Hi all
I need to declare a variable in Node to keep position of that node so I declare it in node.h
like: std::vector<double> exPosition;(public)
then I defined a simple function for getting this variable like:virtual Vector
GetmyPosition (void) const=0;
then in node.cc I wrote this simple function:
node::GetmyPosition (void) const
{
return exPosition;
}
but when i run it it has this error:
cannot allocate an object of abstract type 'ns3::Node'
note: because the following virtual functions are pure within 'ns3::Node';
here is the complete code for node.h:
#ifndef NODE_H
#define NODE_H
#include <vector>
#include "ns3/object.h"
#include "ns3/callback.h"
#include "ns3/ptr.h"
#include "ns3/net-device.h"
namespace ns3 {
class Application;
class Packet;
class Address;
/**
* \ingroup node
*
* \brief A network Node.
*
* This class holds together:
* - a list of NetDevice objects which represent the network interfaces
* of this node which are connected to other Node instances through
* Channel instances.
* - a list of Application objects which represent the userspace
* traffic generation applications which interact with the Node
* through the Socket API.
* - a node Id: a unique per-node identifier.
* - a system Id: a unique Id used for parallel simulations.
*
* Every Node created is added to the NodeList automatically.
*/
class Node : public Object
{
public:
/// exposition of the node
//std::vector<Ptr<Node> > exPosition;
std::vector<double> exPosition;
/// current position of the node
//std::vector<double> cPosition;
static TypeId GetTypeId (void);
/**
* Must be invoked by subclasses only.
*/
Node();
/**
* \param systemId a unique integer used for parallel simulations.
*
* Must be invoked by subclasses only.
*/
Node(uint32_t systemId);
virtual ~Node();
/**
* \returns the unique id of this node.
*
* This unique id happens to be also the index of the Node into
* the NodeList.
*/
uint32_t GetId (void) const;
/**
* \returns the system id for parallel simulations associated
* to this node.
*/
uint32_t GetSystemId (void) const;
/**
* \param device NetDevice to associate to this node.
* \returns the index of the NetDevice into the Node's list of
* NetDevice.
*
* Associate this device to this node.
*/
uint32_t AddDevice (Ptr<NetDevice> device);
/**
* \param index the index of the requested NetDevice
* \returns the requested NetDevice associated to this Node.
*
* The indexes used by the GetDevice method start at one and
* end at GetNDevices ()
*/
Ptr<NetDevice> GetDevice (uint32_t index) const;
/**
* \returns the number of NetDevice instances associated
* to this Node.
*/
uint32_t GetNDevices (void) const;
/**
* \param application Application to associate to this node.
* \returns the index of the Application within the Node's list
* of Application.
*
* Associated this Application to this Node. This method is called
* automatically from Application::Application so the user
* has little reasons to call this method directly.
*/
uint32_t AddApplication (Ptr<Application> application);
/**
* \param index
* \returns the application associated to this requested index
* within this Node.
*/
Ptr<Application> GetApplication (uint32_t index) const;
/**
* \returns the number of applications associated to this Node.
*/
uint32_t GetNApplications (void) const;
/**
* A protocol handler
*
* \param device a pointer to the net device which received the packet
* \param packet the packet received
* \param protocol the 16 bit protocol number associated with this packet.
* This protocol number is expected to be the same protocol number
* given to the Send method by the user on the sender side.
* \param sender the address of the sender
* \param receiver the address of the receiver; Note: this value is
* only valid for promiscuous mode protocol
* handlers. Note: If the L2 protocol does not use L2
* addresses, the address reported here is the value of
* device->GetAddress().
* \param packetType type of packet received
* (broadcast/multicast/unicast/otherhost); Note:
* this value is only valid for promiscuous mode
* protocol handlers.
*/
typedef Callback<void,Ptr<NetDevice>, Ptr<const Packet>,uint16_t,const Address &,
const Address &, NetDevice::PacketType> ProtocolHandler;
/**
* \param handler the handler to register
* \param protocolType the type of protocol this handler is
* interested in. This protocol type is a so-called
* EtherType, as registered here:
* http://standards.ieee.org/regauth/ethertype/eth.txt
* the value zero is interpreted as matching all
* protocols.
* \param device the device attached to this handler. If the
* value is zero, the handler is attached to all
* devices on this node.
* \param promiscuous whether to register a promiscuous mode handler
*/
void RegisterProtocolHandler (ProtocolHandler handler,
uint16_t protocolType,
Ptr<NetDevice> device,
bool promiscuous=false);
/**
* \param handler the handler to unregister
*
* After this call returns, the input handler will never
* be invoked anymore.
*/
void UnregisterProtocolHandler (ProtocolHandler handler);
/**
* \returns true if checksums are enabled, false otherwise.
*/
static bool ChecksumEnabled (void);
protected:
/**
* The dispose method. Subclasses must override this method
* and must chain up to it by calling Node::DoDispose at the
* end of their own DoDispose method.
*/
virtual void DoDispose (void);
virtual void DoStart (void);
private:
/**
* \param device the device added to this Node.
*
* This method is invoked whenever a user calls Node::AddDevice.
*/
virtual void NotifyDeviceAdded (Ptr<NetDevice> device);
bool NonPromiscReceiveFromDevice (Ptr<NetDevice> device, Ptr<const Packet>, uint16_t protocol, const Address &from);
bool PromiscReceiveFromDevice (Ptr<NetDevice> device, Ptr<const Packet>, uint16_t protocol,
const Address &from, const Address &to, NetDevice::PacketType packetType);
bool ReceiveFromDevice (Ptr<NetDevice> device, Ptr<const Packet>, uint16_t protocol,
const Address &from, const Address &to, NetDevice::PacketType packetType, bool promisc);
void Construct (void);
struct ProtocolHandlerEntry {
ProtocolHandler handler;
Ptr<NetDevice> device;
uint16_t protocol;
bool promiscuous;
};
typedef std::vector<struct Node::ProtocolHandlerEntry> ProtocolHandlerList;
uint32_t m_id; // Node id for this node
uint32_t m_sid; // System id for this node
std::vector<Ptr<NetDevice> > m_devices;
std::vector<Ptr<Application> > m_applications;
ProtocolHandlerList m_handlers;
};
} //namespace ns3
#endif /* NODE_H */
and also the complete code for node.cc
/* -*- Mode:C++; c-file-style:"gnu"; indent-tabs-mode:nil; -*- */
/*
* Copyright (c) 2006 Georgia Tech Research Corporation, INRIA
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License version 2 as
* published by the Free Software Foundation;
*
* This program 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 General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program; if not, write to the Free Software
* Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
*
* Authors: George F. Riley<riley#ece.gatech.edu>
* Mathieu Lacage <mathieu.lacage#sophia.inria.fr>
*/
#include "node.h"
#include "node-list.h"
#include "net-device.h"
#include "application.h"
#include "ns3/packet.h"
#include "ns3/simulator.h"
#include "ns3/object-vector.h"
#include "ns3/uinteger.h"
#include "ns3/log.h"
#include "ns3/assert.h"
#include "ns3/global-value.h"
#include "ns3/boolean.h"
#include "ns3/simulator.h"
#include "ns3/vector.h"
NS_LOG_COMPONENT_DEFINE ("Node");
namespace ns3{
NS_OBJECT_ENSURE_REGISTERED (Node);
GlobalValue g_checksumEnabled = GlobalValue ("ChecksumEnabled",
"A global switch to enable all checksums for all protocols",
BooleanValue (false),
MakeBooleanChecker ());
//Vector exposition = (0.0, 0.0, 0.0);
/*.AddAttribute ("exPosition", "The previous position of this node.",
TypeId::ATTR_GET,
VectorValue (Vector (0.0, 0.0, 0.0)), // ignored initial value.
MakeVectorAccessor (&Node::m_exposition),
MakeVectorChecker ())
.AddAttribute ("cPosition", "The current position of this node.",
TypeId::ATTR_GET,
VectorValue (Vector (0.0, 0.0, 0.0)), // ignored initial value.
MakeVectorAccessor (&Node::m_cposition),
MakeVectorChecker ())*/
TypeId
Node::GetTypeId (void)
{
static TypeId tid = TypeId ("ns3::Node")
.SetParent<Object> ()
.AddConstructor<Node> ()
.AddAttribute ("DeviceList", "The list of devices associated to this Node.",
ObjectVectorValue (),
MakeObjectVectorAccessor (&Node::m_devices),
MakeObjectVectorChecker<NetDevice> ())
.AddAttribute ("ApplicationList", "The list of applications associated to this Node.",
ObjectVectorValue (),
MakeObjectVectorAccessor (&Node::m_applications),
MakeObjectVectorChecker<Application> ())
.AddAttribute ("Id", "The id (unique integer) of this Node.",
TypeId::ATTR_GET, // allow only getting it.
UintegerValue (0),
MakeUintegerAccessor (&Node::m_id),
MakeUintegerChecker<uint32_t> ())
;
return tid;
}
Node::Node()
: m_id(0),
m_sid(0)
{
exPosition.at(1)= 0;
exPosition.at(2)= 0;
exPosition.at(3)= 0;
Construct ();
}
Node::Node(uint32_t sid)
: m_id(0),
m_sid(sid)
{
exPosition.at(1)= 0;
exPosition.at(2)= 0;
exPosition.at(3)= 0;
Construct ();
}
void
Node::Construct (void)
{
m_id = NodeList::Add (this);
//exPosition =(0.0,0.0,0.0);
}
Node::~Node ()
{}
uint32_t
Node::GetId (void) const
{
return m_id;
}
uint32_t
Node::GetSystemId (void) const
{
return m_sid;
}
uint32_t
Node::AddDevice (Ptr<NetDevice> device)
{
uint32_t index = m_devices.size ();
m_devices.push_back (device);
device->SetNode (this);
device->SetIfIndex(index);
device->SetReceiveCallback (MakeCallback (&Node::NonPromiscReceiveFromDevice, this));
Simulator::ScheduleWithContext (GetId (), Seconds (0.0),
&NetDevice::Start, device);
NotifyDeviceAdded (device);
return index;
}
Ptr<NetDevice>
Node::GetDevice (uint32_t index) const
{
NS_ASSERT_MSG (index < m_devices.size (), "Device index " << index <<
" is out of range (only have " << m_devices.size () << " devices).");
return m_devices[index];
}
uint32_t
Node::GetNDevices (void) const
{
return m_devices.size ();
}
uint32_t
Node::AddApplication (Ptr<Application> application)
{
uint32_t index = m_applications.size ();
m_applications.push_back (application);
application->SetNode (this);
Simulator::ScheduleWithContext (GetId (), Seconds (0.0),
&Application::Start, application);
return index;
}
Ptr<Application>
Node::GetApplication (uint32_t index) const
{
NS_ASSERT_MSG (index < m_applications.size (), "Application index " << index <<
" is out of range (only have " << m_applications.size () << " applications).");
return m_applications[index];
}
uint32_t
Node::GetNApplications (void) const
{
return m_applications.size ();
}
void
Node::DoDispose()
{
m_handlers.clear ();
for (std::vector<Ptr<NetDevice> >::iterator i = m_devices.begin ();
i != m_devices.end (); i++)
{
Ptr<NetDevice> device = *i;
device->Dispose ();
*i = 0;
}
m_devices.clear ();
for (std::vector<Ptr<Application> >::iterator i = m_applications.begin ();
i != m_applications.end (); i++)
{
Ptr<Application> application = *i;
application->Dispose ();
*i = 0;
}
m_applications.clear ();
Object::DoDispose ();
}
void
Node::DoStart (void)
{
for (std::vector<Ptr<NetDevice> >::iterator i = m_devices.begin ();
i != m_devices.end (); i++)
{
Ptr<NetDevice> device = *i;
device->Start ();
}
for (std::vector<Ptr<Application> >::iterator i = m_applications.begin ();
i != m_applications.end (); i++)
{
Ptr<Application> application = *i;
application->Start ();
}
Object::DoStart ();
}
void
Node::NotifyDeviceAdded (Ptr<NetDevice> device)
{}
void
Node::RegisterProtocolHandler (ProtocolHandler handler,
uint16_t protocolType,
Ptr<NetDevice> device,
bool promiscuous)
{
struct Node::ProtocolHandlerEntry entry;
entry.handler = handler;
entry.protocol = protocolType;
entry.device = device;
entry.promiscuous = promiscuous;
// On demand enable promiscuous mode in netdevices
if (promiscuous)
{
if (device == 0)
{
for (std::vector<Ptr<NetDevice> >::iterator i = m_devices.begin ();
i != m_devices.end (); i++)
{
Ptr<NetDevice> dev = *i;
dev->SetPromiscReceiveCallback (MakeCallback (&Node::PromiscReceiveFromDevice, this));
}
}
else
{
device->SetPromiscReceiveCallback (MakeCallback (&Node::PromiscReceiveFromDevice, this));
}
}
m_handlers.push_back (entry);
}
void
Node::UnregisterProtocolHandler (ProtocolHandler handler)
{
for (ProtocolHandlerList::iterator i = m_handlers.begin ();
i != m_handlers.end (); i++)
{
if (i->handler.IsEqual (handler))
{
m_handlers.erase (i);
break;
}
}
}
bool
Node::ChecksumEnabled (void)
{
BooleanValue val;
g_checksumEnabled.GetValue (val);
return val.Get ();
}
bool
Node::PromiscReceiveFromDevice (Ptr<NetDevice> device, Ptr<const Packet> packet, uint16_t protocol,
const Address &from, const Address &to, NetDevice::PacketType packetType)
{
NS_LOG_FUNCTION(this);
return ReceiveFromDevice (device, packet, protocol, from, to, packetType, true);
}
bool
Node::NonPromiscReceiveFromDevice (Ptr<NetDevice> device, Ptr<const Packet> packet, uint16_t protocol,
const Address &from)
{
NS_LOG_FUNCTION(this);
return ReceiveFromDevice (device, packet, protocol, from, from, NetDevice::PacketType (0), false);
}
bool
Node::ReceiveFromDevice (Ptr<NetDevice> device, Ptr<const Packet> packet, uint16_t protocol,
const Address &from, const Address &to, NetDevice::PacketType packetType, bool promiscuous)
{
NS_ASSERT_MSG (Simulator::GetContext () == GetId (), "Received packet with erroneous context ; " <<
"make sure the channels in use are correctly updating events context " <<
"when transfering events from one node to another.");
NS_LOG_DEBUG("Node " << GetId () << " ReceiveFromDevice: dev "
<< device->GetIfIndex () << " (type=" << device->GetInstanceTypeId ().GetName ()
<< ") Packet UID " << packet->GetUid ());
bool found = false;
for (ProtocolHandlerList::iterator i = m_handlers.begin ();
i != m_handlers.end (); i++)
{
if (i->device == 0 ||
(i->device != 0 && i->device == device))
{
if (i->protocol == 0 ||
i->protocol == protocol)
{
if (promiscuous == i->promiscuous)
{
i->handler (device, packet, protocol, from, to, packetType);
found = true;
}
}
}
}
return found;
}
}//namespace ns3
I'll be thankful if someone help me about it.
Bests
Bahar
You cannot instantiate an object that has a virtual function declared as =0.
You can have pointers to them, but the pointed elements will actually have a type that inherits from your base class.
GetMyPosition is defined, does it need to be pure virtual? Have you tried removing =0 from function declaration?
Related
int step = 100;
for (int i = 0; i < jobs.size(); i += step) {
thread_pool* osgbpool=new thread_pool(mt);
for (size_t k = 0; k < step; ++k) {
if (i + k == jobs.size())
{
break;
}
else {
auto it = jobs[i + k];
osgbpool->push_task(processProjectMainNoTrans, it, out_path);
}
}
osgbpool->wait_for_tasks();
delete osgbpool;
}
The above code is my C++ code. There are some explanations, jobs are my job queue; osgbpool is my thread pool; with time goes by, I will find the cpu is getting down and the application is becoming slow. Hence, I want to know the reason why this situation happend?
Here are my threadpool.h and threadpool.cpp :
#pragma once
/**
* #file thread_pool.hpp
* #author Barak Shoshany (baraksh#gmail.com) (http://baraksh.com)
* #version 2.0.0
* #date 2021-08-14
* #copyright Copyright (c) 2021 Barak Shoshany. Licensed under the MIT license. If you use this library in published research, please cite it as follows:
* - Barak Shoshany, "A C++17 Thread Pool for High-Performance Scientific Computing", doi:10.5281/zenodo.4742687, arXiv:2105.00613 (May 2021)
*
* #brief A C++17 thread pool for high-performance scientific computing.
* #details A modern C++17-compatible thread pool implementation, built from scratch with high-performance scientific computing in mind. The thread pool is implemented as a single lightweight and self-contained class, and does not have any dependencies other than the C++17 standard library, thus allowing a great degree of portability. In particular, this implementation does not utilize OpenMP or any other high-level multithreading APIs, and thus gives the programmer precise low-level control over the details of the parallelization, which permits more robust optimizations. The thread pool was extensively tested on both AMD and Intel CPUs with up to 40 cores and 80 threads. Other features include automatic generation of futures and easy parallelization of loops. Two helper classes enable synchronizing printing to an output stream by different threads and measuring execution time for benchmarking purposes. Please visit the GitHub repository at https://github.com/bshoshany/thread-pool for documentation and updates, or to submit feature requests and bug reports.
*/
#define THREAD_POOL_VERSION "v2.0.0 (2021-08-14)"
#include <atomic> // std::atomic
#include <chrono> // std::chrono
#include <cstdint> // std::int_fast64_t, std::uint_fast32_t
#include <functional> // std::function
#include <future> // std::future, std::promise
#include <iostream> // std::cout, std::ostream
#include <memory> // std::shared_ptr, std::unique_ptr
#include <mutex> // std::mutex, std::scoped_lock
#include <queue> // std::queue
#include <thread> // std::this_thread, std::thread
#include <type_traits> // std::common_type_t, std::decay_t, std::enable_if_t, std::is_void_v, std::invoke_result_t
#include <utility> // std::move
// ============================================================================================= //
// Begin class thread_pool //
/**
* #brief A C++17 thread pool class. The user submits tasks to be executed into a queue. Whenever a thread becomes available, it pops a task from the queue and executes it. Each task is automatically assigned a future, which can be used to wait for the task to finish executing and/or obtain its eventual return value.
*/
class thread_pool
{
typedef std::uint_fast32_t ui32;
typedef std::uint_fast64_t ui64;
public:
// ============================
// Constructors and destructors
// ============================
/**
* #brief Construct a new thread pool.
*
* #param _thread_count The number of threads to use. The default value is the total number of hardware threads available, as reported by the implementation. With a hyperthreaded CPU, this will be twice the number of CPU cores. If the argument is zero, the default value will be used instead.
*/
thread_pool(const ui32 &_thread_count = std::thread::hardware_concurrency())
: thread_count(_thread_count ? _thread_count : std::thread::hardware_concurrency()), threads(new std::thread[_thread_count ? _thread_count : std::thread::hardware_concurrency()])
{
create_threads();
}
/**
* #brief Destruct the thread pool. Waits for all tasks to complete, then destroys all threads. Note that if the variable paused is set to true, then any tasks still in the queue will never be executed.
*/
~thread_pool()
{
wait_for_tasks();
running = false;
destroy_threads();
}
// =======================
// Public member functions
// =======================
/**
* #brief Get the number of tasks currently waiting in the queue to be executed by the threads.
*
* #return The number of queued tasks.
*/
ui64 get_tasks_queued() const
{
const std::scoped_lock lock(queue_mutex);
return tasks.size();
}
/**
* #brief Get the number of tasks currently being executed by the threads.
*
* #return The number of running tasks.
*/
ui32 get_tasks_running() const
{
return tasks_total - (ui32)get_tasks_queued();
}
/**
* #brief Get the total number of unfinished tasks - either still in the queue, or running in a thread.
*
* #return The total number of tasks.
*/
ui32 get_tasks_total() const
{
return tasks_total;
}
/**
* #brief Get the number of threads in the pool.
*
* #return The number of threads.
*/
ui32 get_thread_count() const
{
return thread_count;
}
/**
* #brief Parallelize a loop by splitting it into blocks, submitting each block separately to the thread pool, and waiting for all blocks to finish executing. The user supplies a loop function, which will be called once per block and should iterate over the block's range.
*
* #tparam T1 The type of the first index in the loop. Should be a signed or unsigned integer.
* #tparam T2 The type of the index after the last index in the loop. Should be a signed or unsigned integer. If T1 is not the same as T2, a common type will be automatically inferred.
* #tparam F The type of the function to loop through.
* #param first_index The first index in the loop.
* #param index_after_last The index after the last index in the loop. The loop will iterate from first_index to (index_after_last - 1) inclusive. In other words, it will be equivalent to "for (T i = first_index; i < index_after_last; i++)". Note that if first_index == index_after_last, the function will terminate without doing anything.
* #param loop The function to loop through. Will be called once per block. Should take exactly two arguments: the first index in the block and the index after the last index in the block. loop(start, end) should typically involve a loop of the form "for (T i = start; i < end; i++)".
* #param num_blocks The maximum number of blocks to split the loop into. The default is to use the number of threads in the pool.
*/
template <typename T1, typename T2, typename F>
void parallelize_loop(const T1 &first_index, const T2 &index_after_last, const F &loop, ui32 num_blocks = 0)
{
typedef std::common_type_t<T1, T2> T;
T the_first_index = (T)first_index;
T last_index = (T)index_after_last;
if (the_first_index == last_index)
return;
if (last_index < the_first_index)
{
T temp = last_index;
last_index = the_first_index;
the_first_index = temp;
}
last_index--;
if (num_blocks == 0)
num_blocks = thread_count;
ui64 total_size = (ui64)(last_index - the_first_index + 1);
ui64 block_size = (ui64)(total_size / num_blocks);
if (block_size == 0)
{
block_size = 1;
num_blocks = (ui32)total_size > 1 ? (ui32)total_size : 1;
}
std::atomic<ui32> blocks_running = 0;
for (ui32 t = 0; t < num_blocks; t++)
{
T start = ((T)(t * block_size) + the_first_index);
T end = (t == num_blocks - 1) ? last_index + 1 : ((T)((t + 1) * block_size) + the_first_index);
blocks_running++;
push_task([start, end, &loop, &blocks_running]
{
loop(start, end);
blocks_running--;
});
}
while (blocks_running != 0)
{
sleep_or_yield();
}
}
/**
* #brief Push a function with no arguments or return value into the task queue.
*
* #tparam F The type of the function.
* #param task The function to push.
*/
template <typename F>
void push_task(const F &task)
{
tasks_total++;
{
const std::scoped_lock lock(queue_mutex);
tasks.push(std::function<void()>(task));
}
}
/**
* #brief Push a function with arguments, but no return value, into the task queue.
* #details The function is wrapped inside a lambda in order to hide the arguments, as the tasks in the queue must be of type std::function<void()>, so they cannot have any arguments or return value. If no arguments are provided, the other overload will be used, in order to avoid the (slight) overhead of using a lambda.
*
* #tparam F The type of the function.
* #tparam A The types of the arguments.
* #param task The function to push.
* #param args The arguments to pass to the function.
*/
template <typename F, typename... A>
void push_task(const F &task, const A &...args)
{
push_task([task, args...]
{ task(args...); });
}
/**
* #brief Reset the number of threads in the pool. Waits for all currently running tasks to be completed, then destroys all threads in the pool and creates a new thread pool with the new number of threads. Any tasks that were waiting in the queue before the pool was reset will then be executed by the new threads. If the pool was paused before resetting it, the new pool will be paused as well.
*
* #param _thread_count The number of threads to use. The default value is the total number of hardware threads available, as reported by the implementation. With a hyperthreaded CPU, this will be twice the number of CPU cores. If the argument is zero, the default value will be used instead.
*/
void reset(const ui32 &_thread_count = std::thread::hardware_concurrency())
{
bool was_paused = paused;
paused = true;
wait_for_tasks();
running = false;
destroy_threads();
thread_count = _thread_count ? _thread_count : std::thread::hardware_concurrency();
threads.reset(new std::thread[thread_count]);
paused = was_paused;
running = true;
create_threads();
}
/**
* #brief Submit a function with zero or more arguments and no return value into the task queue, and get an std::future<bool> that will be set to true upon completion of the task.
*
* #tparam F The type of the function.
* #tparam A The types of the zero or more arguments to pass to the function.
* #param task The function to submit.
* #param args The zero or more arguments to pass to the function.
* #return A future to be used later to check if the function has finished its execution.
*/
template <typename F, typename... A, typename = std::enable_if_t<std::is_void_v<std::invoke_result_t<std::decay_t<F>, std::decay_t<A>...>>>>
std::future<bool> submit(const F &task, const A &...args)
{
std::shared_ptr<std::promise<bool>> task_promise(new std::promise<bool>);
std::future<bool> future = task_promise->get_future();
push_task([task, args..., task_promise]
{
try
{
task(args...);
task_promise->set_value(true);
}
catch (...)
{
try
{
task_promise->set_exception(std::current_exception());
}
catch (...)
{
}
}
});
return future;
}
/**
* #brief Submit a function with zero or more arguments and a return value into the task queue, and get a future for its eventual returned value.
*
* #tparam F The type of the function.
* #tparam A The types of the zero or more arguments to pass to the function.
* #tparam R The return type of the function.
* #param task The function to submit.
* #param args The zero or more arguments to pass to the function.
* #return A future to be used later to obtain the function's returned value, waiting for it to finish its execution if needed.
*/
template <typename F, typename... A, typename R = std::invoke_result_t<std::decay_t<F>, std::decay_t<A>...>, typename = std::enable_if_t<!std::is_void_v<R>>>
std::future<R> submit(const F &task, const A &...args)
{
std::shared_ptr<std::promise<R>> task_promise(new std::promise<R>);
std::future<R> future = task_promise->get_future();
push_task([task, args..., task_promise]
{
try
{
task_promise->set_value(task(args...));
}
catch (...)
{
try
{
task_promise->set_exception(std::current_exception());
}
catch (...)
{
}
}
});
return future;
}
/**
* #brief Wait for tasks to be completed. Normally, this function waits for all tasks, both those that are currently running in the threads and those that are still waiting in the queue. However, if the variable paused is set to true, this function only waits for the currently running tasks (otherwise it would wait forever). To wait for a specific task, use submit() instead, and call the wait() member function of the generated future.
*/
void wait_for_tasks()
{
while (true)
{
if (!paused)
{
if (tasks_total == 0)
break;
}
else
{
if (get_tasks_running() == 0)
break;
}
sleep_or_yield();
}
}
// ===========
// Public data
// ===========
/**
* #brief An atomic variable indicating to the workers to pause. When set to true, the workers temporarily stop popping new tasks out of the queue, although any tasks already executed will keep running until they are done. Set to false again to resume popping tasks.
*/
std::atomic<bool> paused = false;
/**
* #brief The duration, in microseconds, that the worker function should sleep for when it cannot find any tasks in the queue. If set to 0, then instead of sleeping, the worker function will execute std::this_thread::yield() if there are no tasks in the queue. The default value is 1000.
*/
ui32 sleep_duration = 1000;
private:
// ========================
// Private member functions
// ========================
/**
* #brief Create the threads in the pool and assign a worker to each thread.
*/
void create_threads()
{
for (ui32 i = 0; i < thread_count; i++)
{
threads[i] = std::thread(&thread_pool::worker, this);
}
}
/**
* #brief Destroy the threads in the pool by joining them.
*/
void destroy_threads()
{
for (ui32 i = 0; i < thread_count; i++)
{
threads[i].join();
}
}
/**
* #brief Try to pop a new task out of the queue.
*
* #param task A reference to the task. Will be populated with a function if the queue is not empty.
* #return true if a task was found, false if the queue is empty.
*/
bool pop_task(std::function<void()> &task)
{
const std::scoped_lock lock(queue_mutex);
if (tasks.empty())
return false;
else
{
task = std::move(tasks.front());
tasks.pop();
return true;
}
}
/**
* #brief Sleep for sleep_duration microseconds. If that variable is set to zero, yield instead.
*
*/
void sleep_or_yield()
{
if (sleep_duration)
std::this_thread::sleep_for(std::chrono::microseconds(sleep_duration));
else
std::this_thread::yield();
}
/**
* #brief A worker function to be assigned to each thread in the pool. Continuously pops tasks out of the queue and executes them, as long as the atomic variable running is set to true.
*/
void worker()
{
while (running)
{
std::function<void()> task;
if (!paused && pop_task(task))
{
task();
tasks_total--;
}
else
{
sleep_or_yield();
}
}
}
// ============
// Private data
// ============
/**
* #brief A mutex to synchronize access to the task queue by different threads.
*/
mutable std::mutex queue_mutex = {};
/**
* #brief An atomic variable indicating to the workers to keep running. When set to false, the workers permanently stop working.
*/
std::atomic<bool> running = true;
/**
* #brief A queue of tasks to be executed by the threads.
*/
std::queue<std::function<void()>> tasks = {};
/**
* #brief The number of threads in the pool.
*/
ui32 thread_count;
/**
* #brief A smart pointer to manage the memory allocated for the threads.
*/
std::unique_ptr<std::thread[]> threads;
/**
* #brief An atomic variable to keep track of the total number of unfinished tasks - either still in the queue, or running in a thread.
*/
std::atomic<ui32> tasks_total = 0;
};
// End class thread_pool //
// ============================================================================================= //
// ============================================================================================= //
// Begin class synced_stream //
/**
* #brief A helper class to synchronize printing to an output stream by different threads.
*/
class synced_stream
{
public:
/**
* #brief Construct a new synced stream.
*
* #param _out_stream The output stream to print to. The default value is std::cout.
*/
synced_stream(std::ostream &_out_stream = std::cout)
: out_stream(_out_stream) {};
/**
* #brief Print any number of items into the output stream. Ensures that no other threads print to this stream simultaneously, as long as they all exclusively use this synced_stream object to print.
*
* #tparam T The types of the items
* #param items The items to print.
*/
template <typename... T>
void print(const T &...items)
{
const std::scoped_lock lock(stream_mutex);
(out_stream << ... << items);
}
/**
* #brief Print any number of items into the output stream, followed by a newline character. Ensures that no other threads print to this stream simultaneously, as long as they all exclusively use this synced_stream object to print.
*
* #tparam T The types of the items
* #param items The items to print.
*/
template <typename... T>
void println(const T &...items)
{
print(items..., '\n');
}
private:
/**
* #brief A mutex to synchronize printing.
*/
mutable std::mutex stream_mutex = {};
/**
* #brief The output stream to print to.
*/
std::ostream &out_stream;
};
// End class synced_stream //
// ============================================================================================= //
// ============================================================================================= //
// Begin class timer //
/**
* #brief A helper class to measure execution time for benchmarking purposes.
*/
class timer
{
typedef std::int_fast64_t i64;
public:
/**
* #brief Start (or restart) measuring time.
*/
void start()
{
start_time = std::chrono::steady_clock::now();
}
/**
* #brief Stop measuring time and store the elapsed time since start().
*/
void stop()
{
elapsed_time = std::chrono::steady_clock::now() - start_time;
}
/**
* #brief Get the number of milliseconds that have elapsed between start() and stop().
*
* #return The number of milliseconds.
*/
i64 ms() const
{
return (std::chrono::duration_cast<std::chrono::milliseconds>(elapsed_time)).count();
}
private:
/**
* #brief The time point when measuring started.
*/
std::chrono::time_point<std::chrono::steady_clock> start_time = std::chrono::steady_clock::now();
/**
* #brief The duration that has elapsed between start() and stop().
*/
std::chrono::duration<double> elapsed_time = std::chrono::duration<double>::zero();
};
// End class timer //
// ============================================================================================= //
I'm the author of the thread pool library you are using.
First of all, your question implicitly presents my work as your own. Please give credit when credit is due...
Second, you seem to be creating a new a thread_pool object and then deleting it in every iteration of the for loop. That's not how you're supposed to use the thread pool, and indeed it defeats the whole purpose of a thread pool.
The point of the thread pool is to avoid the overhead of creating and destroying a new thread for each task, but the way you're using it now, you are forcing your program to create and destroy all the threads all over again in each iteration of the loop, which is most likely the reason your program is being slow.
Instead, create the thread_pool object only once when the program starts, and then use the same object throughout the program by submitting jobs to it.
Also, there is absolutely no reason to use manual memory allocation (new and delete) here. The thread_pool object itself is small, and you only need to create one such object, so it takes a negligible amount of memory in the stack, and you do not need to allocate memory for it on the heap. Furthermore, manual memory allocation in C++ can easily lead to memory leaks if not used correctly. Instead, just create the object as usual (e.g. thread_pool pool).
I suggest that you check out the documentation for the thread pool library for examples of how to use it correctly.
A few things
If the number of threads is greater than or equal to the number of cores and all threads are doing non-blocking work, then of course your computer is going to slow down. Consider having the thread pool use with hardware_concurrency()-1 to save some cycles for the rest of the application or computer.
Your sleep_or_yield function is inefficient. learn how to properly use std::condition_variable with a mutex and lock for the thread to sleep until a condition changes.
I'm not sure if this is the root issue of your performance problems, but I think it will help.
I have created a threading library and it is used in our company program and the program is a customized software which is in C++ based. As shown in the code below, there are some customized variable types, such as FloatM1D, Ints and etc.
In the threading library, i'm using pthread_create (can't use the newer threading option due the C++ based is in old version) to fire threads that do calculation, then store the results in dictionary for later use.
Multiple threads are fired at one-go (DoEVM routine as shown in the code will be called multiple times at upper level) and wait for each thread completion before retrieval of the result. The thread wait completion (WaitThreadCompletion() as shown in the code) will be called only once.
After the retrieval of the results, I will cleanup all the dictionary.
My problem is, the calculation routine (VSA.calculateEvm() as shown in the code) in the thread is throwing segmentation error intermittently and caused my program hang when it tries to check the thread completion in pthread_join. And I have no access/visible to VSA.calculateEvm() routine because this came from third party library, thus can't debugged deeper.
The error:
An invalid memory reference (segmentation violation) has occurred.
The code from the threading library is provided at below and need some help to check if there is any issue with it that can cause the segmentation error?
Any advise is welcome and thanks in advance.
#define PARALLEL_DEVM_VERSION 2.0
#include <iostream>
#include <cstdio>
#include <cstring>
#include <string>
#include <pthread.h>
#include <stdio.h>
#include <stdlib.h>
#include <EVM.h>
#include <cstdlib>
#include <fstream>
ThreadingModule EVMx;
void *thread_EVM(void *arg);
pthread_mutex_t lock_x = PTHREAD_MUTEX_INITIALIZER;
pthread_mutex_t lock_y = PTHREAD_MUTEX_INITIALIZER;
/*THREAD SHARED DATA STRUCTURE*/
struct ThreadData
{
//Warning: Do not use custom variable types in struct, it does not work good with pthread
float if_Freq;
float sample_Rate;
int tnum;
bool enable_Plots;
bool ResequenceIsNeeded;
int if_Bin;
int RAT;
};
typedef struct ThreadData m_ThreadData_Type;
BoolS m_Parallel_EVM;
map<int, FloatM> m_TnumVsDEVM_map;
map<int, FloatM1D> m_Tnum_ReseqArray_map;
map<int, pthread_t> m_TnumVsThreadIds_map;
map<int, FloatM> m_TnumVsResDEVM_map;
map<int, int> m_TnumVsUTPaver_map;
IntS m_tnum;
/*THREAD SHARED DATA - THESE HAVE TO USE WITH MUTEX LOCK */
FloatM1D m_measArrayTmp;
FloatM1D m_refArrayTmp;
UnsignedS m_if_Bin;
UnsignedS m_RAT;
Dot11acChannelType chan_Type;
Dot11PilotTracking pilot_Tracking;
Dot11EqualizerTraining EQ_Train;
Dot11Trigger trig;
StringS wave_ID;
/* RESERVED STRING */
IntS TEST_ID_PREFIX = 1000;
IntS ResequenceTimeDomainToComplex(FloatM1D measArray, FloatM1D complexOutArray, const UnsignedS ifBin, const UnsignedS stride);
IntS Evm (FloatM1D measArray, FloatM1D refArray, const Dot11acChannelType chanType, const Dot11PilotTracking PILOT_TRACK, const FloatS ifFreq, const FloatS sampleRate, const Dot11EqualizerTraining EQ_TRAIN, const Dot11Trigger TRIGGER, OfdmEvmData &evmResults, const BoolS enablePlots = false , const StringS plotTag = "");
BoolS Init(int TestNum, FloatM1D resdevm, IntS resdevmIndex, BoolS enabledataPlots=false, BoolS enableEVMsweep=false, BoolS UTP_AVG=true, IntS NumOfAverage=1);
void WaitThreadCompletion(IntS &totalThreads);
void PostThreadsCleanup();
map<int, FloatM> GetDEVMs();
map<int, FloatM1D> GetResDEVMs();
ThreadingModule::ThreadingModule()
{
//empty constr
}
IntS ThreadingModule::ResequenceTimeDomainToComplex(FloatM1D measArray, FloatM1D &complexOutArray, const UnsignedS ifBin, const UnsignedS stride)
{
/**
* #brief Resequence array wrapper. If in parallel mode, it will do it at EVM thread.
* #param measArray measured array.
* #param complexOutArray array
* #param ifBin argument
* #param stride argument
* #param enableEVMsweep argument
* #return Parallel EVM flag.
*/
IntS rslt;
IntS ret;
/*## PARALLEL MODE ##*/
//Push the array to dictionary, resequence it later
if (!m_Tnum_ReseqArray_map.count(m_tnum)>0){
m_Tnum_ReseqArray_map[m_tnum] = measArray;
m_if_Bin = ifBin;
m_RAT = stride;
} else {
//Fail safe
ERR.ReportError(ERR_GENERIC_WARNING,"Duplicated TNUM found in the dictionary in ResequenceTimeDomainToComplex.",UTL_VOID, NO_SITES, UTL_VOID);
}
return rslt;
}
IntS ThreadingModule::DoEvm(FloatM1D measArray, FloatM1D refArray,const Dot11acChannelType chanType, const Dot11PilotTracking PILOT_TRACK, const FloatS ifFreq, const FloatS sampleRate, const Dot11EqualizerTraining EQ_TRAIN, const Dot11Trigger TRIGGER, OfdmEvmData &evmResults, const BoolS enablePlots, const StringS plotTag)
{
/**
* #brief EVM wrapper.
* #param measArray measured array.
* #param refArray reference array.
* #param chanType channel type.
* #param PILOT_TRACK pilot tracking
* #param ifFreq IF freq.
* #param sampleRate EVM flag sample rate.
* #param Dot11EqualizerTraining equalizer training.
* #param Dot11Trigger trigger.
* #param OfdmEvmData devm results.
* #param enablePlots enable plots
* #param plotTag plot tagging.
* #return Ints not using.
*
*/
IntS rslt, ret;
BoolS m_isResequenceNeeded(false);
FloatM1D m_tmp_meas_Arr;
UnsignedS m_tmpifBin(0);
UnsignedS m_tmpRAT(0);
IntS m_tmp_tnum;
/*## PARALLEL MODE ##*/
/* check if resequence array is needed */
if(m_Tnum_ReseqArray_map.count(m_tnum)>0){
m_isResequenceNeeded=true;
m_tmp_meas_Arr=m_Tnum_ReseqArray_map[m_tnum];
m_tmpifBin= m_if_Bin;
m_tmpRAT=m_RAT;
m_Tnum_ReseqArray_map.erase(m_tnum);
}
/* instantiate the structure */
m_ThreadData_Type *tdata;
tdata =(m_ThreadData_Type *)malloc(sizeof (m_ThreadData_Type));
/* mutex lock to initialize all thread shared variables */
pthread_mutex_lock(&lock_y);
m_measArrayTmp = m_isResequenceNeeded ? m_tmp_meas_Arr : measArray;
m_refArrayTmp = refArray;
chan_Type = chanType;
pilot_Tracking = PILOT_TRACK;
EQ_Train = EQ_TRAIN;
trig = TRIGGER;
wave_ID = plotTag;
pthread_mutex_unlock(&lock_y);
/* thread specific data */
tdata->if_Freq = ifFreq;
tdata->sample_Rate = sampleRate;
tdata->enable_Plots = enablePlots;
tdata->tnum = m_tnum;
tdata->ResequenceIsNeeded=m_isResequenceNeeded;
tdata->if_Bin = m_tmpifBin;
tdata->RAT = m_tmpRAT;
/* create thread */
ret = pthread_create(&m_TnumVsThreadIds_map[m_tnum], NULL, thread_EVM, (void*) tdata);
if (ret!=0) ERR.ReportError(ERR_GENERIC_WARNING , "Error in create EVM thread for TNUM="+ m_tnum.GetText()+".", ret, NO_SITES, UTL_VOID);
return rslt;
}
void* thread_EVM(void* arg)
{
/**
* #brief EVM wrapper.
* #param arg Arguments passed in with structure
*
*/
FloatM1D m_meas_Arr;
FloatM1D m_ref_Arr;
FloatM1D m_cmplx_Arr;
Dot11acChannelType m_chan_Type;
Dot11PilotTracking m_pilot_Tracking;
FloatS m_freq;
FloatS m_sample_Rate;
Dot11EqualizerTraining m_EQ_Train;
Dot11Trigger m_trig;
IntS m_testnum;
BoolS m_enable_Plots;
StringS m_wave_ID;
OfdmEvmData m_evm_Results;
UnsignedS m_Bin;
UnsignedS m_RAT;
BoolS m_isResequencedNeeded(false);
/* initialize all the local variables from arg */
m_ThreadData_Type *tdata;
tdata=(m_ThreadData_Type*)arg;
/* mutex lock to retrieve thread shared data */
pthread_mutex_lock(&lock_y);
m_meas_Arr = m_measArrayTmp;
m_ref_Arr = m_refArrayTmp;
m_pilot_Tracking = pilot_Tracking;
m_EQ_Train = EQ_Train;
m_chan_Type = chan_Type;
m_trig = trig;
m_wave_ID = wave_ID;
pthread_mutex_unlock(&lock_y);
/* initialize thread specific data */
m_freq = tdata->if_Freq;
m_sample_Rate = tdata->sample_Rate;
m_testnum = tdata->tnum;
m_enable_Plots = tdata->enable_Plots;
m_Bin = tdata->if_Bin;
m_RAT = tdata->RAT;
m_isResequencedNeeded = tdata->ResequenceIsNeeded;
/* check if resequence array is needed */
if (m_isResequencedNeeded){
m_cmplx_Arr.Resize(m_ref_Arr.GetSize());
VSA.DOT11.AC.ResequenceTimeDomainToComplex(m_meas_Arr, m_cmplx_Arr, m_Bin, m_RAT);
m_meas_Arr=m_cmplx_Arr;
}
/* perform evm calculation */
VSA.calculateEvm(m_meas_Arr, m_ref_Arr, m_chan_Type, m_pilot_Tracking, m_freq, m_sample_Rate, m_EQ_Train, m_trig, m_evm_Results, m_enable_Plots, m_wave_ID);
pthread_mutex_lock(&lock_x);
m_TnumVsDEVM_map[m_testnum]= m_evm_Results.evm;
pthread_mutex_unlock(&lock_x);
free(tdata);
return NULL;
}
void ThreadingModule::WaitThreadCompletion(IntS &totalThreads)
{
/**
* #brief Check all the created threads are complete.
* #param totalThreads Total number of created threads.
*
*/
int ret;
map<int, pthread_t>::iterator itr;
IntS tnum;
pthread_t threadiD;
/* check each threads if completed */
for (itr = m_TnumVsThreadIds_map.begin(); itr != m_TnumVsThreadIds_map.end(); ++itr) {
tnum = itr->first;
threadiD = itr->second;
ret=pthread_join(threadiD, NULL);
if (ret!=0) ERR.ReportError(ERR_GENERIC_WARNING , "Can't complete the thread for TNUM="+ tnum.GetText()+".", UTL_VOID, NO_SITES, UTL_VOID);
}
totalThreads = m_TnumVsThreadIds_map.size();
}
map<int, FloatM> ThreadingModule::GetDEVMs()
{
return m_TnumVsDEVM_map;
}
map<int, FloatM> ThreadingModule::GetResDEVMs()
{
return m_TnumVsResDEVM_map;
}
void ThreadingModule::PostThreadsCleanup()
{
/**
* #brief Post threads completion cleanup. Make sure to run this only after datalog the results.
* WARNING: Only call this routine when all threads complete executed.
*/
TEST_ID_PREFIX =1000;
m_TnumVsDEVM_map.clear();
m_Tnum_ReseqArray_map.clear();
m_TnumVsThreadIds_map.clear();
m_TnumVsResDEVM_map.clear();
m_TnumVsUTPaver_map.clear();
pthread_mutex_lock(&lock_y);
m_measArrayTmp.Clear();
m_refArrayTmp.Clear();
pthread_mutex_unlock(&lock_y);
}
Was wondering if you guys could provide/assist me with concise solutions to determine whether a BTree underflows, and whether or not you could complete a right rotation on it. Bit confused as to the process of arriving at it. Don think it's as easy as just comparing the two unsigned values (particularly for underflow).
BTreeNode.h:
#ifndef BTREENODE_H
#define BTREENODE_H
#include <string>
#include <vector>
struct BTreeNode {
bool is_leaf_ = true;
std::vector<int> elements_;
std::vector<BTreeNode*> children_;
BTreeNode() {}
BTreeNode(std::vector<int> v) {
this->elements_ = v;
}
/**
* Fix the underflow child node at idx by rotating right
* (borrowing a node from left sibling).
* #param idx The underflow child to be fixed is at children_[idx].
* #return If the rotation can be done.
*/
bool rotateRight(unsigned idx, unsigned order);
};
/**
* Check if the given number of elements in a BTree node underflows.
* #param numElem Number of elements in this node.
* #param order The order of the BTree.
* #return True if it underflows, False otherwise.
*/
bool underflows(unsigned numElem, unsigned order);
/**
* A special case for removing an element from BTree. Assume elem
exists in leaf.
* #param item The element to be removed.
* #param parent The parent node that contains the leaf node as a
child.
* #param leaf_idx The leaf BTreeNode idx that contains the element to
be removed.
* #return If the removal is successful.
*/
bool removeFromLeaf(int item, BTreeNode* parent, unsigned leaf_idx,
unsigned order);
#endif
BTreeNode.cpp:
#include "BTreeNode.h"
#include <assert.h>
#include <algorithm>
#include <iostream>
/**
* Check if the given number of elements in a BTree node underflows.
* #param numElem Number of elements in this node.
* #param order The order of the BTree.
* #return True if it underflows, False otherwise.
*/
bool underflows(unsigned numElem, unsigned order) {
return false;
}
/**
* Fix the underflow child node at idx by rotating right
* (borrowing a node from left sibling).
* #param idx The underflow child to be fixed is at children_[idx].
* #return If the rotation can be done.
*/
bool BTreeNode::rotateRight(unsigned idx, unsigned order) {
/**
* First check if there is a left sibling.
* If there is not, simply return false because rotateRight cannot
be done.
*/
if (idx <= 0) return false;
/**
* Then check if the left sibling leaf contains enough elements
after one being borrowed.
*/
BTreeNode* prev = children_[idx - 1];
if (underflows(prev->elements_.size() - 1, order)) {
/**
* If it's not enough, this case cannot be handled by
rotateRight.
* Simply return false.
*/
return false;
}
/**
* Do the right rotation by stealing one element from left sibling
* and fixing the parent key.
*
* Example: Assume we are doing rotateRight around (40) to fix
right child
* (we are in BTreeNode(40), idx = 1 (the second child)),
* | 40 |
* / \
* | 10 | 20 | 30 | | 60 |
*
* after rotation, the tree should look like
* | 30 |
* / \
* | 10 | 20 | | 40 | 50 |
*
*/
// TODO: do the right rotation here
return true;
}
bool removeFromLeaf(int item, BTreeNode* parent, unsigned leaf_idx,
unsigned order) {
// sanity checks
assert(!parent->is_leaf_);
assert(leaf_idx < parent->children_.size());
BTreeNode* leaf = parent->children_[leaf_idx];
assert(leaf->is_leaf_);
std::vector<int>& elems = leaf->elements_;
std::vector<int>::iterator pos = std::find(elems.begin(),
elems.end(), item);
assert(pos != elems.end());
std::cout << "removing " << item << "..." << std::endl;
// delete item, shift other items, shrink the size
elems.erase(pos);
std::cout << "Does the node underflow? ";
// call rotateRight if current leaf node underflows
if (underflows(elems.size(), order)) {
std::cout << "Yes!" << std::endl;
return parent->rotateRight(leaf_idx, order);
}
std::cout << "No!" << std::endl;
return true;
}
I would like to implement a workaround to use a non-static class as a call-back function. I am working with Eclipse Paho MQTT code. The following type is implemented and used as callback:
typedef void MQTTAsync_onSuccess(void* context, MQTTAsync_successData* response);
MQTTAsync_onSuccess* onSuccess;
onSuccess = myStaticCallback;
void myStaticCallback (void* context, MQTTAsync_successData* response)
{
//...callback actions...
}
I want to wrap this C API (without modifying the existing MQTT C API) and implement non-static / non-centralized callback function that belongs to an object/class.
typedef void MQTTAsync_onSuccess(void* context, MQTTAsync_successData* response);
class myMQTTClass
{
private:
void myCallback (void* context, MQTTAsync_successData* response);
MQTTAsync_onSuccess* onSuccess;
public:
void foo (void)
{
this->onSuccess = this->myCallback;
}
}
As you might guess, the code above causes the error:
cannot convert myCallback from type 'void (myMQTTClass::) (void*, MQTTAsync_successData*)' to type 'void (*)(void*, MQTTAsync_successData*)'.
Any guidance as to how to address this issue or any workaround is greately appreciated. I would be willing to provide any possible missing information. Thanks in advance.
EDIT: Actual code with some omissions
namespace rover
{
typedef struct
{
char * clientID;
char * topic;
char * payload;
int qos; // 1
long int timeout; // Such as 10000L usec
} RoverMQTT_Configure_t;
class RoverPahoMQTT
{
public:
RoverPahoMQTT (char * host_name, int port, RoverMQTT_Configure_t MQTT_Configure);
private:
/**
* #brief Host name used for connecting to the Eclipse Paho MQTT server
*/
char * HOST_NAME;
/**
* #brief Port used for connecting to the Eclipse Paho MQTT server
*/
int PORT;
RoverMQTT_Configure_t rover_MQTT_configure;
/* Internal attributes */
MQTTAsync client;
/**
* #brief Connect options
*/
MQTTAsync_connectOptions conn_opts;
/**
* #brief Disconnect options
*/
MQTTAsync_disconnectOptions disc_opts;
//...
static void onPublisherConnect (void* context, MQTTAsync_successData* response);
void onPublisherConnect_ (MQTTAsync_successData* response);
//...
}
}
int rover::RoverPahoMQTT::publish (void)
{
this->flushFlags ();
this->conn_opts = MQTTAsync_connectOptions_initializer;
this->client = new MQTTAsync;
int rc;
char my_addr[20];
this->constructAddress (my_addr);
printf ("address: %s", my_addr);
MQTTAsync_create ( &(this->client),
my_addr,
this->rover_MQTT_configure.clientID,
MQTTCLIENT_PERSISTENCE_NONE,
NULL);
MQTTAsync_setCallbacks(this->client, NULL, onConnectionLost, NULL, NULL);
conn_opts.keepAliveInterval = 20;
conn_opts.cleansession = 1;
conn_opts.onSuccess = rover::RoverPahoMQTT::onPublisherConnect;
conn_opts.onFailure = onConnectFailure;
conn_opts.context = this->client;
if ((rc = MQTTAsync_connect(this->client, &(this->conn_opts))) != MQTTASYNC_SUCCESS)
{
printf("Failed to start connect, return code %d\n", rc);
return rc;
}
/*printf("Waiting for publication of %s\n"
"on topic %s for client with ClientID: %s\n",
PAYLOAD, TOPIC, CLIENTID);*/
while (!mqtt_finished)
usleep(this->rover_MQTT_configure.timeout);
MQTTAsync_destroy(&client);
return rc;
}
void rover::RoverPahoMQTT::onPublisherConnect_(MQTTAsync_successData* response)
{
MQTTAsync_responseOptions opts = MQTTAsync_responseOptions_initializer;
MQTTAsync_message pubmsg = MQTTAsync_message_initializer;
int rc;
printf("Successful connection\n");
opts.onSuccess = onPublisherSend;
opts.context = client;
pubmsg.payload = &default_MQTT_configure.payload;
pubmsg.payloadlen = strlen(default_MQTT_configure.payload);
pubmsg.qos = default_MQTT_configure.qos;
pubmsg.retained = 0;
deliveredtoken = 0;
if ((rc = MQTTAsync_sendMessage(client, default_MQTT_configure.topic, &pubmsg, &opts)) != MQTTASYNC_SUCCESS)
{
printf("Failed to start sendMessage, return code %d\n", rc);
exit(EXIT_FAILURE);
}
}
void rover::RoverPahoMQTT::onPublisherConnect (void* context, MQTTAsync_successData* response)
{
rover::RoverPahoMQTT* m = (rover::RoverPahoMQTT*) context;
m->onPublisherConnect_(response);
//((rover::RoverPahoMQTT*)context)->onPublisherConnect_(response);
// ^^^HERE IS THE SEGMENTATION FAULT
}
As clearly stated here, the callback has to be
registered with the client library by passing it as an argument in
MQTTAsync_responseOptions
and the context argument is a
pointer to the context value originally passed to
MQTTAsync_responseOptions, which contains any application-specific
context.
I suggest a common interface for your classes, which provides a static method that matches the callback prototype:
class myMQTTClass
{
public:
static void callback(void* context, MQTTAsync_successData* response)
{
myMQTTClass * m = (myMQTTClass*)context;
m->myCallback(response);
}
protected:
virtual void myCallback(MQTTAsync_successData* response) = 0;
};
You can now implement different behaviours in subclasses:
class myMQTTClassImpl : public myMQTTClass
{
protected:
void myCallback(MQTTAsync_successData *response)
{
std::cout << "success!!!" << std::endl;
}
};
Let's see how to use it:
int main()
{
myMQTTClass * m = new myMQTTClassImpl();
MQTTAsync_responseOptions options;
options.onSuccess = myMQTTClass::callback;
options.context = m;
}
Edit (refers to actual code posted):
In your publish method, this is right:
conn_opts.onSuccess = rover::RoverPahoMQTT::onPublisherConnect;
this is wrong:
conn_opts.context = this->client;
it should be:
conn_opts.context = this;
If you can use the context parameter to point to the object you want the callback to fire on you could simply make the callback function static and forward to an instance function. If the context parameter is needed for other data then I would use libffi to generate a closure and associate the object pointer with the closure's saved arguments.
There is no correct way to pass an actual instance function to that callback and be sure that it will work (even if you made the instance function be something like void MyCallback(MQTTAsync_successData*) and then forcibly cast that to the callback type you aren't guaranteed that the calling convention would match.
For the first (where you can use the context arg to point to 'this'):
class MyCallback
{
static void CallbackFunc(void * ptr, MQTTAsync_successData* data)
{
((MyCallback*)ptr)->RealCallback(data);
}
void RealCallback(MQTTAsync_successData*)
{}
};
You would then assign &MyCallback::CallbackFunc to the function pointer.
libffi is quite a bit more complicated.
I've been using this site for a while and so far never needed to ask a new question (found all answers I've needed until now).
I need to push_back multiple objects into a vector but VS throws an error (This may be due to a corruption of the heap, which indicates a bug in PVSS00DataGate.exe or any of the DLLs it has loaded) that I cannot seem to work out for myself.
Here is what I am trying to do, it works to push_back the first object into the vector but when I try to push_back a second object that is when the error occurs.
class HWObject{}
void DataLogger::WriteNewMessages()
{
unsigned int iBattery = 0;
unsigned int iSignal = 0;
TimeVar tTimeStamp;
// I want to store all HWObjects in a temporary vector (loggerData)
std::vector<HWObject> loggerData;
CharString strElement;
strElement.format( "batteryCondition.value" );
SendOneValuePVSS( (const char *)strElement, iBattery, tTimeStamp );
strElement.format( "signalStrength.value" );
SendOneValuePVSS( (const char *)strElement, iSignal, tTimeStamp );
}
void DataLogger::SendOneValuePVSS(const char *szElementName, double dValue, TimeVar, &tValue)
{
HWObject obj;
obj.setOrgTime(tValue); // Current time
obj.setTimeOfPeriphFlag();
CharString address;
address = strID;
address += ".";
address += szElementName;
obj.setAddress(address);
loggerData.reserve( loggerData.size() + 1 );
loggerData.push_back( obj );
obj.cutData();
}
dataLogger is declared in
class DataLogger
{
public:
std::vector<HWObject> loggerData;
...
}
Here is the class HWObject, I didn't want to overwhelm you with code.
public:
/** Default constructor*/
HWObject();
/** Constructor, which sets the periphAddr and transformationType.
* #param addressString address of the HW object
* #param trans type of transformation
*/
HWObject(const char* addressString, TransformationType trans);
/** Destructor. If the date pointer is not NULL, it is deleted.
*/
virtual ~HWObject();
/** Creates a new HWObject
* This function returns an empty HWObject, no properties are duplicated or copied!
* #classification public use, overload, call base
*/
virtual HWObject * clone() const;
/** Reset all pvss2 relevant parts of the HWObject. when overloading this member
* don't forget to call the basic function!
* #classification public use, overload, call base
*/
virtual void clear();
/** Gets pointer to data
* #return pointer to data
*/
const PVSSchar * getDataPtr() const { return dataPtr; }
/** Gets the data buffer pointer
* #return data buffer pointer
*/
PVSSchar * getData() { return dataPtr; }
/** Cut the data buffer out of the HWObject.
* This function is used to avoid the deletion
* of the data buffer, when a new pointer is set using
* setData() or the HWObject is deleted.
* #return pointer to the data of the HWObject
*/
PVSSchar * cutData();
/** Get the data buffer lenght
* #return length of the data buffer
*/
PVSSushort getDlen() const { return dataLen; }
/** Set ptr to the data buffer, pointer is captured.
* The actual data pointer in the HWObject is deleted,
* if it is not NULL. To avoid the deletion use cutData()
* in order to cut out the pointer.
* #param ptr pointer to new data
*/
void setData(PVSSchar *ptr);
/** Set the data length
* #param len length to be set
*/
void setDlen(const PVSSushort len) { dataLen = len; }
/** Get the periph address
* #return periph address string
*/
const CharString & getAddress() const { return address; }
/** Get the transformation type
* #return type of transformation
*/
TransformationType getType() const { return transType; }
/** Set the transformation type
* #param typ type of transformation for setting
*/
void setType(const TransformationType typ) { transType = typ; }
/** Get the subindex
* #return subindex
*/
PVSSushort getSubindex() const { return subindex; }
/** Set the subindex
* #param sx subindex to be set
*/
void setSubindex( const PVSSushort sx) { subindex = sx; }
/** Get the origin time
* #return origin time
*/
const TimeVar& getOrgTime() const { return originTime; }
/** Get the origin time
* #return oriin time
*/
TimeVar& getOrgTime() { return originTime; }
/** Set the origin time
* #param tm origin time to be set
*/
void setOrgTime(const TimeVar& tm) { originTime = tm; }
/** Get HWObject purpose
* #return objSrcType
*/
ObjectSourceType getObjSrcType() const { return objSrcType; }
/** Set HWObject purpose
* #param tp objSrcType
*/
void setObjSrcType(const ObjectSourceType tp) { objSrcType = tp; }
/** Get number of elements in data buffer
* #return number of elements in data buffer
*/
PVSSushort getNumberOfElements() const { return number_of_elements; }
/** Set number of elements in data buffer
* #param var number of elements in data buffer
*/
void setNumberOfElements(const PVSSushort var) { number_of_elements = var; }
/** Prints the basic HWObject information on stderr.
* in case of overloading don't forget to call the base function!
* #classification public use, overload, call base
*/
virtual void debugPrint() const;
/** Prints th basic HWObject info in one CharString for internal debug DP.
* in case of overloading call base function first, then append specific info
* #classification public use, overload, call base
*/
virtual CharString getInfo() const;
/** Set the periph address
* #param adrStr pointer to address string
*/
virtual void setAddress(const char *adrStr);
/** Set the additional data (flag, orig time, valid user byte,etc)
* #param data aditional flags that be set
* #param subix subindex, use subix 0 for setting values by default
*/
virtual void setAdditionalData(const RecVar &data, PVSSushort subix);
/** Set the 'origin time comes from periph' flag
*/
void setTimeOfPeriphFlag()
{
setSbit(DRV_TIME_OF_PERIPH);
}
/** Check whether time comes from periph
* #return PVSS_TRUE if the time is from perip
*/
PVSSboolean isTimeFromPeriph() const
{
// If isTimeOfPeriph is set, it must be valid
return getSbit(DRV_TIME_OF_PERIPH);
}
/** Set the flag if you want to receive callback if event has answered the value change
*/
void setWantAnswerFlag()
{
setSbit(DRV_WANT_ANSWER);
}
/** Get the status of the 'want answer, flag
*/
PVSSboolean getWantAnswerFlag() const
{
// If isTimeOfPeriph is set, it must be valid
return getSbit(DRV_WANT_ANSWER);
}
/** Set the user bit given by input parameter.
* Status bits defined by the enum DriverBits
* #param bitno bit number
* #return PVSS_TRUE if bit was set
*/
PVSSboolean setSbit(PVSSushort bitno)
{
return (status.set(bitno) && status.set(bitno + (PVSSushort)DRV_VALID_INVALID - (PVSSushort)DRV_INVALID));
}
/** Clear the user bit given by input parameter
* #param bitno bit number
* #return PVSS_TRUE if bit was cleared
*/
PVSSboolean clearSbit(PVSSushort bitno)
{
return (status.clear(bitno) && status.set(bitno + (PVSSushort)DRV_VALID_INVALID - (PVSSushort)DRV_INVALID));
}
PVSSboolean isValidSbit(PVSSushort bitno) const
{
return status.get(bitno + (PVSSushort)DRV_VALID_INVALID - (PVSSushort)DRV_INVALID);
}
/** Check any bit
* #param bitno bit number
* #return status of the bit on bitno position
*/
PVSSboolean getSbit(PVSSushort bitno) const {return status.get(bitno);}
/** Clear all status bits
* return status of clear all
*/
PVSSboolean clearStatus() {return status.clearAll();}
/** Get the status of this object
* #return bit vector status
*/
const BitVec & getStatus() const {return status;}
/** Set status of the bit vector
* #param bv deference to bit vector to be set as status
*/
void setStatus(const BitVec &bv) {status = bv;}
/** Set a user byte in the status.
* #param userByteNo number of user byte range 0..3
* #param val value to set
* #return PVSS_TRUE execution OK
* PVSS_FALSE in case of error
*/
PVSSboolean setUserByte (PVSSushort userByteNo, PVSSuchar val);
/** Reads a user byte from the status.
* #param userByteNo number of user byte range 0..3
* #return the requested user byte
*/
PVSSuchar getUserByte (PVSSushort userByteNo) const;
/** Check validity of user byte.
* #param userByteNo number of user byte range 0..3
* #return PVSS_TRUE user byte is valid
* PVSS_FALSE user byte is not valid
*/
PVSSboolean isValidUserByte(PVSSushort userByteNo) const;
/** Format status bits into a string
* #param str status bits converted to string
*/
void formatStatus (CharString & str) const ;
// ------------------------------------------------------------------
// internal ones
/** Read data from bin file
* #param fp file handle
*/
virtual int inFromBinFile( FILE *fp );
/** Write data to bin file
* #param fp file handle
*/
virtual int outToBinFile( FILE *fp );
/** Set data length
* #param dlen data length
*/
void setDlenLlc (PVSSushort dlen) {dataLenLlc = dlen;}
virtual void updateBufferLlc (HWObject * hwo, int offset1 = 0);
virtual int compareLlc (HWObject * hwo, int offset1 = 0, int offset2 = 0, int len = -1);
/** Get dataLenLlc
* #return dataLenLlc
*/
PVSSushort getDlenLlc () const {return dataLenLlc;}
/** Function to delete the data buffer; overload if special deletion must be done
*/
virtual void deleteData ();
/** Set HW identifier
* #param id hw identifier to be set
*/
void setHwoId(PVSSulong id) {hwoId = id;}
/** Get HW identifier
* #return hw identifier
*/
PVSSulong getHwoId() const {return hwoId;}
protected:
/// the dynamic data buffer
PVSSchar* dataPtr;
/// the data buffer len
PVSSushort dataLen;
/// the pvss2 periph address string
CharString address;
/// the start subix for the data buffer
PVSSushort subindex;
/// the datatype of the data in the buffer (i.e. transformationtype)
TransformationType transType;
/// the time of income, normally set by the constructor
TimeVar originTime;
/// the purpose of this HWObject
ObjectSourceType objSrcType;
/// the number of elements in the data buffer, used for arrays and records
PVSSushort number_of_elements; // fuer array!!!
/// the user bits of the original config
BitVec status;
private:
PVSSushort dataLenLlc;
PVSSulong hwoId;
};
You're not showing the important parts. I'd guess that HWObject has
dynamically allocated memory, and doesn't implement the rule of three
(copy constructor, assignment operator and destructor). But it's only a
guess. (Unless you're using special techniques like reference counting
or smart pointers, copy must do a deep copy, and assignment should
probably use the swap idiom.)
Also, there's no point in reserving size() + 1 just before
push_back.