Exception occurring on Windows Service run, only on Vista - c++

I have a third party kernel mode driver that is controlled by a windows service that I wrote in C++. The service initialization code appears to run perfectly, but when the service thread is run it encounters an exception that appears to indicate stack or heap corruption (it varies depending on the circumstances of debugging). The exception occurs in an address that is "not in any loaded module", making it almost impossible to tell what happened. Running in windbg with application verifier as an admin gives this after analyze -v:
Breakpoint 0 hit
eax=001ffb04 ebx=7ffda000 ecx=001ffb04 edx=002e1b34 esi=00000000 edi=001ff8a4
eip=00282051 esp=001ff76c ebp=001ff8b4 iopl=0 nv up ei ng nz na po nc
cs=001b ss=0023 ds=0023 es=0023 fs=003b gs=0000 efl=00000282
MyService!CServiceBase::Run+0x71:
00282051 8bf4 mov esi,esp
0:000> g
Detected memory leaks!
Dumping objects ->
{53} normal block at 0x058BFFF0, 8 bytes long.
Data: < ( > C0 15 28 00 04 FB 1F 00
Object dump complete.
(4e6c.4e9c): Access violation - code c0000005 (first chance)
First chance exceptions are reported before any exception handling.
This exception may be expected and handled.
eax=00000001 ebx=7ffda000 ecx=bffefce9 edx=002f8950 esi=00000000 edi=001ff8a4
eip=bfe1045d esp=001ff778 ebp=001ff8b4 iopl=0 nv up ei pl zr na pe nc
cs=001b ss=0023 ds=0023 es=0023 fs=003b gs=0000 efl=00010246
bfe1045d ?? ???
The _crtdbg "memory leaks" I believe are objects that I allocated during service initialization that will not be freed until service stop.
0:000> !analyze -v
*******************************************************************************
* *
* Exception Analysis *
* *
*******************************************************************************
Current verifier stop:
APPLICATION_VERIFIER_LUAPRIV_OK_OBJECT_DUMP (3317)
The application was able to access the object's security descriptor.
The application was granted the requested access to this object. A standard user should also be able to access this object.
Arguments:
Arg1: 6d101d5c, Object Name
Arg2: 0012019f, Access Requested
Arg3: 05156f60, Security Descriptor
Arg4: 05986fa0, String Security Descriptor
Previous verifier stop:
APPLICATION_VERIFIER_LUAPRIV_OK_OBJECT_GRANT (331e)
Safe Object.
The application opened an object (such as a file or registry key) and requested access that is granted to at least one non-privileged entity (listed). This suggests that the same operation will work when attempted by non-privileged/standard users.
Arguments:
Arg1: 6d101d5c, Object Type
Arg2: 05148fd8, Object Name
Arg3: 05156f8c, Access Control Entry
Arg4: 00000000, N/A
*** WARNING: Unable to verify checksum for ...nfapi.dll
FAULTING_IP:
+0
bfe1045d ?? ???
EXCEPTION_RECORD: ffffffff -- (.exr 0xffffffffffffffff)
ExceptionAddress: bfe1045d
ExceptionCode: c0000005 (Access violation)
ExceptionFlags: 00000000
NumberParameters: 2
Parameter[0]: 00000000
Parameter[1]: bfe1045d
Attempt to read from address bfe1045d
FAULTING_THREAD: 00004e9c
DEFAULT_BUCKET_ID: BAD_INSTRUCTION_PTR
PROCESS_NAME: MyService.exe
ERROR_CODE: (NTSTATUS) 0xc0000005 - The instruction at 0x%08lx referenced memory at 0x%08lx. The memory could not be %s.
EXCEPTION_CODE: (NTSTATUS) 0xc0000005 - The instruction at 0x%08lx referenced memory at 0x%08lx. The memory could not be %s.
EXCEPTION_PARAMETER1: 00000000
EXCEPTION_PARAMETER2: bfe1045d
READ_ADDRESS: bfe1045d
FOLLOWUP_IP:
MyService!main+71 [.....svc.cpp # 51]
00282d81 83c404 add esp,4
FAILED_INSTRUCTION_ADDRESS:
+9
bfe1045d ?? ???
NTGLOBALFLAG: 2000100
APPLICATION_VERIFIER_FLAGS: 80c43267
APP: myservice.exe
IP_ON_HEAP: 7ffda000
The fault address in not in any loaded module, please check your build's rebase
log at <releasedir>\bin\build_logs\timebuild\ntrebase.log for module which may
contain the address if it were loaded.
PRIMARY_PROBLEM_CLASS: BAD_INSTRUCTION_PTR
BUGCHECK_STR: APPLICATION_FAULT_BAD_INSTRUCTION_PTR_INVALID_POINTER_READ
FRAME_ONE_INVALID: 1
LAST_CONTROL_TRANSFER: from 7ffda000 to bfe1045d
STACK_TEXT:
WARNING: Frame IP not in any known module. Following frames may be wrong.
001ff774 7ffda000 cccccccc cccccccc cccccccc 0xbfe1045d
001ff8b4 00282d81 001ffb04 bfe107a5 00000000 0x7ffda000
001ffb4c 002aac69 00000001 05868fa8 0586af60 MyService!main+0x71
001ffb98 002aab2f 001ffbac 75fbd2e9 7ffda000 MyService!__tmainCRTStartup+0x129
001ffba0 75fbd2e9 7ffda000 001ffbec 77691603 MyService!mainCRTStartup+0xf
001ffbac 77691603 7ffda000 762ef334 00000000 KERNEL32!BaseThreadInitThunk+0xe
001ffbec 776915d6 002aab20 7ffda000 00000000 ntdll!__RtlUserThreadStart+0x23
001ffc04 00000000 002aab20 7ffda000 00000000 ntdll!_RtlUserThreadStart+0x1b
FAULTING_SOURCE_LINE: .....svc.cpp
FAULTING_SOURCE_FILE: .....svc.cpp
FAULTING_SOURCE_LINE_NUMBER: 51
FAULTING_SOURCE_CODE:
47: {
48: try
49: {
50: CSampleService service(SVCNAME, TRUE, TRUE, TRUE);
> 51: if (!CServiceBase::Run(service))
52: {
53: printf("Service failed to run w/err 0x%08lx\n", GetLastError());
54: }
55: return 0;
56: } catch (char *str )
SYMBOL_STACK_INDEX: 2
SYMBOL_NAME: myservice!main+71
FOLLOWUP_NAME: MachineOwner
MODULE_NAME: MyService
IMAGE_NAME: MyService.exe
DEBUG_FLR_IMAGE_TIMESTAMP: 530b8caa
STACK_COMMAND: ~0s ; kb
FAILURE_BUCKET_ID: BAD_INSTRUCTION_PTR_c0000005_MyService.exe!main
BUCKET_ID: APPLICATION_FAULT_BAD_INSTRUCTION_PTR_INVALID_POINTER_READ_BAD_IP_myservice!main+71
Followup: MachineOwner
---------
The difficulty here is that windows service startup code cannot be debugged directly, you have to skip the service start code and step into the initialization code. After the service starts you can attach to the process to debug, but this exception is thrown before the service starts successfully but after the initialization code is completed. If I step past the end of the initialization code the call stack is only hex addresses and the code is assembly in ntdll.dll. I can see that something is being corrupted in memory that later is used in service run but I can't detect what that is, all of the local and global variables appear to be correct completely through the initialization code. The third party code runs successfully as an exe, and the service starts successfully if I do not call the third party code.
When run from a command line the service start times out, which can occur if an exception occurs.
int main(int argc, char *argv[])
{
try
{
CSampleService service(SVCNAME, TRUE, TRUE, TRUE);
if (!CServiceBase::Run(service))
{
printf("Service failed to run w/err 0x%08lx\n", GetLastError());
}
return;
} catch (char *str )
{
char sMessage[MAX_PATH];
sprintf_s(sMessage,"Service failed to run, exception=%s", str);
SvcReportEvent(sMessage, 0);
}
catch (...)
{
char sMessage[MAX_PATH];
sprintf_s(sMessage,"Service failed to run, exception occured");
SvcReportEvent(sMessage, 0);
}
}
EDIT: I have now solved the problem. I used Windbg in postportem mode (HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows NT\CurrentVersion\Image File Execution Options\servicename.exe
Debugger (string) set to "C:\Program Files\Debugging Tools for Windows (x86)\windbg.exe"), set my service to be interactive by ORing SERVICE_INTERACTIVE_PROCESS with the service type. I used __debugbreak() statements to break into the debugger in ServiceMain, which brought up an odd fullscreen mode window with only windbg running after an "interactive mode prompt" dialog. Once there, I was able to debug the running service itself, immediately after startup. It turned out the service did have a mistake in it that caused the stack overflow, which I was able to see and correct. Thank you, everyone!

Related

Finding execption using WinDbg

I build my 32 bit application file in windows 2008 server machine using visual studio 2010. this 32 bit application will use IBM wepshere MQ Client ( 7.5.0.3).
while running in the debugging mode i am getting exception so i debugged using WinDbg software
dump info is
*******************************************************************************
* *
* Exception Analysis *
* *
*******************************************************************************
*** ERROR: Symbol file could not be found. Defaulted to export symbols for C:\Program Files (x86)\IBM\WebSphere MQ\bin\mqe.dll -
FAULTING_IP:
ntdll32!RtlpCreateSplitBlock+47b
77d23fcb 8b11 mov edx,dword ptr [ecx]
EXCEPTION_RECORD: ffffffffffffffff -- (.exr 0xffffffffffffffff)
ExceptionAddress: 0000000077d23fcb (ntdll32!RtlpCreateSplitBlock+0x000000000000047b)
ExceptionCode: c0000005 (Access violation)
ExceptionFlags: 00000000
NumberParameters: 2
Parameter[0]: 0000000000000000
Parameter[1]: 000000003b9ac9ff
Attempt to read from address 000000003b9ac9ff
CONTEXT: 0000000000000000 -- (.cxr 0x0;r)
eax=0adaff08 ebx=02960000 ecx=3b9ac9ff edx=00100000 esi=0ae2bf90 edi=02960174
eip=77d23fcb esp=079bd490 ebp=079bd4b4 iopl=0 nv up ei pl zr na pe nc
cs=0023 ss=002b ds=002b es=002b fs=0053 gs=002b efl=00010246
ntdll32!RtlpCreateSplitBlock+0x47b:
77d23fcb 8b11 mov edx,dword ptr [ecx] ds:002b:3b9ac9ff=????????
FAULTING_THREAD: 000000000000046c
PROCESS_NAME: MessageBroker.exe
ERROR_CODE: (NTSTATUS) 0xc0000005 - The instruction at 0x%08lx referenced memory at 0x%08lx. The memory could not be %s.
EXCEPTION_CODE: (NTSTATUS) 0xc0000005 - The instruction at 0x%08lx referenced memory at 0x%08lx. The memory could not be %s.
EXCEPTION_PARAMETER1: 0000000000000000
EXCEPTION_PARAMETER2: 000000003b9ac9ff
READ_ADDRESS: 000000003b9ac9ff
FOLLOWUP_IP:
ntdll32!RtlpCreateSplitBlock+47b
77d23fcb 8b11 mov edx,dword ptr [ecx]
NTGLOBALFLAG: 0
APPLICATION_VERIFIER_FLAGS: 0
APP: messagebroker.exe
ANALYSIS_VERSION: 6.3.9600.17298 (debuggers(dbg).141024-1500) amd64fre
**BUGCHECK_STR: APPLICATION_FAULT_INVALID_POINTER_READ_ZEROED_STACK**
PRIMARY_PROBLEM_CLASS: INVALID_POINTER_READ
DEFAULT_BUCKET_ID: INVALID_POINTER_READ
LAST_CONTROL_TRANSFER: from 0000000077d23c5a to 0000000077d23fcb
**STACK_TEXT:**
079bd4b4 77d23c5a 0ae2bf90 0adaff08 02960150 ntdll32!RtlpCreateSplitBlock+0x47b
079bd5a0 77d23cfe 00000840 00000848 0ae2b74a ntdll32!RtlpAllocateHeap+0xb5d
079bd624 74e04d83 02960000 00000000 00000840 ntdll32!RtlAllocateHeap+0x23a
079bd644 008a8894 00000840 00000002 079bd800 MSVCR80!malloc+0x7a
**WARNING: Stack unwind information not available. Following frames may be wrong.
00000000 00000000 00000000 00000000 00000000 mqe+0x68894**
STACK_COMMAND: .cxr 0x0 ; kb
SYMBOL_STACK_INDEX: 0
SYMBOL_NAME: ntdll32!RtlpCreateSplitBlock+47b
FOLLOWUP_NAME: MachineOwner
MODULE_NAME: ntdll32
IMAGE_NAME: ntdll32.dll
DEBUG_FLR_IMAGE_TIMESTAMP: 51fb1072
FAILURE_BUCKET_ID: INVALID_POINTER_READ_c0000005_ntdll32.dll!RtlpCreateSplitBlock
**BUCKET_ID: X64_APPLICATION_FAULT_INVALID_POINTER_READ_ZEROED_STACK_ntdll32!RtlpCreateSplitBlock+47b**
ANALYSIS_SOURCE: UM
FAILURE_ID_HASH_STRING: um:invalid_pointer_read_c0000005_ntdll32.dll!rtlpcreatesplitblock
FAILURE_ID_HASH: {c09c2d45-666c-6fbd-b6fd-d9b79ba8c4fa}
Above dump show that Mqe.dll(IBM WebSphere MQ dll) is passing invalid pointer
Based on the above dump i am unable find the issue. The stack call very small so i am unable to find where my application calling MQ application ( Mqe.ddl).
Please suggest me how to find the issue and how to debugging Mqe.dll.
MQE.dll is quite possibly an innocent bystander here. It is definitely allocating memory via malloc, and then the crash happens inside of the memory manager. This is typically a result of memory corruption that happened at some point in the past. Buffer overflow, underflow, or use-after-free are the typical reasons for heap corruption.
Options to help track the issue down:
Use !heap -v in WinDbg to perform a heap analysis on the minidump.
Analyze the memory around where the memory manager is attempting to work, you may identify some know data (such as strings) that might give you hints as to who corrupted the memory.
I would walk backwards in the assembly and figure out where the read address (0x3b9ac9ff) came from.
The current registers may also contain pointers inside the heap where things are bad, you also use those as starting points for investigating.
Rerun the application with PageHeap to try and catch an overflow immediately when it happens.
Code review your allocations and uses of any heap memory.

Reading Dump file through Debug engine API

I have written a small program using Debug engine API to read a dump file.
I am executing !analyze -v command through code.
I am able to get almost every detail that could be extracted with above command but not the Process Name and Image name and module name
I really don't know where I'm going wrong.
Things i tried:
copied the dll's ext,exts,Kdexts,kext to the same folder where my exe is
present.
also copied the symsrv.dll.
for symbol path i am using
symbols->SetSymbolPath("srv*http://msdl.microsoft.com/download/symbols") where symbols is an IDebugSymbols pointer
But so far it didn't work.
The result i'm getting is :
*******************************************************************************
* *
* Bugcheck Analysis *
* *
*******************************************************************************
CRITICAL_OBJECT_TERMINATION (f4) A process or thread crucial to system
operation has unexpectedly exited or been terminated. Several
processes and threads are necessary for the operation of the system;
when they are terminated (for any reason), the system can no longer
function. Arguments: Arg1: 00000003, Process Arg2: 84d97860,
Terminating object Arg3: 84d979cc, Process image file name Arg4:
8285cec0, Explanatory message (ascii)
Debugging Details:
------------------
***** Debugger could not find nt in module list, module list might be incorrect, error 0x80070057.
-----------------------------------------------
| NT symbols are not available |
| reduced functionality |
| |
------------------------------------------------
unable to get nt!KiCurrentEtwBufferOffset
unable to get nt!KiCurrentEtwBufferBase
PROCESS_OBJECT: 84d97860
IMAGE_NAME: Unknown_Image
DEBUG_FLR_IMAGE_TIMESTAMP: 0
FAULTING_MODULE: 00000000
CUSTOMER_CRASH_COUNT: 1
DEFAULT_BUCKET_ID: VISTA_DRIVER_FAULT
BUGCHECK_STR: 0xF4
CURRENT_IRQL: 0
STACK_TEXT: WARNING: Frame IP not in any known module. Following
frames may be wrong. 950dbc9c 829223af 000000f4 00000003 84d97860
0x82722bfc 950dbcc0 828a0009 8285cec0 84d979cc 84d97ad0 0x829223af
950dbcf0 8289ff4c 84d97860 8447b030 00000001 0x828a0009 950dbd24
826818c6 000001e0 00000001 001cebb0 0x8289ff4c 950dbd34 77be70f4
badb0d00 001ceba8 00000000 0x826818c6 950dbd38 badb0d00 001ceba8
00000000 00000000 0x77be70f4 950dbd3c 001ceba8 00000000 00000000
00000000 0xbadb0d00 950dbd40 00000000 00000000 00000000 00000000
0x1ceba8
STACK_COMMAND: kb
BUCKET_ID: CORRUPT_MODULELIST
MODULE_NAME: Unknown_Module *** Followup info cannot be found !!!
Please contact "Debugger Team"

How to debug access violation 0xC0000005 in CorExitProcess on exit?

Our application (written in C++, VS 2010 project) has been running fine on all operating systems prior to Windows 8 (and still does). On Windows 8, however, when orderly exiting the application, an access violation occurs:
mfc100.dll!_DllMain#12() <<< Crash here
mfc100.dll!__CRT_INIT#12()
mfc100.dll!__DllMainCRTStartup#12()
ntdll.dll!_LdrxCallInitRoutine#16()
ntdll.dll!LdrpCallInitRoutine()
ntdll.dll!LdrShutdownProcess()
ntdll.dll!RtlExitUserProcess()
kernel32.dll!_ExitProcessImplementation#4()
mscoreei.dll!RuntimeDesc::ShutdownAllActiveRuntimes(unsigned int,class RuntimeDesc *,enum RuntimeDesc::ShutdownCompatMode)
mscoreei.dll!_CorExitProcess#4()
mscoree.dll!_ShellShim_CorExitProcess#4()
msvcr100d.dll!__crtCorExitProcess(int status) line693 C
msvcr100d.dll!__crtExitProcess(int status) line 699 C
msvcr100d.dll!doexit(int code, int quick, int retcaller) line 621 C
msvcr100d.dll!exit(int code) Zeile 393 C
my.exe!__tmainCRTStartup() Zeile 568 C
my.exe!WinMainCRTStartup() Zeile 371 C
kernel32.dll!#BaseThreadInitThunk#12()
ntdll.dll!__RtlUserThreadStart()
ntdll.dll!__RtlUserThreadStart#8()
In an MSDN forum topic it has been suggested to run GC.Collect() before exit, but I couldn't make any difference with such a call shortly before exit.
I am a bit at a loss about how I should debug the problem. As far as I understand, CorExitProcess takes care of cleaning up the managed resources of the application. So could this be a fault in a managed component?
Or is it more likely that some function pointer in _DllMain has been overwritten/corrupted? If so, how would I set a data breakpoint at the address in question? There is a post explaning how to debug a similar issue, but he's having the issue in his own DLL so he can actually peak at the exact source of the problem which I can't.
Any suggestions?
Edit:
Additional information, windbg !analyze -v:
FAULTING_IP:
mfc100+258e6c
64298e6c 8b4654 mov eax,dword ptr [esi+54h]
EXCEPTION_RECORD: ffffffff -- (.exr 0xffffffffffffffff)
ExceptionAddress: 64298e6c (mfc100+0x00258e6c)
ExceptionCode: c0000005 (Access violation)
ExceptionFlags: 00000000
NumberParameters: 2
Parameter[0]: 00000000
Parameter[1]: 53f21f0c
Attempt to read from address 53f21f0c
CONTEXT: 00000000 -- (.cxr 0x0;r)
eax=53f21eb8 ebx=00000000 ecx=64187d2d edx=7fcde000 esi=53f21eb8 edi=00000001
eip=64298e6c esp=00c3f1b8 ebp=00c3f2ec iopl=0 nv up ei pl nz na pe nc
cs=001b ss=0023 ds=0023 es=0023 fs=003b gs=0000 efl=00210206
mfc100+0x258e6c:
64298e6c 8b4654 mov eax,dword ptr [esi+54h] ds:0023:53f21f0c=????????
FAULTING_THREAD: 00000520
DEFAULT_BUCKET_ID: WRONG_SYMBOLS
PROCESS_NAME: ww.exe
ADDITIONAL_DEBUG_TEXT:
You can run '.symfix; .reload' to try to fix the symbol path and load symbols.
MODULE_NAME: mfc100
FAULTING_MODULE: 77bc0000 ntdll
DEBUG_FLR_IMAGE_TIMESTAMP: 4d5f29b8
ERROR_CODE: (NTSTATUS) 0xc0000005 - Die Anweisung in 0x%08lx verweist auf Speicher 0x%08lx. Der Vorgang %s konnte nicht im Speicher durchgef hrt werden.
EXCEPTION_CODE: (NTSTATUS) 0xc0000005 - Die Anweisung in 0x%08lx verweist auf Speicher 0x%08lx. Der Vorgang %s konnte nicht im Speicher durchgef hrt werden.
EXCEPTION_PARAMETER1: 00000000
EXCEPTION_PARAMETER2: 53f21f0c
READ_ADDRESS: 53f21f0c
FOLLOWUP_IP:
mfc100+258e6c
64298e6c 8b4654 mov eax,dword ptr [esi+54h]
APP: ww.exe
ANALYSIS_VERSION: 6.3.9600.17029 (debuggers(dbg).140219-1702) x86fre
MANAGED_STACK: !dumpstack -EE
OS Thread Id: 0x520 (0)
Current frame:
ChildEBP RetAddr Caller, Callee
PRIMARY_PROBLEM_CLASS: WRONG_SYMBOLS
BUGCHECK_STR: APPLICATION_FAULT_WRONG_SYMBOLS
LAST_CONTROL_TRANSFER: from 6429da08 to 64298e6c
STACK_TEXT:
WARNING: Stack unwind information not available. Following frames may be wrong.
00c3f2ec 6429da08 64040000 00000000 00000001 mfc100+0x258e6c
00c3f330 6429dac7 64040000 00c3f35c 77be077a mfc100+0x25da08
00c3f33c 77be077a 64040000 00000000 00000001 mfc100+0x25dac7
00c3f35c 77be07f0 6429daa9 64040000 00000000 ntdll!RtlAddMandatoryAce+0x14e
00c3f3a4 77bfa529 6429daa9 64040000 00000000 ntdll!RtlAddMandatoryAce+0x1c4
00c3f49c 77bfa40e 00000000 00000000 6f2d4890 ntdll!RtlExitUserProcess+0x1e7
00c3f4b0 76ff4231 00000000 77e8f3b0 ffffffff ntdll!RtlExitUserProcess+0xcc
00c3f4c4 6f8b3712 00000000 bd3cbe8b 01f1c054 KERNEL32!ExitProcess+0x15
00c3f74c 6f8c19a2 00000001 00c3f76c 6f1686ad mscoreei!GetFileVersion+0x1835
00c3f758 6f1686ad 00000000 77bdab85 6f8a0000 mscoreei!CorExitProcess+0x27
00c3f76c 70737954 00000000 00c3f784 7073798d mscoree!CorExitProcess+0x94
00c3f778 7073798d 00000000 00c3f7c8 70737ab0 MSVCR100!_query_new_mode+0x159
00c3f784 70737ab0 00000000 a2b843a9 00375f5c MSVCR100!_query_new_mode+0x192
00c3f7c8 70737b1d 00000000 00000000 00000000 MSVCR100!_query_new_mode+0x2b5
00c3f7dc 003274ab 00000000 d1ef1931 00000000 MSVCR100!exit+0x11
00c3f864 76ff173e 7fcdf000 00c3f8b4 77c16911 ww!_enc$textbss$begin+0x64ab
00c3f870 77c16911 7fcdf000 a613e810 00000000 KERNEL32!BaseThreadInitThunk+0x12
00c3f8b4 77c168bd ffffffff 77c8560a 00000000 ntdll!LdrInitializeThunk+0x1f0
00c3f8c4 00000000 003275da 7fcdf000 00000000 ntdll!LdrInitializeThunk+0x19c
STACK_COMMAND: .cxr 0x0 ; kb
SYMBOL_STACK_INDEX: 0
SYMBOL_NAME: mfc100+258e6c
FOLLOWUP_NAME: MachineOwner
IMAGE_NAME: mfc100.dll
BUCKET_ID: WRONG_SYMBOLS
FAILURE_BUCKET_ID: WRONG_SYMBOLS_c0000005_mfc100.dll!Unknown
ANALYSIS_SOURCE: UM
FAILURE_ID_HASH_STRING: um:wrong_symbols_c0000005_mfc100.dll!unknown
FAILURE_ID_HASH: {9e516b68-081f-78d6-cf23-b42f2b3cb573}
Followup: MachineOwner
---------
Screenshot of there the crash occurs:
As discussed in comments, our similar problem was where we had a native C++ application that communicated with a managed C# application running as a COM server. To allow the managed component to communicate events to the C++ app, an event sink was exposed as a simple ATL COM interface from the native side, which on the .NET side was automatically encapsulated in a Runtime Callable Wrapper.
The access violation on application close - which wasn't always visible except in the event logs - was due to the fact that the RCW didn't call Release() on our ATL COM interfaces until it was garbage collected. As this happened when the .NET runtime closed, which was after the native runtime had shut down, it tried to callback into dead code.
The solution for us was to expose a "shutdown" method on the .NET side that disposed of all the communicating objects, then called:
GC.Collect();
GC.WaitForPendingFinalizers();
Ok, I understand that this might not exactly mirror your problem, but the route in to finding out what was causing it was to use the Managed Debugging Assistants, particularly reportAvOnCOMRelease.
We activated the MDA by registry keys and ran the native app via a debugger to see the additional output that identified the COM interfaces that were being held too long. Probably as a first step, it would be wise to activate all of the MDA options to glean as much info as possible from the crash.
I tried debugging this using data breakpoints, but that didn't help a lot. I could see that at some point the data being accessed was overwritten, but that didn't happen in a call stack containing any of my own code.
So I resorted in a simpler method and started removing parts of the program until the error disappeared. In a large application it may be hard to remove some parts without breaking others, but I was able to narrow down the source of the issue.
Basically, the problem stopped occurring after removing a certain call to FreeLibrary. After further investigation it turned out that this call happens during DllMain, which is not allowed:
The entry-point function should perform only simple initialization or termination tasks. It must not call the LoadLibrary or LoadLibraryEx function (or a function that calls these functions), because this may create dependency loops in the DLL load order. This can result in a DLL being used before the system has executed its initialization code. Similarly, the entry-point function must not call the FreeLibrary function (or a function that calls FreeLibrary) during process termination, because this can result in a DLL being used after the system has executed its termination code.
In another SO question, one user apparently noticed a change since Windows 8 in this regard, which would explain why the error only happens on this version of Windows.
We'll now change our application so that FreeLibrary is called at a different point of time.

How can I find the crash line of code from stack?

Now I have the crash info:
TestAPP caused an Access Violation (0xc0000005)
in module XCrashReportTest.exe at 0023:5763c230.
Context:
EDI: 0x002aee0c ESI: 0x002ad7f4 EAX: 0x00000000
EBX: 0x00000000 ECX: 0x57635670 EDX: 0x028c1d27
EIP: 0x5763c230 EBP: 0x002aee18 SegCs: 0x00000023
EFlags: 0x00010246 ESP: 0x002ad554 SegSs: 0x0000002b
Bytes at CS:EIP:
8b 08 52 50 8b 41 18 ff d0 8b 0d c8 e0 66 57 8b
Stack:
0x002ad554: 576504f7 002aee0c 002ad7f4 002ad570 ..eW..*...*.p.*.
0x002ad564: 002aee0c 5763c201 002aee18 002ad59c ..*...cW..*...*.
0x002ad574: 576490b1 5763c201 002aee0c 00000100 ..dW..cW..*.....
0x002ad584: 002ad5f4 57648d9f 75785eaa 57668e70 ..*...dW.^xup.fW
0x002ad594: 002aee0c 00000001 002ad604 5764d4d2 ..*.......*...dW
0x002ad5a4: 002aee0c 57668e70 5763c201 00000000 ..*.p.fW..cW....
0x002ad5b4: 00000100 75785d2a 57668e5c 002aee0c ....*]xu\.fW..*.
0x002ad5c4: 57668e4c 002add84 00000000 00000000 L.fW..*.........
I refer to XCrashReport[XCrashReport : Exception Handling and Crash Reporting - Part 1
But I find that sometimes the EIP differed when I run the same *.exe in the same machine. So, sometimes I can't match the crash line of code with EIP(the methods descriped in XCrashReport : Exception Handling and Crash Reporting - Part 1. I don't know how to use the stack info.
Anyone can help me?
The only one reasonable way for doing this is to load your crash dump into the debugger and provide symbols, i.e. the .pdb files. This can be tricky because pdb files may not have enough info. Companies like Microsoft give out public PDB files that do not have information about the names of the local vars and other details. PDBs that with full information are called public PDBs.
Sometimes the stack gets corrupted and this results in a "lose of control". Code picks up the return address, that is some garbage, loads this value into EIP and starts executing instructions from there. In this scenario the crash happens very soon, after executing just a couple of "instructions".
I mean that you need to figure out if your EIP is pointing to any code first.

Understand Crash Dump from Visual C++ App

UPDATE
Thanks to feedback below I was able to home in on ADPlus.vbs, which is part of the debugging tools for Windows.
Don't forget to set up _NT_SYMBOL_PATH before you run it.
Using this we've been able to see much more clearly in to the application with far greater clarity than we ever have using the regular dumps produced via Windows when the application crashes.
Many thanks to all for the responses.
ORIGINAL QUESTION
We have an server application written in Visual C++ that some times (relatively rarely) crashes on customer sites. We haven't been able to understand why this happens based on looking at our own log files so the next step is to start looking at crash dumps.
We've just purposefully put a bug in to our app (a null pointer) so that we can generate a crash dump and verify that the dumps produced are valuable, but thus far I can't make head or tail of what i'm seeing.
I think my first question is whether i've even got WinDbg set up correctly (the other developer here is loading the dump in to Visual Studio 2010 and seeing the same errors so i'm assuming it's fine, or we're both wrong :) ) - and then next question is, how do I understand what it's telling me.
The main confusion is that the dump seems to be telling me it has reached a break point, which seems odd to me since there was no debugger connected.
The app was running on a Windows Server 2003 system when it crashed. I believe I have pointed WinDbg at the PDB file for the DLL and EXE correctly.
FAULTING_IP:
ntdll!DbgBreakPoint+0
7c81a3e1 cc int 3
EXCEPTION_RECORD: ffffffff -- (.exr 0xffffffffffffffff)
ExceptionAddress: 7c81a3e1 (ntdll!DbgBreakPoint)
ExceptionCode: 80000003 (Break instruction exception)
ExceptionFlags: 00000000
NumberParameters: 3
Parameter[0]: 00000000
Parameter[1]: 8779fdb0
Parameter[2]: 00000003
DEFAULT_BUCKET_ID: STATUS_BREAKPOINT
PROCESS_NAME: CallPlusServerLauncher.exe
ERROR_CODE: (NTSTATUS) 0x80000003 - {EXCEPTION} Breakpoint A breakpoint has been reached.
EXCEPTION_CODE: (HRESULT) 0x80000003 (2147483651) - One or more arguments are invalid
EXCEPTION_PARAMETER1: 00000000
EXCEPTION_PARAMETER2: 8779fdb0
EXCEPTION_PARAMETER3: 00000003
NTGLOBALFLAG: 0
APPLICATION_VERIFIER_FLAGS: 0
ADDITIONAL_DEBUG_TEXT: Followup set based on attribute [Is_ChosenCrashFollowupThread] from Frame:[0] on thread:[ffffffff]
FAULTING_THREAD: ffffffff
PRIMARY_PROBLEM_CLASS: STATUS_BREAKPOINT
BUGCHECK_STR: APPLICATION_FAULT_STATUS_BREAKPOINT
STACK_TEXT:
1bd0ffc8 7c83fe08 00000005 00000004 00000001 ntdll!DbgBreakPoint
1bd0fff4 00000000 00000000 00000000 00000000 ntdll!DbgUiRemoteBreakin+0x36
FOLLOWUP_IP:
ntdll!DbgBreakPoint+0
7c81a3e1 cc int 3
SYMBOL_STACK_INDEX: 0
SYMBOL_NAME: ntdll!DbgBreakPoint+0
FOLLOWUP_NAME: MachineOwner
MODULE_NAME: ntdll
IMAGE_NAME: ntdll.dll
DEBUG_FLR_IMAGE_TIMESTAMP: 49900d60
STACK_COMMAND: ddS 1bd10000 1bd0c000 ; dt ntdll!LdrpLastDllInitializer BaseDllName ; dt ntdll!LdrpFailureData ; ~439s; .ecxr ; kb
BUCKET_ID: MANUAL_BREAKIN
FAILURE_BUCKET_ID: STATUS_BREAKPOINT_80000003_ntdll.dll!DbgBreakPoint
WATSON_STAGEONE_URL: http://watson.microsoft.com/StageOne/CallPlusServerLauncher_exe/0_0_0_0/4df87414/ntdll_dll/5_2_3790_4455/49900d60/80000003/0001a3e1.htm?Retriage=1
Followup: MachineOwner
DbgBreakPoint -- Looks to me like you broke execution using a remote debugger.
If you didn't then I have seen DbgBreakPoint show up when you have code pages (Edit: I meant page heap) turned on (you should know if you did this) and there was a detection of invalid memory access.
Asserts can also trigger a breakpoint exception. For example I have (too often) seen them come out of the heap checking around a delete when the heap has got corrupted by double-delete or overflow. But only with the debug runtime I thought, is that what you have deployed?