I want to implement a function that I create 5 pairs threads(one pair means one thread write and the other read and both of the threads share one list (5 lists in this scenario).
Do I need to create five mutex locks? How to declare them? In global area?
Do I need to create five mutex locks?
Depends on your data structure. If you have five different data objects to access safely from five associated thread pairs, you'll need five, if all threads access only one data object you'll need only one.
How to declare them? In global area?
Encapsulate your data object, the mutex and (writing) thread functions in a class. I'd say you'll not need another reading thread, that's the one that calls run usually, or any other that'll have access to instance of this class.
class MyAsynchDataProvider
{
public:
void run()
{
writeThread = std::thread(writeDataFunc,this);
}
MyDataStruct getSafeDataCopy()
{
std::lock_guard lock(dataGuard);
return data;
}
private:
std::mutex dataGuard;
MyDataStruct data;
std::thread writeThread;
static void writeThreadFunc(MyDataWorker* thisPtr)
{
// ...
std::lock_guard lock(thisPtr->dataGuard);
// Write to thisPtr->data member
}
};
Related
I'm programming a simple 3D rendering engine just to get more familliar with C++. Today I had my first steps with multithreading and already have a problem I cannot wrap my head around. When the application starts it generates a small, minecraft-like terrain consisting of cubes. They're generated withhin the main thread.
Now when I want to generate more chunks
void VoxelWorld::generateChunk(glm::vec2 chunkPosition) {
Chunk* generatedChunk = m_worldGenerator->generateChunk(chunkPosition);
generatedChunk->shader = m_chunkShader;
generatedChunk->generateRenderObject();
m_chunks[chunkPosition.x][chunkPosition.y] = generatedChunk;
m_loadedChunks.push_back(glm::vec2(chunkPosition.x, chunkPosition.y));
}
void VoxelWorld::generateChunkThreaded(glm::vec2 chunkPosition) {
std::thread chunkThread(&VoxelWorld::generateChunk, this, chunkPosition);
chunkThread.detach();
}
void VoxelWorld::draw() {
for(glm::vec2& vec : m_loadedChunks){
Transformation* transformation = new Transformation();
transformation->getPosition().setPosition(glm::vec3(CHUNK_WIDTH*vec.x, 0, CHUNK_WIDTH*vec.y));
m_chunks[vec.x][vec.y]->getRenderObject()->draw(transformation);
delete(transformation); //TODO: Find a better way
}
}
I have my member function (everything is non-static) generateChunk() which generates a Chunk and stores it in the VoxelWorld class. I have a 2D std::map<..> m_chunks which stores every chunk and a std::vector<glm::vec2> m_loadedChunks which stores the positions of the generated chunks.
Calling generateChunk() works fine as expected. But when I try generateChunkThreaded() the application crashes! I tried commenting out the last line of generateChunk(), then it does not crash. Thats what confuses me so much! m_loadedChunks ist just a regular std::vector. I tried making it public, with no effect. Is there anything obvious I miss?
You are accessing m_loadedChunks from several threads without synchronizing it.
You need to lock the usage of shared usages. So few tips here.
Declare a mutex as a member of the class
std::mutex mtx; // mutex for critical section
Use it to lock via a critical section each time you want to access the elements
std::lock_guard lock(mtx);
m_chunks[chunkPosition.x][chunkPosition.y] = generatedChunk;
m_loadedChunks.push_back(glm::vec2(chunkPosition.x, chunkPosition.y));
Hope that helps
When you have many threads access shared resources, you either have those resources available as read-only, atomic, or guarded with a mutex lock.
So, for your m_loadedChunks member variable, you would want to have it wrapped in a lock. For example:
class VoxelWorld
{
// your class members and more ...
private:
std::mutex m_loadedChunksMutex;
}
void VoxelWorld::generateChunk(glm::vec2 chunkPosition)
{
Chunk* generatedChunk = m_worldGenerator->generateChunk(chunkPosition);
generatedChunk->shader = m_chunkShader;
generatedChunk->generateRenderObject();
m_chunks[chunkPosition.x][chunkPosition.y] = generatedChunk;
{
auto&& scopedLock = std::lock_guard< std::mutex >(m_loadedChunksMutex);
(void)scopedLock;
m_loadedChunks.push_back(glm::vec2(chunkPosition.x, chunkPosition.y));
}
}
The scopedLock will automatically wait for a lock and when the code goes out of scope, the lock will be released.
Now note, that I have a mutex for m_loadedChunks and not a generic mutex covering all variables that may be accessed by threads. This is actually a good practice introduced by Herb Sutter in his "Effective Concurrency" courses and on his talks at cppcon.
So, for whatever shared variables you have, use the above example as one means to solve race issues.
I am thinking about how to implement a class that will contain private data that will be eventually be modified by multiple threads through method calls. For synchronization (using the Windows API), I am planning on using a CRITICAL_SECTION object since all the threads will spawn from the same process.
Given the following design, I have a few questions.
template <typename T> class Shareable
{
private:
const LPCRITICAL_SECTION sync; //Can be read and used by multiple threads
T *data;
public:
Shareable(LPCRITICAL_SECTION cs, unsigned elems) : sync{cs}, data{new T[elems]} { }
~Shareable() { delete[] data; }
void sharedModify(unsigned index, T &datum) //<-- Can this be validly called
//by multiple threads with synchronization being implicit?
{
EnterCriticalSection(sync);
/*
The critical section of code involving reads & writes to 'data'
*/
LeaveCriticalSection(sync);
}
};
// Somewhere else ...
DWORD WINAPI ThreadProc(LPVOID lpParameter)
{
Shareable<ActualType> *ptr = static_cast<Shareable<ActualType>*>(lpParameter);
T copyable = /* initialization */;
ptr->sharedModify(validIndex, copyable); //<-- OK, synchronized?
return 0;
}
The way I see it, the API calls will be conducted in the context of the current thread. That is, I assume this is the same as if I had acquired the critical section object from the pointer and called the API from within ThreadProc(). However, I am worried that if the object is created and placed in the main/initial thread, there will be something funky about the API calls.
When sharedModify() is called on the same object concurrently,
from multiple threads, will the synchronization be implicit, in the
way I described it above?
Should I instead get a pointer to the
critical section object and use that instead?
Is there some other
synchronization mechanism that is better suited to this scenario?
When sharedModify() is called on the same object concurrently, from multiple threads, will the synchronization be implicit, in the way I described it above?
It's not implicit, it's explicit. There's only only CRITICAL_SECTION and only one thread can hold it at a time.
Should I instead get a pointer to the critical section object and use that instead?
No. There's no reason to use a pointer here.
Is there some other synchronization mechanism that is better suited to this scenario?
It's hard to say without seeing more code, but this is definitely the "default" solution. It's like a singly-linked list -- you learn it first, it always works, but it's not always the best choice.
When sharedModify() is called on the same object concurrently, from multiple threads, will the synchronization be implicit, in the way I described it above?
Implicit from the caller's perspective, yes.
Should I instead get a pointer to the critical section object and use that instead?
No. In fact, I would suggest giving the Sharable object ownership of its own critical section instead of accepting one from the outside (and embrace RAII concepts to write safer code), eg:
template <typename T>
class Shareable
{
private:
CRITICAL_SECTION sync;
std::vector<T> data;
struct SyncLocker
{
CRITICAL_SECTION &sync;
SyncLocker(CRITICAL_SECTION &cs) : sync(cs) { EnterCriticalSection(&sync); }
~SyncLocker() { LeaveCriticalSection(&sync); }
}
public:
Shareable(unsigned elems) : data(elems)
{
InitializeCriticalSection(&sync);
}
Shareable(const Shareable&) = delete;
Shareable(Shareable&&) = delete;
~Shareable()
{
{
SyncLocker lock(sync);
data.clear();
}
DeleteCriticalSection(&sync);
}
void sharedModify(unsigned index, const T &datum)
{
SyncLocker lock(sync);
data[index] = datum;
}
Shareable& operator=(const Shareable&) = delete;
Shareable& operator=(Shareable&&) = delete;
};
Is there some other synchronization mechanism that is better suited to this scenario?
That depends. Will multiple threads be accessing the same index at the same time? If not, then there is not really a need for the critical section at all. One thread can safely access one index while another thread accesses a different index.
If multiple threads need to access the same index at the same time, a critical section might still not be the best choice. Locking the entire array might be a big bottleneck if you only need to lock portions of the array at a time. Things like the Interlocked API, or Slim Read/Write locks, might make more sense. It really depends on your thread designs and what you are actually trying to protect.
Imagine you have a big function that locks/unlocks a mutex inside and you want to break the function into smaller functions:
#include <pthread.h>
class MyClass : public Uncopyable
{
public:
MyClass() : m_mutexBuffer(PTHREAD_MUTEX_INITIALIZER), m_vecBuffer() {}
~MyClass() {}
void MyBigFunction()
{
pthread_mutex_lock(&m_mutexBuffer);
if (m_vecBuffer.empty())
{
pthread_mutex_unlock(&m_mutexBuffer);
return;
}
// DoSomethingWithBuffer1();
unsigned char ucBcc = CalculateBcc(&m_vecBuffer[0], m_vecBuffer.size());
// DoSomethingWithBuffer2();
pthread_mutex_unlock(&m_mutexBuffer);
}
private:
void DoSomethingWithBuffer1()
{
// Use m_vecBuffer
}
void DoSomethingWithBuffer2()
{
// Use m_vecBuffer
}
private:
pthread_mutex_t m_mutexBuffer;
std::vector<unsigned char> m_vecBuffer;
};
How should I go about locking/unlocking the mutex inside the smaller functions?
Should I unlock the mutex first, then lock it straightaway and finally unlock it before returning?
void DoSomethingWithBuffer1()
{
pthread_mutex_unlock(&m_mutexBuffer);
pthread_mutex_lock(&m_mutexBuffer);
// Use m_vecBuffer
pthread_mutex_unlock(&m_mutexBuffer);
}
How should I go about locking/unlocking the mutex inside the smaller functions?
If your semantics require your mutex to be locked during the whole MyBigFunction() operation then you can't simply unlock it and relock it in the middle of the function.
My best bet would be to ignore the mutex in the smaller DoSomethingWithBuffer...() functions, and simply require that these functions are called with the mutex being already locked. This shouldn't be a problem since those functions are private.
On a side note, your mutex usage is incorrect: it is not exception safe, and you have code paths where you don't release the mutex. You should either use C++11's mutex and lock classes or boost's equivalents if you are using C++03. At worst if you can't use boost, write a small RAII wrapper to hold the lock.
In general, try to keep the regions of code within each lock to a minimum (to avoid contention), but avoid to unlock and immediatly re-lock the same mutex. Thus, if the smaller functions are not mutually exclusive, they should both use their own indepdenent mutices and only when they actually access the shared resource.
Another thing that should consider is to use RAII for locking and unlocking (as in C++11 with std::lock_guard<>), so that returning from a locked region (either directly or via an uncaught exception) does not leave you in a locked state.
There is some exemplary class of container in pseudo code:
class Container
{
public:
Container(){}
~Container(){}
void add(data new)
{
// addition of data
}
data get(size_t which)
{
// returning some data
}
void remove(size_t which)
{
// delete specified object
}
private:
data d;
};
How this container can be made thread safe? I heard about mutexes - where these mutexes should be placed? Should mutex be static for a class or maybe in global scope? What is good library for this task in C++?
First of all mutexes should not be static for a class as long as you going to use more than one instance. There is many cases where you should or shouldn't use use them. So without seeing your code it's hard to say. Just remember, they are used to synchronise access to shared data. So it's wise to place them inside methods that modify or rely on object's state. In your case I would use one mutex to protect whole object and lock all three methods. Like:
class Container
{
public:
Container(){}
~Container(){}
void add(data new)
{
lock_guard<Mutex> lock(mutex);
// addition of data
}
data get(size_t which)
{
lock_guard<Mutex> lock(mutex);
// getting copy of value
// return that value
}
void remove(size_t which)
{
lock_guard<Mutex> lock(mutex);
// delete specified object
}
private:
data d;
Mutex mutex;
};
Intel Thread Building Blocks (TBB) provides a bunch of thread-safe container implementations for C++. It has been open sourced, you can download it from: http://threadingbuildingblocks.org/ver.php?fid=174 .
First: sharing mutable state between threads is hard. You should be using a library that has been audited and debugged.
Now that it is said, there are two different functional issue:
you want a container to provide safe atomic operations
you want a container to provide safe multiple operations
The idea of multiple operations is that multiple accesses to the same container must be executed successively, under the control of a single entity. They require the caller to "hold" the mutex for the duration of the transaction so that only it changes the state.
1. Atomic operations
This one appears simple:
add a mutex to the object
at the start of each method grab a mutex with a RAII lock
Unfortunately it's also plain wrong.
The issue is re-entrancy. It is likely that some methods will call other methods on the same object. If those once again attempt to grab the mutex, you get a dead lock.
It is possible to use re-entrant mutexes. They are a bit slower, but allow the same thread to lock a given mutex as much as it wants. The number of unlocks should match the number of locks, so once again, RAII.
Another approach is to use dispatching methods:
class C {
public:
void foo() { Lock lock(_mutex); foo_impl(); }]
private:
void foo_impl() { /* do something */; foo_impl(); }
};
The public methods are simple forwarders to private work-methods and simply lock. Then one just have to ensure that private methods never take the mutex...
Of course there are risks of accidentally calling a locking method from a work-method, in which case you deadlock. Read on to avoid this ;)
2. Multiple operations
The only way to achieve this is to have the caller hold the mutex.
The general method is simple:
add a mutex to the container
provide a handle on this method
cross your fingers that the caller will never forget to hold the mutex while accessing the class
I personally prefer a much saner approach.
First, I create a "bundle of data", which simply represents the class data (+ a mutex), and then I provide a Proxy, in charge of grabbing the mutex. The data is locked so that the proxy only may access the state.
class ContainerData {
protected:
friend class ContainerProxy;
Mutex _mutex;
void foo();
void bar();
private:
// some data
};
class ContainerProxy {
public:
ContainerProxy(ContainerData& data): _data(data), _lock(data._mutex) {}
void foo() { data.foo(); }
void bar() { foo(); data.bar(); }
};
Note that it is perfectly safe for the Proxy to call its own methods. The mutex will be released automatically by the destructor.
The mutex can still be reentrant if multiple Proxies are desired. But really, when multiple proxies are involved, it generally turns into a mess. In debug mode, it's also possible to add a "check" that the mutex is not already held by this thread (and assert if it is).
3. Reminder
Using locks is error-prone. Deadlocks are a common cause of error and occur as soon as you have two mutexes (or one and re-entrancy). When possible, prefer using higher level alternatives.
Add mutex as an instance variable of class. Initialize it in constructor, and lock it at the very begining of every method, including destructor, and unlock at the end of method. Adding global mutex for all instances of class (static member or just in gloabl scope) may be a performance penalty.
The is also a very nice collection of lock-free containers (including maps) by Max Khiszinsky
LibCDS1 Concurrent Data Structures
Here is the documentation page:
http://libcds.sourceforge.net/doc/index.html
It can be kind of intimidating to get started, because it is fully generic and requires you register a chosen garbage collection strategy and initialize that. Of course, the threading library is configurable and you need to initialize that as well :)
See the following links for some getting started info:
initialization of CDS and the threading manager
http://sourceforge.net/projects/libcds/forums/forum/1034512/topic/4600301/
the unit tests ((cd build && ./build.sh ----debug-test for debug build)
Here is base template for 'main':
#include <cds/threading/model.h> // threading manager
#include <cds/gc/hzp/hzp.h> // Hazard Pointer GC
int main()
{
// Initialize \p CDS library
cds::Initialize();
// Initialize Garbage collector(s) that you use
cds::gc::hzp::GarbageCollector::Construct();
// Attach main thread
// Note: it is needed if main thread can access to libcds containers
cds::threading::Manager::attachThread();
// Do some useful work
...
// Finish main thread - detaches internal control structures
cds::threading::Manager::detachThread();
// Terminate GCs
cds::gc::hzp::GarbageCollector::Destruct();
// Terminate \p CDS library
cds::Terminate();
}
Don't forget to attach any additional threads you are using:
#include <cds/threading/model.h>
int myThreadFunc(void *)
{
// initialize libcds thread control structures
cds::threading::Manager::attachThread();
// Now, you can work with GCs and libcds containers
....
// Finish working thread
cds::threading::Manager::detachThread();
}
1 (not to be confuse with Google's compact datastructures library)
I´m using a class that encapsulates a thread_group, and have some questions about it
class MyGroup{
private:
boost::this_thread::id _id;
boost::thread::thread_group group;
int abc;
//other attributes
public:
void foo();
};
In the class constructor, i launch N threads
for (size_t i=0;i<N;i++){
group.add(new boost::thread(boost::bind(&foo,this)));
}
void foo(){
_id = boost::this_thread::get_id();
//more code.
abc++ //needs to be sync?
}
So, here are my questions.
Do class attributes need to be synchronized?
Do every thread get a different id? For example, if I have
void bar(){
this->id_;
}
will this result in different ids for each thread, or the same for everyone?
Thanks in advance !
Yes, shared data access must be protected even if you use thread creation helpers as boost.
In the end they all will execute the same code at the same time, and there is nothing a library can do to put protection around a variable you own and you manage.
If this->_id prints the current thread id then yes, it will print different values while different threads access it.
I don't know what you are doing with this thread_group so this may or may not apply.
Yes, all threads will have a unique ID.
Yes, you need to protect your shared state, you can do this with synchronization or by 'avoiding' shared state by copying it or with message passing.
A relevant pattern here is the 'actor' pattern.
Essentially rather than just create threads in your constructor, consider either:
a) have a class that derives from boost::thread and store thread specific members there. you can then access the member variables in the thread which won't be global to the group.
e.g.
class MyThreadClass : public boost::thread
{
private:
int thread_local_int;
...
}
b) have a class that contains a boost::thread as a member variable
class MyThreadClass : public boost::thread
{
private:
int thread_local_int;
boost::thread t;
public:
boost::thread& GetThread()
{
return t;
}
...
}
store a collection of either of these in your MyGroup class and use thread_group::add_thread to put the threads in the thread_group.
You can now be incredibly thoughtful about which state is shared in the thread_group (it should be synchronized or read-only) and which state is local to your actor (or thread) and how it's accessible.
Note, I have a personal reluctance against using TLS because I like having some control and guarantees over lifetimes of objects and threads and I just find this easier when I don't use it; YMMV and its great for some uses...