none
0x40000015 {Fatal Application Exit} STATUS_FATAL_APP_EXIT

    Question

  • My app crashed and produced a .mdmp file which displays the following when I run it in Visual Studio:

    Unhandled exception at 0x00454009 (my_app.exe) in my_app.exe.mdmp: 0x40000015:  %hs


    I was able to find little information about the above exception:
    21 0x40000015 {Fatal Application Exit} %hs STATUS_FATAL_APP_EXIT

    I would really like to know what causes the above exception or under what conditions can I possible get it.

    Wednesday, February 27, 2008 8:57 PM

Answers

  • The CRT's heap manager could be bailing with that error code (although looking at the source, I see heap corruption erros but no calls to abort).

     

    However, without a reliable repro, we're just speculating.

     

    Friday, February 29, 2008 8:58 PM
    Moderator

All replies

  •  

    You can use the information on this page to access the public symbol servers.  This page has more information on minidumps. 

     

    Try looking through the callstack to figure out what could have thrown that exception.  Sometimes the crash occurs too late to find the source.  If you can reproduce the crash, try turning on first-chance exceptions to see if the debugger can catch the exception app before it crashes.

     

     

    Walden

    Wednesday, February 27, 2008 9:20 PM
    Moderator
  • Is your application calling the crt's abort method by chance? Unfortunately, that method actually overwrites any existing exception and makes for bogus mini-dumps. To find out, set a function breakpoint in any function called abort.

     

    HTH

     

    Jackson

    Thursday, February 28, 2008 12:10 AM
    Moderator
  • Thank you, I did not know about the symbol server. I have set up my Visual Studio to download the kernel symbols and reran the dump file. Unfortunately, I do not see any red flags there. All the os symbols look normal as expected.

    Thursday, February 28, 2008 12:19 AM
  • Jackson,

    Thanks for pointing this out. Unfortunately, I cannot easily reproduce this core.
    The callstack of this core hints at memory being corrupted. Unfortunately, I cannot find the source of the memory corruption from the callstack I got. I was hoping that knowing under which conditions this exception occurs would help me find the source of the memory corruption and I can go from there.
    I do not believe I am calling crt's abort method from anywhere in my program.

    I looked at the list of various unhandled exceptions, and most of them are pretty specific and have details as to what has happened.

    http://blogs.msdn.com/joshpoley/pages/errors-000-facility-null.aspx

    I was surprised to see that this particular exception had no detailed description, so I posted this thread to get some inside information about it.
    Thursday, February 28, 2008 12:35 AM
  • The CRT's heap manager could be bailing with that error code (although looking at the source, I see heap corruption erros but no calls to abort).

     

    However, without a reliable repro, we're just speculating.

     

    Friday, February 29, 2008 8:58 PM
    Moderator