[parcel]

I have similar problem, it still remains mystery.

It's all pointing towards an internal error of some sort (in my opinion), that isn't being handled correctly (or at all).

I think in the case of the "bookmark crash" we generally saw a failed CApplication::FrameMove() call shortly before Kodi fell over in a heap. There's Leopolds post with crashlog here, but the stacktrace doesn't doesn't match yours or Tobbis (I tried to view some old crashlogs but they've been deleted so if you have new crashlogs that might be useful).

In Tobbi's case there are other CApplication errors but no CApplication::FrameMove() error, but the end result might be the same - that eventually Kodi keels over due to these unhandled exceptions etc.

By the way, I'm currently uploading new debug-enabled #1011x builds so if you're able to use debug-enabled builds (requires a FAT partition > 380MB) then the resulting crashlogs will be much more detailed.

Leopolds post is good hint.

Code:

Thread 42 (Thread 0x752ed000 (LWP 556)):
#0  0x76bd81e0 in memcmp@plt () from /usr/lib/libpython2.7.so.1.0
#1  0x76c13034 in _PyString_Eq (o1=<optimized out>, o2=<optimized out>) at /home/neil/projects/OpenELEC.tv/build.OpenELEC-RPi2.arm-6.0-devel-debug/Python-2.7.10/Objects/stringobject.c:1259
#2  0x76c3e8a8 in lookdict_string.lto_priv.1640 (mp=<optimized out>, key=0x58d9e080, hash=-1363662082) at /home/neil/projects/OpenELEC.tv/build.OpenELEC-RPi2.arm-6.0-devel-debug/Python-2.7.10/Objects/dictobject.c:435
#3  0x76c3d224 in PyDict_GetItem (op=0x6bea84b0, key=0x58d9e080) at /home/neil/projects/OpenELEC.tv/build.OpenELEC-RPi2.arm-6.0-devel-debug/Python-2.7.10/Objects/dictobject.c:743
#4  0x76c61674 in PyDict_GetItemString (v=0x6bea84b0, key=<optimized out>) at /home/neil/projects/OpenELEC.tv/build.OpenELEC-RPi2.arm-6.0-devel-debug/Python-2.7.10/Objects/dictobject.c:2425
#5  0x76c73a84 in PyImport_AddModule (name=0xd887c4 "xbmc") at /home/neil/projects/OpenELEC.tv/build.OpenELEC-RPi2.arm-6.0-devel-debug/Python-2.7.10/Python/import.c:641
#6  0x005e185c in CPythonInvoker::stop (this=0x2c59600, abort=<optimized out>) at PythonInvoker.cpp:457
#7  0x005db1c8 in CLanguageInvokerThread::stop (this=0x2c596d8, wait=<optimized out>) at LanguageInvokerThread.cpp:73
#8  0x005d9cf4 in CScriptInvocationManager::Uninitialize (this=this@entry=0x126ad5c <CScriptInvocationManager::GetInstance()::s_instance>) at ScriptInvocationManager.cpp:109
#9  0x005d9dd4 in CScriptInvocationManager::~CScriptInvocationManager (this=0x126ad5c <CScriptInvocationManager::GetInstance()::s_instance>, __in_chrg=<optimized out>) at ScriptInvocationManager.cpp:44
#10 0x754ddf70 in __run_exit_handlers (status=-1, listp=0x755e34bc <__exit_funcs>, run_list_atexit=run_list_atexit@entry=true) at exit.c:82
#11 0x754ddfc4 in __GI_exit (status=<optimized out>) at exit.c:104
#12 0x754c6d00 in __libc_start_main (main=0x7e9d3664, argc=1969106944, argv=0x754c6d00 <__libc_start_main+276>, init=<optimized out>, fini=0xd7de40 <__libc_csu_fini>, rtld_fini=0x76f41390 <_dl_fini>, stack_end=0x7e9d3664) at libc-start.c:323
#13 0x003a4cb4 in _start () at ../sysdeps/arm/start.S:124
Backtrace stopped: previous frame identical to this frame (corrupt stack?)

This is same part but looks different.

Code:

int
1254 _PyString_Eq(PyObject *o1, PyObject *o2)
1255 {
1256     PyStringObject *a = (PyStringObject*) o1;
1257     PyStringObject *b = (PyStringObject*) o2;
1258     return Py_SIZE(a) == Py_SIZE(b)
1259       && memcmp(a->ob_sval, b->ob_sval, Py_SIZE(a)) == 0;
1260 }

I guess it raise exception at memcmp(line 1259). It's maybe python problem.