash address: 0x0
12:32:14     INFO -  Process uptime: not available
12:32:14     INFO -  Thread 11 (crashed)
12:32:14     INFO -   0  libxul.so!nsWindow::OnGlobalAndroidEvent [nsWindow.cpp:40c42ce8891c : 891 + 0x2]
12:32:14     INFO -       r0 = 0x59282000    r1 = 0x00000081    r2 = 0x00000000    r3 = 0x00000000
12:32:14     INFO -       r4 = 0x5df47700    r5 = 0x5f815fc0    r6 = 0x59298a3c    r7 = 0x4b744794
12:32:14     INFO -       r8 = 0x4b744790    r9 = 0x00000000   r10 = 0x00000001   r12 = 0xfffff6f8
12:32:14     INFO -       fp = 0x00000045    sp = 0x4b7446a0    lr = 0x536654e7    pc = 0x536654ea
12:32:14     INFO -      Found by: given as instruction pointer in context
12:32:14     INFO -   1  libxul.so!nsAppShell::LegacyGeckoEvent::Run [nsAppShell.cpp:40c42ce8891c : 789 + 0x5]
12:32:14     INFO -       r3 = 0x0000001c    r4 = 0x5df47700    r5 = 0x5f815fc0    r6 = 0x59298a3c
12:32:14     INFO -       r7 = 0x4b744794    r8 = 0x4b744790    r9 = 0x00000000   r10 = 0x00000001
12:32:14     INFO -       fp = 0x00000045    sp = 0x4b7446a8    pc = 0x53662fe3
12:32:14     INFO -      Found by: call frame info
12:32:14     INFO -   2  libxul.so!nsAppShell::ProcessNextNativeEvent [nsAppShell.cpp:40c42ce8891c : 352 + 0x3]
12:32:14     INFO -       r4 = 0x00000001    r5 = 0x4b74479c    r6 = 0x59298a3c    r7 = 0x4b744794
12:32:14     INFO -       r8 = 0x4b744790    r9 = 0x00000000   r10 = 0x00000001    fp = 0x00000045
12:32:14     INFO -       sp = 0x4b744790    pc = 0x5365cfbb
12:32:14     INFO -      Found by: call frame info
...

Comment 6

[Joel Maher ( :jmaher ) (UTC -8) (PTO back normal Nov 17)]

Comment on attachment 8669906 [details] [diff] [review]
pass symbolsPath to automation.runApp

Review of attachment 8669906 [details] [diff] [review]:
-----------------------------------------------------------------

this looks nice and neat.

Comment 7

[Pulsebot]

https://hg.mozilla.org/integration/mozilla-inbound/rev/f679b59a7b53

Comment 8

[Carsten Book [:Tomcat]]

https://hg.mozilla.org/mozilla-central/rev/f679b59a7b53