0
votes

I have a code in production which crash one/two times per day, since one week. This crash is due to memory heap corruption

In order to discover what is going wrong in my application, I try to use GFlags.

I have activate page heap with GFlags (I have checked in windows registery that a entry for my application has been well created )

The application run under windows and I have developped it using MSCV2010.

I put the appliction in DEBUG mode in production and I use the feature "attach to process" inside microsoft visual stuio.

When the crash occur I have now the following call stack :

wow64!Wow64NotifyDebugger+0x9
wow64!Wow64KiRaiseException+0x172
wow64!Wow64SystemServiceEx+0xd6
wow64cpu!ServiceNoTurbo+0x28
wow64!RunCpuSimulation+0xa
wow64!Wow64LdrpInitialize+0x2ed
ntdll!LdrpInitialize+0x2aa
ntdll!KiUserApcDispatcher+0x15 (TrapFrame @ 00000000`03e3fb38)
kernel32!BaseThreadStartThunk

glurp...

It seems I am not more advanced in the resolution of my problem..

Any advices concerning the use of GFlags ?

More generally, what can I expect with GFlags which can tell me where is the bug in my application ?

1
@Hans Passant : Sorry but do you know, where I can't find application verifier executable ? - Guillaume07
Type "download application verifier" in a google query and take the first hit. - Hans Passant

1 Answers

1
votes

Maybe this article helps: 32-bit stack traces from x64 complete dumps You can save a dump file using visual studio by using Debug>Save dump as That dump file can then be opened in windbg.