Boost asio async_accept memory leak问题分析

可以看下stackflow上的问题描述：

Using boost::asio i use async_accept to accept connections. This works good, but there is one issue and i need a suggestion how to deal with it. Using typical async_accept:

Listener::Listener(int port)
: acceptor(io, ip::tcp::endpoint(ip::tcp::v4(), port))
, socket(io) {
start_accept();
}

void Listener::start_accept() {
Request *r = new Request(io);
acceptor.async_accept(r->socket(),
boost::bind(&Listener::handle_accept, this, r, placeholders::error));
}

Works fine but there is a issue: Request object is created with plain new so
it can memory "leak". Not really a leak, it leaks only at program stop, but i want to make valgrind happy.

Sure there is an option: i can replace it with shared_ptr, and pass it to every event handler. This will work until program stop, when asio io_service is
stopping, all objects will be destroyed and Requestwill be free'd. But this way i always must have an active asio event for Request,
or it will be destroyed! I think its direct way to crash so i dont like this variant, too.

UPD Third variant: 

Listener

 holds
list of shared_ptr to active connections. Looks great and i prefer to use this unless some better way will be found. The drawback is: since this schema allows to do "garbage collection" on idle connects, its not safe: removing connection pointer from Listener
will immediately destroy it, what can lead to segfault when some of connection's handler is active in other thread. Using mutex cant fix this cus in this case we must lock nearly anything.

Is there a way to make acceptor work with connection management some beautiful and safe way? I will be glad to hear any suggestions.
问题在于如果程序关掉，连接来了，new 出来的tcp::socket没有被释放。
所以官方文档有讲到：

On destruction, the 

io_service

 performs
the following sequence of operations:

For each service object 

svc

 in the 

io_service

 set,
in reverse order of the beginning of service object lifetime, performs 

svc->shutdown_service()

.
Uninvoked handler objects that were scheduled for deferred invocation on the 

io_service

,
or any associated strand, are destroyed.
For each service object 

svc

 in the 

io_service

 set,
in reverse order of the beginning of service object lifetime, performs 

delete static_cast<io_service::service*>(svc)

.


Remarks
The destruction sequence described above permits programs to simplify their resource management by using 

shared_ptr<>

.
Where an object's lifetime is tied to the lifetime of a connection (or some other sequence of asynchronous operations), a 

shared_ptr

 to the object would be bound into the handlers for all asynchronous operations associated
with it. This works as follows:

When a single connection ends, all associated asynchronous operations complete. The corresponding handler objects are destroyed, and all 

shared_ptr

 references to
the objects are destroyed.
To shut down the whole program, the 

io_service

 function 

stop()

 is
called to terminate any 

run()

 calls as soon as possible. The 

io_service

 destructor
defined above destroys all handlers, causing all 

shared_ptr

 references to all connection objects to be destroyed.

提倡用智能指针。其实可以嘛。