generic-library/libzmq
1
0
Fork 0
mirror of https://github.com/zeromq/libzmq.git synced 2024-12-12 10:33:52 +01:00
Code Issues Projects Releases Wiki Activity

Merge pull request #3143 from sigiesec/remove-extra-files

Browse Source 
Problem: temporary files in repo
This commit is contained in:
Luca Boccassi 2018-05-30 08:36:33 +01:00 committed by GitHub
commit 088fd65bf2
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
3 changed files with 0 additions and 1704 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

549
src/router.cpp~RF40cad05.TMP
View File

@ -1,549 +0,0 @@
/*
Copyright (c) 2007-2016 Contributors as noted in the AUTHORS file
This file is part of libzmq, the ZeroMQ core engine in C++.
libzmq is free software; you can redistribute it and/or modify it under
the terms of the GNU Lesser General Public License (LGPL) as published
by the Free Software Foundation; either version 3 of the License, or
(at your option) any later version.
As a special exception, the Contributors give you permission to link
this library with independent modules to produce an executable,
regardless of the license terms of these independent modules, and to
copy and distribute the resulting executable under terms of your choice,
provided that you also meet, for each linked independent module, the
terms and conditions of the license of that module. An independent
module is a module which is not derived from or based on this library.
If you modify this library, you must extend this exception to your
version of the library.
libzmq is distributed in the hope that it will be useful, but WITHOUT
ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public
License for more details.
You should have received a copy of the GNU Lesser General Public License
along with this program. If not, see <http://www.gnu.org/licenses/>.
*/
#include "precompiled.hpp"
#include "macros.hpp"
#include "router.hpp"
#include "pipe.hpp"
#include "wire.hpp"
#include "random.hpp"
#include "likely.hpp"
#include "err.hpp"
zmq::router_t::router_t (class ctx_t *parent_, uint32_t tid_, int sid_) :
socket_base_t (parent_, tid_, sid_),
prefetched (false),
routing_id_sent (false),
current_in (NULL),
terminate_current_in (false),
more_in (false),
current_out (NULL),
more_out (false),
next_integral_routing_id (generate_random ()),
mandatory (false),
// raw_socket functionality in ROUTER is deprecated
raw_socket (false),
probe_router (false),
handover (false)
{
options.type = ZMQ_ROUTER;
options.recv_routing_id = true;
options.raw_socket = false;
prefetched_id.init ();
prefetched_msg.init ();
}
zmq::router_t::~router_t ()
{
zmq_assert (anonymous_pipes.empty ());
;
zmq_assert (outpipes.empty ());
prefetched_id.close ();
prefetched_msg.close ();
}
void zmq::router_t::xattach_pipe (pipe_t *pipe_, bool subscribe_to_all_)
{
LIBZMQ_UNUSED (subscribe_to_all_);
zmq_assert (pipe_);
if (probe_router) {
msg_t probe_msg;
int rc = probe_msg.init ();
errno_assert (rc == 0);
rc = pipe_->write (&probe_msg);
// zmq_assert (rc) is not applicable here, since it is not a bug.
pipe_->flush ();
rc = probe_msg.close ();
errno_assert (rc == 0);
}
bool routing_id_ok = identify_peer (pipe_);
if (routing_id_ok)
fq.attach (pipe_);
else
anonymous_pipes.insert (pipe_);
}
int zmq::router_t::xsetsockopt (int option_,
const void *optval_,
size_t optvallen_)
{
bool is_int = (optvallen_ == sizeof (int));
int value = 0;
if (is_int)
memcpy (&value, optval_, sizeof (int));
switch (option_) {
case ZMQ_CONNECT_ROUTING_ID:
// TODO why isn't it possible to set an empty connect_routing_id
// (which is the default value)
if (optval_ && optvallen_) {
connect_routing_id.assign ((char *) optval_, optvallen_);
return 0;
}
break;
case ZMQ_ROUTER_RAW:
if (is_int && value >= 0) {
raw_socket = (value != 0);
if (raw_socket) {
options.recv_routing_id = false;
options.raw_socket = true;
}
return 0;
}
break;
case ZMQ_ROUTER_MANDATORY:
if (is_int && value >= 0) {
mandatory = (value != 0);
return 0;
}
break;
case ZMQ_PROBE_ROUTER:
if (is_int && value >= 0) {
probe_router = (value != 0);
return 0;
}
break;
case ZMQ_ROUTER_HANDOVER:
if (is_int && value >= 0) {
handover = (value != 0);
return 0;
}
break;
default:
break;
}
errno = EINVAL;
return -1;
}
void zmq::router_t::xpipe_terminated (pipe_t *pipe_)
{
std::set<pipe_t *>::iterator it = anonymous_pipes.find (pipe_);
if (it != anonymous_pipes.end ())
anonymous_pipes.erase (it);
else {
outpipes_t::iterator iter = outpipes.find (pipe_->get_routing_id ());
zmq_assert (iter != outpipes.end ());
outpipes.erase (iter);
fq.pipe_terminated (pipe_);
pipe_->rollback ();
if (pipe_ == current_out)
current_out = NULL;
}
}
void zmq::router_t::xread_activated (pipe_t *pipe_)
{
std::set<pipe_t *>::iterator it = anonymous_pipes.find (pipe_);
if (it == anonymous_pipes.end ())
fq.activated (pipe_);
else {
bool routing_id_ok = identify_peer (pipe_);
if (routing_id_ok) {
anonymous_pipes.erase (it);
fq.attach (pipe_);
}
}
}
void zmq::router_t::xwrite_activated (pipe_t *pipe_)
{
outpipes_t::iterator it;
for (it = outpipes.begin (); it != outpipes.end (); ++it)
if (it->second.pipe == pipe_)
break;
zmq_assert (it != outpipes.end ());
zmq_assert (!it->second.active);
it->second.active = true;
}
int zmq::router_t::xsend (msg_t *msg_)
{
// If this is the first part of the message it's the ID of the
// peer to send the message to.
if (!more_out) {
zmq_assert (!current_out);
// If we have malformed message (prefix with no subsequent message)
// then just silently ignore it.
// TODO: The connections should be killed instead.
if (msg_->flags () & msg_t::more) {
more_out = true;
// Find the pipe associated with the routing id stored in the prefix.
// If there's no such pipe just silently ignore the message, unless
// router_mandatory is set.
blob_t routing_id (static_cast<unsigned char *> (msg_->data ()),
msg_->size (), zmq::reference_tag_t ());
outpipes_t::iterator it = outpipes.find (routing_id);
if (it != outpipes.end ()) {
current_out = it->second.pipe;
// Check whether pipe is closed or not
if (!current_out->check_write ()) {
// Check whether pipe is full or not
bool pipe_full = !current_out->check_hwm ();
it->second.active = false;
current_out = NULL;
if (mandatory) {
more_out = false;
if (pipe_full)
errno = EAGAIN;
else
errno = EHOSTUNREACH;
return -1;
}
}
} else if (mandatory) {
more_out = false;
errno = EHOSTUNREACH;
return -1;
}
}
int rc = msg_->close ();
errno_assert (rc == 0);
rc = msg_->init ();
errno_assert (rc == 0);
return 0;
}
// Ignore the MORE flag for raw-sock or assert?
if (options.raw_socket)
msg_->reset_flags (msg_t::more);
// Check whether this is the last part of the message.
more_out = (msg_->flags () & msg_t::more) != 0;
// Push the message into the pipe. If there's no out pipe, just drop it.
if (current_out) {
// Close the remote connection if user has asked to do so
// by sending zero length message.
// Pending messages in the pipe will be dropped (on receiving term- ack)
if (raw_socket && msg_->size () == 0) {
current_out->terminate (false);
int rc = msg_->close ();
errno_assert (rc == 0);
rc = msg_->init ();
errno_assert (rc == 0);
current_out = NULL;
return 0;
}
bool ok = current_out->write (msg_);
if (unlikely (!ok)) {
// Message failed to send - we must close it ourselves.
int rc = msg_->close ();
errno_assert (rc == 0);
// HWM was checked before, so the pipe must be gone. Roll back
// messages that were piped, for example REP labels.
current_out->rollback ();
current_out = NULL;
} else {
if (!more_out) {
current_out->flush ();
current_out = NULL;
}
}
} else {
int rc = msg_->close ();
errno_assert (rc == 0);
}
// Detach the message from the data buffer.
int rc = msg_->init ();
errno_assert (rc == 0);
return 0;
}
int zmq::router_t::xrecv (msg_t *msg_)
{
if (prefetched) {
if (!routing_id_sent) {
int rc = msg_->move (prefetched_id);
errno_assert (rc == 0);
routing_id_sent = true;
} else {
int rc = msg_->move (prefetched_msg);
errno_assert (rc == 0);
prefetched = false;
}
more_in = (msg_->flags () & msg_t::more) != 0;
if (!more_in) {
if (terminate_current_in) {
current_in->terminate (true);
terminate_current_in = false;
}
current_in = NULL;
}
return 0;
}
pipe_t *pipe = NULL;
int rc = fq.recvpipe (msg_, &pipe);
// It's possible that we receive peer's routing id. That happens
// after reconnection. The current implementation assumes that
// the peer always uses the same routing id.
while (rc == 0 && msg_->is_routing_id ())
rc = fq.recvpipe (msg_, &pipe);
if (rc != 0)
return -1;
zmq_assert (pipe != NULL);
// If we are in the middle of reading a message, just return the next part.
if (more_in) {
more_in = (msg_->flags () & msg_t::more) != 0;
if (!more_in) {
if (terminate_current_in) {
current_in->terminate (true);
terminate_current_in = false;
}
current_in = NULL;
}
} else {
// We are at the beginning of a message.
// Keep the message part we have in the prefetch buffer
// and return the ID of the peer instead.
rc = prefetched_msg.move (*msg_);
errno_assert (rc == 0);
prefetched = true;
current_in = pipe;
const blob_t &routing_id = pipe->get_routing_id ();
rc = msg_->init_size (routing_id.size ());
errno_assert (rc == 0);
memcpy (msg_->data (), routing_id.data (), routing_id.size ());
msg_->set_flags (msg_t::more);
if (prefetched_msg.metadata ())
msg_->set_metadata (prefetched_msg.metadata ());
routing_id_sent = true;
}
return 0;
}
int zmq::router_t::rollback ()
{
if (current_out) {
current_out->rollback ();
current_out = NULL;
more_out = false;
}
return 0;
}
bool zmq::router_t::xhas_in ()
{
// If we are in the middle of reading the messages, there are
// definitely more parts available.
if (more_in)
return true;
// We may already have a message pre-fetched.
if (prefetched)
return true;
// Try to read the next message.
// The message, if read, is kept in the pre-fetch buffer.
pipe_t *pipe = NULL;
int rc = fq.recvpipe (&prefetched_msg, &pipe);
// It's possible that we receive peer's routing id. That happens
// after reconnection. The current implementation assumes that
// the peer always uses the same routing id.
// TODO: handle the situation when the peer changes its routing id.
while (rc == 0 && prefetched_msg.is_routing_id ())
rc = fq.recvpipe (&prefetched_msg, &pipe);
if (rc != 0)
return false;
zmq_assert (pipe != NULL);
const blob_t &routing_id = pipe->get_routing_id ();
rc = prefetched_id.init_size (routing_id.size ());
errno_assert (rc == 0);
memcpy (prefetched_id.data (), routing_id.data (), routing_id.size ());
prefetched_id.set_flags (msg_t::more);
prefetched = true;
routing_id_sent = false;
current_in = pipe;
return true;
}
bool zmq::router_t::xhas_out ()
{
// In theory, ROUTER socket is always ready for writing (except when
// MANDATORY is set). Whether actual attempt to write succeeds depends
// on whitch pipe the message is going to be routed to.
if (!mandatory)
return true;
bool has_out = false;
outpipes_t::iterator it;
for (it = outpipes.begin (); it != outpipes.end (); ++it)
has_out |= it->second.pipe->check_hwm ();
return has_out;
}
const zmq::blob_t &zmq::router_t::get_credential () const
{
return fq.get_credential ();
}
int zmq::router_t::get_peer_state (const void *routing_id_,
size_t routing_id_size_) const
{
int res = 0;
blob_t routing_id_blob ((unsigned char *) routing_id_, routing_id_size_);
outpipes_t::const_iterator it = outpipes.find (routing_id_blob);
if (it == outpipes.end ()) {
errno = EHOSTUNREACH;
return -1;
}
const outpipe_t &outpipe = it->second;
if (outpipe.pipe->check_hwm ())
res |= ZMQ_POLLOUT;
/** \todo does it make any sense to check the inpipe as well? */
return res;
}
bool zmq::router_t::identify_peer (pipe_t *pipe_)
{
msg_t msg;
bool ok;
blob_t routing_id;
if (connect_routing_id.length ()) {
routing_id.set ((unsigned char *) connect_routing_id.c_str (),
connect_routing_id.length ());
connect_routing_id.clear ();
outpipes_t::iterator it = outpipes.find (routing_id);
if (it != outpipes.end ())
zmq_assert (false); // Not allowed to duplicate an existing rid
} else if (
options
.raw_socket) { // Always assign an integral routing id for raw-socket
unsigned char buf[5];
buf[0] = 0;
put_uint32 (buf + 1, next_integral_routing_id++);
routing_id.set (buf, sizeof buf);
} else if (!options.raw_socket) {
// Pick up handshake cases and also case where next integral routing id is set
msg.init ();
ok = pipe_->read (&msg);
if (!ok)
return false;
if (msg.size () == 0) {
// Fall back on the auto-generation
unsigned char buf[5];
buf[0] = 0;
put_uint32 (buf + 1, next_integral_routing_id++);
routing_id.set (buf, sizeof buf);
msg.close ();
} else {
routing_id.set (static_cast<unsigned char *> (msg.data ()),
msg.size ());
outpipes_t::iterator it = outpipes.find (routing_id);
msg.close ();
if (it != outpipes.end ()) {
if (!handover)
// Ignore peers with duplicate ID
return false;
// We will allow the new connection to take over this
// routing id. Temporarily assign a new routing id to the
// existing pipe so we can terminate it asynchronously.
unsigned char buf[5];
buf[0] = 0;
put_uint32 (buf + 1, next_integral_routing_id++);
blob_t new_routing_id (buf, sizeof buf);
it->second.pipe->set_router_socket_routing_id (new_routing_id);
outpipe_t existing_outpipe = {it->second.pipe,
it->second.active};
ok = outpipes
.ZMQ_MAP_INSERT_OR_EMPLACE (ZMQ_MOVE (new_routing_id),
existing_outpipe)
.second;
zmq_assert (ok);
// Remove the existing routing id entry to allow the new
// connection to take the routing id.
outpipes.erase (it);
if (existing_outpipe.pipe == current_in)
terminate_current_in = true;
else
existing_outpipe.pipe->terminate (true);
}
}
}
pipe_->set_router_socket_routing_id (routing_id);
// Add the record into output pipes lookup table
outpipe_t outpipe = {pipe_, true};
ok = outpipes.ZMQ_MAP_INSERT_OR_EMPLACE (ZMQ_MOVE (routing_id), outpipe)
.second;
zmq_assert (ok);
return true;
}

711
src/session_base.cpp~RF4069b78.TMP
View File

@ -1,711 +0,0 @@
/*
Copyright (c) 2007-2016 Contributors as noted in the AUTHORS file
This file is part of libzmq, the ZeroMQ core engine in C++.
libzmq is free software; you can redistribute it and/or modify it under
the terms of the GNU Lesser General Public License (LGPL) as published
by the Free Software Foundation; either version 3 of the License, or
(at your option) any later version.
As a special exception, the Contributors give you permission to link
this library with independent modules to produce an executable,
regardless of the license terms of these independent modules, and to
copy and distribute the resulting executable under terms of your choice,
provided that you also meet, for each linked independent module, the
terms and conditions of the license of that module. An independent
module is a module which is not derived from or based on this library.
If you modify this library, you must extend this exception to your
version of the library.
libzmq is distributed in the hope that it will be useful, but WITHOUT
ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public
License for more details.
You should have received a copy of the GNU Lesser General Public License
along with this program. If not, see <http://www.gnu.org/licenses/>.
*/
#include "precompiled.hpp"
#include "macros.hpp"
#include "session_base.hpp"
#include "i_engine.hpp"
#include "err.hpp"
#include "pipe.hpp"
#include "likely.hpp"
#include "tcp_connecter.hpp"
#include "ipc_connecter.hpp"
#include "tipc_connecter.hpp"
#include "socks_connecter.hpp"
#include "vmci_connecter.hpp"
#include "pgm_sender.hpp"
#include "pgm_receiver.hpp"
#include "address.hpp"
#include "norm_engine.hpp"
#include "udp_engine.hpp"
#include "ctx.hpp"
#include "req.hpp"
#include "radio.hpp"
#include "dish.hpp"
zmq::session_base_t *zmq::session_base_t::create (class io_thread_t *io_thread_,
bool active_,
class socket_base_t *socket_,
const options_t &options_,
address_t *addr_)
{
session_base_t *s = NULL;
switch (options_.type) {
case ZMQ_REQ:
s = new (std::nothrow)
req_session_t (io_thread_, active_, socket_, options_, addr_);
break;
case ZMQ_RADIO:
s = new (std::nothrow)
radio_session_t (io_thread_, active_, socket_, options_, addr_);
break;
case ZMQ_DISH:
s = new (std::nothrow)
dish_session_t (io_thread_, active_, socket_, options_, addr_);
break;
case ZMQ_DEALER:
case ZMQ_REP:
case ZMQ_ROUTER:
case ZMQ_PUB:
case ZMQ_XPUB:
case ZMQ_SUB:
case ZMQ_XSUB:
case ZMQ_PUSH:
case ZMQ_PULL:
case ZMQ_PAIR:
case ZMQ_STREAM:
case ZMQ_SERVER:
case ZMQ_CLIENT:
case ZMQ_GATHER:
case ZMQ_SCATTER:
case ZMQ_DGRAM:
s = new (std::nothrow)
session_base_t (io_thread_, active_, socket_, options_, addr_);
break;
default:
errno = EINVAL;
return NULL;
}
alloc_assert (s);
return s;
}
zmq::session_base_t::session_base_t (class io_thread_t *io_thread_,
bool active_,
class socket_base_t *socket_,
const options_t &options_,
address_t *addr_) :
own_t (io_thread_, options_),
io_object_t (io_thread_),
active (active_),
pipe (NULL),
zap_pipe (NULL),
incomplete_in (false),
pending (false),
engine (NULL),
socket (socket_),
io_thread (io_thread_),
has_linger_timer (false),
addr (addr_)
{
}
const char *zmq::session_base_t::get_endpoint () const
{
return engine->get_endpoint ();
}
zmq::session_base_t::~session_base_t ()
{
zmq_assert (!pipe);
zmq_assert (!zap_pipe);
// If there's still a pending linger timer, remove it.
if (has_linger_timer) {
cancel_timer (linger_timer_id);
has_linger_timer = false;
}
// Close the engine.
if (engine)
engine->terminate ();
LIBZMQ_DELETE (addr);
}
void zmq::session_base_t::attach_pipe (pipe_t *pipe_)
{
zmq_assert (!is_terminating ());
zmq_assert (!pipe);
zmq_assert (pipe_);
pipe = pipe_;
pipe->set_event_sink (this);
}
int zmq::session_base_t::pull_msg (msg_t *msg_)
{
if (!pipe || !pipe->read (msg_)) {
errno = EAGAIN;
return -1;
}
incomplete_in = (msg_->flags () & msg_t::more) != 0;
return 0;
}
int zmq::session_base_t::push_msg (msg_t *msg_)
{
if (msg_->flags () & msg_t::command)
return 0;
if (pipe && pipe->write (msg_)) {
int rc = msg_->init ();
errno_assert (rc == 0);
return 0;
}
errno = EAGAIN;
return -1;
}
int zmq::session_base_t::read_zap_msg (msg_t *msg_)
{
if (zap_pipe == NULL) {
errno = ENOTCONN;
return -1;
}
if (!zap_pipe->read (msg_)) {
errno = EAGAIN;
return -1;
}
return 0;
}
int zmq::session_base_t::write_zap_msg (msg_t *msg_)
{
if (zap_pipe == NULL || !zap_pipe->write (msg_)) {
errno = ENOTCONN;
return -1;
}
if ((msg_->flags () & msg_t::more) == 0)
zap_pipe->flush ();
const int rc = msg_->init ();
errno_assert (rc == 0);
return 0;
}
void zmq::session_base_t::reset ()
{
}
void zmq::session_base_t::flush ()
{
if (pipe)
pipe->flush ();
}
void zmq::session_base_t::clean_pipes ()
{
zmq_assert (pipe != NULL);
// Get rid of half-processed messages in the out pipe. Flush any
// unflushed messages upstream.
pipe->rollback ();
pipe->flush ();
// Remove any half-read message from the in pipe.
while (incomplete_in) {
msg_t msg;
int rc = msg.init ();
errno_assert (rc == 0);
rc = pull_msg (&msg);
errno_assert (rc == 0);
rc = msg.close ();
errno_assert (rc == 0);
}
}
void zmq::session_base_t::pipe_terminated (pipe_t *pipe_)
{
// Drop the reference to the deallocated pipe if required.
zmq_assert (pipe_ == pipe || pipe_ == zap_pipe
|| terminating_pipes.count (pipe_) == 1);
if (pipe_ == pipe) {
// If this is our current pipe, remove it
pipe = NULL;
if (has_linger_timer) {
cancel_timer (linger_timer_id);
has_linger_timer = false;
}
} else if (pipe_ == zap_pipe)
zap_pipe = NULL;
else
// Remove the pipe from the detached pipes set
terminating_pipes.erase (pipe_);
if (!is_terminating () && options.raw_socket) {
if (engine) {
engine->terminate ();
engine = NULL;
}
terminate ();
}
// If we are waiting for pending messages to be sent, at this point
// we are sure that there will be no more messages and we can proceed
// with termination safely.
if (pending && !pipe && !zap_pipe && terminating_pipes.empty ()) {
pending = false;
own_t::process_term (0);
}
}
void zmq::session_base_t::read_activated (pipe_t *pipe_)
{
// Skip activating if we're detaching this pipe
if (unlikely (pipe_ != pipe && pipe_ != zap_pipe)) {
zmq_assert (terminating_pipes.count (pipe_) == 1);
return;
}
if (unlikely (engine == NULL)) {
pipe->check_read ();
return;
}
if (likely (pipe_ == pipe))
engine->restart_output ();
else {
// i.e. pipe_ == zap_pipe
engine->zap_msg_available ();
}
}
void zmq::session_base_t::write_activated (pipe_t *pipe_)
{
// Skip activating if we're detaching this pipe
if (pipe != pipe_) {
zmq_assert (terminating_pipes.count (pipe_) == 1);
return;
}
if (engine)
engine->restart_input ();
}
void zmq::session_base_t::hiccuped (pipe_t *)
{
// Hiccups are always sent from session to socket, not the other
// way round.
zmq_assert (false);
}
zmq::socket_base_t *zmq::session_base_t::get_socket ()
{
return socket;
}
void zmq::session_base_t::process_plug ()
{
if (active)
start_connecting (false);
}
// This functions can return 0 on success or -1 and errno=ECONNREFUSED if ZAP
// is not setup (IE: inproc://zeromq.zap.01 does not exist in the same context)
// or it aborts on any other error. In other words, either ZAP is not
// configured or if it is configured it MUST be configured correctly and it
// MUST work, otherwise authentication cannot be guaranteed and it would be a
// security flaw.
int zmq::session_base_t::zap_connect ()
{
if (zap_pipe != NULL)
return 0;
endpoint_t peer = find_endpoint ("inproc://zeromq.zap.01");
if (peer.socket == NULL) {
errno = ECONNREFUSED;
return -1;
}
zmq_assert (peer.options.type == ZMQ_REP || peer.options.type == ZMQ_ROUTER
|| peer.options.type == ZMQ_SERVER);
// Create a bi-directional pipe that will connect
// session with zap socket.
object_t *parents[2] = {this, peer.socket};
pipe_t *new_pipes[2] = {NULL, NULL};
int hwms[2] = {0, 0};
bool conflates[2] = {false, false};
int rc = pipepair (parents, new_pipes, hwms, conflates);
errno_assert (rc == 0);
// Attach local end of the pipe to this socket object.
zap_pipe = new_pipes[0];
zap_pipe->set_nodelay ();
zap_pipe->set_event_sink (this);
send_bind (peer.socket, new_pipes[1], false);
// Send empty routing id if required by the peer.
if (peer.options.recv_routing_id) {
msg_t id;
rc = id.init ();
errno_assert (rc == 0);
id.set_flags (msg_t::routing_id);
bool ok = zap_pipe->write (&id);
zmq_assert (ok);
zap_pipe->flush ();
}
return 0;
}
bool zmq::session_base_t::zap_enabled ()
{
return (options.mechanism != ZMQ_NULL || !options.zap_domain.empty ());
}
void zmq::session_base_t::process_attach (i_engine *engine_)
{
zmq_assert (engine_ != NULL);
// Create the pipe if it does not exist yet.
if (!pipe && !is_terminating ()) {
object_t *parents[2] = {this, socket};
pipe_t *pipes[2] = {NULL, NULL};
bool conflate =
options.conflate
&& (options.type == ZMQ_DEALER || options.type == ZMQ_PULL
|| options.type == ZMQ_PUSH || options.type == ZMQ_PUB
|| options.type == ZMQ_SUB);
int hwms[2] = {conflate ? -1 : options.rcvhwm,
conflate ? -1 : options.sndhwm};
bool conflates[2] = {conflate, conflate};
int rc = pipepair (parents, pipes, hwms, conflates);
errno_assert (rc == 0);
// Plug the local end of the pipe.
pipes[0]->set_event_sink (this);
// Remember the local end of the pipe.
zmq_assert (!pipe);
pipe = pipes[0];
// Ask socket to plug into the remote end of the pipe.
send_bind (socket, pipes[1]);
}
// Plug in the engine.
zmq_assert (!engine);
engine = engine_;
engine->plug (io_thread, this);
}
void zmq::session_base_t::engine_error (
zmq::stream_engine_t::error_reason_t reason_)
{
// Engine is dead. Let's forget about it.
engine = NULL;
// Remove any half-done messages from the pipes.
if (pipe)
clean_pipes ();
zmq_assert (reason_ == stream_engine_t::connection_error
|| reason_ == stream_engine_t::timeout_error
|| reason_ == stream_engine_t::protocol_error);
switch (reason_) {
case stream_engine_t::timeout_error:
/* FALLTHROUGH */
case stream_engine_t::connection_error:
if (active) {
reconnect ();
break;
}
/* FALLTHROUGH */
case stream_engine_t::protocol_error:
if (pending) {
if (pipe)
pipe->terminate (false);
if (zap_pipe)
zap_pipe->terminate (false);
} else {
terminate ();
}
break;
}
// Just in case there's only a delimiter in the pipe.
if (pipe)
pipe->check_read ();
if (zap_pipe)
zap_pipe->check_read ();
}
void zmq::session_base_t::process_term (int linger_)
{
zmq_assert (!pending);
// If the termination of the pipe happens before the term command is
// delivered there's nothing much to do. We can proceed with the
// standard termination immediately.
if (!pipe && !zap_pipe && terminating_pipes.empty ()) {
own_t::process_term (0);
return;
}
pending = true;
if (pipe != NULL) {
// If there's finite linger value, delay the termination.
// If linger is infinite (negative) we don't even have to set
// the timer.
if (linger_ > 0) {
zmq_assert (!has_linger_timer);
add_timer (linger_, linger_timer_id);
has_linger_timer = true;
}
// Start pipe termination process. Delay the termination till all messages
// are processed in case the linger time is non-zero.
pipe->terminate (linger_ != 0);
// TODO: Should this go into pipe_t::terminate ?
// In case there's no engine and there's only delimiter in the
// pipe it wouldn't be ever read. Thus we check for it explicitly.
if (!engine)
pipe->check_read ();
}
if (zap_pipe != NULL)
zap_pipe->terminate (false);
}
void zmq::session_base_t::timer_event (int id_)
{
// Linger period expired. We can proceed with termination even though
// there are still pending messages to be sent.
zmq_assert (id_ == linger_timer_id);
has_linger_timer = false;
// Ask pipe to terminate even though there may be pending messages in it.
zmq_assert (pipe);
pipe->terminate (false);
}
void zmq::session_base_t::reconnect ()
{
// For delayed connect situations, terminate the pipe
// and reestablish later on
if (pipe && options.immediate == 1 && addr->protocol != "pgm"
&& addr->protocol != "epgm" && addr->protocol != "norm"
&& addr->protocol != "udp") {
pipe->hiccup ();
pipe->terminate (false);
terminating_pipes.insert (pipe);
pipe = NULL;
if (has_linger_timer) {
cancel_timer (linger_timer_id);
has_linger_timer = false;
}
}
reset ();
// Reconnect.
if (options.reconnect_ivl != -1)
start_connecting (true);
else {
std::string *ep = new (std::string);
addr->to_string (*ep);
send_term_endpoint (socket, ep);
}
// For subscriber sockets we hiccup the inbound pipe, which will cause
// the socket object to resend all the subscriptions.
if (pipe
&& (options.type == ZMQ_SUB || options.type == ZMQ_XSUB
|| options.type == ZMQ_DISH))
pipe->hiccup ();
}
void zmq::session_base_t::start_connecting (bool wait_)
{
zmq_assert (active);
// Choose I/O thread to run connecter in. Given that we are already
// running in an I/O thread, there must be at least one available.
io_thread_t *io_thread = choose_io_thread (options.affinity);
zmq_assert (io_thread);
// Create the connecter object.
if (addr->protocol == "tcp") {
if (!options.socks_proxy_address.empty ()) {
address_t *proxy_address = new (std::nothrow)
address_t ("tcp", options.socks_proxy_address, this->get_ctx ());
alloc_assert (proxy_address);
socks_connecter_t *connecter =
new (std::nothrow) socks_connecter_t (io_thread, this, options,
addr, proxy_address, wait_);
alloc_assert (connecter);
launch_child (connecter);
} else {
tcp_connecter_t *connecter = new (std::nothrow)
tcp_connecter_t (io_thread, this, options, addr, wait_);
alloc_assert (connecter);
launch_child (connecter);
}
return;
}
#if !defined ZMQ_HAVE_WINDOWS && !defined ZMQ_HAVE_OPENVMS \
&& !defined ZMQ_HAVE_VXWORKS
if (addr->protocol == "ipc") {
ipc_connecter_t *connecter = new (std::nothrow)
ipc_connecter_t (io_thread, this, options, addr, wait_);
alloc_assert (connecter);
launch_child (connecter);
return;
}
#endif
#if defined ZMQ_HAVE_TIPC
if (addr->protocol == "tipc") {
tipc_connecter_t *connecter = new (std::nothrow)
tipc_connecter_t (io_thread, this, options, addr, wait_);
alloc_assert (connecter);
launch_child (connecter);
return;
}
#endif
if (addr->protocol == "udp") {
zmq_assert (options.type == ZMQ_DISH || options.type == ZMQ_RADIO
|| options.type == ZMQ_DGRAM);
udp_engine_t *engine = new (std::nothrow) udp_engine_t (options);
alloc_assert (engine);
bool recv = false;
bool send = false;
if (options.type == ZMQ_RADIO) {
send = true;
recv = false;
} else if (options.type == ZMQ_DISH) {
send = false;
recv = true;
} else if (options.type == ZMQ_DGRAM) {
send = true;
recv = true;
}
int rc = engine->init (addr, send, recv);
errno_assert (rc == 0);
send_attach (this, engine);
return;
}
#ifdef ZMQ_HAVE_OPENPGM
// Both PGM and EPGM transports are using the same infrastructure.
if (addr->protocol == "pgm" || addr->protocol == "epgm") {
zmq_assert (options.type == ZMQ_PUB || options.type == ZMQ_XPUB
|| options.type == ZMQ_SUB || options.type == ZMQ_XSUB);
// For EPGM transport with UDP encapsulation of PGM is used.
bool const udp_encapsulation = addr->protocol == "epgm";
// At this point we'll create message pipes to the session straight
// away. There's no point in delaying it as no concept of 'connect'
// exists with PGM anyway.
if (options.type == ZMQ_PUB || options.type == ZMQ_XPUB) {
// PGM sender.
pgm_sender_t *pgm_sender =
new (std::nothrow) pgm_sender_t (io_thread, options);
alloc_assert (pgm_sender);
int rc =
pgm_sender->init (udp_encapsulation, addr->address.c_str ());
errno_assert (rc == 0);
send_attach (this, pgm_sender);
} else {
// PGM receiver.
pgm_receiver_t *pgm_receiver =
new (std::nothrow) pgm_receiver_t (io_thread, options);
alloc_assert (pgm_receiver);
int rc =
pgm_receiver->init (udp_encapsulation, addr->address.c_str ());
errno_assert (rc == 0);
send_attach (this, pgm_receiver);
}
return;
}
#endif
#ifdef ZMQ_HAVE_NORM
if (addr->protocol == "norm") {
// At this point we'll create message pipes to the session straight
// away. There's no point in delaying it as no concept of 'connect'
// exists with NORM anyway.
if (options.type == ZMQ_PUB || options.type == ZMQ_XPUB) {
// NORM sender.
norm_engine_t *norm_sender =
new (std::nothrow) norm_engine_t (io_thread, options);
alloc_assert (norm_sender);
int rc = norm_sender->init (addr->address.c_str (), true, false);
errno_assert (rc == 0);
send_attach (this, norm_sender);
} else { // ZMQ_SUB or ZMQ_XSUB
// NORM receiver.
norm_engine_t *norm_receiver =
new (std::nothrow) norm_engine_t (io_thread, options);
alloc_assert (norm_receiver);
int rc = norm_receiver->init (addr->address.c_str (), false, true);
errno_assert (rc == 0);
send_attach (this, norm_receiver);
}
return;
}
#endif // ZMQ_HAVE_NORM
#if defined ZMQ_HAVE_VMCI
if (addr->protocol == "vmci") {
vmci_connecter_t *connecter = new (std::nothrow)
vmci_connecter_t (io_thread, this, options, addr, wait_);
alloc_assert (connecter);
launch_child (connecter);
return;
}
#endif
zmq_assert (false);
}

444
src/tcp_connecter.cpp.orig
View File

@ -1,444 +0,0 @@
/*
Copyright (c) 2007-2016 Contributors as noted in the AUTHORS file
This file is part of libzmq, the ZeroMQ core engine in C++.
libzmq is free software; you can redistribute it and/or modify it under
the terms of the GNU Lesser General Public License (LGPL) as published
by the Free Software Foundation; either version 3 of the License, or
(at your option) any later version.
As a special exception, the Contributors give you permission to link
this library with independent modules to produce an executable,
regardless of the license terms of these independent modules, and to
copy and distribute the resulting executable under terms of your choice,
provided that you also meet, for each linked independent module, the
terms and conditions of the license of that module. An independent
module is a module which is not derived from or based on this library.
If you modify this library, you must extend this exception to your
version of the library.
libzmq is distributed in the hope that it will be useful, but WITHOUT
ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public
License for more details.
You should have received a copy of the GNU Lesser General Public License
along with this program. If not, see <http://www.gnu.org/licenses/>.
*/
#include "precompiled.hpp"
#include <new>
#include <string>
#include "macros.hpp"
#include "tcp_connecter.hpp"
#include "stream_engine.hpp"
#include "io_thread.hpp"
#include "random.hpp"
#include "err.hpp"
#include "ip.hpp"
#include "tcp.hpp"
#include "address.hpp"
#include "tcp_address.hpp"
#include "session_base.hpp"
#if !defined ZMQ_HAVE_WINDOWS
#include <unistd.h>
#include <sys/types.h>
#include <sys/socket.h>
#include <arpa/inet.h>
#include <netinet/tcp.h>
#include <netinet/in.h>
#include <netdb.h>
#include <fcntl.h>
#ifdef ZMQ_HAVE_VXWORKS
#include <sockLib.h>
#endif
#ifdef ZMQ_HAVE_OPENVMS
#include <ioctl.h>
#endif
#endif
#ifdef __APPLE__
#include <TargetConditionals.h>
#endif
zmq::tcp_connecter_t::tcp_connecter_t (class io_thread_t *io_thread_,
class session_base_t *session_,
const options_t &options_,
address_t *addr_,
bool delayed_start_) :
own_t (io_thread_, options_),
io_object_t (io_thread_),
_addr (addr_),
_s (retired_fd),
_handle (static_cast<handle_t> (NULL)),
_delayed_start (delayed_start_),
_connect_timer_started (false),
_reconnect_timer_started (false),
_session (session_),
_current_reconnect_ivl (options.reconnect_ivl),
_socket (_session->get_socket ())
{
zmq_assert (_addr);
zmq_assert (_addr->protocol == "tcp");
_addr->to_string (_endpoint);
// TODO the return value is unused! what if it fails? if this is impossible
// or does not matter, change such that endpoint in initialized using an
// initializer, and make endpoint const
}
zmq::tcp_connecter_t::~tcp_connecter_t ()
{
zmq_assert (!_connect_timer_started);
zmq_assert (!_reconnect_timer_started);
zmq_assert (!_handle);
zmq_assert (_s == retired_fd);
}
void zmq::tcp_connecter_t::process_plug ()
{
if (_delayed_start)
add_reconnect_timer ();
else
start_connecting ();
}
void zmq::tcp_connecter_t::process_term (int linger_)
{
if (_connect_timer_started) {
cancel_timer (connect_timer_id);
_connect_timer_started = false;
}
if (_reconnect_timer_started) {
cancel_timer (reconnect_timer_id);
_reconnect_timer_started = false;
}
if (_handle) {
rm_handle ();
}
if (_s != retired_fd)
close ();
own_t::process_term (linger_);
}
void zmq::tcp_connecter_t::in_event ()
{
// We are not polling for incoming data, so we are actually called
// because of error here. However, we can get error on out event as well
// on some platforms, so we'll simply handle both events in the same way.
out_event ();
}
void zmq::tcp_connecter_t::out_event ()
{
if (_connect_timer_started) {
cancel_timer (connect_timer_id);
_connect_timer_started = false;
}
rm_handle ();
const fd_t fd = connect ();
// Handle the error condition by attempt to reconnect.
if (fd == retired_fd || !tune_socket (fd)) {
close ();
add_reconnect_timer ();
return;
}
// Create the engine object for this connection.
stream_engine_t *engine =
new (std::nothrow) stream_engine_t (fd, options, _endpoint);
alloc_assert (engine);
// Attach the engine to the corresponding session object.
send_attach (_session, engine);
// Shut the connecter down.
terminate ();
_socket->event_connected (_endpoint, fd);
}
void zmq::tcp_connecter_t::rm_handle ()
{
rm_fd (_handle);
_handle = static_cast<handle_t> (NULL);
}
void zmq::tcp_connecter_t::timer_event (int id_)
{
zmq_assert (id_ == reconnect_timer_id || id_ == connect_timer_id);
if (id_ == connect_timer_id) {
_connect_timer_started = false;
rm_handle ();
close ();
add_reconnect_timer ();
} else if (id_ == reconnect_timer_id) {
_reconnect_timer_started = false;
start_connecting ();
}
}
void zmq::tcp_connecter_t::start_connecting ()
{
// Open the connecting socket.
const int rc = open ();
// Connect may succeed in synchronous manner.
if (rc == 0) {
_handle = add_fd (_s);
out_event ();
}
// Connection establishment may be delayed. Poll for its completion.
else if (rc == -1 && errno == EINPROGRESS) {
_handle = add_fd (_s);
set_pollout (_handle);
_socket->event_connect_delayed (_endpoint, zmq_errno ());
// add userspace connect timeout
add_connect_timer ();
}
// Handle any other error condition by eventual reconnect.
else {
if (_s != retired_fd)
close ();
add_reconnect_timer ();
}
}
void zmq::tcp_connecter_t::add_connect_timer ()
{
if (options.connect_timeout > 0) {
add_timer (options.connect_timeout, connect_timer_id);
_connect_timer_started = true;
}
}
void zmq::tcp_connecter_t::add_reconnect_timer ()
{
const int interval = get_new_reconnect_ivl ();
add_timer (interval, reconnect_timer_id);
_socket->event_connect_retried (_endpoint, interval);
_reconnect_timer_started = true;
}
int zmq::tcp_connecter_t::get_new_reconnect_ivl ()
{
// The new interval is the current interval + random value.
const int interval =
_current_reconnect_ivl + generate_random () % options.reconnect_ivl;
// Only change the current reconnect interval if the maximum reconnect
// interval was set and if it's larger than the reconnect interval.
if (options.reconnect_ivl_max > 0
&& options.reconnect_ivl_max > options.reconnect_ivl)
// Calculate the next interval
_current_reconnect_ivl =
std::min (_current_reconnect_ivl * 2, options.reconnect_ivl_max);
return interval;
}
int zmq::tcp_connecter_t::open ()
{
zmq_assert (_s == retired_fd);
// Resolve the address
if (_addr->resolved.tcp_addr != NULL) {
LIBZMQ_DELETE (_addr->resolved.tcp_addr);
}
_addr->resolved.tcp_addr = new (std::nothrow) tcp_address_t ();
alloc_assert (_addr->resolved.tcp_addr);
int rc = _addr->resolved.tcp_addr->resolve (_addr->address.c_str (), false,
options.ipv6);
if (rc != 0) {
LIBZMQ_DELETE (_addr->resolved.tcp_addr);
return -1;
}
zmq_assert (_addr->resolved.tcp_addr != NULL);
const tcp_address_t *const tcp_addr = _addr->resolved.tcp_addr;
// Create the socket.
_s = open_socket (tcp_addr->family (), SOCK_STREAM, IPPROTO_TCP);
// IPv6 address family not supported, try automatic downgrade to IPv4.
if (_s == zmq::retired_fd && tcp_addr->family () == AF_INET6
&& errno == EAFNOSUPPORT && options.ipv6) {
rc = _addr->resolved.tcp_addr->resolve (_addr->address.c_str (), false,
false);
if (rc != 0) {
LIBZMQ_DELETE (_addr->resolved.tcp_addr);
return -1;
}
_s = open_socket (AF_INET, SOCK_STREAM, IPPROTO_TCP);
}
if (_s == retired_fd) {
return -1;
}
// On some systems, IPv4 mapping in IPv6 sockets is disabled by default.
// Switch it on in such cases.
if (tcp_addr->family () == AF_INET6)
enable_ipv4_mapping (_s);
// Set the IP Type-Of-Service priority for this socket
if (options.tos != 0)
set_ip_type_of_service (_s, options.tos);
// Bind the socket to a device if applicable
if (!options.bound_device.empty ())
bind_to_device (_s, options.bound_device);
// Set the socket to non-blocking mode so that we get async connect().
unblock_socket (_s);
// Set the socket to loopback fastpath if configured.
if (options.loopback_fastpath)
tcp_tune_loopback_fast_path (_s);
// Set the socket buffer limits for the underlying socket.
if (options.sndbuf >= 0)
set_tcp_send_buffer (_s, options.sndbuf);
if (options.rcvbuf >= 0)
set_tcp_receive_buffer (_s, options.rcvbuf);
// Set the IP Type-Of-Service for the underlying socket
if (options.tos != 0)
set_ip_type_of_service (_s, options.tos);
// Set a source address for conversations
if (tcp_addr->has_src_addr ()) {
// Allow reusing of the address, to connect to different servers
// using the same source port on the client.
int flag = 1;
#ifdef ZMQ_HAVE_WINDOWS
rc = setsockopt (_s, SOL_SOCKET, SO_REUSEADDR,
reinterpret_cast<const char *> (&flag), sizeof (int));
wsa_assert (rc != SOCKET_ERROR);
#elif defined ZMQ_HAVE_VXWORKS
rc = setsockopt (_s, SOL_SOCKET, SO_REUSEADDR, (char *) &flag,
sizeof (int));
errno_assert (rc == 0);
#else
rc = setsockopt (_s, SOL_SOCKET, SO_REUSEADDR, &flag, sizeof (int));
errno_assert (rc == 0);
#endif
#if defined ZMQ_HAVE_VXWORKS
rc = ::bind (_s, (sockaddr *) tcp_addr->src_addr (),
tcp_addr->src_addrlen ());
#else
rc = ::bind (_s, tcp_addr->src_addr (), tcp_addr->src_addrlen ());
#endif
if (rc == -1)
return -1;
}
// Connect to the remote peer.
#if defined ZMQ_HAVE_VXWORKS
rc = ::connect (_s, (sockaddr *) tcp_addr->addr (), tcp_addr->addrlen ());
#else
rc = ::connect (_s, tcp_addr->addr (), tcp_addr->addrlen ());
#endif
// Connect was successful immediately.
if (rc == 0) {
return 0;
}
// Translate error codes indicating asynchronous connect has been
// launched to a uniform EINPROGRESS.
#ifdef ZMQ_HAVE_WINDOWS
const int last_error = WSAGetLastError ();
if (last_error == WSAEINPROGRESS || last_error == WSAEWOULDBLOCK)
errno = EINPROGRESS;
else
errno = wsa_error_to_errno (last_error);
#else
if (errno == EINTR)
errno = EINPROGRESS;
#endif
return -1;
}
zmq::fd_t zmq::tcp_connecter_t::connect ()
{
// Async connect has finished. Check whether an error occurred
int err = 0;
#if defined ZMQ_HAVE_HPUX || defined ZMQ_HAVE_VXWORKS
int len = sizeof err;
#else
socklen_t len = sizeof err;
#endif
const int rc = getsockopt (_s, SOL_SOCKET, SO_ERROR,
reinterpret_cast<char *> (&err), &len);
// Assert if the error was caused by 0MQ bug.
// Networking problems are OK. No need to assert.
#ifdef ZMQ_HAVE_WINDOWS
zmq_assert (rc == 0);
if (err != 0) {
if (err == WSAEBADF || err == WSAENOPROTOOPT || err == WSAENOTSOCK
|| err == WSAENOBUFS) {
wsa_assert_no (err);
}
return retired_fd;
}
#else
// Following code should handle both Berkeley-derived socket
// implementations and Solaris.
if (rc == -1)
err = errno;
if (err != 0) {
errno = err;
#if !defined(TARGET_OS_IPHONE) || !TARGET_OS_IPHONE
errno_assert (errno != EBADF && errno != ENOPROTOOPT
&& errno != ENOTSOCK && errno != ENOBUFS);
#else
errno_assert (errno != ENOPROTOOPT && errno != ENOTSOCK
&& errno != ENOBUFS);
#endif
return retired_fd;
}
#endif
// Return the newly connected socket.
const fd_t result = _s;
_s = retired_fd;
return result;
}
bool zmq::tcp_connecter_t::tune_socket (const fd_t fd_)
{
const int rc = tune_tcp_socket (fd_)
| tune_tcp_keepalives (
fd_, options.tcp_keepalive, options.tcp_keepalive_cnt,
options.tcp_keepalive_idle, options.tcp_keepalive_intvl)
| tune_tcp_maxrt (fd_, options.tcp_maxrt);
return rc == 0;
}
void zmq::tcp_connecter_t::close ()
{
zmq_assert (_s != retired_fd);
#ifdef ZMQ_HAVE_WINDOWS
const int rc = closesocket (_s);
wsa_assert (rc != SOCKET_ERROR);
#else
const int rc = ::close (_s);
errno_assert (rc == 0);
#endif
_socket->event_closed (_endpoint, _s);
_s = retired_fd;
}