JVM Error Codes

101 Previous startup failed The device has been reset during the JVM boot process. This error indicates that the JVM found the "boot in progress" flag was set on startup. The screen is intended to break continuous reset loops so that corrective action can be taken.
102 Invalid code in filesystem The system has checked the .COD files in the handheld for modification, and determined that there is a problem with one or more .COD files. If all loads fail, it could indicate a build process error where there was a problem signing the codfiles. If some user action to the device caused this problem, the reset cycle will be continuous since the code in the filesystem has been corrupted. The only recovery method is to wipe the device and restore a new system.
103 Cannot find starting address The starting address for the boot .COD file cannot be found. This may indicate that a boot .COD file has not been installed on the device, or that its format is invalid or corrupt.
104 Uncaught: <Java-type-name> An uncaught Java exception was thrown by the Java code and diagnosed by the JVM. Execution can be continued, or the handheld can be attached to a debugger on a desktop through a serial or USB cable. The event log should contain the traceback of the thrown exception.
105 Example, DbRecSize( %d ) -> %d The file system API has returned an error status for a certain operation. This could indicate a corrupt filesystem or an error in the JVM.

106 Graphics system error An error has been detected in the graphics system of the device.
107 operator new() called A C++ class in the JVM has not been coded correctly to inherit from VMRamObject that has the correct override for operator new. Extract the current (post-reset) BUGDISP
108 operator delete() called A C++ class in the JVM has not been coded correctly to inherit from VMRamObject that has the correct override for operator delete. Extract the current (post-reset) BUGDISP.
109 PriorityMessageCount error: <priority-count> The value returned by RimPriorityMessageCount is negative when it should always be greater than or equal to zero. This indicates an error in the OS code. Extract the current (post-reset) BUGDISP and EVENTLOG.
110 Non-idle event downtime error: <down-time> <idle-down-time> A problem has been detected in the accumulation of JVM down time, which represents how long the JVM has been idle. This usually indicates an error in the handheld firmware or the JVM. This could also occur if the tick count rolls over after 400+ days of device time.
111 Font engine error An error has been detected in the font engine system of the device. Extract the current (post-reset) BUGDISP and EVENTLOG.
112 Java Native Assertion Failure An error has been detected in the Java native code. Extract the current (post-reset) BUGDISP and EVENTLOG.
200 Application manager threw an uncaught exception The application manager event thread threw an uncaught exception and so cannot continue execution.
201 Crypto initialization code failed The initialization of the crypto system failed and the handheld cannot continue execution.
202 An attack on the key store has been detected The attack has been detected and execution cannot continue.
203 Console process died The application manager console process (usually the Ribbon) has died. This is likely due to an uncaught exception during execution.
204 Persistent Content Exception An application tried to commit a plaintext object to the Persistent Store. This will only happen if Content Protection is on and a process tries to save something in the PersistentStore that is marked as plaintext. Since this exception was not handled, the persistent store is in a bad state. We need to reset to roll back to the last good commit point. NOTE: This JVM error reflects an error in the Java code, the JVM is simply diagnosing the error. This is not a JVM problem, the eventlog will contain information about the erroneous Java code.
501 VM_THREAD_SWITCHED: Internal Error This is an error return used internally in the VM. It should never be reported as a device error.
502 VM_PROCESS_DEATH: All processes exited The last Java process has terminated. There is nothing left to execute.
503 VM_THREAD_DEATH: Internal Error This is an error return used internally in the VM. It should never be reported as a device error.
504 VM_THREAD_SWITCH: Internal Error This is an error return used internally in the VM. It should never be reported as a device error.
505 VM_BAD_CODE: Bad Byte Code An error has occurred in the JIT compiler.
506 Uncaught Exception An uncaught Java exception was thrown in the initial VM Java thread thus ending the only live thread in the system. The eventlog contains the traceback for the exception.
507 Unsatisfied Link A dependency on a .COD file could not be satisfied because the .COD file is missing.
508 Invalid object A problem has been detected with a debugger command to the VM.
509 VM_PPO_INFINITE_LOOP: infinite loop in PPO phase of GC The maximum iteration count for the PPO phase of a GC must be the maximum number of file handles in the system. This error indicates that the iteration count has exceeded that count and hence there is a flaw in the PPO loop or a corrupted file system. The extra hex integer in the error string is the flash id of the current record where the infinite loop was detected.
510 Deadlock All threads are waiting on objects, resulting in deadlock. The system cannot recover from this state because no thread can release a lock.
511 Debug connection died A problem has occurred while debugging that may be caused by a VM problem or an incorrect debugging command being sent to the VM.
512 GC Aborted An idle garbage collection has been interrupted by a user event, such as pressing the keyboard or moving the trackwheel.
513 <clinit> needs running An opcode requires that a class <clinit> execute before it can continue execution.
514 <init> needs running A new instance of a class has been allocated and it must be initialized by the default constructor before it can be used.
515 Object group too big The reachable objects form a group that cannot be represented properly by the JVM because either there are too many objects or the total size of the objects is too large.
516 Persistent ids exhausted When committing a persistent object, the JVM found that the persistent store id counter has reached its limit. The object was not committed and a critical error was reported. This error should never happen unless a device is heavily used for years.
517 Filesystem corrupt An inconsistency has been detected in the JVM persistent object store.
518 Unexpected longjmp A garbage collection marking phase was terminated via a longjmp. This indicates that the marking phase was interrupted when it should have completed without interruption. This should never happen because these actions are executed when the device is not idle, and GCs can only be interrupted when the device is idle.
519 Internal Error The JVM host is missing or has been disabled.
520 Internal Return An internal state that indicates a Java method return needs to be executed.
521 Dangerous Wait An Object.wait() has been executed by a thread that holds a lock on another object. This is only checked in the simulator under the control of the JvmDebugWaits application switch.
522 Interlaced synchronization A thread has acquired two locks on objects in an order that doesn�t match the order that previous locks for the two types were acquired. This indicates a future potential deadlock situation and is reported. The check is only available in the simulator under the control of the JvmDebugLocks application switch.
523 System process died A critical Java process has terminated, and the device cannot continue to operate in a normal manner.
524 LMM error An object has been marked as recovered by the Low Memory Manager, but it was not freed during a garbage collection. This is only checked in the simulator under the control of the JvmDebugLMM application switch.
525 Bad persistent object An auto-commit operation during a garbage collection has detected a non-persistent object reachable from the persistent store root. The type of the object has been output into the eventlog.
526 java.lang.Object not found The class definition for java.lang.Object cannot be found.
527 java.lang.String not found The class definition for java.lang.String cannot be found.
528 Corrupt filesystem. Unrecoverable. All data will be lost All data will be lost when execution is continued. The error message screen contains a number representing an internal "reason" for the corruption. This error is not diagnosed if a .COD file has been removed, because the JVM must delete objects that were defined in the removed .COD file; therefore, this error is not expected in normal operation of the device.
Reason codes:
1. root array reference is not a valid array reference
2. root array type is not Object[]
3. root array size < 1 (i.e., Object[0])
4. contents of root[0] is not a valid ref
5. type of root[0] is not a LongIntHashtable
6. persistent segmented array header contains an invalid reference
7. an entry in a persistent Object[] contains an invalid reference
8. an Object�s type refers to an unknown codfile
9. an Object�s type description in the codfile doesn�t match the size in the store
10. a reference type field in an Object has an invalid reference in it
11. a reference type field in an object points to an object of the wrong type
12. a persistent Object[] is missing its descriptor
13. object in persistent store is not marked as persistable
14. root array is segmented and one of the segments is invalid
529 Corrupt filesystem. About to attempt recovery. Some data may be lost
Some data will be lost when execution is continued. The error message screen contains a number representing an internal "reason" for the corruption. This error is not diagnosed if a .COD file has been removed, because the VM must delete objects that were defined in the removed .COD file; therefore, this error is not expected in normal operation of the device.
Reason codes:
1. root array reference is not a valid array reference
2. root array type is not Object[]
3. root array size < 1 (i.e., Object[0])
4. contents of root[0] is not a valid ref
5. type of root[0] is not a LongIntHashtable
6. persistent segmented array header contains an invalid reference
7. an entry in a persistent Object[] contains an invalid reference
8. an Object�s type refers to an unknown codfile
9. an Object�s type description in the codfile doesn�t match the size in the store
10. a reference type field in an Object has an invalid reference in it
11. a reference type field in an object points to an object of the wrong type
12. a persistent Object[] is missing its descriptor
13. object in persistent store is not marked as persistable
14. root array is segmented and one of the segments is invalid
530 VM_PREVENT_GC_OVERFLOW: _preventGC overflow A fixed number of native objects can be protected from being garbage collected. This error indicates that a native has exceeded the fixed limit of objects that can be protected. If the device is reset or thread tracebacks are logged, the name of the actual native can be extracted.
531 Flash exhausted There are certain operations where the JVM cannot tolerate running out of flash space. In these circumstances, this error will be reported if the JVM cannot allocate a required amount of flash space.
532 VM_ASSERTION_FAILED: Assertion failed Normally this JVM error should never be reported since the device is not shipped with assertions enabled. The simulator may report this error in debug mode indicating a VM assertion has been violated. Try typing BKPT to activate the debugger and dump the native call stack for forwarding to the VM team.
533 VM_RUN_METHOD: <method> needs running Used internally for ECMAScript to call Java methods.
534 VM_FAST_RESET_DISABLED: Fast Reset Disabled Used internally to indicate that fast reset capability is not available. Often used in platform-specific code.
535 VM_UNUSED_535: Unused Unused vm error.
Error 535 means out of memory. A reason for error 535 could be that the virtual memory thread scheduler ceased when a thread ran out of memory. This error is known to occur during a start-up sequence or if the main event thread ceases.
Resolution
First, hard reset the handheld. If this does not resolve the issue, obtain the latest version of the BlackBerry handheld software through your carrier. Then, run Application Loader from BlackBerry Desktop Manager to update the handheld.
Delete some themes/ringtones etc to make space.
536 VM_FAST_RESET_BAD_INSTANCE: VM Instance Check Failed Used internally to indicate that the VM structure passed in is at the wrong address or has been corrupted.
537 VM_FAST_RESET_BAD_HEAP: Heap Check Failed Used internally to indicate that the VM heap has been corrupted or pointers into the heap have been corrupted.
538 VM_FAST_RESET_BAD_IRAM: IRAM Check Failed Used internally to indicate that the VM IRAM checks have detected corruption of VM data structures (threads + local stacks) that reside in IRAM.
539 VM_FAST_RESET_NOT_IDLE: Not Idle Used internally to indicate that the VM was not idle when the reset occurred and, as such, cannot continue with a fast reset.
540 VM_FAST_RESET_MULTIPLE_RESETS: Multiple Resets Used internally to indicate that the time since the last fast reset is less than a minimum time. By disallowing multiple fast resets in a short amount of time, this should prevent fast reset loops.
541 VM_HEAP_COMPACT_INFINITE_LOOP: infinite loop detected in heap compaction The VM has detected a problem in its RAM heap that indicates its RAM has been corrupted. The problem was detected by identifying a possible infinite loop during RAM heap compaction. A bugdisp log and eventlog should be extracted quickly when the device is in this condition. If possible, images of RAM should be saved.
542 Transient memory leak The JVM has detected that some memory has not been freed, which indicates that a memory leak has occurred. This condition is detected as early as possible in the hope that the cause can be isolated.
543 VM_FS_MISMATCH: Incompatible Java filesystem installed The VM has detected that the OS binary is different than the OS binary used to create the Java file system. This means that the Java native methods may not be linked properly and as such, the integrity of the system cannot be guaranteed. The system can be recovered by using the VM DLFX and DLPS commands to delete the fixups and persistent store. This will clear all data and fixups and let the filesystem relink to match the new OS binary. The recovery order is important: 1) Delete Fixups 2) Delete Persistent Store 3) Reset device.
544 VM_SECTION_MAP_OVERFLOW: a module references more than 255 other modules The VM has detected that a module is trying to reference more than 255 other modules. The filesystem should be extracted immediately when this error is detected.
545 VM_INCOMPATIBLE_FILESYS: an incompatible or corrupt filesystem was found The VM has detected an incompatible or corrupt filesystem. The filesystem should be extracted immediately when this error is detected.
546 VM_UNUSED_546: unused The VM has detected that the RAM image of its filesystem is corrupted (failed CRC check) Better to reset than to duplice the corruption into flash.
547 VM_UNUSED_547: unused Unused vm error.
548 VM_UNUSED_548: unused Unused vm error.
549 VM_UNUSED_549: unused Unused vm error.

Source: FAQ: Explanation of All BlackBerry Java Virtual Machine Error Codes | BerryReview.com �
RIM KB: What do JVM error codes mean?
eVeek Post: Guide to Blackberry JVM Error Codes - eVeek | The Free Blackberry Download Community