I will sketch the scenario I would like to get working below.
I have one main application.
That application, based on user interactions, can load other applications inside a secure environment/shell. This means these child applications cannot interact with the OS anymore, nor with each other.
The parent program can at any time call functions of these child programs.
The child program can at any time call functions of these parent programs.
Does anyone know how to implement this in C++? Preferably both parent and child should be written in C++.
The performance of loading the child applications inside the parent application doesn't matter. The only thing that matters is the performance of the communication between child and parent when calling functions of each other.
You will have to write your own compiler.
Consider: No normal OS supports what you want. You want both executables to run inside a single process, yet that process may or may not make OS calls depending on some weirdness inside the process which the OS doesn't understand at all.
This is no longer a problem with your custom compiler, as it simply will not create the offending instructions. It's similar to Java and .Net, which also prevent such OS calls outside their control.
A portable solution: Google Native Client
One possible Linux solution:
Make AppArmor profile with "hats" (a "hat" is a sandboxing configuration to which the application can switch programmatically with libapparmor),
have the main application create a "pipe",
have the main application "fork",
change into a "hat" corresponding to the child application,
"exec" the child application,
the main application and the child application communicate via the "pipe" created earlier.
If you want a (semi)crossplatform way to do this you can use RPC to call functions in another process. It's going to work on anything that supports the distributed computing environment. It's been around for some time and the msdn documentation states that parts of windows use it for inter process communication so it's probably fast enough. Here's a tutorial on msdn that should get you up and running http://msdn.microsoft.com/en-us/library/windows/desktop/aa379010.aspx The bad part is that I haven't been able to find a tutorial about using it on linux.
If you don't want to use RPC or find it too hard to find good documentation on the subject, you can use the standard IPC(Inter Process Communication) mechanisms from unix systems to signal your process that should call a certain function. I'd recommend a message queue because it's very fast and lightweight. You can find a tutorial here: http://www.cs.cf.ac.uk/Dave/C/node25.html
I am not familiar with OS restrictions in above answers. However, I found an easy way to solve this problem. I hope it helps and does not have a technical issue. I used Linux OS. Suppose I want to call C++ program B inside another C++ program A. I wrote a perl script (such as PerlScript.pl) that contains a system call to run program B. Then in A, I did a system call like system("perl PerlScript.pl") that ask perl to run B for me.
Related
I run a process thanks to the Boost library with this code using the boost::process::child class. My program works well. I am launching a process done by someone else, and I only need to send things on its stdin, and read its stdout. But I consider I cannot trust the child process owner, so I would like the process to be able to communicate with my main process with standard input/output, but I would like to prevent him from creating files, starting other processes (running command in terminal), or making other unsafe things.
I saw some ways to achieve that, but they use other APIs, and I cannot apply that to my boost process. I want my program to be cross-platform (Windows + Linux).
So, is there any way to achieve that, thanks to the boost library, or with other libraries (but which would be appropriate using the child class from the boost::process library) ?
EDIT: I would like my program to be easily given to a "ordinary" person. So I cannot use a solution which would imply the user to install other things than my executable (and possibly some libraries/data in a folder)
Its been a while since I've had to do this and in the past I've used "spawn" to create processes.
Now I want to launch processes from my application asynchronously so my application continues to execute in the background and does not get held up by launching the process.
I also want to be able to communicate with the launched processes. When I launch the process I will send it the launchers process id so that the launched process can communicate with the launcher using it's pid.
What is the best method to use that is not specific to any platform / operating system, I'm looking for a solution that is multi-platform?
I'm writing this in C++, I don't want a solution that ties me to any third party licensed product.
I don't want to use threads, the solution must be for creating new processes.
A portable to launch a new process is std::system.
#include <cstdlib>
int main() {
std::system("./myapp");
return 0;
}
if you use linux and you want to share handles/memory between processes, fork is what you are looking for
Try Boost.Process.
Boost.Process provides a flexible framework for the C++ programming language to manage running programs, also known as processes. It empowers C++ developers to do what Java developers can do with java.lang.Runtime/java.lang.Process and .NET developers can do with System.Diagnostics.Process. Among other functionality, this includes the ability to manage the execution context of the currently running process, the ability to spawn new child processes, and a way to communicate with them them using standard C++ streams and asynchronous I/O.
The library is designed in a way to transparently abstract all process management details to the user, allowing for painless development of cross-platform applications. However, as such abstractions often restrict what the developer can do, the framework allows direct access to operating system specific functionality - obviously losing the portability features of the library.
Example code to run and wait to finish for child process from the site:
bp::child c(bp::search_path("g++"), "main.cpp");
while (c.running())
do_some_stuff();
c.wait(); //wait for the process to exit
int result = c.exit_code();
I'll plug my own little (single header) library:
PStreams allows you to run another program from your C++ application and to transfer data between the two programs similar to shell pipelines.
In the simplest case, a PStreams class is like a C++ wrapper for the POSIX.2 functions popen(3) and pclose(3), using C++ iostreams instead of C's stdio library.
The library provides class templates in the style of the standard iostreams that can be used with any ISO C++ compiler on a POSIX platform. The classes use a streambuf class that uses fork(2) and the exec(2) family of functions to create a new process and creates up to three pipes to write/read data to/from the process.
I have a DLL which invokes an application running underneath. This DLL is loaded by several other applications/processes simultaneously.
So, Basically Architecture is:
My Problem is if Application(s) using DLL is crashed, I want to execute an exit sequence in my Base Application and Exit it.
How can I detect that this DLL is no longer used by any application?
Is there any thing like Load Count of DLL which I can keep track of?
Another glitch is I may have to monitor this via a C# application but that is a further thing.
If you can shell an external program to do the check, you can use this:
http://technet.microsoft.com/en-us/sysinternals/bb896656
Your image is not a good model for what really happens in Windows. Every process gets its own copy of the DLL. The code inside the DLL is shared in RAM but not its data. There are ways to share data as well but that's not otherwise common, a memory mapped file is the far more typical approach.
Windows doesn't give cheap way to find out if a DLL is loaded into a process. There is no notification mechanism either. Whatever you do, it has to start with the processes first. That works in C# too, you could use the Process.Modules property.
Just keeping track of the processes you know that load the DLL is probably sufficient, when the process no longer runs then you can safely assume it doesn't have the DLL loaded anymore either. Use the Process.Exited event or use WMI as shown in this answer.
You could host the DLL in the Base Application either and implement your own ref count using shared sections, easy, and often used. As already mentioned, one possibility would be to implemented your ref counting inside your DLL entry point and detect DLL_PROCESS_DETACH, DLL_THREAD_DETACH, DLL_PROCESS_DETACH, etc according to you specifications.
I have a C++ program that captures videos, and I would like to be able create a command-line program to update its frame rate, image format, etc on the fly.
How can I do this without halting the entire program? I need it to be able to wait for user input, but still capture videos at the same time. I know this will probably involve some kind of multi-threading, which I am entirely new to. Some suggestions/links would be nice.
Than you all,
Are you developing this for a specific platform or does it need to be platform independent?
If you are developing for windows you should look into the win32 API. specifically beginthread or _beginthreadex on msdn
I'm not too familiar with *nix development but pthreads i believe would do the trick and can be used in Windows and *nix
Another option would be to use the BOOST libraries. BOOST can be used on Windows and *nix systems. Below is a link to the BOOST Thread documentation.
http://www.boost.org/doc/libs/1_44_0/doc/html/thread.html
I find BOOST Threads a lot easier to use than WIN32 Threads and at the same time you're not tied down to a specific platforms API.
Create a thread to handle video, whilst using the main thread to wait for input. Thread creation depends on platform, and can be a little overwhelming to those who are new. You will need a mutex on variables that can be altered through the command line, and you'll need to look up of how to make your code "thread safe".
In the days before multi-threading it was possible to solve this as well by regularly peeking into the keyboard buffer from time to time. I mention this just as an alternative to opening the multi-threading box which often gives you more than you bargain for.
EDIT: I read now a bit more carefully what you want to achieve, having a console program to update another program with new settings. I think what you then need is for the programs to communicate with one another. Look at boost::interprocess for that.
I've already saw many tutorials and articles about hooking, yet I don't quite understand it. Mainly because every single example uses different solution.
I know I will have to implement something that will keep the hook alive. Usually it's some kind of while cycle. Q1: If this loop was in some class with callbacks, will it prevent the thread from executing them?
I know it will take a while, but I would highly appreciate some well explained example of global keyboard hook. Or simply link me to some working example with binaries. (Trust me, I've been trying to google it last few hours).
Thank you
I know I will have to implement something that will keep the hook alive
No, that's not a concern. A global hook requires a DLL with the callback. That DLL gets injected in all running processes. It will stay loaded in the process until you call UnHookWindowsHookEx() or the process terminates, whichever comes first.
Do note that you can also hook the keyboard with WH_KEYBOARD_LL. That's not a global hook, Windows will switch context to your program and make the callback. It is much easier to use since you don't need an IPC mechanism with the injected DLL that the global hook requires. The low level hook stays active until you unhook, the thread that owns the message queue terminates or your process terminates, whichever comes first.
There's a skeletal keyboard hook with downloadable code on my web site here (note the "download" button at the bottom of the page). The article discusses some things you need to understand like shared sections, and informs you of Kyle Marsh's excellent article on Win32 hooks (if some idiot at MSDN hasn't taken it down by now, for being insufficiently .netty).
The source code is in the example, it just needs a makefile/sln building for it (I don't know what compiler/version you'll be using). Code spookily similar to that has been in a shipping commercial product for a decade, so I know it works.
Note that integrity level issues can reduce the utility of hooking in Vista and W7.
your program will need to stay alive during the hook. Your program is the one that calls SetWindowsHookEx / UnSetWindowsHookEx (or whatever it's called). Between those calls, you will indeed have a message loop (this is probably the while loop you're talking about) just like any typical windows program.
But because your program is a different process than the ones you're hooking, your message loop will not cause other processes to hang. It's called multitasking :)