How to safely pass string reference from c# to c++? - c#

I am developing a C# dll project with C++ dll project.
Let's say that C++ dll logins to a certain web site, and do some query on the web server.
And C++ dll has to return that html code of a web site.
In the same time, C++ dll must save the cookie data from the web site.
So, I passed StringBuilder object to C++ function.
I already know how to get html code from a web site using HttpWebRequest and HttpWebResponse classed in C#, but unfortunately I have to do it in C++ dll project.
So bear in mind, I don't need any C# codes.
I have tried Passing a string variable as a ref between a c# dll and c++ dll.
Passing StringBuilder from C# and get it as LPTSTR.
It works fine, but some strings were missing from the result.
I couldn't find out the reason.
Anyway, here is my code.
C++
extern "C" __declspec(dllexport) BSTR LoginQuery(const char* UserID, const char* UserPW, char Cookies[])
{
std::string html;
try
{
std::map<std::string, std::string> cookies;
MyClass *myclass = new MyClass();
html = myclass->LoginQuery(UserID, UserPW, cookies);
// Response cookies
std::string cookieData;
for (std::map<std::string, std::string>::iterator iterator = cookies.begin(); iterator != cookies.end(); iterator++)
{
cookieData += iterator->first;
cookieData += "=";
cookieData += iterator->second;
cookieData += ";";
}
sprintf(Cookies, cookieData.c_str(), 0);
delete myclass;
}
catch (...)
{
}
return ::SysAllocString(CComBSTR(html.c_str()).Detach());
}
C#
[DllImport(#"MyDll.dll", EntryPoint="LoginQuery", CallingConvention = CallingConvention.Cdecl)]
[return: MarshalAs(UnmanagedType.BStr)]
private static extern void LoginQuery(string UserID, string UserPW, StringBuilder Cookies);
void SomeThing()
{
StringBuilder _cookies = new StringBuilder(1024);
string result = LoginQuery("test", "1234", _cookies);
}
It works fine.
With the StringBuilder as cookie, I can carry on the next url of the web site.
(I am using libcurl in C++ project.)
But the problem is that I have about 100 ids.
When it runs about 3~40, it returns heap error.
Debug Assertion Failed!
Program: ~~~~mics\dbgheap.c
Line: 1424
Expression: _pFirstBlock == pHead
I cannot click the abort, retry or ignore button.
It looks like C# application hangs.
I read so many articles about debug assertion failed with dbgheap.
Mostly like free memory object from another heap.
I am newbie to C++.
I read Edson's question on http://bytes.com/topic/c-sharp/answers/812465-best-way-interop-c-system-string-c-std-string.
But the error does not always comes out in certain time.
So I came across with a guess, that it happens when .NET runs garbage collector.
.NET garbage collector tries to free some memory which created from C++ dll and I get the heap error.
Am I right?
I think he's suffering same problem as mine.
What is the best way to avoid heap error and return correct string from C++ dll to C# dll?
P/S: And the heap error occurs only when I run it debug mode. I don't get the error when I run release mode.
EDIT
According to WhozCraig answer, I changed my code as below.
//return ::SysAllocString(CComBSTR(html.c_str()).Detach());
CComBSTR res(html.c_str());
return res.Detach();
But no luck, I still get the heap error.

Your question is title asks about passing a string reference from c# to c++, but later in the text you ask how to return a string from C++ to C#. Also, you tell you are new to C++. With this in mind, I'll tell how I did this sort of interaction last time I had to do that. I just made C++ side allocate and free all the memory, passing out to C# only IntPtrs to be Marshal.PtrToStringAnsi(p)ed. In my case, storing recerences thread-local in C++ and freeing them on each function call was enough, but you can make a C++ function that frees whatever ref it is given. Not very intellectual and not necessarily the most efficient way, but it works.
upd:
It does just what they say it does. Some quick googling comes up with this article. I think it's pretty good, so you can refer to it instead of my suggestion. Passing raw IntPtrs is good if the pointer is not okay to be freed by itself (like old Delphi/C++Builder style strings, for example) and you prefer to be bothered more on the managed side than on the native side.
As an example, piece of code doing Delphi interaction (good for C++ Builder as well):
// function Get_TI_TC(AuraFileName:PAnsiChar; var TI,TC:PAnsiChar):Boolean; stdcall; external 'AuraToIec104.dll' name 'Get_TI_TC';
[DllImport("AuraToIec104")]
[return: MarshalAs(UnmanagedType.I1)]
private static extern bool Get_TI_TC(string AuraFileName, out IntPtr TI, out IntPtr TC);
public static bool Get_TI_TC(string AuraFileName, out string TI, out string TC)
{
IntPtr pTI, pTC;
bool result = Get_TI_TC(AuraFileName, out pTI, out pTC);
TI = Marshal.PtrToStringAnsi(pTI);
TC = Marshal.PtrToStringAnsi(pTC);
return result;
}

It looks like your problem is rather simple. You are creating a StringBuilder that can hold as much as 1024 chars. If your C++ function returns more than that, your application will crash (sooner or later).
So to fix your problem, increase the StringBuilder's size to the maximum possible output length. More details: Passing StringBuilder to PInvoke function which quotes:
The only
caveat is that the StringBuilder must
be allocated enough space for the
return value, or the text will
overflow, causing an exception to be
thrown by P/Invoke.
It might actually be better in your case with dynamic string lengths to use a BSTR parameter in the C++ function. You can then use [MarshalAs(UnmanagedType.AnsiBStr), Out] ref string ... in C# and BSTR* in C++.

Related

How does passing field as argument differ from passing method call as argument?

I am writing a .Net CE application on a Smart Device that has a printer on it. I collect my data in a StringBuilder object, and then try to print it. Here's how I print it
var receipt = new StringBuilder();
// ...
Printer.getInstance().print(receipt.ToString(), (int) Printer.TextAlign.Left, 0, 24, false);
Printer class is imported from a DLL. The application throws an unmanaged exception on the print line and crashes. But when I change my code to this
var receipt = new StringBuilder();
// ...
var str = receipt.ToString();
Printer.getInstance().print(str, (int) Printer.TextAlign.Left, 0, 24, false);
everything works fine. How is it even possible for the evalution of the StringBuilder affect the flow?
Here's a the methods from my Printer.dll (decompiled)
public int print(string text, int textAlign, int fontWeight, int fontSize, bool endLineFeed)
{
open();
Int32 prnReturn;
Printer.Prn_SetLang(1);
Printer.PRN_SetFont((byte) fontSize, (byte) fontSize, 0);
Printer.PRN_SetAlign(textAlign);
Printer.PRN_SetBold(fontWeight);
Prn_String(text.TrimEnd());
prnReturn = PRN_PrintAndWaitComplete();
if(endLineFeed)
printEndingLineFeed();
close();
return prnReturn;
}
public void open()
{
PRN_Open();
}
public void close()
{
PRN_Close();
}
private void printEndingLineFeed()
{
open();
//lineFeed(ENDING_LINE_FEED);
PRN_FeedLine(ENDING_LINE_FEED);
close();
}
and here are the methods that it calls from another DLL. Unfortunately, DotPeek doesn't decompile this.
[DllImport(PrinterDllName, SetLastError = true, EntryPoint = "PRN_FeedLine")]
public static extern Int32 PRN_FeedLine(Int32 pszData);
[DllImport(PrinterDllName, SetLastError = true, EntryPoint = "PRN_PrintAndWaitComplete")]
public static extern Int32 PRN_PrintAndWaitComplete();
Edit: Thanks to Kevin Gosse, I found out that the problem is only there in Debug mode. So my question now is, how does the debug mode evaluation differ from normal execution. Although I do understand that this might be off-topic, I'd be glad if someone could share a relevant piece of documentation.
In release mode, both versions of your code are identical.
In debug mode, there is a subtle difference as the lifetime of str will be extended until the end of the current method (for debugging purpose).
So this code in debug mode:
var receipt = new StringBuilder();
// ...
var str = receipt.ToString();
Printer.getInstance().print(str, (int) Printer.TextAlign.Left, 0, 24, false);
Is equivalent to this in release mode:
var receipt = new StringBuilder();
// ...
var str = receipt.ToString();
Printer.getInstance().print(str, (int) Printer.TextAlign.Left, 0, 24, false);
GC.KeepAlive(str);
When your string is given to the native code, it's not tracked by the GC anymore. Therefore, it could be collected which will cause errors in the native part.
In theory, the marshaller automatically protects you from such situations when using common types (such as string), as described here: https://learn.microsoft.com/en-us/dotnet/framework/interop/copying-and-pinning?redirectedfrom=MSDN. That's another puzzle, but it's possible that .net compact framework has a different marshaller and doesn't protect you automatically. Unfortunately it's hard to find specific documentation on the subject.
What surprised me when you decompiled the method is how the native call doesn't actually receive the original string but text.TrimEnd(). It means that the lifetime of the original value shouldn't have any impact (since the native code receives a different string). However, it turns out that .TrimEnd returns the original string when there's nothing to trim. When you added a space, it started crashing even with the version of the code that extends the lifetime of the string. That's because now you're extending the lifetime of the wrong string (since TrimEnd() will return a different instance, and that's the one that will be used by the native code).
I believe that the code working in Release is pure luck. Maybe it just changes the timing of the garbage collector and you don't run into that specific issue, but it could cause problems in the future. Out of caution, I would suggest you to:
Trim the string before calling Printer.getInstance().print
Call GC.KeepAlive on the trimmed string after the call to print
If you want to be extra safe, you can pin the string instead of calling GC.KeepAlive
I wish I could give more than theories, but I believe you're running into specificities of .net compact framework. If somebody with more experience on the subject reads this and could give more information, that would be much appreciated.

C# environment stack overflows when starting dll

I have a piece of C# program loading a dll function:
[DllImport("/Users/frk/Workspaces/MySharedLibrary.dll", CallingConvention = CallingConvention.Cdecl, CharSet = CharSet.Auto, EntryPoint = "MyCFunction")]
public static extern int MyFunction( [In][MarshalAs(UnmanagedType.I4)]MyFormat format, [In][MarshalAs(UnmanagedType.LPArray)] byte[] myString, [In][MarshalAs(UnmanagedType.I4)] int myStringLength, [MarshalAs(UnmanagedType.LPArray)] byte[] output, ref UIntPtr outputLength);
and calling it
int result = MyFunction(format, inPut, inputLength, outPut, ref outputLength);
on the C++ side, I have:
MyCPPFunction that works perfectly when called from a C test executable. That MyCPPFunction contains somewhere deep in its dependencies a global const variable declared and initialized in an anonymous namespace:
namespace
{
constexpr unsigned RandTileSize = 256;
std::array<unsigned, RandTileSize * RandTileSize> GenerateSamples()
{
std::array<unsigned, RandTileSize * RandTileSize> samples;
std::mt19937 rd(0);
std::uniform_int_distribution<unsigned> distribution(0, 255);
for (unsigned i = 0; i < RandTileSize * RandTileSize; ++i)
{
samples[i] = distribution(rd);
}
return samples;
};
const auto samples = GenerateSamples();<-- Option#1 this causes a stack overflow when loading the dll in C# environment
unsigned Sample(unsigned index)
{
static const auto samples = GenerateSamples();<-- Option#2 this works and dll loads correctly
return samples[index];
}
}
I am confused here since afaik, the option 1 should allocate memory in the code part of the dll, which the C# environment should deal with right ?
How can we have option #1 not to cause memory allocation problems while loading the dll ?
The lifetime of a static variable in a function within a DLL is from the first time the statement is encountered, to the time the DLL unloads.
The lifetime of a class or file scoped variable is from the time the DLL loads until the time the DLL unloads.
The consequence of this is that in the failing case, your initialisation code is running while the DLL is in the process of loading.
It is not generally a good idea to run nontrivial code in class constructors, as there are limits to what can safely be done inside the loader lock.
In particular if you perform any action which requires dynamically loading another DLL (such as LoadLibrary or calling a delay-load linked function) this is likely to cause difficult-to-diagnose issues.
Without diagnosing exactly what has gone wrong in your case the answer is simple: Use option2 or option 3.
Option 3:
void MyDLLInitialize(){
// Initialize the dll here
}
void MyDLLUninitialize(){
// Uninitialize the dll here
}
Then call these functions from C# before you use any other DLL function, and after you have finished with it, respectively.

P/Invoke sometimes cause Win32 1008 Error with StringBuilder parameters

I have a DLL from which I need to P/Invoke the following C method:
int DAStart(
HANDLE hOpen,
char* IPAddress,
int IPPort,
int threadPriority,
char* name,
char* password,
char* userName)
Using the P/Invoke Assistant and my own research, I've come up with the following C# signature:
[DllImportAttribute("<libName>", EntryPoint="DAStart")]
static extern int DAStart(
IntPtr hOpen,
[MarshalAs(Unmanaged.LPStr)] StringBuilder IPAddress,
int IPPort,
int threadPriority,
[MarshalAs(Unmanaged.LPStr)] StringBuilder name,
[MarshalAs(Unmanaged.LPStr)] StringBuilder password,
[MarshalAs(Unmanaged.LPStr)] StringBuilder userName);
Now, I'm doing the call in the following way:
int port = 3000;
int threadPriority = 20;
DAStart(
this.nativeDllHandle, // an IntPtr class field
new StringBuilder("10.0.10.1"),
int port,
int threadPriority,
new StringBuilder("admin"),
new StringBuilder("admin"),
new StringBuilder("admin"));
Now, sometimes this works just fine, but something I get a Win32 Error 1008 - An attempt was made to reference a token that does not exist on following calls to this library.
Could it be that my StringBuilder objects get garbage collected so that the reference no longer exist if the native code tries to use it? Should I keep a reference for each one of them?
Would an IntPtr be better solution for passing my strings in that case?
** UPDATE **
This is all the API documentation I have for DAStart:
Inputs
HANDLE hInfo The handle returned by the DAOpen
char *IPAdress_in IP Address of the TMEE Server
int IPPort Console Port of the TMEE Server (default is port 3000)
int threadPriority The thread priority for the send file thread.
char *name Not used in Hardware DLL
char *password Not used in Hardware DLL
char *username Not used in Hardware DLL
Returns
ERROR_SUCCESS 0
ERROR_BAD_HANDLE -1
ERROR_BIND_FAILED -10
Comments
The DAStart API connects the client dll to the active TMEE Server service. The client thread is started with a priority set to the threadPriority parameter. The IP address parameter must be set to the address of the TMEE Server. The port parameter must be set to the port the TMEE Server listens on for Console connections (the default port the Consoles use is 3010). The Console thread is started with a priority set to the threadPriority parameter.
In the comments you indicate that the DLL takes a copy of the char* pointers (and not the contents of the strings) and then modifies the contents of the strings after DAStart returns.
In the face of this very unconventional interface design your only option is to take charge of the marshaling of the string parameters. You cannot use StringBuilder since the char* passed to the native code is valid only during the pinvoke call. It cannot be relied upon to be valid after the pinvoke call returns.
So your only solution is to pass the string parameters as IntPtr. Allocate the memory with Marshal.StringToHGlobalAnsi. Pass the resulting IntPtr to DAStart. When you are sure the DLL is done with the pointer, deallocate with Marshal.FreeHGlobal.
References outside the managed environment are unknown to the Garbage collector.
Thus you need to keep an additional reference somewhere as long as you need to the object - easiest is via a static property in the calling class.
An IntPtr does not make a difference - it will get collected as well.

Creating C# to C++ bridge: Why do I get a AccessViolationException when calling DLL?

I'm exploring the idea of building a bridge between a DLL plugin for a 3rd party app and a C# app. I'm writing both the plugin DLL and the C# application. The plugin will be loaded into the 3rd party app and then I want to use call the plugin from C# to indirectly get data from the 3rd party app.
I am able to successfully call an exported function from the DLL from C#. For example:
C++ DLL:
extern "C" __declspec(dllexport) char * HelloFromDll()
{
char *result;
result = "Hello from my DLL";
return result;
}
C#:
using System.Runtime.InteropServices;
[DllImport(#"MyDll.dll")]
private static extern string HelloFromDll();
I can then call this DLL function from C# and display the string in the UI. However, as soon as I create an export function that calls a function from my 3rd party app, I get an AccessViolationException. For example,
extern "C" __declspec(dllexport) char * GetData()
{
char *result;
result = 3rdPartyLibrary::SomeFunction();
return result;
}
Through some testing, the error seems to occur as soon as I make a call to a 3rd party function. How can I fix this?
This function is very difficult to use in a C program as well. Returning strings from functions is a poorly supported scenario. There's a memory management problem, it isn't clear who owns the string. In most cases the caller is expected to take ownership of the string and free it after using it. That's not going to work out well for your function, the program will crash since you returned a string literal.
The .NET pinvoke marshaller needs to solve this problem as well. With the extra problem that it cannot use the allocator that's used by the C code. It is going to call CoTaskMemFree (the COM allocator). That causes an undiagnosable memory leak on XP, a crash on Vista and Win7.
Just don't write C code like this. Always let the caller pass the buffer for the string. Now there's no guessing who owns the memory. Like this:
extern "C" __declspec(dllexport) void HelloFromDll(char* buffer, int bufferSize)
{
strcpy_s(result, bufferSize, "Hello from my DLL");
}
With your C# code like this:
[DllImport("foo.dll", CharSet = CharSet.Ansi)]
private static extern void HelloFromDll(StringBuilder buffer, int bufferSize);
...
var sb = new StringBuilder(666);
HelloFromDll(sb, sb.Capacity);
string result = sb.ToString();
From your question it seems that this is the scenario:
ProcessA (3rd party App) --> loads X.DLL --> initializes the plugin --> does other stuff.
ProcessB (Your C# App) --> loads X.DLL --> calls GetData();
Does X.DLL loaded in ProcessA have any mechanism to talk to X.DLL loaded in ProcessB?
if not then this approach is flawed. Your code probbably crashes because "3rdPartyLibrary" class hasn't been initialised in your C# app as it is completely different copy of the DLL.
For you to extract this data you need a query interface defined by X.DLL which can talk across processes, maybe sockets?
Then ProcessB talks to this interface and extracts the data. if using sockets, then your X.DLL would implement both server and client code, where your GetData() would use this mechanism (maybe sockets) and query the data and return it.
So : X.DLL in ProcessA should act like a server.
And: X.DLL (or write a Y.DLL) in ProcessB should act like a client and get this information from ProcessA.
btw, if the query is only needed to be done once, just hard code this is in X.DLL and dump to disk, and then explore at your convinience :-)
Generally, a returned char* needs to be returned as an IntPtr:
[DllImport(#"MyDll.dll")]
private static IntPtr HelloFromDll();
Then, you'll need to convert that IntPtr into a string:
string retVal=Marshal.PtrToStringAnsi(HelloFromDll());
Strings are a bit difficult in P/Invoke. My general rule of thumb is:
Input char* parameter = c# string
Return char * = IntPtr (use PtrToStringAnsi)
Output char* parameter = c# StringBuilder - and be sure to pre-allocate it large enough
before (ie = new StringBuilder(size)) calling the function.

Memory AccessViolationException Error Calling DLL From C#

I'm encountering a strange memory read/write error while calling a compiled DLL from C#. I use DllImport to get a handle to the function we need, which writes a return value to a parametric pointer to an int (i.e., int* out). This function is called multiple times within a thread, and runs successfully over the execution life of the first thread. However, if one launches another thread after the first has completed, the call to the external dll throws an AccessViolationException. Since the multiple calls from the first thread execute successfully, I'm thinking this is somehow related to the first thread not releasing the pointers to the relevant integer parameters(?). If so, how can I explicitly release that memory? Or, perhaps someone has a different insight into what might be going on here? Thank you very much.
EDIT: Danny has requested more specifics, and I'm happy to oblige. Here is where the external routine is invoked:
int success = -1;
unsafe
{
int res = 0;
int* res_ptr = &res;
int len = cmd.ToCharArray().Length;
int* len_ptr = &len;
CmdInterpreter(ref cmd, len_ptr, res_ptr);
success = *res_ptr;
}
Where CmdInterpreter is defined as:
[DllImport("brugs.dll", EntryPoint="CmdInterpreter",
ExactSpelling=false, CallingConvention = CallingConvention.StdCall)]
public static unsafe extern void CmdInterpreter(ref string cmd, int *len, int *res);
Please let me know if there is any additional info that would be helpful. Thank you!
Given that the problem only occurs when multiple threads are involved, it may be that the command interpreter DLL is using some sort of thread-local storage and doing it incorrectly. It could also have to do with the COM initialization state of the second thread (the one that generates the error).
It would be interesting to know what happens if you launch your new thread and have it call into the DLL before making any calls into the DLL on your first/main thread. If it works, that might support the thread-local storage theory. If it fails, that would support the COM state theory.
It may be the [DllImport]. If you post the [DllImport] signature, and the DLL's ptototype, maybe we can spot a problem.
I read that the Managed, Native, and COM Interop Team released the PInvoke Interop Assistant on CodePlex. http://www.codeplex.com/clrinterop/Release/ProjectReleases.aspx?ReleaseId=14120

Categories

Resources