Question

I’m trying to understand why the second example below works with no issues, but the first example gives me the exception below. It seems to me that both examples should give an exception based on the description. Can anyone enlighten me?

Unhandled Exception:
System.TypeLoadException: Could not
load type ‘StructTest.OuterType’ from
assembly ‘StructTest, Version=1.0.0.0,
Culture=neutral, PublicKeyToken=null’
because it contains an object field at
offset 0 that is incorrectly aligned
or overlapped by a non-object field.
at StructTest.Program.Main(String[]
args) Press any key to continue . . .

Example 1

using System;
using System.Collections.Generic;
using System.Linq;
using System.Runtime.InteropServices;
using System.Text;

namespace StructTest
{
    [StructLayout(LayoutKind.Sequential, Pack = 1)]
    struct InnerType
    {
        [MarshalAs(UnmanagedType.ByValArray, SizeConst = 100)]
        char[] buffer;
    }

    [StructLayout(LayoutKind.Explicit)]
    struct OuterType
    {
        [FieldOffset(0)]
        int someValue;

        [FieldOffset(0)]
        InnerType someOtherValue;
    }

    class Program
    {
        static void Main(string[] args)
        {
            OuterType t = new OuterType();
            System.Console.WriteLine(t);
        }
    }
}

---

Example 2

using System;
using System.Collections.Generic;
using System.Linq;
using System.Runtime.InteropServices;
using System.Text;

namespace StructTest
{
    [StructLayout(LayoutKind.Sequential, Pack = 1)]
    struct InnerType
    {
        [MarshalAs(UnmanagedType.ByValArray, SizeConst = 100)]
        char[] buffer;
    }

    [StructLayout(LayoutKind.Explicit)]
    struct OuterType
    {
        [FieldOffset(4)]
        private int someValue;

        [FieldOffset(0)]
        InnerType someOtherValue;

    }

    class Program
    {
        static void Main(string[] args)
        {
            OuterType t = new OuterType();
            System.Console.WriteLine(t);
        }
    }
}

Answer 1

The common language runtime contains a verifier that makes sure the running code (verifiable IL) cannot possibly corrupt memory in the managed environment. This prevents you to declare such a structure in which fields overlap. Basically, your struct contains two data members. One integer (which is 4 bytes) and a native integer (pointer size). On a 32 bit CLR, in which you are probably running your code, the char[] will take 4 bytes so if you put the integer less than four bytes away from the beginning of the struct, you’ll have overlapping fields. It’s interesting to note that both of your code snippets with fail on a 64 bit runtime, as the pointer size is 8 bytes.