VMMap crash with stack buffer overflow after long run RRS feed

  • Question

  • Hi there,

    I'm running our 32-bit MS C console application through VMMap 3.25 to try to identify a long-term memory leak; it takes several hours for the app to reach steady state and the gradual leak to become evident.

    Several times now, VMMap has simply disappeared after a few hours runtime, leaving the app running. I have sampling set to 10s (which doesn't show a checkmark or any indication on win 10 x64 1803). I'd really like to sample at 1 or even 5 minute intervals.

    Last night I paused tracing overnight, having been logging for about an hour. After resuming tracing this morning, VMMap crashed after an hour with the following info in the event log (the previous crash had the same code and offset):

    Faulting application name: vmmap.exe, version:, time stamp: 0x5be345cd
    Faulting module name: vmmap.exe, version:, time stamp: 0x5be345cd
    Exception code: 0xc0000409
    Fault offset: 0x0003f5cf
    Faulting process id: 0x28d8
    Faulting application start time: 0x01d4b2c68d915e83
    Faulting application path: C:\SysInternals\vmmap.exe
    Faulting module path: C:\SysInternals\vmmap.exe
    Report Id: 40e992f1-2ca4-4ef9-a39e-5cbb3d570be5
    Wednesday, January 23, 2019 10:40 PM

All replies

  • Hi Len

    Do you have a memory dump that you could share with us? If so could you contact me offline at and I will provide you with an upload location.

    If dump files are not configured you can enable this by creating the  HKLM\SOFTWARE\Microsoft\Windows\Windows\Error Reporting\LocalDumps key (if it doesn't already exist) then creating a DWORD value called DumpType and setting this to 2 (FullDumps).

    The dump file will the  be created at %LOCALAPPDATA%\CrashDumps

    See for more details.

    BTW if you are trying to trace a memory leak I find usermode stack database massively helpful (Gflags +ust)

    MarkC (MSFT)

    Monday, January 28, 2019 4:20 PM