I have a c# .net 2.0 CF application that interfaces with a native DLL implementing a function like this:
struct NATIVE_METHOD_REPLY
{
int other_irrelevant_data;
int data_size;
void* data;
}
// reply_buffer will contain an array of NATIVE_METHOD_REPLY structures
// and their data.
//
// returns an error code
int Foo(NATIVE_METHOD_REPLY* reply_buffer, int reply_size);
I've implemented it in C# like this:
[StructLayout(LayoutKind.Sequential)]
internal struct NATIVE_METHOD_REPLY
{
public Int32 OtherIrrelevantData;
public Int16 DataSize;
public IntPtr DataPtr;
}
[DllImport("my_lib.dll", SetLastError = true)]
internal static extern Int32 Foo(byte[] replyBuffer, Int32 replySize);
public byte[] void Bar()
{
// data returned to the user. May be an arbitrary size.
byte[] result_buffer = new byte[256];
// data sent to Foo()
byte[] reply_buffer =
new byte[Marshal.SizeOf(typeof(NativeMethods.NATIVE_METHOD_REPLY)) +
result_buffer.Length];
NativeMethods.Foo(reply_buffer, reply_buffer.Length);
// is there a better way of doing this?
NativeMethods.NATIVE_METHOD_REPLY reply;
GCHandle pinned_reply = GCHandle.Alloc(reply_buffer,
GCHandleType.Pinned);
try
{
reply = (NativeMethods.NATIVE_METHOD_REPLY)Marshal.PtrToStructure(
pinned_reply.AddrOfPinnedObject(),
typeof(NativeMethods.NATIVE_METHOD_REPLY));
Marshal.Copy(reply.DataPtr, result_buffer, 0, reply.DataSize);
}
finally
{
pinned_reply.Free();
}
// bonus point*: is this okay to do after the Free() call?
int test = reply.OtherIrrelevantData;
return result_buffer;
}
While this works correctly, I would like to know if this is the most efficient / most correct way of implementing this function.
Is there some method converting a managed byte array to a managed structure that doesn't involve an intermediate native handle and a copy? For instance, in C++, I would just do this:
NATIVE_METHOD_REPLY* reply = reinterpret_cast< NATIVE_METHOD_REPLY* >( reply.DataPtr );
*For a bonus point, is it okay to use data in the structure after the native handle has been freed?
Thanks,
PaulH
Edit: Updated solution
[DllImport("my_lib.dll", SetLastError = true)]
internal static extern Int32 Foo(IntPtr replyBuffer, Int32 replySize);
public byte[] void Bar()
{
byte[] result_buffer = new byte[256];
int reply_buffer_len = Marshal.SizeOf(typeof(NativeMethods.NATIVE_METHOD_REPLY)) + result_buffer.Length;
IntPtr reply_buffer = Marshal.AllocCoTaskMem(reply_buffer_len);
NativeMethods.NATIVE_METHOD_REPLY reply;
try
{
NativeMethods.Foo(reply_buffer, reply_buffer_len);
reply = (NativeMethods.NATIVE_METHOD_REPLY)Marshal.PtrToStructure(
reply_buffer,
typeof(NativeMethods.NATIVE_METHOD_REPLY));
Marshal.Copy(reply.DataPtr, result_buffer, 0, reply.DataSize);
}
finally
{
Marshal.FreeCoTaskMem(reply_buffer);
}
return result_buffer;
}
The structure has a fixed size. There's no point in passing an array, just pass the structure:
[DllImport("my_lib.dll", SetLastError = true)]
internal static extern Int32 Foo(out NATIVE_METHOD_REPLY replyBuffer, Int32 replySize);
You do have a memory management problem. Who owns the pointer?
Okay, the structure is actually variable sized and the pointer points into the array. You need nother approach. Simply allocate a chunk of unmanaged memory up front instead of letting the P/Invoke marshaller copy the data into a managed array. Which is in fact a hard requirement since the garbage collector can move the array, invalidating the pointer. Call Marshal.CoTaskMemAlloc() to reserve the memory, you'll have to free it later. And change the first argument of the function to IntPtr (not out).
You'll find that marshaling the structure a lot easier too, no need to pin the memory. Don't forget to Marshal.FreeCoTaskMem() when you're done.
In C# under the full framework, you can marshal the array directly. See Default Marshaling for Arrays. I don't know what the limitations are on the Compact Framework.
Related
I've been banging my head all day and hope someone can help. I need to marshal a managed data structure to an unmanaged C dll. When I look at all the memory it appears that what I'm doing is working, but the C dll (a black box to me) is returning an error indicating the data is corrupt. Can anyone point out my error?
C declarations
typedef struct _TAG_Data
{
void *data; // Binary data
uint32_t size; // Data size bytes
} Data;
// Parse binary data, extract int value
ParseData(Data ∗ result, uint64_t parameter, int32_t ∗ value)
Managed object:
public class MyData
{
public byte[] data;
public UInt32 size;
}
Packed equivalent for moving to unmanaged memory:
[StructLayout(LayoutKind.Sequential, Pack = 1)]
struct MyData_Packed
{
public IntPtr data;
public UInt32 size;
}
At this point I have a managed "MyData" struct called MyResult with valid data that needs to go into the dll. Here's what I'm doing:
[DllImport("Some.dll", EntryPoint = "ParseData", SetLastError = true, CharSet = CharSet.Ansi)]
private static extern Int32 ParseData_Native(IntPtr result, UInt64 parameter, ref Int32 value);
IntPtr MyResultPackedPtr = new IntPtr();
MyData_Packed MyResultPacked = new MyData_Packed();
// Copy "MyResult" into un-managed memory so it can be passed to the C library.
// Allocate un-managed memory for the data buffer
MyResultPacked.data = Marshal.AllocHGlobal((int)MyResult.size);
// Copy data from managed "MyResult.data" into the unmanaged "MyResultPacked.data"
Marshal.Copy(MyResult.data, 0, MyResultPacked.data, (int)MyResult.size);
MyResultPacked.size = MyResult.size;
// Allocate unmanaged memory for the structure itself
MyResultPackedPtr = Marshal.AllocHGlobal(Marshal.SizeOf(typeof(MyData_Packed)));
// Copy the packed struct into the unmanaged memory and get our pointer
Marshal.StructureToPtr(MyResultPacked, MyResultPackedPtr, false);
// Pass our pointer to the unmanaged struct, which points to the unmanaged data buffer, to the C dll
Int32 tmp = 0;
ErrorCode = ParseData_Native(MyResultPackedPtr, parameter, ref tmp);
When I look at the unmanaged memory pointed to by MyResultPacked.data, the data is correct so the copy was good. And when I look at MyResultPackedPtr in memory, the first 8 bytes are the address of the same unmanaged memory pointed to by MyResultPacked.data (64-bit machine), and the next 4 bytes are the proper size of the data. So it appears that MyResultPackedPtr points to a valid copy of MyResultPacked. But the return value from ParseData() indicates my data must be corrupt, so I must be doing something wrong.
To take it a step further, I wrote the same code 100% in C and it works. And the data in the binary buffer in C matched the data in the binary buffer in C#, going by the memory watch feature in Visual Studio, so it appears my data handling is correct. Which makes me think something is wrong with the way I'm passing MyResultPackedPtr to the dll. Unfortunately I don't have the source for the dll and cannot step into it. Can anyone offer a suggestion on what to try next?
I don't see why you need any of this custom marshalling code in the first place. You should be able to pass the struct with the byte[] array directly, and the marshaller will sort out the copying.
You also need to set the calling convention correctly.
[StructLayout(LayoutKind.Sequential, Pack = 1)]
struct MyData_Packed
{
public byte[] data;
public UInt32 size;
}
[DllImport("Some.dll", EntryPoint = "ParseData", SetLastError = true, CallingConvention = CallingConvention.CDecl)]
private static extern int ParseData_Native(ref MyData_Packed result, ulong parameter, out int value);
var MyResultPacked = new MyData_Packed
{
data = MyResult,
size = MyResult.size,
};
ErrorCode = ParseData_Native(ref MyResultPacked, parameter, out var tmp);
Basically I have a DLL and a header file (C++) for a data acquisition device that I use for my research. The device has 8 sensors and each sensor has x,y,z,etc data points. I'd like to create a UI program in WPF for this device. So far I've successfully been able to communicate with the device in .NET via an unsafe code implementation and read data into an array of structs (where each struct corresponds to one of the sensors). However I'd like to see if I can get it working with a strictly safe code implementation, but I've just been hitting brick walls for a while now. This is my first time working with unmanaged vs. managed code so please bear with me. I've searched countless threads online and the rabbit hole just keeps getting deeper so I'd like some advice from someone with experience. Basically the API header file has a function with the following definition:
int GetSynchronousRecord(USHORT sensorID, void *pRecord, int recordSize);
Essentially we pass the function a buffer by reference, and it fills it up. I have the option of getting either a single sensor's data, or all sensors at once depending on the sensorID argument I pass. What HAS worked for me so far (in the managed world) is if I do the following:
[StructLayout(LayoutKind.Sequential)]
public struct DOUBLE_POSITION_ANGLES_TIME_Q_RECORD
{
public double x;
public double y;
public double z;
public double a;
public double e;
public double r;
public double time;
public ushort quality;
};
...
[DllImport("ATC3DG64.DLL", CallingConvention = CallingConvention.Cdecl)]
public static extern int GetSynchronousRecord(ushort sensorID, ref DOUBLE_POSITION_ANGLES_TIME_Q_RECORD record, int recordSize);
...
DOUBLE_POSITION_ANGLES_TIME_Q_RECORD record = new DOUBLE_POSITION_ANGLES_TIME_Q_RECORD();
// Get the data from SENSOR_1 only
errorCode = GetSynchronousRecord(1, ref record, Marshal.SizeOf(record));
So this implementation works fine, I can get all the coordinate data and at a really good speed. However, I'd like to get ALL the sensors at once. In the C++ API code samples, they pass the GetSynchronousRecord function an ARRAY of STRUCTS, one struct for each sensor. I tried to do the same in C# as follows:
[DllImport("ATC3DG64.DLL", CallingConvention = CallingConvention.Cdecl)]
public static extern int GetSynchronousRecord(ushort sensorID, ref DOUBLE_POSITION_ANGLES_TIME_Q_RECORD[] record, int recordSize);
// Define Array of Struct for each sensor
DOUBLE_POSITION_ANGLES_TIME_Q_RECORD[] record = new DOUBLE_POSITION_ANGLES_TIME_Q_RECORD[8];
while(recording) {
...
// Get data from ALL sensors (0xffff)
errorCode = GetSynchronousRecord(0xffff, ref record, Marshal.SizeOf(record)*8);
...
}
But this straight up crashes my program with an System.ExecutionEngineException error. I've read that since my function is expecting a void* pointer, that I should use an IntPtr argument, but this approach seemed quite confusing to be honest. Another thing I tried is to actually loop over each sensor and call the function for the sensor, but this dropped the speed INSANELY almost to 1 record/second (instead of 100 records/second). Many other similar threads on stackexchange say to use out parameter, or to use [In, Out] attribute on the function definition, but none of these suggestions worked.
TL;DR: If I understand my situation correctly, I have a MANAGED array of structs that I need to correctly pass to a C++ function as an argument (pass by reference), and then the function will fill my structs with data from a data acquisition device.
I apologize for the wall of text/code, any information for me from someone with experience would be much appreciated.
EDIT: Just to clarify, the GetSynchronousRecord function is INSIDE a while loop where on each iteration I'm getting new data points for each struct.
Your second p/invoke declaration is wrong. You had
[DllImport("ATC3DG64.DLL", CallingConvention = CallingConvention.Cdecl)]
public static extern int GetSynchronousRecord(
ushort sensorID,
ref DOUBLE_POSITION_ANGLES_TIME_Q_RECORD[] record,
int recordSize
);
The problem is the array parameter. Because you pass that array by ref that actually makes it a double pointer. Instead you want to simply remove the ref and declare the import like so:
[DllImport("ATC3DG64.DLL", CallingConvention = CallingConvention.Cdecl)]
public static extern int GetSynchronousRecord(
ushort sensorID,
[Out] DOUBLE_POSITION_ANGLES_TIME_Q_RECORD[] record,
int recordSize
);
The [Out] attribute tells the marshaler that the data is flowing out of the function. Without it the default assumption is that the data flows in.
When you call the function do so like this:
errorCode = GetSynchronousRecord(0xffff, record, Marshal.SizeOf(record)*record.Length);
The best way is probably using IntPtr rather than ref DOUBLE_POSITION_ANGLES_TIME_Q_RECORD.
I'd change the P/Invoke signature to:
[DllImport("ATC3DG64.DLL", CallingConvention = CallingConvention.Cdecl)]
public static extern int GetSynchronousRecord(ushort sensorID, IntPtr record, int recordSize);
Create an IntPtr that points to the needed memory space:
IntPtr ptr = Marshal.AllocHGlobal(Marshal.SizeOf<DOUBLE_POSITION_ANGLES_TIME_Q_RECORD>() * 8);
Call the P/Invoke function (which should fill the memory with the structs):
errorCode = GetSynchronousRecord(0xffff, ptr, Marshal.SizeOf<DOUBLE_POSITION_ANGLES_TIME_Q_RECORD>() * 8);
Get the structures from the memory block:
DOUBLE_POSITION_ANGLES_TIME_Q_RECORD[] records = new DOUBLE_POSITION_ANGLES_TIME_Q_RECORD[8];
for (i = 0; i < 8; i++) {
records[i] = Marshal.PtrToStructure<DOUBLE_POSITION_ANGLES_TIME_Q_RECORD>(IntPtr.Add(ptr, i * Marshal.SizeOf<DOUBLE_POSITION_ANGLES_TIME_Q_RECORD>()));
}
Marshal.FreeHGlobal(ptr);
I want to use c# interop to call a function from a dll written in c. I have the header files.
Take a look at this:
enum CTMBeginTransactionError {
CTM_BEGIN_TRX_SUCCESS = 0,
CTM_BEGIN_TRX_ERROR_ALREADY_IN_PROGRESS,
CTM_BEGIN_TRX_ERROR_NOT_CONNECTED
};
#pragma pack(push)
#pragma pack(1)
struct CTMBeginTransactionResult {
char * szTransactionID;
enum CTMBeginTransactionError error;
};
struct CTMBeginTransactionResult ctm_begin_customer_transaction(const char * szTransactionID);
How do I call ctm_begin_customer_transaction from c#. The const char * mapps well to string, but despite various attempts (looking at stackoverflow and other sites), I fail to marshal the return structure. If I define the function to return IntPtr it works ok...
Edit
I changed the return type to IntPtr and use:
CTMBeginTransactionResult structure = (CTMBeginTransactionResult)Marshal.PtrToStructure(ptr, typeof(CTMBeginTransactionResult));
but it throws AccessViolationException
I also tried:
IntPtr ptr = Transactions.ctm_begin_customer_transaction("");
int size = 50;
byte[] byteArray = new byte[size];
Marshal.Copy(ptr, byteArray, 0, size);
string stringData = Encoding.ASCII.GetString(byteArray);
stringData == "70e3589b-2de0-4d1e-978d-55e22225be95\0\"\0\0\a\0\0\b\b?" at this point. The "70e3589b-2de0-4d1e-978d-55e22225be95" is the szTransactionID from the struct. Where is the Enum? Is it the next byte?
There's a memory management problem hidden in this struct. Who owns the C string pointer? The pinvoke marshaller will always assume that the caller owns it so it will try to release the string. And passes the pointer to CoTaskMemFree(), same function as the one called by Marshal.FreeCoTaskMem(). These functions use the COM memory allocator, the universal interop memory manager in Windows.
This rarely comes to a good end, C code does not typically use that allocator unless the programmer designed his code with interop in mind. In which case he'd never have used a struct as a return value, interop always works much less trouble-free when the caller supplies buffers.
So you cannot afford to let the marshaller do its normal duty. You must declare the return value type as IntPtr so it doesn't try to release the string. And you must marshal it yourself with Marshal.PtrToStructure().
That however still leaves the question unanswered, who owns the string? There is nothing you can do to release the string buffer, you don't have access to the allocator used in the C code. The only hope you have is that the string wasn't actually allocated on the heap. That's possible, the C program might be using string literals. You need to verify that guess. Call the function a billion times in a test program. If that doesn't explode the program then you're good. If not then only C++/CLI can solve your problem. Given the nature of the string, a "transaction ID" ought to change a lot, I'd say you do have a problem.
I hate to answer my own question, but I found the solution to marshal the resulting struct. The struct is 8 bytes long (4 bytes for the char * and 4 bytes for enum). Marshalling the string does not work automatically, but the following works:
// Native (unmanaged)
public enum CTMBeginTransactionError
{
CTM_BEGIN_TRX_SUCCESS = 0,
CTM_BEGIN_TRX_ERROR_ALREADY_IN_PROGRESS,
CTM_BEGIN_TRX_ERROR_NOT_CONNECTED
};
// Native (unmanaged)
[StructLayout(LayoutKind.Sequential, Pack = 1, CharSet = CharSet.Ansi)]
internal struct CTMBeginTransactionResult
{
public IntPtr szTransactionID;
public CTMBeginTransactionError error;
};
// Managed wrapper around native struct
public class BeginTransactionResult
{
public string TransactionID;
public CTMBeginTransactionError Error;
internal BeginTransactionResult(CTMBeginTransactionResult nativeStruct)
{
// Manually marshal the string
if (nativeStruct.szTransactionID == IntPtr.Zero) this.TransactionID = "";
else this.TransactionID = Marshal.PtrToStringAnsi(nativeStruct.szTransactionID);
this.Error = nativeStruct.error;
}
}
[DllImport("libctmclient-0.dll")]
internal static extern CTMBeginTransactionResult ctm_begin_customer_transaction(string ptr);
public static BeginTransactionResult BeginCustomerTransaction(string transactionId)
{
CTMBeginTransactionResult nativeResult = Transactions.ctm_begin_customer_transaction(transactionId);
return new BeginTransactionResult(nativeResult);
}
The code works, but I still need to investigate, if calling the unmanaged code results in memory leaks.
So, there have been many variants of this question, and after looking at several I still can't figure it out.
This is the C code:
typedef struct
{
unsigned long Identifier;
char Name[128];
} Frame;
Frame GetFrame(int index);
This is the C# code:
struct Frame
{
public ulong Identifier;
[MarshalAs(UnmanagedType.ByValArray, ArraySubType = UnmanagedType.I1, SizeConst = 128)]
public char[] Name;
}
[DllImport("XNETDB.dll", CallingConvention = CallingConvention.Cdecl, CharSet = CharSet.Ansi)]
private static extern Frame GetFrame(int index);
This is the last attempt I tried in C#, and it seems pretty logical, but I get the error "Method's signature is not PInvoke compatible." So, I'm kind of lost on what to try next. Any help is appreciated.
Thanks,
Kevin
Updated Kevin added this as an edit to my answer
I should instead change my C code:
void GetFrame(int index, Frame * f);
and use instead for C#:
struct Frame
{
public uint Identifier;
[MarshalAsAttribute(UnmanagedType.ByValTStr, SizeConst = 128)]
public string Name;
}
[DllImport("XNETDB.dll", CallingConvention = CallingConvention.Cdecl, CharSet = CharSet.Ansi)]
private static extern void GetFrame(int index, ref Frame f);
There are two problems with the PInvoke signature that you've chosen.
The first is easy to fix. You have a mistranslation of unsigned long. In C an unsigned long is typically only 4 bytes. You chose the C# type long which is 8 bytes. Changing the C# code to use uint will fix this.
The second is a bit harder. As Tergiver pointed out the CLR Marshaller only supports a struct in the return position if it's blittable. Blittable is a fancy way of saying that it has the exact same memory representation in native and managed code. The struct definition you've chosen isn't blittable because it has a nested array.
This can be worked around though if you remember that PInvoke is a very simple process. The CLR Marshaller really just needs you to answer 2 questions with the signature of your types and pinvoke methods
How many bytes am I copying?
In which direction do they need to go?
In this case the number of bytes is sizeof(unsigned long) + 128 == 132. So all we need to do is build up a managed type that is blittable and has a size of 132 bytes. The easiest way to do this is to define a blob to handle the array portion
[StructLayout(LayoutKind.Sequential, Size = 128)]
struct Blob
{
// Intentionally left empty. It's just a blob
}
This is a struct with no members that will appear to the marshaller as having a size of 128 bytes (and as a bonus it's blittable!). Now we can easily define the Frame structure as a combination of an uint and this type
struct Frame
{
public int Identifier;
public Blob NameBlob;
...
}
Now we have a blittable type with a size the marshaller will see as 132 bytes. This means it will work just fine with the GetFrame signature you've defined
The only part left is giving you access to the actual char[] for the name. This is a bit tricky but can be solved with a bit of marshal magic.
public string GetName()
{
IntPtr ptr = IntPtr.Zero;
try
{
ptr = Marshal.AllocHGlobal(128);
Marshal.StructureToPtr(NameBlob, ptr, false);
return Marshal.PtrToStringAnsi(ptr, 128);
}
finally
{
if (ptr != IntPtr.Zero)
{
Marshal.FreeHGlobal(ptr);
}
}
}
Note: I can't comment on the calling convention portion because I'm unfamiliar with the GetFrame API but that's something i would definitely check on.
The problem is that the native function returns a non-blittable type as a return value.
http://msdn.microsoft.com/en-us/library/ef4c3t39.aspx
P/Invoke cannot have non-blittable types as a return value.
You cannot p/Invoke that method. [EDIT It is actually possible, see JaredPar's answer]
Returning 132 bytes by value is a bad idea. If this native code is yours, I would fix it. You can fix it by allocating the 132 bytes and returning a pointer. Then add a FreeFrame method to release that memory. Now it can be p/Invoked.
Alternately, you could change it to accept a pointer to the Frame memory that it will fill in.
Another option to JaredPar's is to utilize the C# fixed size buffer feature. This does however require you to turn on the setting to allow unsafe code, but it avoids having 2 structs.
class Program
{
private const int SIZE = 128;
unsafe public struct Frame
{
public uint Identifier;
public fixed byte Name[SIZE];
}
[DllImport("PinvokeTest2.DLL", CallingConvention = CallingConvention.Cdecl, CharSet = CharSet.Ansi)]
private static extern Frame GetFrame(int index);
static unsafe string GetNameFromFrame(Frame frame)
{
//Option 1: Use if the string in the buffer is always null terminated
//return Marshal.PtrToStringAnsi(new IntPtr(frame.Name));
//Option 2: Use if the string might not be null terminated for any reason,
//like if were 128 non-null characters, or the buffer has corrupt data.
return Marshal.PtrToStringAnsi(new IntPtr(frame.Name), SIZE).Split('\0')[0];
}
static void Main()
{
Frame a = GetFrame(0);
Console.WriteLine(GetNameFromFrame(a));
}
}
I'm using C# with P/Invoke to access to a DLL method. The definition of the method is the following:
[DllImport("userManager.dll")]
static extern int GetUsers(out IntPtr userList);
Original structs:
typedef struct user_list {
unsigned short NumUsers;
USER_LIST_ITEM List[VARLEN];
} USER_LIST
typedef struct user_list_item {
char name[260];
unsigned char address[256];
} USER_LIST_ITEM
And the struct layout I've done is the following:
[StructLayout(LayoutKind.Sequential)]
public class USER_LIST
{
public uint NumUsers;
[MarshalAs(UnmanagedType.ByValArray)]
public USER_LIST_ITEM [] List;
}
[StructLayout(LayoutKind.Sequential)]
public class USER_LIST_ITEM
{
[MarshalAs(UnmanagedType.ByValTStr, SizeConst = 260)]
public string name;
[MarshalAs(UnmanagedType.ByValTStr, SizeConst = 256)]
public string address;
};
But I get an error when I try to unmarshall it:
USER_LIST userList = new USER_LIST();
// Prepare pointer
IntPtr uList = Marshal.AllocHGlobal(Marshal.SizeOf(userList));
Marshal.StructureToPtr(userList, uList, false);
result = GetUsers(out uList);
Marshal.PtrToStructure(uList, userList); <--
The runtime has encountered a fatal error. The address of the error was at 0x79f82af6, on thread 0x464. The error code is 0xc0000005. This error may be a bug in the CLR or in the unsafe or non-verifiable portions of user code. Common sources of this bug include user marshaling errors for COM-interop or PInvoke, which may corrupt the stack.
I get the NumUsers property right, but it seems the error occurs when unmarshalling the array. Any thoughts?
If you specify an array in a structure used as an out parameter, you need to tell the marshaler what length is the array going to be. With your code, the marshaler is probably allocating a zero-length array or just using null, which produces the crash. Unfortunately there seems to be no way to specify a variable-length out array as a member of a structure, because MarshalAs.SizeParamIndex only works for methods. You might get away with specifying a large, constant-size array using MarshalAs.SizeConst, but generally you'd have to parse the (presumably callee-allocated) return buffer like this:
var count = Marshal.ReadInt32 (uList) ;
var users = new List<USER_LIST_ITEM> () ;
var ptr = (long)uList + 4 ;
for (int i = 0 ; i < count ; ++i)
{
users.Add (Marshal.PtrToStructure (typeof (USER_LIST_ITEM),
new IntPtr (ptr))) ;
ptr += Marshal.SizeOf (typeof (USER_LIST_ITEM)) ;
}
You'll have to pay extra attention to alignment&padding and 32/64 bit issues.
That is because List has not been allocated yet.
You will need initialize all the fields.
Another problem I see is with the following:
IntPtr uList = Marshal.AllocHGlobal(Marshal.SizeOf(userList));
...
result = GetUsers(out uList);
Are you sure that out should not be ref? Else there is no point (not sure if ref is correct either).
Update: Looking at your code again, you should be doing this (and avoid that memory leak poking your eye).
IntPtr uList;
var result = GetUsers(out uList);
var userlist = (USER_LIST) Marshal.PtrToStructure(ulist, typeof(USER_LIST));
Marshal.FreeHGlobal(ulist); // pray here or shoot the author of the C function
Update again:
Your p/invoke signature is likely wrong or you are interpreting it wrong.
I can guess it probably something like:
int GetUsers(USER_LIST* ulist);
And that what you have is not the same thing.
If this is case, the solution is easy.
Change USER_LIST to a class (but keep sequential layout) and use
// pinvoke sig
int GetUsers(USER_LIST ulist);
var ulist = new USER_LIST();
// initialize fields
var r = GetUsers(ulist);
-- or --
Call it by ref.
// pinvoke sig
int GetUsers(ref USER_LIST ulist);
var ulist = new USER_LIST();
// initialize fields
var r = GetUsers(ref ulist);
This way, you dont have to mess with manual marshalling, and I cant see anymore potential for memory leaks.
Final update:
Given the signature you posted, it looks like GetUsers returns a pointer to a list of USER_LIST with the return value being the count. Nice memory leak there.
Anyways, I would probably experiment with an unsafe approach here, and just walk thru the result , and make sure everything gets freed. (I still think you should shoot the author).
I think your original code isn't probably so wrong.
You've probably just used the wrong overload of Marshal.PtrToStructure.
Have you tried this?
[DllImport("userManager.dll")]
static extern int GetUsers(out IntPtr userList);
[DllImport("userManager.dll")]
static extern void UMFree(IntPtr userList);
static void Main()
{
IntPtr userList; // no need to allocate memory in managed code;
GetUsers(out userList); // memory is allocated by native function
USER_LIST u = (USER_LIST)Marshal.PtrToStructure(userList, typeof(USER_LIST));
UMFree(userList);
}
Using unsafe code:
public unsafe struct USER_LIST
{
public uint numUsers;
public USER_LIST_ITEM* list;
}
public unsafe struct USER_LIST_ITEM
{
public fixed byte name[260];
public fixed byte address[256];
}
class Program
{
[DllImport("userManager.dll")]
static unsafe extern int GetUsers(USER_LIST** userList);
[DllImport("userManager.dll")]
static unsafe extern int UMFree(USER_LIST* userList);
private static unsafe void Main()
{
USER_LIST* list;
GetUsers(&list);
UMFree(list);
}
}