...one of the most highly
regarded and expertly designed C++ library projects in the
world.
— Herb Sutter and Andrei
Alexandrescu, C++
Coding Standards
You can just allocate a portion of a shared memory segment, copy the message to that buffer, send the offset of that portion of shared memory to another process, and you are done. Let's see the example:
#include <boost/interprocess/managed_shared_memory.hpp> #include <cstdlib> //std::system #include <sstream> int main (int argc, char *argv[]) { using namespace boost::interprocess; if(argc == 1){ //Parent process //Remove shared memory on construction and destruction struct shm_remove { shm_remove() { shared_memory_object::remove("MySharedMemory"); } ~shm_remove(){ shared_memory_object::remove("MySharedMemory"); } } remover; //Create a managed shared memory segment managed_shared_memory segment(create_only, "MySharedMemory", 65536); //Allocate a portion of the segment (raw memory) managed_shared_memory::size_type free_memory = segment.get_free_memory(); void * shptr = segment.allocate(1024/*bytes to allocate*/); //Check invariant if(free_memory <= segment.get_free_memory()) return 1; //An handle from the base address can identify any byte of the shared //memory segment even if it is mapped in different base addresses managed_shared_memory::handle_t handle = segment.get_handle_from_address(shptr); std::stringstream s; s << argv[0] << " " << handle; s << std::ends; //Launch child process if(0 != std::system(s.str().c_str())) return 1; //Check memory has been freed if(free_memory != segment.get_free_memory()) return 1; } else{ //Open managed segment managed_shared_memory segment(open_only, "MySharedMemory"); //An handle from the base address can identify any byte of the shared //memory segment even if it is mapped in different base addresses managed_shared_memory::handle_t handle = 0; //Obtain handle value std::stringstream s; s << argv[1]; s >> handle; //Get buffer local address from handle void *msg = segment.get_address_from_handle(handle); //Deallocate previously allocated memory segment.deallocate(msg); } return 0; }
You want to create objects in a shared memory segment, giving a string name to them so that any other process can find, use and delete them from the segment when the objects are not needed anymore. Example:
#include <boost/interprocess/managed_shared_memory.hpp> #include <cstdlib> //std::system #include <cstddef> #include <cassert> #include <utility> int main(int argc, char *argv[]) { using namespace boost::interprocess; typedef std::pair<double, int> MyType; if(argc == 1){ //Parent process //Remove shared memory on construction and destruction struct shm_remove { shm_remove() { shared_memory_object::remove("MySharedMemory"); } ~shm_remove(){ shared_memory_object::remove("MySharedMemory"); } } remover; //Construct managed shared memory managed_shared_memory segment(create_only, "MySharedMemory", 65536); //Create an object of MyType initialized to {0.0, 0} MyType *instance = segment.construct<MyType> ("MyType instance") //name of the object (0.0, 0); //ctor first argument //Create an array of 10 elements of MyType initialized to {0.0, 0} MyType *array = segment.construct<MyType> ("MyType array") //name of the object [10] //number of elements (0.0, 0); //Same two ctor arguments for all objects //Create an array of 3 elements of MyType initializing each one //to a different value {0.0, 0}, {1.0, 1}, {2.0, 2}... float float_initializer[3] = { 0.0, 1.0, 2.0 }; int int_initializer[3] = { 0, 1, 2 }; MyType *array_it = segment.construct_it<MyType> ("MyType array from it") //name of the object [3] //number of elements ( &float_initializer[0] //Iterator for the 1st ctor argument , &int_initializer[0]); //Iterator for the 2nd ctor argument //Launch child process std::string s(argv[0]); s += " child "; if(0 != std::system(s.c_str())) return 1; //Check child has destroyed all objects if(segment.find<MyType>("MyType array").first || segment.find<MyType>("MyType instance").first || segment.find<MyType>("MyType array from it").first) return 1; } else{ //Open managed shared memory managed_shared_memory segment(open_only, "MySharedMemory"); std::pair<MyType*, managed_shared_memory::size_type> res; //Find the array res = segment.find<MyType> ("MyType array"); //Length should be 10 if(res.second != 10) return 1; //Find the object res = segment.find<MyType> ("MyType instance"); //Length should be 1 if(res.second != 1) return 1; //Find the array constructed from iterators res = segment.find<MyType> ("MyType array from it"); //Length should be 3 if(res.second != 3) return 1; //We're done, delete all the objects segment.destroy<MyType>("MyType array"); segment.destroy<MyType>("MyType instance"); segment.destroy<MyType>("MyType array from it"); } return 0; }
Boost.Interprocess offers offset_ptr smart pointer family as an offset pointer that stores the distance between the address of the offset pointer itself and the address of the pointed object. When offset_ptr is placed in a shared memory segment, it can point safely objects stored in the same shared memory segment, even if the segment is mapped in different base addresses in different processes.
This allows placing objects with pointer members in shared memory. For example, if we want to create a linked list in shared memory:
#include <boost/interprocess/managed_shared_memory.hpp> #include <boost/interprocess/offset_ptr.hpp> using namespace boost::interprocess; //Shared memory linked list node struct list_node { offset_ptr<list_node> next; int value; }; int main () { //Remove shared memory on construction and destruction struct shm_remove { shm_remove() { shared_memory_object::remove("MySharedMemory"); } ~shm_remove(){ shared_memory_object::remove("MySharedMemory"); } } remover; //Create shared memory managed_shared_memory segment(create_only, "MySharedMemory", //segment name 65536); //Create linked list with 10 nodes in shared memory offset_ptr<list_node> prev = 0, current, first; int i; for(i = 0; i < 10; ++i, prev = current){ current = static_cast<list_node*>(segment.allocate(sizeof(list_node))); current->value = i; current->next = 0; if(!prev) first = current; else prev->next = current; } //Communicate list to other processes //. . . //When done, destroy list for(current = first; current; /**/){ prev = current; current = current->next; segment.deallocate(prev.get()); } return 0; }
To help with basic data structures, Boost.Interprocess offers containers like vector, list, map, so you can avoid these manual data structures just like with standard containers.
Boost.Interprocess allows creating complex objects in shared memory and memory mapped files. For example, we can construct STL-like containers in shared memory. To do this, we just need to create a special (managed) shared memory segment, declare a Boost.Interprocess allocator and construct the vector in shared memory just if it was any other object.
The class that allows this complex structures in shared memory is called
boost::interprocess::managed_shared_memory
and it's easy to use. Just execute this example without arguments:
#include <boost/interprocess/managed_shared_memory.hpp> #include <boost/interprocess/containers/vector.hpp> #include <boost/interprocess/allocators/allocator.hpp> #include <string> #include <cstdlib> //std::system using namespace boost::interprocess; //Define an STL compatible allocator of ints that allocates from the managed_shared_memory. //This allocator will allow placing containers in the segment typedef allocator<int, managed_shared_memory::segment_manager> ShmemAllocator; //Alias a vector that uses the previous STL-like allocator so that allocates //its values from the segment typedef vector<int, ShmemAllocator> MyVector; //Main function. For parent process argc == 1, for child process argc == 2 int main(int argc, char *argv[]) { if(argc == 1){ //Parent process //Remove shared memory on construction and destruction struct shm_remove { shm_remove() { shared_memory_object::remove("MySharedMemory"); } ~shm_remove(){ shared_memory_object::remove("MySharedMemory"); } } remover; //Create a new segment with given name and size managed_shared_memory segment(create_only, "MySharedMemory", 65536); //Initialize shared memory STL-compatible allocator const ShmemAllocator alloc_inst (segment.get_segment_manager()); //Construct a vector named "MyVector" in shared memory with argument alloc_inst MyVector *myvector = segment.construct<MyVector>("MyVector")(alloc_inst); for(int i = 0; i < 100; ++i) //Insert data in the vector myvector->push_back(i); //Launch child process std::string s(argv[0]); s += " child "; if(0 != std::system(s.c_str())) return 1; //Check child has destroyed the vector if(segment.find<MyVector>("MyVector").first) return 1; } else{ //Child process //Open the managed segment managed_shared_memory segment(open_only, "MySharedMemory"); //Find the vector using the c-string name MyVector *myvector = segment.find<MyVector>("MyVector").first; //Use vector in reverse order std::sort(myvector->rbegin(), myvector->rend()); //When done, destroy the vector from the segment segment.destroy<MyVector>("MyVector"); } return 0; };
The parent process will create an special shared memory class that allows easy construction of many complex data structures associated with a name. The parent process executes the same program with an additional argument so the child process opens the shared memory and uses the vector and erases it.
Just like a vector, Boost.Interprocess allows creating maps in shared memory and memory mapped files. The only difference is that like standard associative containers, Boost.Interprocess's map needs also the comparison functor when an allocator is passed in the constructor:
#include <boost/interprocess/managed_shared_memory.hpp> #include <boost/interprocess/containers/map.hpp> #include <boost/interprocess/allocators/allocator.hpp> #include <functional> #include <utility> int main () { using namespace boost::interprocess; //Remove shared memory on construction and destruction struct shm_remove { shm_remove() { shared_memory_object::remove("MySharedMemory"); } ~shm_remove(){ shared_memory_object::remove("MySharedMemory"); } } remover; //Shared memory front-end that is able to construct objects //associated with a c-string. Erase previous shared memory with the name //to be used and create the memory segment at the specified address and initialize resources managed_shared_memory segment (create_only ,"MySharedMemory" //segment name ,65536); //segment size in bytes //Note that map<Key, MappedType>'s value_type is std::pair<const Key, MappedType>, //so the allocator must allocate that pair. typedef int KeyType; typedef float MappedType; typedef std::pair<const int, float> ValueType; //Alias an STL compatible allocator of for the map. //This allocator will allow to place containers //in managed shared memory segments typedef allocator<ValueType, managed_shared_memory::segment_manager> ShmemAllocator; //Alias a map of ints that uses the previous STL-like allocator. //Note that the third parameter argument is the ordering function //of the map, just like with std::map, used to compare the keys. typedef map<KeyType, MappedType, std::less<KeyType>, ShmemAllocator> MyMap; //Initialize the shared memory STL-compatible allocator ShmemAllocator alloc_inst (segment.get_segment_manager()); //Construct a shared memory map. //Note that the first parameter is the comparison function, //and the second one the allocator. //This the same signature as std::map's constructor taking an allocator MyMap *mymap = segment.construct<MyMap>("MyMap") //object name (std::less<int>() //first ctor parameter ,alloc_inst); //second ctor parameter //Insert data in the map for(int i = 0; i < 100; ++i){ mymap->insert(std::pair<const int, float>(i, (float)i)); } return 0; }
For a more advanced example including containers of containers, see the section Containers of containers.