Question

We are working with some legacy DLLs on our C#/.NET project. Why some exception on the legacy code cannot be caught and the application crashes? What makes the difference with standard .NET exceptions?

Thanks.

Answer 1

Jeffrey Ritcher, in his magnificent “CLR via C#”, tells us: some unmanaged-code failures are considered “corrupted state exceptions” (CSEs) by the .Net Runtime (CLR). Usually (see later), these exceptions cannot be caught by us mere mortals. Even finally blocks aren’t executed upon one of these failures, which include:

• Access violations
• Illegal instructions
• Stack overflows
• Page errors

You can, however, apply the HandleProcessCorruptedStateExceptionsAttribute and the SecurityCriticalAttribute attributes to the specific method in which you expect CSEs to happen. Inside this method you can code a try...catch block which will catch the CSE.

Much more detailed information can be found in this article: http://msdn.microsoft.com/en-us/magazine/dd419661.aspx