threads.h

/*
 * cugar
 * Copyright (c) 2011-2018, NVIDIA CORPORATION. All rights reserved.
 * 
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions are met:
 *    * Redistributions of source code must retain the above copyright
 *      notice, this list of conditions and the following disclaimer.
 *    * Redistributions in binary form must reproduce the above copyright
 *      notice, this list of conditions and the following disclaimer in the
 *      documentation and/or other materials provided with the distribution.
 *    * Neither the name of the NVIDIA CORPORATION nor the
 *      names of its contributors may be used to endorse or promote products
 *      derived from this software without specific prior written permission.
 * 
 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND
 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
 * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
 * DISCLAIMED. IN NO EVENT SHALL NVIDIA CORPORATION BE LIABLE FOR ANY
 * DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
 * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
 * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
 * ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
 * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 */

#pragma once

#include <cugar/basic/types.h>
#include <cugar/basic/numbers.h>
#include <cugar/basic/atomics.h>
#include <cugar/basic/shared_pointer.h>
#include <queue>

namespace cugar {




CUGAR_API uint32 num_physical_cores();
CUGAR_API uint32 num_logical_cores();

struct ThreadBaseImpl;
struct MutexImpl;

class ThreadBase
{
public:
    CUGAR_API  ThreadBase();
    CUGAR_API ~ThreadBase();

    void   set_id(const uint32 id) { m_id = id; }
    uint32 get_id() const { return m_id; }

    CUGAR_API void create(void* (*func)(void*), void* arg);

    CUGAR_API void join();

private:
    uint32                                      m_id;
    SharedPointer<ThreadBaseImpl, AtomicInt32>  m_impl;
};

template <typename DerivedThreadType>
class Thread : public ThreadBase
{
public:
    void create() { ThreadBase::create( DerivedThreadType::execute, static_cast<DerivedThreadType*>(this) ); }

    void join() { ThreadBase::join(); }

private:
    static void* execute(void* arg)
    {
        DerivedThreadType* data = reinterpret_cast<DerivedThreadType*>( arg );
        data->run();
        return NULL;
    }
};

//template class CUGAR_API internals::SharedCount<AtomicInt32>;
//template class CUGAR_API SharedPointer<MutexImpl, AtomicInt32>;

class CUGAR_API Mutex
{
public:
     Mutex();
    ~Mutex();

    void lock();
    void unlock();

private:
    SharedPointer<MutexImpl, AtomicInt32>  m_impl;
};

class ScopedLock
{
public:
     ScopedLock(Mutex* mutex) : m_mutex( mutex ) { m_mutex->lock(); }
    ~ScopedLock() { m_mutex->unlock(); }

private:
    Mutex* m_mutex;
};

template <typename WorkItemT, typename ProgressCallbackT>
class WorkQueue
{
public:
    typedef WorkItemT           WorkItem;
    typedef ProgressCallbackT   ProgressCallback;

    WorkQueue() : m_callback(), m_size(0u) {}

    void push(const WorkItem work) { m_queue.push( work ); m_size++; }

    void locked_push(const WorkItem work)
    {
        ScopedLock block( &m_lock );
        m_queue.push( work ); m_size++;
    }

    bool pop(WorkItem& work)
    {
        ScopedLock block( &m_lock );
        if (m_queue.empty())
            return false;

        work = m_queue.front();
        m_queue.pop();

        m_callback( m_size - (uint32)m_queue.size() - 1u, m_size );
        return true;
    }

    void set_callback(const ProgressCallback callback) { m_callback = callback; }

private:
    ProgressCallback      m_callback;
    std::queue<WorkItem>  m_queue;
    Mutex                 m_lock;
    uint32                m_size;
};

inline uint32 balance_batch_size(uint32 batch_size, uint32 total_count, uint32 thread_count)
{
    // How many batches we'd get with the proposed batch_size
    const uint32 batch_count = divide_ri(total_count, batch_size);
    // How many rounds we'd need for those batches
    const uint32 rounds      = divide_ri(batch_count, thread_count);
    // Might as well assume all threads should work, and see how many batches
    // they would consume
    const uint32 bal_batches = rounds * thread_count;
    // So that the batch size that will attain it, is computed as follows
    return divide_ri(total_count, bal_batches);
}

CUGAR_API void yield();


} // namespace cugar