WIP: Socket migration between threads, new zmq_close() semantics

Sockets may now be migrated between OS threads; sockets may not be used by
more than one thread at any time. To migrate a socket to another thread the
caller must ensure that a full memory barrier is called before using the
socket from the target thread.

The new zmq_close() semantics implement the behaviour discussed at:

http://lists.zeromq.org/pipermail/zeromq-dev/2010-July/004244.html

Specifically, zmq_close() is now deterministic and while it still returns
immediately, it does not discard any data that may still be queued for
sending. Further, zmq_term() will now block until all outstanding data has
been sent.

TODO: Many bugs have been introduced, needs testing. Further, SO_LINGER or
an equivalent mechanism (possibly a configurable timeout to zmq_term())
needs to be implemented.
This commit is contained in:
Martin Sustrik
2010-08-06 17:49:37 +02:00
parent b7e0fa972f
commit 05d908492d
47 changed files with 1432 additions and 1411 deletions

View File

@@ -33,7 +33,7 @@ namespace zmq
{
public:
sub_t (class app_thread_t *parent_);
sub_t (class ctx_t *parent_, uint32_t slot_);
~sub_t ();
protected:
@@ -41,16 +41,11 @@ namespace zmq
// Overloads of functions from socket_base_t.
void xattach_pipes (class reader_t *inpipe_, class writer_t *outpipe_,
const blob_t &peer_identity_);
void xdetach_inpipe (class reader_t *pipe_);
void xdetach_outpipe (class writer_t *pipe_);
void xkill (class reader_t *pipe_);
void xrevive (class reader_t *pipe_);
void xrevive (class writer_t *pipe_);
void xterm_pipes ();
bool xhas_pipes ();
int xsetsockopt (int option_, const void *optval_, size_t optvallen_);
int xsend (zmq_msg_t *msg_, int flags_);
int xrecv (zmq_msg_t *msg_, int flags_);
bool xhas_in ();
bool xhas_out ();
private: