src/common/SynchronizedConfig.h

/***************************************************************************
 *                                                                         *
 *   Copyright (C) 2006-2012 Andreas Persson                               *
 *                                                                         *
 *   This program is free software; you can redistribute it and/or modify  *
 *   it under the terms of the GNU General Public License as published by  *
 *   the Free Software Foundation; either version 2 of the License, or     *
 *   (at your option) any later version.                                   *
 *                                                                         *
 *   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., 51 Franklin St, Fifth Floor, Boston,                *
 *   MA  02110-1301  USA                                                   *
 ***************************************************************************/

#ifndef SYNCHRONIZEDCONFIG_H
#define SYNCHRONIZEDCONFIG_H

#include <set>
#include <unistd.h>
#include "lsatomic.h"
#include "Mutex.h"

namespace LinuxSampler {

    /**
     * Thread-safe management of configuration data, where the data is
     * updated by a single non real time thread and read by a number
     * of real time threads.
     *
     * The synchronization is achieved by using two instances of the
     * configuration data. The non real time thread gets access to the
     * instance not currently in use by the real time threads by
     * calling GetConfigForUpdate(). After the data is updated, the
     * non real time thread must call SwitchConfig() and redo the
     * update on the other instance. SwitchConfig() blocks until it is
     * safe to modify the other instance.
     *
     * The real time threads need one Reader object each to access the
     * configuration data. This object must be created outside the
     * real time thread. The Lock() function returns a reference to
     * the data to be read, and Unlock() must be called when finished
     * reading the data. (Neither Lock nor Unlock will block the real
     * time thread, or use any system calls.)
     *
     * Note that the non real time side isn't safe for concurrent
     * access, so if there are multiple non real time threads that
     * update the configuration data, a mutex has to be used.
     *
     * Implementation note: the memory order parameters and fences are
     * very carefully chosen to make the code fast but still safe for
     * memory access reordering made by the CPU.
     */
    template<class T>
    class SynchronizedConfig {
        public:
            SynchronizedConfig();

            // methods for the real time thread

            class Reader {
                public:
                    /**
                     * Gets the configuration object for use by the
                     * real time thread. The object is safe to use
                     * (read only) until Unlock() is called.
                     *
                     * @returns a reference to the configuration
                     *          object to be read by the real time
                     *          thread
                     */
                    /*const*/ T& Lock() { //TODO const currently commented for the DoubleBuffer usage below
                        lock.store(lockCount += 2, memory_order_relaxed);
                        atomic_thread_fence(memory_order_seq_cst);
                        return parent->config[parent->indexAtomic.load(
                                memory_order_acquire)];
                    }

                    /**
                     * Unlock the configuration object. Unlock() must
                     * be called by the real time thread after it has
                     * finished reading the configuration object. If
                     * the non real time thread is waiting in
                     * SwitchConfig() it will be awaken when no real
                     * time threads are locked anymore.
                     */
                    void Unlock() {
                        lock.store(0, memory_order_release);
                    }

                    Reader(SynchronizedConfig& config);
                    Reader(SynchronizedConfig* config);
                    virtual ~Reader();
                private:
                    friend class SynchronizedConfig;
                    SynchronizedConfig* parent;
                    int lockCount; // increased in every Lock(),
                                   // lowest bit is always set.
                    atomic<int> lock; // equals lockCount when inside
                                      // critical region, otherwise 0
                    Reader* next; // only used locally in SwitchConfig
                    int prevLock; // only used locally in SwitchConfig
            };


            // methods for the non real time thread

            /**
             * Gets the configuration object for use by the non real
             * time thread. The object returned is not in use by the
             * real time thread, so it can safely be updated. After
             * the update is done, the non real time thread must call
             * SwitchConfig() and the same update must be done again.
             *
             * @returns a reference to the configuration object to be
             *          updated by the non real time thread
             */
            T& GetConfigForUpdate();

            /**
             * Atomically switch the newly updated configuration
             * object with the one used by the real time thread, then
             * wait for the real time thread to finish working with
             * the old object before returning the old object.
             * SwitchConfig() must be called by the non real time
             * thread after an update has been done, and the object
             * returned must be updated in the same way as the first.
             *
             * @returns a reference to the configuration object to be
             *          updated by the non real time thread
             */
            T& SwitchConfig();

        private:
            atomic<int> indexAtomic;
            int updateIndex;
            T config[2];
            std::set<Reader*> readers;
    };

    template<class T> SynchronizedConfig<T>::SynchronizedConfig() :
        indexAtomic(0) {
        updateIndex = 1;
    }

    template<class T> T& SynchronizedConfig<T>::GetConfigForUpdate() {
        return config[updateIndex];
    }

    template<class T> T& SynchronizedConfig<T>::SwitchConfig() {
        indexAtomic.store(updateIndex, memory_order_release);
        atomic_thread_fence(memory_order_seq_cst);

        // first put all locking readers in a linked list
        Reader* lockingReaders = 0;
        for (typename std::set<Reader*>::iterator iter = readers.begin() ;
             iter != readers.end() ;
             iter++) {
            (*iter)->prevLock = (*iter)->lock.load(memory_order_acquire);
            if ((*iter)->prevLock) {
                (*iter)->next = lockingReaders;
                lockingReaders = *iter;
            }
        }

        // wait until there are no locking readers left
        while (lockingReaders) {
            usleep(50000);
            Reader** prev = &lockingReaders;
            for (Reader* p = lockingReaders ; p ; p = p->next) {
                if (p->lock.load(memory_order_acquire) == p->prevLock) {
                    prev = &p->next;
                } else {
                    *prev = p->next; // unlink
                }
            }
        }

        updateIndex ^= 1;
        return config[updateIndex];
    }


    // ----- Reader ----

    template <class T>
    SynchronizedConfig<T>::Reader::Reader(SynchronizedConfig& config) :
        parent(&config), lock(0), lockCount(1) {
        parent->readers.insert(this);
    }
    
    template <class T>
    SynchronizedConfig<T>::Reader::Reader(SynchronizedConfig* config) :
        parent(config), lock(0), lockCount(1) {
        parent->readers.insert(this);
    }

    template <class T>
    SynchronizedConfig<T>::Reader::~Reader() {
        parent->readers.erase(this);
    }
    
    
    // ----- Convenience classes on top of SynchronizedConfig ----


    /**
     * Base interface class for classes that implement synchronization of data
     * shared between multiple threads.
     */
    template<class T>
    class Synchronizer {
    public:
        /**
         * Signal intention to enter a synchronized code block. Depending
         * on the actual implementation, this call may block the calling
         * thread until it is safe to actually use the protected data. After
         * this call returns, it is safe for the calling thread to access and
         * modify the shared data. As soon as the thread is done with accessing
         * the shared data, it MUST call endSync().
         *
         * @return the shared protected data
         */
        virtual T& beginSync() = 0; //TODO: or call it lock() instead ?
            
        /**
         * Signal that the synchronized code block has been left. Depending
         * on the actual implementation, this call may block the calling
         * thread for a certain amount of time.
         */
        virtual void endSync() = 0; //TODO: or call it unlock() instead ?
    };

    /**
     * Wraps as a kind of pointer class some data object shared with other
     * threads, to protect / synchronize the shared data against
     * undeterministic concurrent access. It does so by locking the shared
     * data in the Sync constructor and unlocking the shared data in the Sync
     * destructor. Accordingly it can always be considered safe to access the
     * shared data during the whole life time of the Sync object. Due to
     * this design, a Sync object MUST only be accessed and destroyed
     * by exactly one and the same thread which created that same Sync object.
     */
    template<class T>
    class Sync {
    public:
        Sync(Synchronizer<T>* syncer) {
            this->syncer = syncer;
            this->data = &syncer->beginSync();
        }
        
        virtual ~Sync() {
            syncer->endSync();
        }
        
        Sync& operator =(const Sync& arg) {
            *this->data = *arg.data;
            return *this;
        }

        Sync& operator =(const T& arg) {
            *this->data = arg;
            return *this;
        }
        
        const T& operator *() const { return *data; }
        T&       operator *()       { return *data; }

        const T* operator ->() const { return data; }
        T*       operator ->()       { return data; }

    private:
        Synchronizer<T>* syncer; ///< Points to the object that shall be responsible to protect the shared data.
        T* data; ///< Points to the shared data that should be protected.
    };

    /**
     * BackBuffer object to be accessed by multiple non-real-time threads.
     *
     * Since a back buffer is designed for being accessed by non-real-time
     * threads, its methods involved may block the calling thread for a long
     * amount of time.
     */
    template<class T>
    class BackBuffer : public SynchronizedConfig<T>, public Synchronizer<T> {
    public:
        virtual T& beginSync() OVERRIDE {
            mutex.Lock();
            data = &SynchronizedConfig<T>::GetConfigForUpdate();
            return *data;
        }

        virtual void endSync() OVERRIDE {
            const T clone = *data;
            SynchronizedConfig<T>::SwitchConfig() = clone;
            mutex.Unlock();
        }

    private:
        T* data;
        Mutex mutex;
    };

    /**
     * FrontBuffer object to be accessed by exactly ONE real-time thread.
     * A front buffer is designed for real-time access. That is, its methods
     * involved are lock free, that is none of them block the calling thread
     * for a long time.
     *
     * If you need the front buffer's data to be accessed by multiple real-time
     * threads instead, then you need to create multiple instances of the
     * FrontBuffer object. They would point to the same data, but ensure
     * protection against concurrent access among those real-time threads.
     */
    template<class T>
    class FrontBuffer : public SynchronizedConfig<T>::Reader, public Synchronizer<T> {
    public:
        FrontBuffer(BackBuffer<T>& backBuffer) : SynchronizedConfig<T>::Reader::Reader(&backBuffer) {}
        virtual T& beginSync() OVERRIDE { return SynchronizedConfig<T>::Reader::Lock(); }
        virtual void endSync() OVERRIDE { SynchronizedConfig<T>::Reader::Unlock(); }
    };

    /**
     * Synchronization / protection of data shared between multiple threads by
     * using a double buffer design. The FrontBuffer is meant to be accessed by
     * exactly one real-time thread, whereas the BackBuffer is meant to be
     * accessed by multiple non-real-time threads.
     *
     * This class is built on top of SynchronizedConfig as convenient API to
     * reduce the amount of code required to protect shared data.
     */
    template<class T>
    class DoubleBuffer {
    public:
        DoubleBuffer() : m_front(m_back) {}
        
        /**
         * Synchronized access of the shared data for EXACTLY one real-time
         * thread.
         *
         * The returned shared data is wrapped into a Sync object, which
         * ensures that the shared data is protected against concurrent access
         * during the life time of the returned Sync object.
         */
        inline
        Sync<T> front() { return Sync<T>(&m_front); }
        
        /**
         * Synchronized access of the shared data for multiple non-real-time
         * threads.
         *
         * The returned shared data is wrapped into a Sync object, which
         * ensures that the shared data is protected against concurrent access
         * during the life time of the returned Sync object.
         *
         * As soon as the returned Sync object is destroyed, the FrontBuffer
         * will automatically be exchanged by the hereby modified BackBuffer.
         */
        inline
        Sync<T> back() { return Sync<T>(&m_back); }

    private:
        BackBuffer<T> m_back; ///< Back buffer (non real-time thread(s) side).
        FrontBuffer<T> m_front; ///< Front buffer (real-time thread side).
    };

} // namespace LinuxSampler

#endif
1	persson	840	/***************************************************************************
2			* *
3	persson	2343	* Copyright (C) 2006-2012 Andreas Persson *
4	persson	840	* *
5			* This program is free software; you can redistribute it and/or modify *
6			* it under the terms of the GNU General Public License as published by *
7			* the Free Software Foundation; either version 2 of the License, or *
8			* (at your option) any later version. *
9			* *
10			* This program is distributed in the hope that it will be useful, *
11			* but WITHOUT ANY WARRANTY; without even the implied warranty of *
12			* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the *
13			* GNU General Public License for more details. *
14			* *
15			* You should have received a copy of the GNU General Public License *
16			* along with this program; if not, write to the Free Software *
17			* Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, *
18			* MA 02110-1301 USA *
19			***************************************************************************/
20
21	persson	1790	#ifndef SYNCHRONIZEDCONFIG_H
22			#define SYNCHRONIZEDCONFIG_H
23	persson	840
24	persson	846	#include <set>
25	persson	1790	#include <unistd.h>
26			#include "lsatomic.h"
27	schoenebeck	2453	#include "Mutex.h"
28	persson	840
29			namespace LinuxSampler {
30
31			/**
32	persson	2343	* Thread-safe management of configuration data, where the data is
33	persson	846	* updated by a single non real time thread and read by a number
34			* of real time threads.
35	persson	840	*
36			* The synchronization is achieved by using two instances of the
37			* configuration data. The non real time thread gets access to the
38	persson	846	* instance not currently in use by the real time threads by
39	persson	840	* calling GetConfigForUpdate(). After the data is updated, the
40			* non real time thread must call SwitchConfig() and redo the
41			* update on the other instance. SwitchConfig() blocks until it is
42			* safe to modify the other instance.
43			*
44	persson	846	* The real time threads need one Reader object each to access the
45	persson	1038	* configuration data. This object must be created outside the
46			* real time thread. The Lock() function returns a reference to
47			* the data to be read, and Unlock() must be called when finished
48	persson	846	* reading the data. (Neither Lock nor Unlock will block the real
49			* time thread, or use any system calls.)
50	persson	2343	*
51			* Note that the non real time side isn't safe for concurrent
52			* access, so if there are multiple non real time threads that
53			* update the configuration data, a mutex has to be used.
54			*
55			* Implementation note: the memory order parameters and fences are
56			* very carefully chosen to make the code fast but still safe for
57			* memory access reordering made by the CPU.
58	persson	840	*/
59			template<class T>
60			class SynchronizedConfig {
61			public:
62			SynchronizedConfig();
63
64			// methods for the real time thread
65
66	persson	846	class Reader {
67			public:
68			/**
69			* Gets the configuration object for use by the
70			* real time thread. The object is safe to use
71			* (read only) until Unlock() is called.
72			*
73			* @returns a reference to the configuration
74			* object to be read by the real time
75			* thread
76			*/
77	schoenebeck	2453	/const/ T& Lock() { //TODO const currently commented for the DoubleBuffer usage below
78	persson	1887	lock.store(lockCount += 2, memory_order_relaxed);
79	persson	1790	atomic_thread_fence(memory_order_seq_cst);
80	schoenebeck	2453	return parent->config[parent->indexAtomic.load(
81	persson	1790	memory_order_acquire)];
82	persson	846	}
83	persson	840
84	persson	846	/**
85			* Unlock the configuration object. Unlock() must
86			* be called by the real time thread after it has
87			* finished reading the configuration object. If
88			* the non real time thread is waiting in
89			* SwitchConfig() it will be awaken when no real
90			* time threads are locked anymore.
91			*/
92			void Unlock() {
93	persson	1887	lock.store(0, memory_order_release);
94	persson	846	}
95	persson	840
96	persson	846	Reader(SynchronizedConfig& config);
97	schoenebeck	2453	Reader(SynchronizedConfig* config);
98			virtual ~Reader();
99	persson	846	private:
100			friend class SynchronizedConfig;
101	schoenebeck	2453	SynchronizedConfig* parent;
102	persson	1887	int lockCount; // increased in every Lock(),
103			// lowest bit is always set.
104			atomic<int> lock; // equals lockCount when inside
105			// critical region, otherwise 0
106			Reader* next; // only used locally in SwitchConfig
107			int prevLock; // only used locally in SwitchConfig
108	persson	846	};
109	persson	840
110	persson	846
111	persson	840	// methods for the non real time thread
112
113			/**
114			* Gets the configuration object for use by the non real
115			* time thread. The object returned is not in use by the
116			* real time thread, so it can safely be updated. After
117			* the update is done, the non real time thread must call
118			* SwitchConfig() and the same update must be done again.
119			*
120			* @returns a reference to the configuration object to be
121			* updated by the non real time thread
122			*/
123			T& GetConfigForUpdate();
124
125			/**
126			* Atomically switch the newly updated configuration
127			* object with the one used by the real time thread, then
128			* wait for the real time thread to finish working with
129			* the old object before returning the old object.
130			* SwitchConfig() must be called by the non real time
131			* thread after an update has been done, and the object
132			* returned must be updated in the same way as the first.
133			*
134			* @returns a reference to the configuration object to be
135			* updated by the non real time thread
136			*/
137			T& SwitchConfig();
138
139			private:
140	persson	1790	atomic<int> indexAtomic;
141	persson	840	int updateIndex;
142			T config[2];
143	persson	846	std::set<Reader*> readers;
144	persson	840	};
145
146	persson	1790	template<class T> SynchronizedConfig<T>::SynchronizedConfig() :
147			indexAtomic(0) {
148	persson	846	updateIndex = 1;
149	persson	840	}
150
151			template<class T> T& SynchronizedConfig<T>::GetConfigForUpdate() {
152			return config[updateIndex];
153			}
154
155			template<class T> T& SynchronizedConfig<T>::SwitchConfig() {
156	persson	1790	indexAtomic.store(updateIndex, memory_order_release);
157			atomic_thread_fence(memory_order_seq_cst);
158	persson	846
159			// first put all locking readers in a linked list
160			Reader* lockingReaders = 0;
161			for (typename std::set<Reader*>::iterator iter = readers.begin() ;
162			iter != readers.end() ;
163			iter++) {
164	persson	1887	(iter)->prevLock = (iter)->lock.load(memory_order_acquire);
165			if ((*iter)->prevLock) {
166	persson	846	(*iter)->next = lockingReaders;
167			lockingReaders = *iter;
168			}
169			}
170
171			// wait until there are no locking readers left
172			while (lockingReaders) {
173	persson	840	usleep(50000);
174	persson	846	Reader** prev = &lockingReaders;
175			for (Reader* p = lockingReaders ; p ; p = p->next) {
176	persson	1887	if (p->lock.load(memory_order_acquire) == p->prevLock) {
177			prev = &p->next;
178			} else {
179			*prev = p->next; // unlink
180			}
181	persson	846	}
182			}
183
184			updateIndex ^= 1;
185			return config[updateIndex];
186	persson	840	}
187
188	persson	846
189			// ----- Reader ----
190
191			template <class T>
192	persson	1790	SynchronizedConfig<T>::Reader::Reader(SynchronizedConfig& config) :
193	schoenebeck	2453	parent(&config), lock(0), lockCount(1) {
194			parent->readers.insert(this);
195			}
196
197			template <class T>
198			SynchronizedConfig<T>::Reader::Reader(SynchronizedConfig* config) :
199	persson	1887	parent(config), lock(0), lockCount(1) {
200	schoenebeck	2453	parent->readers.insert(this);
201	persson	846	}
202
203			template <class T>
204			SynchronizedConfig<T>::Reader::~Reader() {
205	schoenebeck	2453	parent->readers.erase(this);
206	persson	846	}
207	schoenebeck	2453
208
209			// ----- Convenience classes on top of SynchronizedConfig ----
210	persson	846
211	schoenebeck	2453
212			/**
213			* Base interface class for classes that implement synchronization of data
214			* shared between multiple threads.
215			*/
216			template<class T>
217			class Synchronizer {
218			public:
219			/**
220			* Signal intention to enter a synchronized code block. Depending
221			* on the actual implementation, this call may block the calling
222			* thread until it is safe to actually use the protected data. After
223			* this call returns, it is safe for the calling thread to access and
224			* modify the shared data. As soon as the thread is done with accessing
225			* the shared data, it MUST call endSync().
226			*
227			* @return the shared protected data
228			*/
229			virtual T& beginSync() = 0; //TODO: or call it lock() instead ?
230
231			/**
232			* Signal that the synchronized code block has been left. Depending
233			* on the actual implementation, this call may block the calling
234			* thread for a certain amount of time.
235			*/
236			virtual void endSync() = 0; //TODO: or call it unlock() instead ?
237			};
238
239			/**
240			* Wraps as a kind of pointer class some data object shared with other
241			* threads, to protect / synchronize the shared data against
242			* undeterministic concurrent access. It does so by locking the shared
243			* data in the Sync constructor and unlocking the shared data in the Sync
244			* destructor. Accordingly it can always be considered safe to access the
245			* shared data during the whole life time of the Sync object. Due to
246			* this design, a Sync object MUST only be accessed and destroyed
247			* by exactly one and the same thread which created that same Sync object.
248			*/
249			template<class T>
250			class Sync {
251			public:
252			Sync(Synchronizer<T>* syncer) {
253			this->syncer = syncer;
254			this->data = &syncer->beginSync();
255			}
256
257			virtual ~Sync() {
258			syncer->endSync();
259			}
260
261			Sync& operator =(const Sync& arg) {
262			this->data = arg.data;
263			return *this;
264			}
265
266			Sync& operator =(const T& arg) {
267			*this->data = arg;
268			return *this;
269			}
270
271			const T& operator () const { return data; }
272			T& operator () { return data; }
273
274			const T* operator ->() const { return data; }
275			T* operator ->() { return data; }
276
277			private:
278			Synchronizer<T>* syncer; ///< Points to the object that shall be responsible to protect the shared data.
279			T* data; ///< Points to the shared data that should be protected.
280			};
281
282			/**
283			* BackBuffer object to be accessed by multiple non-real-time threads.
284			*
285			* Since a back buffer is designed for being accessed by non-real-time
286			* threads, its methods involved may block the calling thread for a long
287			* amount of time.
288			*/
289			template<class T>
290			class BackBuffer : public SynchronizedConfig<T>, public Synchronizer<T> {
291			public:
292			virtual T& beginSync() OVERRIDE {
293			mutex.Lock();
294			data = &SynchronizedConfig<T>::GetConfigForUpdate();
295			return *data;
296			}
297
298			virtual void endSync() OVERRIDE {
299			const T clone = *data;
300			SynchronizedConfig<T>::SwitchConfig() = clone;
301			mutex.Unlock();
302			}
303
304			private:
305			T* data;
306			Mutex mutex;
307			};
308
309			/**
310			* FrontBuffer object to be accessed by exactly ONE real-time thread.
311			* A front buffer is designed for real-time access. That is, its methods
312			* involved are lock free, that is none of them block the calling thread
313			* for a long time.
314			*
315			* If you need the front buffer's data to be accessed by multiple real-time
316			* threads instead, then you need to create multiple instances of the
317			* FrontBuffer object. They would point to the same data, but ensure
318			* protection against concurrent access among those real-time threads.
319			*/
320			template<class T>
321			class FrontBuffer : public SynchronizedConfig<T>::Reader, public Synchronizer<T> {
322			public:
323			FrontBuffer(BackBuffer<T>& backBuffer) : SynchronizedConfig<T>::Reader::Reader(&backBuffer) {}
324			virtual T& beginSync() OVERRIDE { return SynchronizedConfig<T>::Reader::Lock(); }
325			virtual void endSync() OVERRIDE { SynchronizedConfig<T>::Reader::Unlock(); }
326			};
327
328			/**
329			* Synchronization / protection of data shared between multiple threads by
330			* using a double buffer design. The FrontBuffer is meant to be accessed by
331			* exactly one real-time thread, whereas the BackBuffer is meant to be
332			* accessed by multiple non-real-time threads.
333			*
334			* This class is built on top of SynchronizedConfig as convenient API to
335			* reduce the amount of code required to protect shared data.
336			*/
337			template<class T>
338			class DoubleBuffer {
339			public:
340			DoubleBuffer() : m_front(m_back) {}
341
342			/**
343			* Synchronized access of the shared data for EXACTLY one real-time
344			* thread.
345			*
346			* The returned shared data is wrapped into a Sync object, which
347			* ensures that the shared data is protected against concurrent access
348			* during the life time of the returned Sync object.
349			*/
350			inline
351			Sync<T> front() { return Sync<T>(&m_front); }
352
353			/**
354			* Synchronized access of the shared data for multiple non-real-time
355			* threads.
356			*
357			* The returned shared data is wrapped into a Sync object, which
358			* ensures that the shared data is protected against concurrent access
359			* during the life time of the returned Sync object.
360			*
361			* As soon as the returned Sync object is destroyed, the FrontBuffer
362			* will automatically be exchanged by the hereby modified BackBuffer.
363			*/
364			inline
365			Sync<T> back() { return Sync<T>(&m_back); }
366
367			private:
368			BackBuffer<T> m_back; ///< Back buffer (non real-time thread(s) side).
369			FrontBuffer<T> m_front; ///< Front buffer (real-time thread side).
370			};
371
372	persson	840	} // namespace LinuxSampler
373
374			#endif