[bugreport] crash when adding audio track to video file with jit.qt.movie editing commands


    Feb 19 2008 | 3:29 pm
    Hi all,
    Description:
    I'm experiencing an occasional (but reproducable) crash when I copy or insert a track of a movie from disk into an existing movie with the editing commands. The patch below demonstrates this.
    Expected behaviour:
    The patch should continue adding A and V together for ever..
    Steps to reproduce:
    Are in the patch. Example files are here: http://www.smadsteck.nl/Sources.zip (1 MB)
    Max 4.6.3, Jitter 1.6.3, Mac OS 10.4.11, Mac Pro Quad 2.66
    If someone could confirm I'd be very thankful.
    Mattijs
    Patch:
    Crash report:
    First I got the following:
    Thread 0 Crashed:
    0 com.apple.QuickTime 0x9445aad3 InternalDisposeTrack_priv + 24
    1 com.apple.QuickTime 0x9445024f DisposeMovieEditState_priv + 124
    2 com.apple.QuickTime 0x944501cd DisposeMovieEditState + 22
    3 com.cycling74.JitterAPI 0x262def0a jit_qt_movie_inserttrack + 70
    4 com.cycling74.JitterAPI 0x262e1121 jit_qt_movie_insert + 830
    But now I get this in the console:
    Invalid memory access of location 0000008f eip=9445aad3
    Feb 19 16:13:20 SmadS-Xeon-2 /Applications/MaxMSP 4.6/MaxMSP copy.app/Contents/MacOS/MaxMSP: An unexpected Java error has been detected by HotSpot Virtual Machine.
    Feb 19 16:13:20 SmadS-Xeon-2 /Applications/MaxMSP 4.6/MaxMSP copy.app/Contents/MacOS/MaxMSP: If this error is reproducible, please report it with the following information:
    Feb 19 16:13:20 SmadS-Xeon-2 /Applications/MaxMSP 4.6/MaxMSP copy.app/Contents/MacOS/MaxMSP: 1. Provide the steps to reproduce, a test case, and any relevant information
    Feb 19 16:13:20 SmadS-Xeon-2 /Applications/MaxMSP 4.6/MaxMSP copy.app/Contents/MacOS/MaxMSP: 2. The corresponding JavaNativeCrash_pid.crash.log (Java state)
    Feb 19 16:13:20 SmadS-Xeon-2 /Applications/MaxMSP 4.6/MaxMSP copy.app/Contents/MacOS/MaxMSP: 3. The corresponding .crash.log (native state; generated by CrashReporter)
    Feb 19 16:13:20 SmadS-Xeon-2 /Applications/MaxMSP 4.6/MaxMSP copy.app/Contents/MacOS/MaxMSP: 4. This data:
    Feb 19 16:13:20 SmadS-Xeon-2 /Applications/MaxMSP 4.6/MaxMSP copy.app/Contents/MacOS/MaxMSP: Java VM: Java HotSpot(TM) Client VM (1.5.0_13-121 mixed mode)
    Feb 19 16:13:20 SmadS-Xeon-2 /Applications/MaxMSP 4.6/MaxMSP copy.app/Contents/MacOS/MaxMSP: Bus Error (0xa) at pc=0x9445aad3
    Feb 19 16:13:20 SmadS-Xeon-2 /Applications/MaxMSP 4.6/MaxMSP copy.app/Contents/MacOS/MaxMSP: Process ID: 534, Current Thread: 34562048
    Feb 19 16:13:20 SmadS-Xeon-2 /Applications/MaxMSP 4.6/MaxMSP copy.app/Contents/MacOS/MaxMSP: File report at: http://bugreport.apple.com/
    Feb 19 16:13:20 SmadS-Xeon-2 /Applications/MaxMSP 4.6/MaxMSP copy.app/Contents/MacOS/MaxMSP: An error report file has been written to:
    Feb 19 16:13:20 SmadS-Xeon-2 /Applications/MaxMSP 4.6/MaxMSP copy.app/Contents/MacOS/MaxMSP: /Users/mattijskneppers/Library/Logs/Java/JavaNativeCrash_pid534.crash.log
    Feb 19 16:13:21 SmadS-Xeon-2 crashdump[537]: MaxMSP crashed
    Feb 19 16:13:22 SmadS-Xeon-2 crashdump[537]: crash report written to: /Users/mattijskneppers/Library/Logs/CrashReporter/MaxMSP.crash.log
    And the log starts with:
    Thread 0 Crashed:
    0 libstdc++.6.dylib 0x90b0d2d9 std::_Rb_tree_rebalance_for_erase(std::_Rb_tree_node_base*, std::_Rb_tree_node_base&) + 401
    1 com.apple.QuickTime 0x94630cb9 std::_Rb_tree, std::_Select1st >, AudioContextCompareProxy, std::allocator > >::erase(std::_Rb_tree_iterator >, std::_Rb_tree_iterator >) + 61
    2 com.apple.QuickTime 0x94630d97 std::_Rb_tree, std::_Select1st >, AudioContextCompareProxy, std::allocator > >::erase(AudioContextProxy* const&) + 103
    3 com.apple.QuickTime 0x9445a953 AudioContextManager::UnregisterContext(void*) + 381
    4 ...ickTimeComponents.component 0x99cf1db5 AudioContextChannel::Release(long*) + 89
    5 ...ickTimeComponents.component 0x99ced5c2 AudioContextBase::ComponentEntryDispatch(ComponentParameters*, AudioContextBase*) + 134
    6 ...ple.CoreServices.CarbonCore 0x90cd0734 CallComponentDispatch + 34
    7 com.apple.QuickTime 0x943864df AudioContextRelease + 43
    8 com.apple.QuickTime 0x943869f0 myDeallocate + 27
    9 com.apple.CoreFoundation 0x9080e127 _CFRelease + 253
    10 com.apple.QuickTime 0x94386927 QTAudioContextRelease + 43
    11 com.apple.QuickTime 0x9438705d PrivateSetMovieAudioContext_priv + 1476
    12 com.apple.QuickTime 0x944503db DisposeMovie_priv + 343
    13 com.cycling74.JitterAPI 0x262d6a38 jit_qt_movie_dispose + 411
    14 com.cycling74.JitterAPI 0x262e8dfd jit_qt_movie_method_typed + 90
    15 com.cycling74.JitterAPI 0x262e8e61 jit_qt_movie_proxy_method + 93
    16 com.cycling74.JitterAPI 0x262a726f max_jit_usurp_qfn + 192
    17 com.cycling74.JitterAPI 0x262aa3d4 max_jit_usurp_method + 156
    18 com.cycling74.MaxMSP46 0x00030755 typedmess_fun + 1487 (message.c:631)
    19 com.cycling74.MaxMSP46 0x000e8585 outlet_anything + 355 (inletoutlet.c:967)

    • Feb 19 2008 | 3:51 pm
    • Feb 19 2008 | 8:00 pm
      Just a reminder:
      Do not EVER crosspost messages to both the forums and Cycling '74
      support, especially when they were already logged and confirmed several
      months ago.
      Thanks,
      Andrew B.
    • Feb 19 2008 | 9:37 pm
      Ah, I thought I'd share this problem with other users first to make sure it was a valid bug before bothering Cycling '74. I'm still not certain as to the preferred method of bug reporting. I'd say if other users can confirm erroneous behavior or even add some information to the report, that's only a good thing, no?
      But as it seems I personally reported this bug a few months before. Heh, not a smart move, admitted. But I honestly forgot. I'll be more careful next time.
      Mattijs
    • Feb 20 2008 | 11:30 am
      Just in case I forget and report this bug again, here is a little reminder for myself: by removing the dispose message and creating a new track instead of reading the first one, the crashes appear much less frequent (only once in 100 times). I hope I'll search the archives..
      Quote: Mattijs wrote on Tue, 19 February 2008 22:37
      ----------------------------------------------------
      > Ah, I thought I'd share this problem with other users first to make sure it was a valid bug before bothering Cycling '74. I'm still not certain as to the preferred method of bug reporting. I'd say if other users can confirm erroneous behavior or even add some information to the report, that's only a good thing, no?
      >
      > But as it seems I personally reported this bug a few months before. Heh, not a smart move, admitted. But I honestly forgot. I'll be more careful next time.
      >
      > Mattijs
      ----------------------------------------------------