We have *.exe application which is also out-process COM server.
The main thread is doing some network routine: it receives data packets and puts them into queue.
COM client, VBA for example, uses COM server and wants to use queue too. Despite the fact that they are in the same address space the question is:
How can we provide an opportunity for COM client to use the queue simultaneously with exe process.
There was an idea to use shared memory, but with no success
UPD:
I've tryed to use boost::interprocess.
Due to the same address space I've wanted just share object pointer.
std::vector<int> //just example of MyType
exe part:
main()
...
using namespace boost::interprocess;
struct shm_remove
{
shm_remove() { shared_memory_object::remove("SharedMemory"); }
~shm_remove(){ shared_memory_object::remove("SharedMemory"); }
} remover;
managed_shared_memory segment(open_or_create, "SharedMemory", 65536);
std::vector<int>** instance = segment.construct<std::vector<int>* >
("my_instance") //name of the object
(); //ctor first argument
*instance = new std::vector<int>();
(*instance)->push_back(1);
// initialize the COM library
::CoInitialize(NULL);`enter code here`
COM part:
HRESULT __stdcall CoMyCOMServer::Add(int *value)
{
cout << "Add()\n";
// this line goes out of debug, then VBA get error
managed_shared_memory segment(open_only, "SharedMemory");
std::vector<int>* *res = segment.find<std::vector<int>* > ("my_instance").first;
(*res)->push_back(*value);
return S_OK;
}
COM client(VBA) tells
Method "ADD" of object "IMyCOMServer" failed
Dim obj As IMyCOMServer
Set obj = CreateObject("MyCOMServer.object")
obj.Add (2)
UPD2:
I've just surrounded Com part with try{}catch{} and found out that the exception with the message "File not found"
Related
There are other similar questions on this site, but they either do not related to winsock2 or they are suitable only for use with ipv4 address spaces. The default compiler for Visual Studio 2019 produces an error when the ntoa function is used, hence an ipv4 and ipv6 solution is required.
I did once produce the code to do this for a Linux system however I am currently at work and do not have access to that. It may or may not be "copy and paste"-able into a windows environment with winsock2. (Edit: I will of course add that code later this evening, but of course it might not be useful.)
The following contains an example, however this is an example for client side code, not server side code.
https://www.winsocketdotnetworkprogramming.com/winsock2programming/winsock2advancedInternet3c.html
Here, the getaddrinfo() function is used to obtain a structure containing matching ipv4 and ipv6 addresses. To obtain this information there is some interaction with DNS, which is not required in this case.
I have some server code which calls accept() (after bind and listen) to accept a client connection. I want to be able to print the client ip address and port to stdout.
The most closely related question on this site is here. However the answer uses ntoa and is only ipv4 compatible.
What I have so far:
So far I have something sketched out like this:
SOCKET acceptSocket = INVALID_SOCKET;
SOCKADDR_IN addr; // both of these are NOT like standard unix sockets
// I don't know how they differ and if they can be used with standard
// unix like function calls (eg: inet_ntop)
int addrlen = sizeof addr;
acceptSocket = accept(listenSocket, (SOCKADDR*)&addr, &addrlen);
if(acceptSocket == INVALID_SOCKET)
{
// some stuff
}
else
{
const std::size_t addrbuflen = INET6_ADDRSRTLEN;
char addrbuf[addrbuflen] = '\0'
inet_ntop(AF_INET, (void*)addr.sin_addr, (PSTR)addrbuf, addrbuflen);
// above line does not compile and mixes unix style function calls
// with winsock2 structures
std::cout << addrbuf << ':' << addr.sin_port << std::endl;
}
getpeername()
int ret = getpeername(acceptSocket, addrbuf, &addrbuflen);
// addrbuf cannot convert from char[65] to sockaddr*
if(ret == ???)
{
// TODO
}
You need to access the SOCKADDR. This is effectively a discriminated union. The first field tells you whether its an IPv4 (==AF_INET) or IPv6 (==AF_INET6) address. Depending on that you cast the addr pointer to be either struct sockaddr_in* or struct sockaddr_in6*, and then read off the IP address from the relevant field.
C++ code snippet in vs2019:
char* CPortListener::get_ip_str(struct sockaddr* sa, char* s, size_t maxlen)
{
switch (sa->sa_family) {
case AF_INET:
inet_ntop(AF_INET, &(((struct sockaddr_in*)sa)->sin_addr),
s, maxlen);
break;
case AF_INET6:
inet_ntop(AF_INET6, &(((struct sockaddr_in6*)sa)->sin6_addr),
s, maxlen);
break;
default:
strncpy(s, "Unknown AF", maxlen);
return NULL;
}
return s;
}
Example:
{
...
char s[INET6_ADDRSTRLEN];
sockaddr_storage ca;
socklen_t al = sizeof(ca);
SOCKET recv = accept(sd, (sockaddr*)&ca, &al);
pObj->m_ip = get_ip_str(((sockaddr*)&ca),s,sizeof(s));
}
I’m developing a system for our application to get data from an external device. As soon as I send it a specific message, it sends back short messages to us 10x/second (so about 1 message per 100 milliseconds). I’m using Boost for this communication.
The process is rather simple: I create the socket, send the message, giving it a handler for the message receive:
// Header file:
...
std::unique_ptr<boost::asio::io_service> _theIOService;
std::unique_ptr<boost::asio::ip::tcp::socket> _theSocket;
int size_of_the_data = 100;
std::vector<char> _raw_buffer = std::vector<char>(size_of_the_data);
boost::asio::mutable_buffers_1 _data_buffer = boost::asio::buffer(_raw_buffer, size_of_the_data);
...
// Implementation file:
...
void DeviceDataListener::initiateTransfer() {
// create and connect the socket up here
...
// send the message
boost::system::error_code error;
boost::asio::write(*_theSocket,
boost::asio::buffer(beginMessage),
boost::asio::transfer_all(), error);
// start the receive
auto handler = boost::bind(&SCUDataListener::dataHandler, this, _1, _2);
_theSocket->async_receive( _data_buffer, handler );
std::thread run_thread([&]{ _theIOService->run(); });
...
}
void DeviceDataListener::dataHandler (
const boost::system::error_code& error, // Result of operation.
std::size_t bytes_transferred // Number of bytes received.
) {
int foo = bytes_transferred;
// this line crashes application
char* pData = static_cast<char*>(_data_buffer.data());
}
It works, my handler gets called immediately, as it should. The problem is, I can’t get the data out of _data_buffer. This:
auto it = _data_buffer.begin();
causes a crash, even though _data_buffer is valid. This:
const char* pData = static_cast<char*>(_data_buffer.data());
won’t compile. The error is “Method 'data' could not be resolved”. The mutable_buffer_1 API says data() is a completely valid method that returns the beginning of the memory range.
Inspecting via a debugger, I can see that there is no error and I can see data as a member of _data_buffer and the memory address it contains does contain the data we’re expecting. The thing is, I can’t get to it via code. Does anyone know how to get to the data in a Boost mutable_buffers_1?
We’re using Eclipse CDT, C++11 and gcc running on Linux.
“Method 'data' could not be resolved”.
this error may be true, but it depends on what version of Boost you use. data() is member of mutable_buffer since >= 1.66 version. Because mutable_buffer is the base class for mutable_buffers_1 your code should compile if you use at least 1.66 version of Boost.
If your version is < 1.66 you should use
char* p1 = boost::asio::buffer_cast<char*>(_data_buffer);
to get the pointer to data in the buffer.
_data_buffer.begin();
you should not use begin() method, it returns pointer to mutable_buffer_1 itself. This method is used by internal functions of asio-boost library, for instance to copy sequence of buffers, then begin() points the particular buffer to be copied.
I manual map dll and i can't get MODULEINFO for it's working region with GetModuleInformation (it's always answer for me with "Unable to obtain module")?. That happens because that function tries to get data from the module list in the process environment block. But a manually mapped dll is usually not linked in that list unless of course you manually add a new list entry.
It doesn't use the info from the header (or at least not directly).
So i already has dllBase that is hModule. So now i only need to get it's size.
Is any way to get it without GetModuleInformation?
static void someFunc(HINSTANCE hModule)
{
// all the vars we need for the GetModuleInformation call
MODULEINFO modInfo;
HANDLE hProcess = GetCurrentProcess();
if (GetModuleInformation(hProcess, hModule, &modInfo, sizeof(MODULEINFO)))
{
// some work
}
else {
std::cout << "Unable to obtain module" << std::endl;
}
}
if we want get image size for mapped image in self process - we can read it from SizeOfImage member of IMAGE_OPTIONAL_HEADER - this is size of the mapped as image image in memory (not size on disk)
ULONG GetImageSize(PVOID ImageBase = &__ImageBase)
{
if (PIMAGE_NT_HEADERS pinth = RtlImageNtHeader(ImageBase))
{
return pinth->OptionalHeader.SizeOfImage;
}
return 0;
}
the device driver I'm working on is implementing a virtual device. The logic
is as follows:
static struct net_device_ops virt_net_ops = {
.ndo_init = virt_net_init,
.ndo_open = virt_net_open,
.ndo_stop = virt_net_stop,
.ndo_do_ioctl = virt_net_ioctl,
.ndo_get_stats = virt_net_get_stats,
.ndo_start_xmit = virt_net_start_xmit,
};
...
struct net_device *dev;
struct my_dev *virt;
dev = alloc_netdev(..);
/* check for NULL */
virt = netdev_priv(dev);
dev->netdev_ops = &virt_net_ops;
SET_ETHTOOL_OPS(dev, &virt_ethtool_ops);
dev_net_set(dev, net);
virt->magic = MY_VIRT_DEV_MAGIC;
ret = register_netdev(dev);
if (ret) {
printk("register_netdev failed\n");
free_netdev(dev);
return ret;
}
...
What happens is that somewhere somehow the pointer net_device_ops in
'net_dev' gets corrupted, i.e.
1) create the device the first time (allocated net_dev, init the fields
including net_device_ops,which is
initialized with a static structure containing function pointers), register
the device with the kernel invoking register_netdev() - OK
2) attempt to create the device with the same name again, repeat the above
steps, call register_netdev() which will return negative and we
free_netdev(dev) and return error to the caller.
And between these two events the pointer to net_device_ops has changed,
although nowhere in the code it is done explicitly except the initialization
phase.
The kernel version is 2.6.31.8, platform MIPS. Communication channel between the user space and the kernel is implemented via netlink sockets.
Could anybody suggest what possibly can go wrong?
Appreciate any advices, thanks.
Mark
"The bug is somewhere else. "
The second device should not interact with the existing one. If you register_netdev with an existing name, nevertheless the ndo_init virtual function is called first before the condition is detected and -EEXIST is returned. Maybe your init function does something nasty involving some global variables. (For example, does the code assume there is one device, and stash a global pointer to it during initialization?)
I'm working an application of which only one instance must exist at any given time. There are several possibilities to accomplish this:
Check running processes for one matching our EXE's name (unreliable)
Find the main window (unreliable, and I don't always have a main window)
Create a mutex with a unique name (GUID)
The mutex option seems to me the most reliable and elegant.
However, before my second instance terminates, I want to post a message to the already running instance. For this, I need a handle to the thread (or the process) that owns the mutex.
However, there seems to be no API function to get the creator/owner of a given mutex. Am I just overlooking it? Is there another way to get to this thread/process? Is there another way to go about this?
Update: This guy simply broadcast a message to all running processes. I guess that's possible, but I don't really like it...
This should get you started on the original request to get a process that owns a mutex.
It's in C#, but the Win32 calls are the same.
class HandleInfo
{
[DllImport("ntdll.dll", CharSet = CharSet.Auto)]
public static extern uint NtQuerySystemInformation(int SystemInformationClass, IntPtr SystemInformation, int SystemInformationLength, out int ReturnLength);
[DllImport("kernel32.dll", SetLastError = true)]
internal static extern IntPtr VirtualAlloc(IntPtr address, uint numBytes, uint commitOrReserve, uint pageProtectionMode);
[DllImport("kernel32.dll", SetLastError=true)]
internal static extern bool VirtualFree(IntPtr address, uint numBytes, uint pageFreeMode);
[StructLayout(LayoutKind.Sequential)]
public struct SYSTEM_HANDLE_INFORMATION
{
public int ProcessId;
public byte ObjectTypeNumber;
public byte Flags; // 1 = PROTECT_FROM_CLOSE, 2 = INHERIT
public short Handle;
public int Object;
public int GrantedAccess;
}
static uint MEM_COMMIT = 0x1000;
static uint PAGE_READWRITE = 0x04;
static uint MEM_DECOMMIT = 0x4000;
static int SystemHandleInformation = 16;
static uint STATUS_INFO_LENGTH_MISMATCH = 0xC0000004;
public HandleInfo()
{
IntPtr memptr = VirtualAlloc(IntPtr.Zero, 100, MEM_COMMIT, PAGE_READWRITE);
int returnLength = 0;
bool success = false;
uint result = NtQuerySystemInformation(SystemHandleInformation, memptr, 100, out returnLength);
if (result == STATUS_INFO_LENGTH_MISMATCH)
{
success = VirtualFree(memptr, 0, MEM_DECOMMIT);
memptr = VirtualAlloc(IntPtr.Zero, (uint)(returnLength + 256), MEM_COMMIT, PAGE_READWRITE);
result = NtQuerySystemInformation(SystemHandleInformation, memptr, returnLength, out returnLength);
}
int handleCount = Marshal.ReadInt32(memptr);
SYSTEM_HANDLE_INFORMATION[] returnHandles = new SYSTEM_HANDLE_INFORMATION[handleCount];
using (StreamWriter sw = new StreamWriter(#"C:\NtQueryDbg.txt"))
{
sw.WriteLine("# Offset\tProcess Id\tHandle Id\tHandleType");
for (int i = 0; i < handleCount; i++)
{
SYSTEM_HANDLE_INFORMATION thisHandle = (SYSTEM_HANDLE_INFORMATION)Marshal.PtrToStructure(
new IntPtr(memptr.ToInt32() + 4 + i * Marshal.SizeOf(typeof(SYSTEM_HANDLE_INFORMATION))),
typeof(SYSTEM_HANDLE_INFORMATION));
sw.WriteLine("{0}\t{1}\t{2}\t{3}", i.ToString(), thisHandle.ProcessId.ToString(), thisHandle.Handle.ToString(), thisHandle.ObjectTypeNumber.ToString());
}
}
success = VirtualFree(memptr, 0, MEM_DECOMMIT);
}
}
I don't think there is a trivial way to resolve the actual owner of a Mutex, but the process that owns it can create other secondary items whose lifetimes are tied to it. There are plenty of mechanisms that are suitable for calling back across-process without having a main window.
Register an object in the COM Running Object Table. Clients that are unable to take ownership of the Mutex can lookup the owner via the ROT and call back to the owner. A File Moniker should be suitable for registration here.
Create a chunk of shared memory containing location details for the owner process. From there, write into the buffer the process handle and thread handle of a thread that can receive windows messages, and then use PostThreadMessage() to send a notification. Any other competing process may open the shared memory for read-only to determine where to send a windows message.
Listen in the owner process on a Socket or Named Pipe. Probably overkill and not a good match for your needs.
Use a shared file with locking. I'm not fond of this because the owner will need to poll, and it won't gracefully handle N potential other processes that could be trying to contact the owner at the same time.
Here are reference links for the first two options.
IRunningObjectTable # MSDN ,
File Monikers # MSDN
Creating Named Shared Memory # MSDN
I have never really understood the rational behind using a Mutex which has no signaling capability. I would instead create an event (using CreateEvent) which has the same properties as creating a mutex (i.e. with a name it can return that the object already existed) but you can set the event flag in the new process, as long as the original process is waiting on the event flag it can be notified when it needs to wake itself up.
You could always do it the UNIX way and create a "pid" file, putting the process id of the currently running instance into that file. Then have the app delete the file when it exits.
When a new instance starts up it should verify that the process in the PID file is actually alive as well (in case the app exits abnormally and the file doesn't get deleted)
Create a shared memory area with the fixed name:
http://msdn.microsoft.com/en-us/library/aa366551%28VS.85%29.aspx
Then you can put any structure you like inside, including process id, HWND etc.
There's a portable option: create a socket on a port (with a fixed number) and wait (accept) on it. The second instance of the app will fail since the port is already taken. Then the second instance can connect to the socket of the primary instance and send any information desired.
I hope this helps...
I had similar problems. I am want a function that returns if a single instance of an app is running. Then another function to bring the app to the front. In which I must first deduce the HWND of the already running window.
FindWindow sucks big time. Window titles can change, another window could be using the same class and title, etc.
Then I thought maybe extra data could be stored with a mutex. But I dont see where user data can be stored in a mutex object or event object. But a mutex knows which thread it belongs to and thus which process it belongs to. But as you said, the api doesnt seem to exist.
Many new and complicated looking methods have been suggested here; with the exeception of simply using a file. So I want to add another method, temporary registry keys.
This method is easiest for me as I already built an hkey library. But the win32 registry api is pretty straight forward compared to the horrifying looking shared memory method.