How to Call Schannel Functions from .Net/C# - c#

I am trying to re-order/remove cipher suites due to compliance reasons (I want to use 256 bit AES and ephemeral keys) in .Net. However, using WCF TCP Transport Security, I cede all control over the security to Windows' TLS implementation and its preferred ciphers. I don't want to change the system-level ciphers.
I found this great Microsoft page that lists how to do it at the application level using BCryptEnumContextFunctions, BCryptAddContextFunction, BCryptRemoveContextFunction, but it's all C and I don't have enough P/Invoke experience to even know where to begin. I googled but didn't find anyone doing it.
The C++ code snippets on the MSDN page are below, and I need help converting them to .Net P/Invoke calls:
BCryptEnumContextFunctions
#include <stdio.h>
#include <windows.h>
#include <bcrypt.h>
void main()
{
HRESULT Status = ERROR_SUCCESS;
DWORD cbBuffer = 0;
PCRYPT_CONTEXT_FUNCTIONS pBuffer = NULL;
Status = BCryptEnumContextFunctions(
CRYPT_LOCAL,
L"SSL",
NCRYPT_SCHANNEL_INTERFACE,
&cbBuffer,
&pBuffer);
if(FAILED(Status))
{
printf_s("\n**** Error 0x%x returned by BCryptEnumContextFunctions\n", Status);
goto Cleanup;
}
if(pBuffer == NULL)
{
printf_s("\n**** Error pBuffer returned from BCryptEnumContextFunctions is null");
goto Cleanup;
}
printf_s("\n\n Listing Cipher Suites ");
for(UINT index = 0; index < pBuffer->cFunctions; ++index)
{
printf_s("\n%S", pBuffer->rgpszFunctions[index]);
}
Cleanup:
if (pBuffer != NULL)
{
BCryptFreeBuffer(pBuffer);
}
}
BCryptAddContextFunction
#include <stdio.h>
#include <windows.h>
#include <bcrypt.h>
void main()
{
SECURITY_STATUS Status = ERROR_SUCCESS;
LPWSTR wszCipher = (L"RSA_EXPORT1024_DES_CBC_SHA");
Status = BCryptAddContextFunction(
CRYPT_LOCAL,
L"SSL",
NCRYPT_SCHANNEL_INTERFACE,
wszCipher,
CRYPT_PRIORITY_TOP);
}
BCryptRemoveContextFunction
#include <stdio.h>
#include <windows.h>
#include <bcrypt.h>
void main()
{
SECURITY_STATUS Status = ERROR_SUCCESS;
LPWSTR wszCipher = (L"TLS_RSA_WITH_RC4_128_SHA");
Status = BCryptRemoveContextFunction(
CRYPT_LOCAL,
L"SSL",
NCRYPT_SCHANNEL_INTERFACE,
wszCipher);
}
Could someone please help me convert these to .Net so I can call them from managed code to adjust the ciphers? Thanks!
Edit:
Later last night, I tried the following in a test program (still have no idea what I'm doing in P/Invoke):
// I found this in bcrypt.h
const uint CRYPT_LOCAL = 0x00000001;
// I can't find this anywhere in Microsoft's headers in my SDK,
// but I found some random .c file with it in there. No idea
// what this constant actually is according to Microsoft
const uint NCRYPT_SCHANNEL_INTERFACE = 0x00010002;
public static void DoStuff()
{
PCRYPT_CONTEXT_FUNCTIONS pBuffer = new PCRYPT_CONTEXT_FUNCTIONS();
pBuffer.rgpszFunctions = String.Empty.PadRight(1500);
uint cbBuffer = (uint)Marshal.SizeOf(typeof(PCRYPT_CONTEXT_FUNCTIONS));
uint Status = BCryptEnumContextFunctions(
CRYPT_LOCAL,
"SSL",
NCRYPT_SCHANNEL_INTERFACE,
ref cbBuffer,
ref pBuffer);
Console.WriteLine(Status);
Console.WriteLine(pBuffer);
Console.WriteLine(cbBuffer);
Console.WriteLine(pBuffer.cFunctions);
Console.WriteLine(pBuffer.rgpszFunctions);
}
/*
typedef struct _CRYPT_CONTEXT_FUNCTIONS {
ULONG cFunctions;
PWSTR rgpszFunctions;
} CRYPT_CONTEXT_FUNCTIONS, *PCRYPT_CONTEXT_FUNCTIONS;
*/
[StructLayout(LayoutKind.Sequential, CharSet = CharSet.Unicode)]
public struct PCRYPT_CONTEXT_FUNCTIONS
{
public uint cFunctions;
public string rgpszFunctions;
}
/*
NTSTATUS WINAPI BCryptEnumContextFunctions(
ULONG dwTable,
LPCWSTR pszContext,
ULONG dwInterface,
ULONG *pcbBuffer,
PCRYPT_CONTEXT_FUNCTIONS *ppBuffer
);
*/
[DllImport("Bcrypt.dll", CharSet = CharSet.Unicode, SetLastError = true)]
static extern uint BCryptEnumContextFunctions(uint dwTable, string pszContext, uint dwInterface, ref uint pcbBuffer, ref PCRYPT_CONTEXT_FUNCTIONS ppBuffer);
The output right now of the only method I got even this far with is:
0
MyClass+PCRYPT_CONTEXT_FUNCTIONS
2400
8934576
[1500 spaces that I initialized rgpszFunctions with, not the cipher functions]

This is a poorly documented library. For instance, the declaration of CRYPT_CONTEXT_FUNCTION_PROVIDERS is in fact:
typedef struct _CRYPT_CONTEXT_FUNCTION_PROVIDERS
{
ULONG cProviders;
PWSTR *rgpszProviders;
}
CRYPT_CONTEXT_FUNCTION_PROVIDERS, *PCRYPT_CONTEXT_FUNCTION_PROVIDERS;
This is lifted directly from bcrypt.h.
Anyway, here's a translation of the C++ code for you:
using System;
using System.Text;
using System.Runtime.InteropServices;
namespace ConsoleApplication1
{
class Program
{
[DllImport("Bcrypt.dll", CharSet = CharSet.Unicode)]
static extern uint BCryptEnumContextFunctions(uint dwTable, string pszContext, uint dwInterface, ref uint pcbBuffer, ref IntPtr ppBuffer);
[DllImport("Bcrypt.dll")]
static extern void BCryptFreeBuffer(IntPtr pvBuffer);
[DllImport("Bcrypt.dll", CharSet = CharSet.Unicode)]
static extern uint BCryptAddContextFunction(uint dwTable, string pszContext, uint dwInterface, string pszFunction, uint dwPosition);
[DllImport("Bcrypt.dll", CharSet = CharSet.Unicode)]
static extern uint BCryptRemoveContextFunction(uint dwTable, string pszContext, uint dwInterface, string pszFunction);
[StructLayout(LayoutKind.Sequential)]
public struct CRYPT_CONTEXT_FUNCTIONS
{
public uint cFunctions;
public IntPtr rgpszFunctions;
}
const uint CRYPT_LOCAL = 0x00000001;
const uint NCRYPT_SCHANNEL_INTERFACE = 0x00010002;
const uint CRYPT_PRIORITY_TOP = 0x00000000;
const uint CRYPT_PRIORITY_BOTTOM = 0xFFFFFFFF;
public static void DoStuff()
{
uint cbBuffer = 0;
IntPtr ppBuffer = IntPtr.Zero;
uint Status = BCryptEnumContextFunctions(
CRYPT_LOCAL,
"SSL",
NCRYPT_SCHANNEL_INTERFACE,
ref cbBuffer,
ref ppBuffer);
if (Status == 0)
{
CRYPT_CONTEXT_FUNCTIONS functions = (CRYPT_CONTEXT_FUNCTIONS)Marshal.PtrToStructure(ppBuffer, typeof(CRYPT_CONTEXT_FUNCTIONS));
Console.WriteLine(functions.cFunctions);
IntPtr pStr = functions.rgpszFunctions;
for (int i = 0; i < functions.cFunctions; i++)
{
Console.WriteLine(Marshal.PtrToStringUni(Marshal.ReadIntPtr(pStr)));
pStr += IntPtr.Size;
}
BCryptFreeBuffer(ppBuffer);
}
}
static void Main(string[] args)
{
DoStuff();
Console.ReadLine();
}
}
}
On my machine the output is:
30
TLS_RSA_WITH_AES_128_CBC_SHA256
TLS_RSA_WITH_AES_128_CBC_SHA
TLS_RSA_WITH_AES_256_CBC_SHA256
TLS_RSA_WITH_AES_256_CBC_SHA
TLS_RSA_WITH_RC4_128_SHA
TLS_RSA_WITH_3DES_EDE_CBC_SHA
TLS_ECDHE_RSA_WITH_AES_128_CBC_SHA256_P256
TLS_ECDHE_RSA_WITH_AES_128_CBC_SHA256_P384
TLS_ECDHE_RSA_WITH_AES_128_CBC_SHA_P256
TLS_ECDHE_RSA_WITH_AES_128_CBC_SHA_P384
TLS_ECDHE_RSA_WITH_AES_256_CBC_SHA_P256
TLS_ECDHE_RSA_WITH_AES_256_CBC_SHA_P384
TLS_ECDHE_ECDSA_WITH_AES_128_GCM_SHA256_P256
TLS_ECDHE_ECDSA_WITH_AES_128_CBC_SHA256_P256
TLS_ECDHE_ECDSA_WITH_AES_256_GCM_SHA384_P384
TLS_ECDHE_ECDSA_WITH_AES_256_CBC_SHA384_P384
TLS_ECDHE_ECDSA_WITH_AES_128_CBC_SHA_P256
TLS_ECDHE_ECDSA_WITH_AES_128_CBC_SHA_P384
TLS_ECDHE_ECDSA_WITH_AES_256_CBC_SHA_P256
TLS_ECDHE_ECDSA_WITH_AES_256_CBC_SHA_P384
TLS_DHE_DSS_WITH_AES_128_CBC_SHA256
TLS_DHE_DSS_WITH_AES_128_CBC_SHA
TLS_DHE_DSS_WITH_AES_256_CBC_SHA256
TLS_DHE_DSS_WITH_AES_256_CBC_SHA
TLS_DHE_DSS_WITH_3DES_EDE_CBC_SHA
TLS_RSA_WITH_RC4_128_MD5
SSL_CK_RC4_128_WITH_MD5
SSL_CK_DES_192_EDE3_CBC_WITH_MD5
TLS_RSA_WITH_NULL_SHA256
TLS_RSA_WITH_NULL_SHA
I urge you to get to grips with some simple C++ to make progress. Take the example from MSDN and compile it. Run it under the debugger and get to know it. Use Visual Studio to locate definitions and declarations. For instance, Visual Studio took me straight to the definition of NCRYPT_SCHANNEL_INTERFACE.

Related

P/Invoke: OREnumValue returning ERROR_INVALID_PARAMETER

I'm in the process of writing a basic .NET wrapper for the unmanaged Offline Registry Library. Currently, I'm working to implement a class, OfflineRegistryKey, which is modelled on Microsoft.Win32.RegistryKey but intended for use with offline Registry Hives rather than the live system Registry.
While trying to implement an equivalent to the GetValueNames() method, which enumerates values within a Registry key, I ran into some trouble trying to P/Invoke the unmanaged function OREnumValue from Offreg.dll. The only parameter I'm interested in in this case is the name of the value, or lpValueName.
My P/Invoke code for OREnumValue:
[DllImport("Offreg.dll", CharSet = CharSet.Auto, SetLastError = false)]
public static extern uint OREnumValue(
SafeRegistryHandle Handle,
uint dwIndex,
IntPtr lpValueName,
ref uint lpcValueName,
IntPtr lpType,
IntPtr lpData,
ref uint lpcbData);
I call the function like this in my code:
public class OfflineRegistryKey : IDisposable
{
private readonly SafeRegistryHandle hKey;
public string[] GetValueNames()
{
uint dwIndex = 0;
while (true)
{
uint lpcValueName = 0;
uint lpcbData = 0;
uint returnValue = OREnumValue(hKey, dwIndex, IntPtr.Zero, ref lpcValueName, IntPtr.Zero, IntPtr.Zero, ref lpcbData);
if (returnValue == 0x0103) // ERROR_NO_MORE_ITEMS
{
break;
}
else if (returnValue == 0x00EA) // ERROR_MORE_DATA
{
// Do stuff...
}
else
{
throw new Win32Exception(returnValue);
}
dwIndex++;
}
}
}
Ideally, The call to OREnumKey would return 0xEA (ERROR_MORE_DATA), after which it would be called again until it reaches the end index and returns 0x103 (ERROR_NO_MORE_ITEMS).
The return value I'm getting (which throws an exception at runtime and halts execution on the first iteration of the loop) is 0x57 (ERROR_INVALID_PARAMETER), which leads me to believe that I've somehow screwed up the P/Invoke call. I would very much appreciate if someone with a greater knowledge of P/Invoke and the Windows API could provide me with some guidance.
Thanks!
Your P/Invoke declaration is not correct, because lpValueName does not accept NULL according to the documentation.
It's also defined as Unicode, so you should specify that.
[DllImport("Offreg.dll", CharSet = CharSet.Unicode, SetLastError = false)]
public static extern uint OREnumValue(
SafeRegistryHandle Handle,
uint dwIndex,
[Out, MarshalAs(UnmanagedType.LPWStr)] StringBuilder lpValueName,
ref uint lpcValueName,
IntPtr lpType,
IntPtr lpData,
ref uint lpcbData);
Then you simply pass a pre-assigned buffer. Theoretically it could be 16383 characters, you could use ORQueryInfoKey to get the max size, or just allocate the whole thing. I suggest you allocate it only once, outside the loop.
public string[] GetValueNames()
{
const int ERROR_SUCCESS = 0;
const int ERROR_NO_MORE_ITEMS = 0x103;
const int ERROR_MORE_DATA = 0x00EA;
uint dwIndex = 0;
string lpValueName = new StringBuilder(16384);
while (true)
{
uint lpcValueName = 0;
uint lpcbData = 0;
uint returnValue = OREnumValue(hKey, dwIndex, IntPtr.Zero, ref lpcValueName, IntPtr.Zero, IntPtr.Zero, ref lpcbData);
if (returnValue == ERROR_NO_MORE_ITEMS)
{
break;
}
else if (returnValue == ERROR_MORE_DATA)
{
// Do stuff...
}
else if (returnValue != ERROR_SUCCESS)
{
throw new Win32Exception(returnValue);
}
dwIndex++;
}
}

Marshaling an IntPtr[] inside a struct causes midiStream functions to bug out, but unrolling the array to a bunch of fields works

I'm trying to use the Windows multimedia MIDI functions from C#. Specifically:
MMRESULT midiOutPrepareHeader( HMIDIOUT hmo, LPMIDIHDR lpMidiOutHdr, UINT cbMidiOutHdr );
MMRESULT midiOutUnprepareHeader( HMIDIOUT hmo, LPMIDIHDR lpMidiOutHdr, UINT cbMidiOutHdr );
MMRESULT midiStreamOut( HMIDISTRM hMidiStream, LPMIDIHDR lpMidiHdr, UINT cbMidiHdr );
MMRESULT midiStreamRestart( HMIDISTRM hms );
/* MIDI data block header */
typedef struct midihdr_tag {
LPSTR lpData; /* pointer to locked data block */
DWORD dwBufferLength; /* length of data in data block */
DWORD dwBytesRecorded; /* used for input only */
DWORD_PTR dwUser; /* for client's use */
DWORD dwFlags; /* assorted flags (see defines) */
struct midihdr_tag far *lpNext; /* reserved for driver */
DWORD_PTR reserved; /* reserved for driver */
#if (WINVER >= 0x0400)
DWORD dwOffset; /* Callback offset into buffer */
DWORD_PTR dwReserved[8]; /* Reserved for MMSYSTEM */
#endif
} MIDIHDR, *PMIDIHDR, NEAR *NPMIDIHDR, FAR *LPMIDIHDR;
From a C program, I can successfully use these functions, by doing the following:
HMIDISTRM hms;
midiStreamOpen(&hms, ...);
MIDIHDR hdr;
hdr.this = that; ...
midiStreamRestart(hms);
midiOutPrepareHeader(hms, &hdr, sizeof(MIDIHDR)); // sizeof(MIDIHDR) == 64
midiStreamOut(hms, &hdr, sizeof(MIDIHDR));
// wait for an event that is set from the midi callback when the playback has finished
WaitForSingleObject(...);
midiOutUnprepareHeader(hms, &hdr, sizeof(MIDIHDR));
The above calling sequence works and produces no errors (error checks have been omitted for readability).
For the purpose of using those in C#, I have created some P/Invoke code:
[DllImport("winmm.dll")]
public static extern int midiOutPrepareHeader(IntPtr handle, ref MidiHeader header, uint headerSize);
[DllImport("winmm.dll")]
public static extern int midiOutUnprepareHeader(IntPtr handle, ref MidiHeader header, uint headerSize);
[DllImport("winmm.dll")]
public static extern int midiStreamOut(IntPtr handle, ref MidiHeader header, uint headerSize);
[DllImport("winmm.dll")]
public static extern int midiStreamRestart(IntPtr handle);
[StructLayout(LayoutKind.Sequential)]
public struct MidiHeader
{
public IntPtr Data;
public uint BufferLength;
public uint BytesRecorded;
public IntPtr UserData;
public uint Flags;
public IntPtr Next;
public IntPtr Reserved;
public uint Offset;
//[MarshalAs(UnmanagedType.ByValArray, SizeConst = 8)]
//public IntPtr[] Reserved2;
public IntPtr Reserved0;
public IntPtr Reserved1;
public IntPtr Reserved2;
public IntPtr Reserved3;
public IntPtr Reserved4;
public IntPtr Reserved5;
public IntPtr Reserved6;
public IntPtr Reserved7;
}
The call sequence is the same as in C:
var hdr = new MidiHeader();
hdr.this = that;
midiStreamRestart(handle);
midiOutPrepareHeader(handle, ref header, headerSize); // headerSize == 64
midiStreamOut(handle, ref header, headerSize);
mre.WaitOne(); // wait until the midi playback has finished.
midiOutUnprepareHeader(handle, ref header, headerSize);
MIDI output works and the code produces no error (error checks are again omitted).
As soon as I uncomment the two lines with the array in MidiHeader, and instead remove the Reserved0 to Reserved7 fields, it doesn't work anymore. What happens is the following:
Everything is normal until and including midiStreamOut. I can hear the midi output. The playback length is correct. However, the event callback is never called when the playback ends.
At this point the value of MidiHeader.Flags is 0xe, indicating that the stream is still playing (even though the callback has been notified with the message that playback has finished). The value of MidiHeader.Flags should be 9, indicating that the stream has finished playing.
The call to midiOutUnprepareHeader fails with the error code 0x41 ("Cannot perform this operation while media data is still playing. Reset the device, or wait until the data is finished playing."). Note that resetting the device, as suggested in the error message, does in fact not fix the problem (neither does waiting or trying it multiple times).
Another variant that works correctly is if I use IntPtr instead of ref MidiHeader in the signatures of the C# declarations, and then manually allocate unmanaged memory, copying my MidiHeader structure onto that memory, and then using the allocated memory to call the functions.
Furthermore, I tried decreasing the size I'm passing to the headerSize argument to 32. Since the fields are reserved (and, in fact, didn't exist in previous version of the Windows API), they seemed to have no particular purpose anyway. However, this does not fix the problem, even though Windows should not even know that the array exists, and therefore it should not do anything. Commenting out the array entirely yet again fixes the problem (i.e., both the array as well as the 8 Reserved* fields are commented out, and the headerSize is 32).
This hints to me that the IntPtr[] Reserved2 cannot be marshaled correctly, and attempting to do so is corrupting other values. To verify that, I have created a Platform Invoke test project:
WIN32PROJECT1_API void __stdcall test_function(struct test_struct_t *s)
{
printf("%u %u %u %u %u %u %u %u\n", s->test0, s->test1, s->test2, s->test3, s->test4, s->test5, s->test6, s->test7);
for (int i = 0; i < sizeof(s->pointer_array) / sizeof(s->pointer_array[0]); ++i)
{
printf("%u ", ((uint32_t)s->pointer_array[i]) >> 16);
}
printf("\n");
}
typedef int32_t *test_ptr;
struct test_struct_t
{
test_ptr test0;
uint32_t test1;
uint32_t test2;
test_ptr test3;
uint32_t test4;
test_ptr test5;
uint32_t test6;
uint32_t test7;
test_ptr pointer_array[8];
};
Which is called from C#:
[StructLayout(LayoutKind.Sequential)]
struct TestStruct
{
public IntPtr test0;
public uint test1;
public uint test2;
public IntPtr test3;
public uint test4;
public IntPtr test5;
public uint test6;
public uint test7;
[MarshalAs(UnmanagedType.ByValArray, SizeConst = 8)]
public IntPtr[] pointer_array;
}
[DllImport("Win32Project1.dll")]
static extern void test_function(ref TestStruct s);
static void Main(string[] args)
{
TestStruct s = new TestStruct();
s.test0 = IntPtr.Zero;
s.test1 = 1;
s.test2 = 2;
s.test3 = IntPtr.Add(IntPtr.Zero, 3);
s.test4 = 4;
s.test5 = IntPtr.Add(IntPtr.Zero, 5);
s.test6 = 6;
s.test7 = 7;
s.pointer_array = new IntPtr[8];
for (int i = 0; i < s.pointer_array.Length; ++i)
{
s.pointer_array[i] = IntPtr.Add(IntPtr.Zero, i << 16);
}
test_function(ref s);
Console.ReadLine();
}
And the output is as expected, hence the marshaling of the IntPtr[] pointer_array worked in this program.
I am aware that not using SafeHandle is suboptimal, however, when using that, the behavior of the MIDI functions when using the array is even weirder, so I chose to maybe tackle one problem at a time.
Why does using IntPtr[] Reserved2 cause an error?
Here is some more code to produce a complete example:
C Code
/*
* example9.c
*
* Created on: Dec 21, 2011
* Author: David J. Rager
* Email: djrager#fourthwoods.com
*
* This code is hereby released into the public domain per the Creative Commons
* Public Domain dedication.
*
* http://http://creativecommons.org/publicdomain/zero/1.0/
*/
#include <windows.h>
#include <mmsystem.h>
#include <stdio.h>
HANDLE event;
static void CALLBACK example9_callback(HMIDIOUT out, UINT msg, DWORD dwInstance, DWORD dwParam1, DWORD dwParam2)
{
switch (msg)
{
case MOM_DONE:
SetEvent(event);
break;
case MOM_POSITIONCB:
case MOM_OPEN:
case MOM_CLOSE:
break;
}
}
int main()
{
unsigned int streambufsize = 24;
char* streambuf = NULL;
HMIDISTRM out;
MIDIPROPTIMEDIV prop;
MIDIHDR mhdr;
unsigned int device = 0;
streambuf = (char*)malloc(streambufsize);
if (streambuf == NULL)
goto error2;
memset(streambuf, 0, streambufsize);
if ((event = CreateEvent(0, FALSE, FALSE, 0)) == NULL)
goto error3;
memset(&mhdr, 0, sizeof(mhdr));
mhdr.lpData = streambuf;
mhdr.dwBufferLength = mhdr.dwBytesRecorded = streambufsize;
mhdr.dwFlags = 0;
// flags and event code
mhdr.lpData[8] = (char)0x90;
mhdr.lpData[9] = 63;
mhdr.lpData[10] = 0x55;
mhdr.lpData[11] = 0;
// next event
mhdr.lpData[12] = 96; // delta time?
mhdr.lpData[20] = (char)0x80;
mhdr.lpData[21] = 63;
mhdr.lpData[22] = 0x55;
mhdr.lpData[23] = 0;
if (midiStreamOpen(&out, &device, 1, (DWORD)example9_callback, 0, CALLBACK_FUNCTION) != MMSYSERR_NOERROR)
goto error4;
//printf("sizeof midiheader = %d\n", sizeof(MIDIHDR));
if (midiOutPrepareHeader((HMIDIOUT)out, &mhdr, sizeof(MIDIHDR)) != MMSYSERR_NOERROR)
goto error5;
if (midiStreamRestart(out) != MMSYSERR_NOERROR)
goto error6;
if (midiStreamOut(out, &mhdr, sizeof(MIDIHDR)) != MMSYSERR_NOERROR)
goto error7;
WaitForSingleObject(event, INFINITE);
error7:
//midiOutReset((HMIDIOUT)out);
error6:
MMRESULT blah = midiOutUnprepareHeader((HMIDIOUT)out, &mhdr, sizeof(MIDIHDR));
printf("stuff: %d\n", blah);
error5:
midiStreamClose(out);
error4:
CloseHandle(event);
error3:
free(streambuf);
error2:
//free(tracks);
error1:
//free(hdr);
return(0);
}
C# Code
using System;
using System.Collections.Generic;
using System.Diagnostics;
using System.Linq;
using System.Runtime.InteropServices;
using System.Text;
using System.Threading;
namespace MidiOutTest
{
class Program
{
[DllImport("winmm.dll")]
public static extern int midiStreamOpen(out IntPtr handle, ref uint deviceId, uint cMidi, MidiCallback callback, IntPtr userData, uint flags);
[DllImport("winmm.dll")]
public static extern int midiStreamOut(IntPtr handle, ref MidiHeader header, uint headerSize);
[DllImport("winmm.dll")]
public static extern int midiStreamRestart(IntPtr handle);
[DllImport("winmm.dll")]
public static extern int midiOutPrepareHeader(IntPtr handle, ref MidiHeader header, uint headerSize);
[DllImport("winmm.dll")]
public static extern int midiOutUnprepareHeader(IntPtr handle, ref MidiHeader header, uint headerSize);
[DllImport("winmm.dll", CharSet = CharSet.Unicode)]
public static extern int midiOutGetErrorText(int mmsyserr, StringBuilder errMsg, int capacity);
[DllImport("winmm.dll")]
public static extern int midiStreamClose(IntPtr handle);
public delegate void MidiCallback(IntPtr handle, uint msg, IntPtr instance, IntPtr param1, IntPtr param2);
private static readonly ManualResetEvent mre = new ManualResetEvent(false);
private static void TestMidiCallback(IntPtr handle, uint msg, IntPtr instance, IntPtr param1, IntPtr param2)
{
Debug.WriteLine(msg.ToString());
if (msg == MOM_DONE)
{
Debug.WriteLine("MOM_DONE");
mre.Set();
}
}
public const uint MOM_DONE = 0x3C9;
public const int MMSYSERR_NOERROR = 0;
public const int MAXERRORLENGTH = 256;
public const uint CALLBACK_FUNCTION = 0x30000;
public const uint MidiHeaderSize = 64;
public static void CheckMidiOutMmsyserr(int mmsyserr)
{
if (mmsyserr != MMSYSERR_NOERROR)
{
var sb = new StringBuilder(MAXERRORLENGTH);
var errorResult = midiOutGetErrorText(mmsyserr, sb, sb.Capacity);
if (errorResult != MMSYSERR_NOERROR)
{
throw new /*Midi*/Exception("An error occurred and there was another error while attempting to retrieve the error message."/*, mmsyserr*/);
}
throw new /*Midi*/Exception(sb.ToString()/*, mmsyserr*/);
}
}
static void Main(string[] args)
{
IntPtr handle;
uint deviceId = 0;
CheckMidiOutMmsyserr(midiStreamOpen(out handle, ref deviceId, 1, TestMidiCallback, IntPtr.Zero, CALLBACK_FUNCTION));
try
{
var bytes = new byte[24];
IntPtr buffer = Marshal.AllocHGlobal(bytes.Length);
try
{
MidiHeader header = new MidiHeader();
header.Data = buffer;
header.BufferLength = 24;
header.BytesRecorded = 24;
header.UserData = IntPtr.Zero;
header.Flags = 0;
header.Next = IntPtr.Zero;
header.Reserved = IntPtr.Zero;
header.Offset = 0;
#warning uncomment if using array
//header.Reserved2 = new IntPtr[8];
// flags and event code
bytes[8] = 0x90;
bytes[9] = 63;
bytes[10] = 0x55;
bytes[11] = 0;
// next event
bytes[12] = 96;
bytes[20] = 0x80;
bytes[21] = 63;
bytes[22] = 0x55;
bytes[23] = 0;
Marshal.Copy(bytes, 0, buffer, bytes.Length);
CheckMidiOutMmsyserr(midiStreamRestart(handle));
CheckMidiOutMmsyserr(midiOutPrepareHeader(handle, ref header, MidiHeaderSize));
CheckMidiOutMmsyserr(midiStreamOut(handle, ref header, MidiHeaderSize));
mre.WaitOne();
CheckMidiOutMmsyserr(midiOutUnprepareHeader(handle, ref header, MidiHeaderSize));
}
finally
{
Marshal.FreeHGlobal(buffer);
}
}
finally
{
midiStreamClose(handle);
}
}
}
[StructLayout(LayoutKind.Sequential)]
public struct MidiHeader
{
public IntPtr Data;
public uint BufferLength;
public uint BytesRecorded;
public IntPtr UserData;
public uint Flags;
public IntPtr Next;
public IntPtr Reserved;
public uint Offset;
#if false
[MarshalAs(UnmanagedType.ByValArray, SizeConst = 8)]
public IntPtr[] Reserved2;
#else
public IntPtr Reserved0;
public IntPtr Reserved1;
public IntPtr Reserved2;
public IntPtr Reserved3;
public IntPtr Reserved4;
public IntPtr Reserved5;
public IntPtr Reserved6;
public IntPtr Reserved7;
#endif
}
}
From the documentation of midiOutPrepareHeader:
Before you pass a MIDI data block to a device driver, you must prepare the buffer by passing it to the midiOutPrepareHeader function. After the header has been prepared, do not modify the buffer. After the driver is done using the buffer, call the midiOutUnprepareHeader function.
You are not adhering to this. The marshaller creates a temporary native version of your struct which lives for the duration of the call to midiOutPrepareHeader. Once midiOutPrepareHeader returns, the temporary native struct is destroyed. But the MIDI code still has a reference to it. And that's the key point, the MIDI code holds a reference to your struct and needs to be able to access it.
The version with the separately written fields works because that struct is blittable. And so the p/invoke marshaller optimises the call by pinning the managed structure which is binary compatible with the native structure. There's still a window of opportunity for the GC to relocate the struct before you call midiOutUnprepareHeader but it seems that you've not been caught out by that yet. Were you to persevere with the bittable struct you would need to pin it until you called midiOutUnprepareHeader.
So, the bottom line is that you need to provide a structure that lives until you call midiOutUnprepareHeader. Personally, I suggest that you use Marshal.AllocHGlobal, Marshal.StructureToPtr, and then Marshal.FreeHGlobal once midiOutUnprepareHeader returns. And obviously switch the parameters from ref MidiHeader to IntPtr.
I don't think I need to show you any code because it is clear from your question that you know how to do this stuff. Indeed the solution I propose is one that you've already tried and observed work. But now you know why!

C# call C++ DLL passing pointer-to-pointer argument

Could you guys please help me solve the following issue?
I have a C++ function dll, and it will be called by another C# application.
One of the functions I needed is as follow:
struct DataStruct
{
unsigned char* data;
int len;
};
DLLAPI int API_ReadFile(const wchar_t* filename, DataStruct** outData);
I wrote the following code in C#:
class CS_DataStruct
{
public byte[] data;
public int len;
}
[DllImport("ReadFile.dll", CallingConvention = CallingConvention.Cdecl, CharSet = CharSet.Unicode)]
private static extern int API_ReadFile([MarshalAs(UnmanagedType.LPWStr)]string filename, ref CS_DataStruct data);
Unfortunately, the above code is not working... I guess that is due to the C++ func takes a pointer-to-pointer of DataStruct, while I just passed a reference of CS_DataStruct in.
May I know how can I pass a pointer-to-pointer to the C++ func? If it is not possible, is there any workaround? (the C++ API is fixed, so changing API to pointer is not possible)
Edit:
Memory of DataStruct will be allocated by c++ function. Before that, I have no idea how large the data array should be.
(Thanks for the comments below)
I used the following test implementation:
int API_ReadFile(const wchar_t* filename, DataStruct** outData)
{
*outData = new DataStruct();
(*outData)->data = (unsigned char*)_strdup("hello");
(*outData)->len = 5;
return 0;
}
void API_Free(DataStruct** pp)
{
free((*pp)->data);
delete *pp;
*pp = NULL;
}
The C# code to access those functions are as follows:
[StructLayout(LayoutKind.Sequential)]
struct DataStruct
{
public IntPtr data;
public int len;
};
[DllImport("ReadFile.dll", CallingConvention = CallingConvention.Cdecl, CharSet = CharSet.Unicode)]
unsafe private static extern int API_ReadFile([MarshalAs(UnmanagedType.LPWStr)]string filename, DataStruct** outData);
[DllImport("ReadFile.dll", CallingConvention = CallingConvention.Cdecl)]
unsafe private static extern void API_Free(DataStruct** handle);
unsafe static int ReadFile(string filename, out byte[] buffer)
{
DataStruct* outData;
int result = API_ReadFile(filename, &outData);
buffer = new byte[outData->len];
Marshal.Copy((IntPtr)outData->data, buffer, 0, outData->len);
API_Free(&outData);
return result;
}
static void Main(string[] args)
{
byte[] buffer;
ReadFile("test.txt", out buffer);
foreach (byte ch in buffer)
{
Console.Write("{0} ", ch);
}
Console.Write("\n");
}
The data is now transferred to buffer safely, and there should be no memory leaks. I wish it would help.
It isn't necessary to use unsafe to pass a pointer to an array from a DLL. Here is an example (see the 'results' parameter). The key is to use the ref attribute. It also shows how to pass several other types of data.
As defined in C++/C:
#ifdef __cplusplus
extern "C" {
#endif
#ifdef BUILDING_DLL
#define DLLCALL __declspec(dllexport)
#else
#define DLLCALL __declspec(dllimport)
#endif
static const int DataLength = 10;
static const int StrLen = 16;
static const int MaxResults = 30;
enum Status { on = 0, off = 1 };
struct Result {
char name[StrLen]; //!< Up to StrLen-1 char null-terminated name
float location;
Status status;
};
/**
* Analyze Data
* #param data [in] array of doubles
* #param dataLength [in] number of floats in data
* #param weight [in]
* #param status [in] enum with data status
* #param results [out] array of MaxResults (pre-allocated) DLLResult structs.
* Up to MaxResults results will be returned.
* #param nResults [out] the actual number of results being returned.
*/
void DLLCALL __stdcall analyzeData(
const double *data, int dataLength, float weight, Status status, Result **results, int *nResults);
#ifdef __cplusplus
}
#endif
As used in C#:
private const int DataLength = 10;
private const int StrLen = 16;
private const int MaxThreatPeaks = 30;
public enum Status { on = 0, off = 1 };
[StructLayout(LayoutKind.Sequential, CharSet = CharSet.Ansi)]
public struct Result
{
[MarshalAs(UnmanagedType.ByValTStr, SizeConst = StrLen)] public string name; //!< Up to StrLen-1 char null-terminated name
public float location;
public Status status;
}
[DllImport("dllname.dll", CallingConvention = CallingConvention.StdCall, EntryPoint = "analyzeData#32")] // "#32" is only used in the 32-bit version.
public static extern void analyzeData(
double[] data,
int dataLength,
float weight,
Status status,
[MarshalAs(UnmanagedType.LPArray, SizeConst = MaxResults)] ref Result[] results,
out int nResults
);
Without the extern "C" part, the C++ compiler would mangle the export name in a compiler dependent way. I noticed that the EntryPoint / Exported function name matches the function name exactly in a 64-bit DLL, but has an appended '#32' (the number may vary) when compiled into a 32-bit DLL. Run dumpbin /exports dllname.dll to find the exported name for sure. In some cases you may also need to use the DLLImport parameter ExactSpelling = true. Note that this function is declared __stdcall. If it were not specified, it would be __cdecl and you'd need CallingConvention.Cdecl.
Here is how it might be used in C#:
Status status = Status.on;
double[] data = { -0.034, -0.05, -0.039, -0.034, -0.057, -0.084, -0.105, -0.146, -0.174, -0.167};
Result[] results = new Result[MaxResults];
int nResults = -1; // just to see that it changes (input value is ignored)
analyzeData(data, DataLength, 1.0f, status, ref results, out nResults);
If you do call native code, make sure your structs are alligned in the memory. CLR does not guarantee alignment unless you push it.
Try
[StructLayout(LayoutKind.Explicit)]
struct DataStruct
{
string data;
int len;
};
More info:
http://www.developerfusion.com/article/84519/mastering-structs-in-c/

Illegal parameter in converting C SDK to C#

I am trying to conver a C SDK to C# and am running into a "Illegal Parameter" error on the conversion of a C function.
The details of the C SDK function are listed below
#ifndef LLONG
#ifdef WIN32
#define LLONG LONG
#else //WIN64
#define LLONG INT64
#endif
#endif
#ifndef CLIENT_API
#define CLIENT_API __declspec(dllexport)
#endif
#else
#ifndef CLIENT_API
#define CLIENT_API __declspec(dllimport)
#endif
#endif
#define CALLBACK __stdcall
#define CALL_METHOD __stdcall //__cdecl
// Configuration type,corresponding to CLIENT_GetDevConfig and CLIENT_SetDevConfig
#define DH_DEV_DEVICECFG 0x0001 // Device property setup
#define DH_DEV_NETCFG 0x0002 // Network setup
#define DH_DEV_CHANNELCFG 0x0003 // Video channel setup
#define DH_DEV_PREVIEWCFG 0x0004 // Preview parameter setup
#define DH_DEV_RECORDCFG 0x0005 // Record setup
#define DH_DEV_COMMCFG 0x0006 // COM property setup
#define DH_DEV_ALARMCFG 0x0007 // Alarm property setup
#define DH_DEV_TIMECFG 0x0008 // DVR time setup
#define DH_DEV_TALKCFG 0x0009 // Audio talk parameter setup
#define DH_DEV_AUTOMTCFG 0x000A // Auto matrix setup
#define DH_DEV_VEDIO_MARTIX 0x000B // Local matrix control strategy setup
#define DH_DEV_MULTI_DDNS 0x000C // Multiple ddns setup
#define DH_DEV_SNAP_CFG 0x000D // Snapshot corresponding setup
#define DH_DEV_WEB_URL_CFG 0x000E // HTTP path setup
#define DH_DEV_FTP_PROTO_CFG 0x000F // FTP upload setup
#define DH_DEV_INTERVIDEO_CFG 0x0010 // Plaform embedded setup. Now the channel parameter represents the platform type.
// Search configuration information
CLIENT_API BOOL CALL_METHOD CLIENT_GetDevConfig(LLONG lLoginID, DWORD dwCommand, LONG lChannel, LPVOID lpOutBuffer, DWORD dwOutBufferSize, LPDWORD lpBytesReturned,int waittime=500);
The C# info is as follows:
// [DllImport("dhnetsdk.dll", CallingConvention = CallingConvention.StdCall)]
[DllImport("dhnetsdk.dll", CallingConvention = CallingConvention.Cdecl)]
// [DllImport("dhnetsdk.dll")]
public static extern bool CLIENT_GetDevConfig(long lLoginID,
uint dwCommand,
long lChannel,
IntPtr lpBuffer,
uint dwOutBufferSize,
uint lpBytesReturned,
int waittime = 500);
And I am calling the method as follows:
int t = 500;
uint BytesReturned = 0;
uint c = 8;
var lpOutBuffer = Marshal.AllocHGlobal(Marshal.SizeOf(typeof(NET_TIME)));
if (CLIENT_GetDevConfig(lLogin, c, 0, lpOutBuffer, (uint)Marshal.SizeOf(typeof(NET_TIME)), BytesReturned, t) == false)
{
Console.WriteLine("GetDevConfig FAILED");
}
[StructLayout(LayoutKind.Sequential)]
public struct NET_TIME
{
// [FieldOffset(0)]
uint dwYear;
// [FieldOffset(4)]
uint dwMonth;
// [FieldOffset(4)]
uint dwDay;
// [FieldOffset(4)]
uint dwHour;
// [FieldOffset(4)]
uint dwMinute;
// [FieldOffset(4)]
uint dwSecond;
}
I am positive the lLogin is correct since I successfully logged into the device using it.
But when I check GetLastError immediately after the call to GetDevConfig fails, it indicates a illegal parameter. So, can anybody point out the illegal parameter in the above code?
The following is my C# code with the illegal parameter issues...
using System;
using System.Runtime.InteropServices;
class PlatformInvokeTest
{
static public int lLogin;
public delegate void fDisConnect(long lLoginID, IntPtr pchDVRIP, long nDVRPort, uint dwUser);
public delegate void fHaveReConnect(long lLoginID, IntPtr pchDVRIP, long nDVRPort, uint dwUser);
[DllImport("dhnetsdk.dll", CallingConvention = CallingConvention.Cdecl)]
public static extern bool CLIENT_Init(fDisConnect cbDisConnect, uint dwUser);
[DllImport("dhnetsdk.dll", CallingConvention = CallingConvention.Cdecl)]
public static extern void CLIENT_SetAutoReconnect(fHaveReConnect cbHaveReconnt, uint dwUser);
[DllImport("dhnetsdk.dll", CallingConvention = CallingConvention.Cdecl)]
public static extern int CLIENT_Login(string pchDVRIP, ushort wDVRPort, string pchUserName, string pchPassword, NET_DEVICEINFO lpDeviceInfo, IntPtr error);
[DllImport("dhnetsdk.dll", CallingConvention = CallingConvention.Cdecl)]
[return: MarshalAs(UnmanagedType.Bool)]
public static extern bool CLIENT_GetDevConfig(
int loginId,
uint command,
int channel,
out NET_TIME buffer,
out uint bufferSize,
IntPtr lpBytesReturned,
int waittime = 500);
[DllImport("dhnetsdk.dll", CallingConvention = CallingConvention.Cdecl)]
public static extern bool CLIENT_Logout(long lID);
[DllImport("dhnetsdk.dll", CallingConvention = CallingConvention.Cdecl)]
public static extern void CLIENT_Cleanup();
[DllImport("dhnetsdk.dll", CallingConvention = CallingConvention.Cdecl)]
public static extern uint CLIENT_GetLastError();
public static void fDisConnectMethod(long lLoginID, IntPtr pchDVRIP, long nDVRPort, uint dwUser)
{
System.Console.WriteLine("Disconnect");
return;
}
public static void fHaveReConnectMethod(long lLoginID, IntPtr pchDVRIP, long nDVRPort, uint dwUser)
{
System.Console.WriteLine("Reconnect success");
return;
}
[StructLayout(LayoutKind.Sequential)]
public class NET_DEVICEINFO
{
[MarshalAs(UnmanagedType.ByValArray, SizeConst = 48)]
public byte[] sSerialNumber;
public byte byAlarmInPortNum;
public byte byAlarmOutPortNum;
public byte byDiskNum;
public byte byDVRType;
public byte byChanNum;
}
[StructLayout(LayoutKind.Sequential)]
public struct NET_TIME
{
uint dwYear;
uint dwMonth;
uint dwDay;
uint dwHour;
uint dwMinute;
uint dwSecond;
}
public static void Main()
{
fDisConnect fDisConnecthandler = fDisConnectMethod;
fHaveReConnect fHaveReConnecthandler = fHaveReConnectMethod;
NET_DEVICEINFO deviceinfo = new NET_DEVICEINFO();
IntPtr iRet = new IntPtr(0);
CLIENT_Init(fDisConnectMethod, 0);
CLIENT_SetAutoReconnect(fHaveReConnecthandler, 0);
lLogin = CLIENT_Login("192.168.1.198", 31111, "user", "password", deviceinfo, iRet);
if (lLogin <= 0)
Console.WriteLine("Login device failed");
else
{
Console.WriteLine("Login device successful");
byte[] byteout = new byte[20];
const int t = 500;
IntPtr BytesReturned;
BytesReturned = IntPtr.Zero;
const uint c = 8;
NET_TIME nt;
uint sizeofnt = (uint)Marshal.SizeOf(typeof(NET_TIME));
if (CLIENT_GetDevConfig(lLogin, c, 0, out nt, out sizeofnt, BytesReturned, t) == false)
{
uint gle = CLIENT_GetLastError();
Console.WriteLine("getDevConfig failed");
}
CLIENT_Logout(lLogin);
CLIENT_Cleanup();
}
}
}
And here is my C code that I'm trying to port to C#. It works without any issues..
#pragma comment(lib,"dhnetsdk.lib")
#include <Windows.h>
#include <stdio.h>
#include <conio.h>
#include "dhnetsdk.h"
void CALLBACK DisConnectFunc(LONG lLoginID, char *pchDVRIP, LONG nDVRPort, DWORD dwUser)
{
printf("Disconnect.");
return;
}
void CALLBACK AutoConnectFunc(LONG lLoginID,char *pchDVRIP,LONG nDVRPort,DWORD dwUser)
{
printf("Reconnect success.");
return;
}
int main(void)
{
NET_TIME nt = {0};
NET_DEVICEINFO deviceInfo = {0};
unsigned long lLogin = 0;
LPVOID OutBuffer;
int iRet = 0;
DWORD dwRet = 0;
//Initialize the SDK, set the disconnection callback functions
CLIENT_Init(DisConnectFunc,0);
//Setting disconnection reconnection success of callback functions. If don't call this interface, the SDK will not break reconnection.
CLIENT_SetAutoReconnect(AutoConnectFunc,0);
lLogin = CLIENT_Login("192.168.1.108",31111,"user","password",&deviceInfo, &iRet);
if(lLogin <= 0)
{
printf("Login device failed");
}
else
{
OutBuffer = (LPVOID)malloc(sizeof(NET_TIME));
memset(OutBuffer, 0, sizeof(NET_TIME));
if(CLIENT_GetDevConfig( lLogin, 8 /* DH_DEV_TIMECFG */, 0, OutBuffer, sizeof(NET_TIME), &dwRet, 500) == FALSE)
{
printf("Failed\n");
}
else
{
memcpy(&nt, OutBuffer, sizeof(nt));
printf("Time %d %d %d %d %d %d\n", nt.dwYear,nt.dwMonth,nt.dwDay, nt.dwHour,nt.dwMinute, nt.dwSecond);
}
_getch();
}
CLIENT_Logout(lLogin);
CLIENT_Cleanup();
return 0;
}
Your extern function is wrongly defined. Let's take your C call example.
// Search configuration information
CLIENT_API BOOL CALL_METHOD CLIENT_GetDevConfig(LLONG lLoginID, DWORD dwCommand, LONG lChannel, LPVOID lpOutBuffer, DWORD dwOutBufferSize, LPDWORD lpBytesReturned,int waittime=500);
As stated in comment of your post, the length of LONG is 32-bit in Win32, so you have to use an int. You can also use the keyword out to get your structure without manually use the Mashaller. I would define your function as that.
[DllImport("dhnetsdk.dll", CallingConvention = CallingConvention.Cdecl)]
[return: MarshalAs(UnmanagedType.Bool)]
public static extern bool CLIENT_GetDevConfig(int loginId, uint command, int channel, out NET_TIME buffer, out uint bufferSize, IntPtr lpBytesReturned, int waittime = 500);
Note the presence of the additional attribute MarshalAs. It indicates how the managed code should consider the return value of the pinvoke'd function.
Here are the differences that I can see:
The C++ code:
CLIENT_API BOOL CALL_METHOD CLIENT_GetDevConfig(
LLONG lLoginID,
DWORD dwCommand,
LONG lChannel,
LPVOID lpOutBuffer,
DWORD dwOutBufferSize,
LPDWORD lpBytesReturned,
int waittime
);
Now, LLONG is pointer sized, 32 bit under x86, 64 bit under x64. That translates to IntPtr for C#. I'd declare the p/invoke like this:
[DllImport("dhnetsdk.dll", CallingConvention = CallingConvention.Cdecl)]
[return: MarshalAs(UnmanagedType.Bool)]
public static extern bool CLIENT_GetDevConfig(
IntPtr lLoginID,
uint dwCommand,
int lChannel,
out NET_TIME lpOutBuffer,
uint dwOutBufferSize,
out uint lpBytesReturned,
int waittime
);
The main problems that I can see are that:
You are translated dwOutBufferSize incorrectly. It's an IN parameter but you are passing it by reference. This is the most likely explanation for the failure.
In the C++ code you pass a pointer to a DWORD variable for lpBytesReturned. In your C# code you pass IntPtr.Zero which is the null pointer.
So, I'd have the call to CLIENT_GetDevConfig looking like this:
NET_TIME nt;
uint sizeofnt = (uint)Marshal.SizeOf(typeof(NET_TIME));
uint BytesReturned;
if (!CLIENT_GetDevConfig(lLogin, 8, 0, out nt, sizeofnt, out BytesReturned, 500))
....
It may be that you can indeed pass IntPtr.Zero to the BytesReturned parameter. Perhaps it's an optional parameter. But I cannot tell that from here. However, for sure the dwOutBufferSize mis-declaration is an error.

C++ to C# conversion of SendMessage using COPYDATASTRUCT

I'm converting a C++ application into C# which has generally been fairly straight forward, but now I'm dealing with pointers and running into problems.
This is the original C++ code
ShockVideoInfo* pVideoInfo = new ShockVideoInfo;
COPYDATASTRUCT cd;
cd.dwData = bSelf ? SHOCK_REQUEST_SELFVIEW_WINDOW : SHOCK_REQUEST_MAINVIEW_WINDOW;
cd.lpData = pVideoInfo;
cd.cbData = sizeof(ShockVideoInfo);
pVideoInfo->dwLeft = 10;
pVideoInfo->dwTop = 10;
pVideoInfo->dwWidth = 100;
pVideoInfo->dwHeight = 100;
pVideoInfo->dwFlags = SHVIDSHOW | SHVIDSIZE | SHVIDTOPMOST | SHVIDMOVE | SHVIDSIZABLE;
if (::SendMessage(m_hShockWnd, WM_COPYDATA, (WPARAM)GetSafeHwnd(), (LPARAM)&cd))
Log(eResult, _T("Window updated"));
else
Log(eError, _T("Window not updated"));
with the structure definition
typedef struct tagShockVideoInfo
{
DWORD dwLeft;
DWORD dwTop;
DWORD dwWidth;
DWORD dwHeight;
DWORD dwFlags;
} ShockVideoInfo;
Here's my conversion to C#, using Marshalling to handle the pointers:
ShockVideoInfo videoInfo = new ShockVideoInfo();
COPYDATASTRUCT cd = new COPYDATASTRUCT();
cd.dwData = (IntPtr)_shockRequestMainviewWindow;
cd.lpData = Marshal.AllocCoTaskMem(cd.cbData); // Pointer to videoInfo
Marshal.StructureToPtr(videoInfo, cd.lpData, true);
cd.cbData = Marshal.SizeOf(videoInfo);
// Set-up video window position
videoInfo.dwFlags = _shvidShow | _shvidSize | _shvidTopmost | _shvidMove | _shvidSizable;
videoInfo.dwTop = 10;
videoInfo.dwLeft = 10;
videoInfo.dwHeight = 100;
videoInfo.dwWidth = 100;
IntPtr cdPointer = Marshal.AllocCoTaskMem(Marshal.SizeOf(videoInfo));
Marshal.StructureToPtr(videoInfo, cdPointer, true);
if ( (int) SendMessage(_hWnd, WM_COPYDATA, _formHandle, cdPointer) == 1 )
{
Debug.WriteLine("Window updated");
}
else
{
Debug.WriteLine("Window not updated");
}
Marshal.FreeCoTaskMem(cdPointer);
Marshal.FreeCoTaskMem(cd.lpData);
In addition I'm using the following in my C# code:
[DllImport("user32.dll", CharSet = CharSet.Auto)]
static extern IntPtr SendMessage(IntPtr hWnd, UInt32 Msg, IntPtr wParam, IntPtr lParam);
[StructLayout(LayoutKind.Sequential)]
struct ShockVideoInfo
{
public uint dwLeft;
public uint dwTop;
public uint dwWidth;
public uint dwHeight;
public uint dwFlags;
}
[StructLayout(LayoutKind.Sequential)]
struct COPYDATASTRUCT
{
public IntPtr dwData;
public int cbData;
public IntPtr lpData;
}
My code isn't getting the response I want (which is moving a video window), so I've done something wrong.
Can anyone with experience in this area please show me what I'm doing wrong?
I think I've found my issue:
IntPtr cdPointer = Marshal.AllocCoTaskMem(Marshal.SizeOf(videoInfo));
Marshal.StructureToPtr(videoInfo, cdPointer, true);
Should have been:
IntPtr cdPointer = Marshal.AllocCoTaskMem(Marshal.SizeOf(cd));
Marshal.StructureToPtr(cd, cdPointer, false);
In other words, I wasn't passing the pointer to the videoInfo structure! Which helps :)

Categories