Memory strategy when using IOCP

Question

Memory strategy when using IOCP

I am new to Windows IOCP, I am currently rewriting the network part of our server with IOCP. I'm trying to figure out how to handle memory (Overlapped / WSABUF objects). I have a small structure that comes from OVERLAPPED and contains WSABUF and a few other fields.

I tried to save a ring buffer for these objects and reuse them, but it didn't work. I thought that when the completion routine was called, one of the objects would be passed back to me and I could mark it as available again.

When I allocate them on the heap (and store them in a vector), how can I know when it is safe to delete / reuse it?

Thanks Michael

+3

c ++ windows networking

MUXCAH June 24. '17 at 3:38

source to share

No one has answered this question yet

Check out similar questions:

2299

When should static_cast, dynamic_cast, const_cast, and reinterpret_cast be used?

1783

C ++ 11 introduced a standardized memory model. What does it mean? And how will this affect C ++ programming?

1699

What is a smart pointer and when should I use it?

1305

When to use virtual destructors?

951

Can a local memory variable be accessed outside of its scope?

five

When using IOCP, should WSAOVERLAPPED hEvent be set to NULL, or a valid handle to the WSAEVENT object?

3

Quick design question about C ++ container classes in shared memory

3

Caching OVERLAPPED Structures Using IOCP on Windows

2

WSARecv sometimes returns an "invalid descriptor (error # 6)" for a socket associated with an IOCP port. (C ++)

1

Preventing automatic destruction of variables through dynamic memory management

Memory strategy when using IOCP

More articles: