I have a TestNet wrapper for a native component. The native component exposes a blocking TestNative::Foo() that communicates with managed part through calling managed callbacks and a weak GCHandle that is used to retrieve the reference to the .NET wrapper and provides a context. The GCHandle is weak since the .NET wrapper is meant to hide the fact that is handling unmanaged resources to user and deliberately doesn't implement the IDisposable interface: being non weak it would prevent TestNet instances from being collected at all, creating a memory leak. What's happening is that in Release build only the garbage collector will collect reference to .NET wrapper while executing the managed callback, even before both TestNative::Foo() and surprisingly TestNet::Foo() unblocks. I understood the problem my self and I can fix it by issuing a GC.KeepAlive(this) after the P/Invoke call but since the knowledge of this is not very widespread, it seems a lot of people are doing it wrong. I have few questions:
GC.KeepAlive(this) always needed in a managed method if last instruction is a P/Invoke call on unmanaged resources or it's just needed in this special case, namely the switch to managed execution context while marshaling the managed callback from native code? The question could be: should I put GC.KeepAlive(this) everywhere? This old microsoft blog (original link is 404, here is cached) seems to suggest so! But this would be game changer and basically it would mean that most people never did P/Invoke correctly, because this would require reviewing most P/Invoke calls in wrappers. Is there for example a rule that say that garbage collector (EDIT: or better the finalizer) can't run for objects that belong to the current thread while execution context is unamanaged (native)?GC.KeepAlive(this) any time a unmanaged resource is accessed with P/Invoke. In general GC.KeepAlive(this) seems to be very rarely needed when dealing with finalizers.NOTE: I have no problem with delegates being collected, that is a different issue which I know how to handle properly. The issue here is with objects holding unmanaged resources being collected when P/Invoke calls are not finished yet.
It follows code that clearly manifest the problem. Creates a C# console application and a C++ Dll1 project and build them in Release mode:
Program.cs:
using System;
using System.Runtime.InteropServices;
namespace ConsoleApp1
{
class Program
{
static void Main(string[] args)
{
var test = new TestNet();
try
{
test.Foo();
}
catch (Exception ex)
{
Console.WriteLine(ex);
}
}
}
class TestNet
{
[UnmanagedFunctionPointer(CallingConvention.Cdecl)]
delegate void Callback(IntPtr data);
static Callback _callback;
IntPtr _nativeHandle;
GCHandle _thisHandle;
static TestNet()
{
// NOTE: Keep delegates references so they can be
// stored persistently in unmanaged resources
_callback = callback;
}
public TestNet()
{
_nativeHandle = CreateTestNative();
// Keep a weak handle to self. Weak is necessary
// to not prevent garbage collection of TestNet instances
_thisHandle = GCHandle.Alloc(this, GCHandleType.Weak);
TestNativeSetCallback(_nativeHandle, _callback, GCHandle.ToIntPtr(_thisHandle));
}
~TestNet()
{
Console.WriteLine("this.~TestNet()");
FreeTestNative(_nativeHandle);
_thisHandle.Free();
}
public void Foo()
{
Console.WriteLine("this.Foo() begins");
TestNativeFoo(_nativeHandle);
// This is never printed when the object is collected!
Console.WriteLine("this.Foo() ends");
// Without the following GC.KeepAlive(this) call
// in Release build the program will consistently collect
// the object in callback() and crash on next iteration
//GC.KeepAlive(this);
}
static void callback(IntPtr data)
{
Console.WriteLine("TestNet.callback() begins");
// Retrieve the weak reference to self. As soon as the istance
// of TestNet exists.
var self = (TestNet)GCHandle.FromIntPtr(data).Target;
self.callback();
// Enforce garbage collection. On release build
self = null;
GC.Collect();
GC.WaitForPendingFinalizers();
Console.WriteLine("TestNet.callback() ends");
}
void callback()
{
Console.WriteLine("this.callback()");
}
[DllImport("Dll1", CallingConvention = CallingConvention.Cdecl)]
static extern IntPtr CreateTestNative();
[DllImport("Dll1", CallingConvention = CallingConvention.Cdecl)]
static extern void FreeTestNative(IntPtr obj);
[DllImport("Dll1", CallingConvention = CallingConvention.Cdecl)]
static extern void TestNativeSetCallback(IntPtr obj, Callback callback, IntPtr data);
[DllImport("Dll1", CallingConvention = CallingConvention.Cdecl)]
static extern void TestNativeFoo(IntPtr obj);
}
}
Dll1.cpp:
#include <iostream>
extern "C" typedef void (*Callback)(void *data);
class TestNative
{
public:
void SetCallback(Callback callback1, void *data);
void Foo();
private:
Callback m_callback;
void *m_data;
};
void TestNative::SetCallback(Callback callback, void * data)
{
m_callback = callback;
m_data = data;
}
void TestNative::Foo()
{
// Foo() will never end
while (true)
{
m_callback(m_data);
}
}
extern "C"
{
__declspec(dllexport) TestNative * CreateTestNative()
{
return new TestNative();
}
__declspec(dllexport) void FreeTestNative(TestNative *obj)
{
delete obj;
}
__declspec(dllexport) void TestNativeSetCallback(TestNative *obj, Callback callback1, void * data)
{
obj->SetCallback(callback1, data);
}
__declspec(dllexport) void TestNativeFoo(TestNative *obj)
{
obj->Foo();
}
}
The output is consistently:
this.Foo() begins
TestNet.callback() begins
this.callback()
this.~TestNet()
TestNet.callback() ends
TestNet.callback() begins
System.NullReferenceException: Object reference not set to an instance of an object.
If one uncomment the GC.KeepAlive(this) call in TestNet.Foo() the program correctly never ends.
Summarizing very useful comments and research done:
1) Is GC.KeepAlive(this) always needed in a managed instance method if last instruction is a P/Invoke call using unmanaged resources hold by the instance?
Yes, if you don't want the user of the API to have last responsibility of holding a non-collectible reference for the instance of the managed object in pathological cases, look the example below. But it's not the only way: HandleRef or SafeHandle techiniques can also be used to prolong the lifetime of a managed object when doing P/Invoke Interop.
The example will subsequently call native methods through managed instances holding native resources:
using System;
using System.Diagnostics;
using System.Runtime.InteropServices;
using System.Threading;
namespace ConsoleApp1
{
class Program
{
static void Main(string[] args)
{
new Thread(delegate()
{
// Run a separate thread enforcing GC collections every second
while(true)
{
GC.Collect();
Thread.Sleep(1000);
}
}).Start();
while (true)
{
var test = new TestNet();
test.Foo();
TestNet.Dump();
}
}
}
class TestNet
{
static ManualResetEvent _closed;
static long _closeTime;
static long _fooEndTime;
IntPtr _nativeHandle;
public TestNet()
{
_closed = new ManualResetEvent(false);
_closeTime = -1;
_fooEndTime = -1;
_nativeHandle = CreateTestNative();
}
public static void Dump()
{
// Ensure the now the object will now be garbage collected
GC.Collect();
GC.WaitForPendingFinalizers();
// Wait for current object to be garbage collected
_closed.WaitOne();
Trace.Assert(_closeTime != -1);
Trace.Assert(_fooEndTime != -1);
if (_closeTime <= _fooEndTime)
Console.WriteLine("WARN: Finalize() commenced before Foo() return");
else
Console.WriteLine("Finalize() commenced after Foo() return");
}
~TestNet()
{
_closeTime = Stopwatch.GetTimestamp();
FreeTestNative(_nativeHandle);
_closed.Set();
}
public void Foo()
{
// The native implementation just sleeps for 250ms
TestNativeFoo(_nativeHandle);
// Uncomment to have all Finalize() to commence after Foo()
//GC.KeepAlive(this);
_fooEndTime = Stopwatch.GetTimestamp();
}
[DllImport("Dll1", CallingConvention = CallingConvention.Cdecl)]
static extern IntPtr CreateTestNative();
[DllImport("Dll1", CallingConvention = CallingConvention.Cdecl)]
static extern void FreeTestNative(IntPtr obj);
[DllImport("Dll1", CallingConvention = CallingConvention.Cdecl)]
static extern void TestNativeFoo(IntPtr obj);
}
}
For the native call to be always safe we expect finalizer to be called only after Foo() return. Instead we can easily enforce violations by manually invoking garbage collection in a background thread. Output follows:
Finalize() commenced after Foo() return
WARN: Finalize() commenced before Foo() return
Finalize() commenced after Foo() return
Finalize() commenced after Foo() return
Finalize() commenced after Foo() return
WARN: Finalize() commenced before Foo() return
Finalize() commenced after Foo() return
2) Where I can find documentation?
Documentation of GC.KeepAlive() provides an example very similar to the managed callback in the original question. HandleRef has also very interesting considerations about lifecycle of managed objects and Interop:
If you use platform invoke to call a managed object, and the object is not referenced elsewhere after the platform invoke call, it is possible for the garbage collector to finalize the managed object. This action releases the resource and invalidates the handle, causing the platform invoke call to fail. Wrapping a handle with HandleRef guarantees that the managed object is not garbage collected until the platform invoke call completes.
Also link[1] found by @GSerg explains when an object is eligible for collection, pointing that this reference is not in the root set, allowing it to be collected also when instance method has not returned.
3) Why is this happening only in Release build?
It's an optimization and can happen also in Debug build, with optimization enabled, as pointed by @SimonMourier. It's not enabled by default also in Debug because it could prevent debugging of variables in the current method scope, as explained in these other answers.
[1] https://devblogs.microsoft.com/oldnewthing/20100810-00/?p=13193?