[svn-upgrade] new version strongswan (4.5.0)

7 files changed, 1296 insertions, 0 deletions

diff --git a/src/libstrongswan/processing/jobs/callback_job.c b/src/libstrongswan/processing/jobs/callback_job.c
new file mode 100644
index 000000000..556cbd907
--- /dev/null
+++ b/src/libstrongswan/processing/jobs/callback_job.c
@@ -0,0 +1,271 @@
+/*
+ * Copyright (C) 2009 Tobias Brunner
+ * Copyright (C) 2007 Martin Willi
+ * Hochschule fuer Technik Rapperswil
+ *
+ * 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.  See <http://www.fsf.org/copyleft/gpl.txt>.
+ *
+ * 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.
+ */
+
+#include "callback_job.h"
+
+#include <semaphore.h>
+
+#include <threading/thread.h>
+#include <threading/condvar.h>
+#include <threading/mutex.h>
+#include <utils/linked_list.h>
+
+typedef struct private_callback_job_t private_callback_job_t;
+
+/**
+ * Private data of an callback_job_t Object.
+ */
+struct private_callback_job_t {
+	/**
+	 * Public callback_job_t interface.
+	 */
+	callback_job_t public;
+
+	/**
+	 * Callback to call on execution
+	 */
+	callback_job_cb_t callback;
+
+	/**
+	 * parameter to supply to callback
+	 */
+	void *data;
+
+	/**
+	 * cleanup function for data
+	 */
+	callback_job_cleanup_t cleanup;
+
+	/**
+	 * thread of the job, if running
+	 */
+	thread_t *thread;
+
+	/**
+	 * mutex to access jobs interna
+	 */
+	mutex_t *mutex;
+
+	/**
+	 * list of asociated child jobs
+	 */
+	linked_list_t *children;
+
+	/**
+	 * parent of this job, or NULL
+	 */
+	private_callback_job_t *parent;
+
+	/**
+	 * TRUE if the job got cancelled
+	 */
+	bool cancelled;
+
+	/**
+	 * condvar to synchronize the cancellation/destruction of the job
+	 */
+	condvar_t *destroyable;
+
+	/**
+	 * semaphore to synchronize the termination of the assigned thread.
+	 *
+	 * separately allocated during cancellation, so that we can wait on it
+	 * without risking that it gets freed too early during destruction.
+	 */
+	sem_t *terminated;
+};
+
+/**
+ * unregister a child from its parent, if any.
+ * note: this->mutex has to be locked
+ */
+static void unregister(private_callback_job_t *this)
+{
+	if (this->parent)
+	{
+		this->parent->mutex->lock(this->parent->mutex);
+		if (this->parent->cancelled && !this->cancelled)
+		{
+			/* if the parent has been cancelled but we have not yet, we do not
+			 * unregister until we got cancelled by the parent. */
+			this->parent->mutex->unlock(this->parent->mutex);
+			this->destroyable->wait(this->destroyable, this->mutex);
+			this->parent->mutex->lock(this->parent->mutex);
+		}
+		this->parent->children->remove(this->parent->children, this, NULL);
+		this->parent->mutex->unlock(this->parent->mutex);
+		this->parent = NULL;
+	}
+}
+
+/**
+ * Implements job_t.destroy.
+ */
+static void destroy(private_callback_job_t *this)
+{
+	this->mutex->lock(this->mutex);
+	unregister(this);
+	if (this->cleanup)
+	{
+		this->cleanup(this->data);
+	}
+	if (this->terminated)
+	{
+		sem_post(this->terminated);
+	}
+	this->children->destroy(this->children);
+	this->destroyable->destroy(this->destroyable);
+	this->mutex->unlock(this->mutex);
+	this->mutex->destroy(this->mutex);
+	free(this);
+}
+
+/**
+ * Implementation of callback_job_t.cancel.
+ */
+static void cancel(private_callback_job_t *this)
+{
+	callback_job_t *child;
+	sem_t *terminated = NULL;
+
+	this->mutex->lock(this->mutex);
+	this->cancelled = TRUE;
+	/* terminate children */
+	while (this->children->get_first(this->children, (void**)&child) == SUCCESS)
+	{
+		this->mutex->unlock(this->mutex);
+		child->cancel(child);
+		this->mutex->lock(this->mutex);
+	}
+	if (this->thread)
+	{
+		/* terminate the thread, if there is currently one executing the job.
+		 * we wait for its termination using a semaphore */
+		this->thread->cancel(this->thread);
+		terminated = this->terminated = malloc_thing(sem_t);
+		sem_init(terminated, 0, 0);
+	}
+	else
+	{
+		/* if the job is currently queued, it gets terminated later.
+		 * we can't wait, because it might not get executed at all.
+		 * we also unregister the queued job manually from its parent (the
+		 * others get unregistered during destruction) */
+		unregister(this);
+	}
+	this->destroyable->signal(this->destroyable);
+	this->mutex->unlock(this->mutex);
+
+	if (terminated)
+	{
+		sem_wait(terminated);
+		sem_destroy(terminated);
+		free(terminated);
+	}
+}
+
+/**
+ * Implementation of job_t.execute.
+ */
+static void execute(private_callback_job_t *this)
+{
+	bool cleanup = FALSE, requeue = FALSE;
+
+	thread_cleanup_push((thread_cleanup_t)destroy, this);
+
+	this->mutex->lock(this->mutex);
+	this->thread = thread_current();
+	this->mutex->unlock(this->mutex);
+
+	while (TRUE)
+	{
+		this->mutex->lock(this->mutex);
+		if (this->cancelled)
+		{
+			this->mutex->unlock(this->mutex);
+			cleanup = TRUE;
+			break;
+		}
+		this->mutex->unlock(this->mutex);
+		switch (this->callback(this->data))
+		{
+			case JOB_REQUEUE_DIRECT:
+				continue;
+			case JOB_REQUEUE_FAIR:
+			{
+				requeue = TRUE;
+				break;
+			}
+			case JOB_REQUEUE_NONE:
+			default:
+			{
+				cleanup = TRUE;
+				break;
+			}
+		}
+		break;
+	}
+	this->mutex->lock(this->mutex);
+	this->thread = NULL;
+	this->mutex->unlock(this->mutex);
+	/* manually create a cancellation point to avoid that a cancelled thread
+	 * goes back into the thread pool */
+	thread_cancellation_point();
+	if (requeue)
+	{
+		lib->processor->queue_job(lib->processor,
+								  &this->public.job_interface);
+	}
+	thread_cleanup_pop(cleanup);
+}
+
+/*
+ * Described in header.
+ */
+callback_job_t *callback_job_create(callback_job_cb_t cb, void *data,
+									callback_job_cleanup_t cleanup,
+									callback_job_t *parent)
+{
+	private_callback_job_t *this = malloc_thing(private_callback_job_t);
+
+	/* interface functions */
+	this->public.job_interface.execute = (void (*) (job_t *)) execute;
+	this->public.job_interface.destroy = (void (*) (job_t *)) destroy;
+	this->public.cancel = (void(*)(callback_job_t*))cancel;
+
+	/* private variables */
+	this->mutex = mutex_create(MUTEX_TYPE_DEFAULT);
+	this->callback = cb;
+	this->data = data;
+	this->cleanup = cleanup;
+	this->thread = 0;
+	this->children = linked_list_create();
+	this->parent = (private_callback_job_t*)parent;
+	this->cancelled = FALSE;
+	this->destroyable = condvar_create(CONDVAR_TYPE_DEFAULT);
+	this->terminated = NULL;
+
+	/* register us at parent */
+	if (parent)
+	{
+		this->parent->mutex->lock(this->parent->mutex);
+		this->parent->children->insert_last(this->parent->children, this);
+		this->parent->mutex->unlock(this->parent->mutex);
+	}
+
+	return &this->public;
+}
+
diff --git a/src/libstrongswan/processing/jobs/callback_job.h b/src/libstrongswan/processing/jobs/callback_job.h
new file mode 100644
index 000000000..62da1edd1
--- /dev/null
+++ b/src/libstrongswan/processing/jobs/callback_job.h
@@ -0,0 +1,118 @@
+/*
+ * Copyright (C) 2007 Martin Willi
+ * Hochschule fuer Technik Rapperswil
+ *
+ * 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.  See <http://www.fsf.org/copyleft/gpl.txt>.
+ *
+ * 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.
+ */
+
+/**
+ * @defgroup callback_job callback_job
+ * @{ @ingroup jobs
+ */
+
+#ifndef CALLBACK_JOB_H_
+#define CALLBACK_JOB_H_
+
+typedef struct callback_job_t callback_job_t;
+
+#include <library.h>
+#include <processing/jobs/job.h>
+
+
+typedef enum job_requeue_t job_requeue_t;
+
+/**
+ * Job requeueing policy
+ *
+ * The job requeueing policy defines how a job is handled when the callback
+ * function returns.
+ */
+enum job_requeue_t {
+
+	/**
+	 * Do not requeue job, destroy it
+	 */
+	JOB_REQUEUE_NONE,
+
+	/**
+	 * Reque the job fairly, meaning it has to requeue as any other job
+	 */
+	JOB_REQUEUE_FAIR,
+
+	/**
+	 * Reexecute the job directly, without the need of requeueing it
+	 */
+	JOB_REQUEUE_DIRECT,
+};
+
+/**
+ * The callback function to use for the callback job.
+ *
+ * This is the function to use as callback for a callback job. It receives
+ * a parameter supplied to the callback jobs constructor.
+ *
+ * @param data			param supplied to job
+ * @return				requeing policy how to requeue the job
+ */
+typedef job_requeue_t (*callback_job_cb_t)(void *data);
+
+/**
+ * Cleanup function to use for data cleanup.
+ *
+ * The callback has an optional user argument which receives data. However,
+ * this data may be cleaned up if it is allocated. This is the function
+ * to supply to the constructor.
+ *
+ * @param data			param supplied to job
+ * @return				requeing policy how to requeue the job
+ */
+typedef void (*callback_job_cleanup_t)(void *data);
+
+/**
+ * Class representing an callback Job.
+ *
+ * This is a special job which allows a simple callback function to
+ * be executed by a thread of the thread pool. This allows simple execution
+ * of asynchronous methods, without to manage threads.
+ */
+struct callback_job_t {
+	/**
+	 * The job_t interface.
+	 */
+	job_t job_interface;
+
+	/**
+	 * Cancel the job's thread and wait for its termination. This only works
+	 * reliably for jobs that always use JOB_REQUEUE_FAIR or JOB_REQUEUE_DIRECT,
+	 * otherwise the job may already be destroyed when cancel is called. */
+	void (*cancel)(callback_job_t *this);
+};
+
+/**
+ * Creates a callback job.
+ *
+ * The cleanup function is called when the job gets destroyed to destroy
+ * the associated data.
+ * If parent is not NULL, the specified job gets an association. Whenever
+ * the parent gets cancelled (or runs out), all of its children are cancelled,
+ * too.
+ *
+ * @param cb				callback to call from the processor
+ * @param data				user data to supply to callback
+ * @param cleanup			destructor for data on destruction, or NULL
+ * @param parent			parent of this job
+ * @return					callback_job_t object
+ */
+callback_job_t *callback_job_create(callback_job_cb_t cb, void *data,
+									callback_job_cleanup_t cleanup,
+									callback_job_t *parent);
+
+#endif /** CALLBACK_JOB_H_ @}*/
diff --git a/src/libstrongswan/processing/jobs/job.h b/src/libstrongswan/processing/jobs/job.h
new file mode 100644
index 000000000..0f1c16ebe
--- /dev/null
+++ b/src/libstrongswan/processing/jobs/job.h
@@ -0,0 +1,52 @@
+/*
+ * Copyright (C) 2005-2006 Martin Willi
+ * Copyright (C) 2005 Jan Hutter
+ * Hochschule fuer Technik Rapperswil
+ *
+ * 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.  See <http://www.fsf.org/copyleft/gpl.txt>.
+ *
+ * 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.
+ */
+
+/**
+ * @defgroup job job
+ * @{ @ingroup jobs
+ */
+
+#ifndef JOB_H_
+#define JOB_H_
+
+typedef struct job_t job_t;
+
+#include <library.h>
+
+/**
+ * Job-Interface as it is stored in the job queue.
+ */
+struct job_t {
+
+	/**
+	 * Execute a job.
+	 *
+	 * The processing facility executes a job using this method. Jobs are
+	 * one-shot, they destroy themself after execution, so don't use a job
+	 * once it has been executed.
+	 */
+	void (*execute) (job_t *this);
+
+	/**
+	 * Destroy a job.
+	 *
+	 * Is only called whenever a job was not executed (e.g. due daemon shutdown).
+	 * After execution, jobs destroy themself.
+	 */
+	void (*destroy) (job_t *job);
+};
+
+#endif /** JOB_H_ @}*/
diff --git a/src/libstrongswan/processing/processor.c b/src/libstrongswan/processing/processor.c
new file mode 100644
index 000000000..2a44f61e8
--- /dev/null
+++ b/src/libstrongswan/processing/processor.c
@@ -0,0 +1,273 @@
+/*
+ * Copyright (C) 2005-2007 Martin Willi
+ * Copyright (C) 2005 Jan Hutter
+ * Hochschule fuer Technik Rapperswil
+ *
+ * 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.  See <http://www.fsf.org/copyleft/gpl.txt>.
+ *
+ * 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.
+ */
+
+#include <stdlib.h>
+#include <string.h>
+#include <errno.h>
+
+#include "processor.h"
+
+#include <debug.h>
+#include <threading/thread.h>
+#include <threading/condvar.h>
+#include <threading/mutex.h>
+#include <utils/linked_list.h>
+
+
+typedef struct private_processor_t private_processor_t;
+
+/**
+ * Private data of processor_t class.
+ */
+struct private_processor_t {
+	/**
+	 * Public processor_t interface.
+	 */
+	processor_t public;
+
+	/**
+	 * Number of running threads
+	 */
+	u_int total_threads;
+
+	/**
+	 * Desired number of threads
+	 */
+	u_int desired_threads;
+
+	/**
+	 * Number of threads waiting for work
+	 */
+	u_int idle_threads;
+
+	/**
+	 * All threads managed in the pool (including threads that have been
+	 * cancelled, this allows to join them during destruction)
+	 */
+	linked_list_t *threads;
+
+	/**
+	 * The jobs are stored in a linked list
+	 */
+	linked_list_t *list;
+
+	/**
+	 * access to linked_list is locked through this mutex
+	 */
+	mutex_t *mutex;
+
+	/**
+	 * Condvar to wait for new jobs
+	 */
+	condvar_t *job_added;
+
+	/**
+	 * Condvar to wait for terminated threads
+	 */
+	condvar_t *thread_terminated;
+};
+
+static void process_jobs(private_processor_t *this);
+
+/**
+ * restart a terminated thread
+ */
+static void restart(private_processor_t *this)
+{
+	thread_t *thread;
+
+	DBG2(DBG_JOB, "terminated worker thread, ID: %u", thread_current_id());
+
+	/* respawn thread if required */
+	this->mutex->lock(this->mutex);
+	if (this->desired_threads < this->total_threads ||
+		(thread = thread_create((thread_main_t)process_jobs, this)) == NULL)
+	{
+		this->total_threads--;
+		this->thread_terminated->signal(this->thread_terminated);
+	}
+	else
+	{
+		this->threads->insert_last(this->threads, thread);
+	}
+	this->mutex->unlock(this->mutex);
+}
+
+/**
+ * Process queued jobs, called by the worker threads
+ */
+static void process_jobs(private_processor_t *this)
+{
+	/* worker threads are not cancellable by default */
+	thread_cancelability(FALSE);
+
+	DBG2(DBG_JOB, "started worker thread, ID: %u", thread_current_id());
+
+	this->mutex->lock(this->mutex);
+	while (this->desired_threads >= this->total_threads)
+	{
+		job_t *job;
+
+		if (this->list->get_count(this->list) == 0)
+		{
+			this->idle_threads++;
+			this->job_added->wait(this->job_added, this->mutex);
+			this->idle_threads--;
+			continue;
+		}
+		this->list->remove_first(this->list, (void**)&job);
+		this->mutex->unlock(this->mutex);
+		/* terminated threads are restarted, so we have a constant pool */
+		thread_cleanup_push((thread_cleanup_t)restart, this);
+		job->execute(job);
+		thread_cleanup_pop(FALSE);
+		this->mutex->lock(this->mutex);
+	}
+	this->mutex->unlock(this->mutex);
+	restart(this);
+}
+
+/**
+ * Implementation of processor_t.get_total_threads.
+ */
+static u_int get_total_threads(private_processor_t *this)
+{
+	u_int count;
+	this->mutex->lock(this->mutex);
+	count = this->total_threads;
+	this->mutex->unlock(this->mutex);
+	return count;
+}
+
+/**
+ * Implementation of processor_t.get_idle_threads.
+ */
+static u_int get_idle_threads(private_processor_t *this)
+{
+	u_int count;
+	this->mutex->lock(this->mutex);
+	count = this->idle_threads;
+	this->mutex->unlock(this->mutex);
+	return count;
+}
+
+/**
+ * implements processor_t.get_job_load
+ */
+static u_int get_job_load(private_processor_t *this)
+{
+	u_int load;
+	this->mutex->lock(this->mutex);
+	load = this->list->get_count(this->list);
+	this->mutex->unlock(this->mutex);
+	return load;
+}
+
+/**
+ * implements function processor_t.queue_job
+ */
+static void queue_job(private_processor_t *this, job_t *job)
+{
+	this->mutex->lock(this->mutex);
+	this->list->insert_last(this->list, job);
+	this->job_added->signal(this->job_added);
+	this->mutex->unlock(this->mutex);
+}
+
+/**
+ * Implementation of processor_t.set_threads.
+ */
+static void set_threads(private_processor_t *this, u_int count)
+{
+	this->mutex->lock(this->mutex);
+	if (count > this->total_threads)
+	{	/* increase thread count */
+		int i;
+		thread_t *current;
+
+		this->desired_threads = count;
+		DBG1(DBG_JOB, "spawning %d worker threads", count - this->total_threads);
+		for (i = this->total_threads; i < count; i++)
+		{
+			current = thread_create((thread_main_t)process_jobs, this);
+			if (current)
+			{
+				this->threads->insert_last(this->threads, current);
+				this->total_threads++;
+			}
+		}
+	}
+	else if (count < this->total_threads)
+	{	/* decrease thread count */
+		this->desired_threads = count;
+	}
+	this->job_added->broadcast(this->job_added);
+	this->mutex->unlock(this->mutex);
+}
+
+/**
+ * Implementation of processor_t.destroy.
+ */
+static void destroy(private_processor_t *this)
+{
+	thread_t *current;
+	set_threads(this, 0);
+	this->mutex->lock(this->mutex);
+	while (this->total_threads > 0)
+	{
+		this->job_added->broadcast(this->job_added);
+		this->thread_terminated->wait(this->thread_terminated, this->mutex);
+	}
+	while (this->threads->remove_first(this->threads,
+									   (void**)&current) == SUCCESS)
+	{
+		current->join(current);
+	}
+	this->mutex->unlock(this->mutex);
+	this->thread_terminated->destroy(this->thread_terminated);
+	this->job_added->destroy(this->job_added);
+	this->mutex->destroy(this->mutex);
+	this->list->destroy_offset(this->list, offsetof(job_t, destroy));
+	this->threads->destroy(this->threads);
+	free(this);
+}
+
+/*
+ * Described in header.
+ */
+processor_t *processor_create(size_t pool_size)
+{
+	private_processor_t *this = malloc_thing(private_processor_t);
+
+	this->public.get_total_threads = (u_int(*)(processor_t*))get_total_threads;
+	this->public.get_idle_threads = (u_int(*)(processor_t*))get_idle_threads;
+	this->public.get_job_load = (u_int(*)(processor_t*))get_job_load;
+	this->public.queue_job = (void(*)(processor_t*, job_t*))queue_job;
+	this->public.set_threads = (void(*)(processor_t*, u_int))set_threads;
+	this->public.destroy = (void(*)(processor_t*))destroy;
+
+	this->list = linked_list_create();
+	this->threads = linked_list_create();
+	this->mutex = mutex_create(MUTEX_TYPE_DEFAULT);
+	this->job_added = condvar_create(CONDVAR_TYPE_DEFAULT);
+	this->thread_terminated = condvar_create(CONDVAR_TYPE_DEFAULT);
+	this->total_threads = 0;
+	this->desired_threads = 0;
+	this->idle_threads = 0;
+
+	return &this->public;
+}
+
diff --git a/src/libstrongswan/processing/processor.h b/src/libstrongswan/processing/processor.h
new file mode 100644
index 000000000..bebbe3a15
--- /dev/null
+++ b/src/libstrongswan/processing/processor.h
@@ -0,0 +1,94 @@
+/*
+ * Copyright (C) 2005-2007 Martin Willi
+ * Copyright (C) 2005 Jan Hutter
+ * Hochschule fuer Technik Rapperswil
+ *
+ * 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.  See <http://www.fsf.org/copyleft/gpl.txt>.
+ *
+ * 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.
+ */
+
+/**
+ * @defgroup processor processor
+ * @{ @ingroup processing
+ */
+
+#ifndef PROCESSOR_H_
+#define PROCESSOR_H_
+
+typedef struct processor_t processor_t;
+
+#include <stdlib.h>
+
+#include <library.h>
+#include <processing/jobs/job.h>
+
+/**
+ * The processor uses threads to process queued jobs.
+ */
+struct processor_t {
+
+	/**
+	 * Get the total number of threads used by the processor.
+	 *
+	 * @return				size of thread pool
+	 */
+	u_int (*get_total_threads) (processor_t *this);
+
+	/**
+	 * Get the number of threads currently waiting.
+	 *
+	 * @return				number of idle threads
+	 */
+	u_int (*get_idle_threads) (processor_t *this);
+
+	/**
+	 * Get the number of queued jobs.
+	 *
+	 * @return				number of items in queue
+	 */
+	u_int (*get_job_load) (processor_t *this);
+
+	/**
+	 * Adds a job to the queue.
+	 *
+	 * This function is non blocking and adds a job_t to the queue.
+	 *
+	 * @param job			job to add to the queue
+	 */
+	void (*queue_job) (processor_t *this, job_t *job);
+
+	/**
+	 * Set the number of threads to use in the processor.
+	 *
+	 * If the number of threads is smaller than number of currently running
+	 * threads, thread count is decreased. Use 0 to disable the processor.
+	 * This call blocks if it decreases thread count until threads have
+	 * terminated, so make sure there are not too many blocking jobs.
+	 *
+	 * @param count			number of threads to allocate
+	 */
+	void (*set_threads)(processor_t *this, u_int count);
+
+	/**
+	 * Destroy a processor object.
+	 */
+	void (*destroy) (processor_t *processor);
+};
+
+/**
+ * Create the thread pool without any threads.
+ *
+ * Use the set_threads method to start processing jobs.
+ *
+ * @return					processor_t object
+ */
+processor_t *processor_create();
+
+#endif /** PROCESSOR_H_ @}*/
diff --git a/src/libstrongswan/processing/scheduler.c b/src/libstrongswan/processing/scheduler.c
new file mode 100644
index 000000000..e23f04598
--- /dev/null
+++ b/src/libstrongswan/processing/scheduler.c
@@ -0,0 +1,358 @@
+/*
+ * Copyright (C) 2008 Tobias Brunner
+ * Copyright (C) 2005-2006 Martin Willi
+ * Copyright (C) 2005 Jan Hutter
+ * Hochschule fuer Technik Rapperswil
+ *
+ * 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.  See <http://www.fsf.org/copyleft/gpl.txt>.
+ *
+ * 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.
+ */
+
+#include <stdlib.h>
+
+#include "scheduler.h"
+
+#include <debug.h>
+#include <processing/processor.h>
+#include <processing/jobs/callback_job.h>
+#include <threading/thread.h>
+#include <threading/condvar.h>
+#include <threading/mutex.h>
+
+/* the initial size of the heap */
+#define HEAP_SIZE_DEFAULT 64
+
+typedef struct event_t event_t;
+
+/**
+ * Event containing a job and a schedule time
+ */
+struct event_t {
+	/**
+	 * Time to fire the event.
+	 */
+	timeval_t time;
+
+	/**
+	 * Every event has its assigned job.
+	 */
+	job_t *job;
+};
+
+/**
+ * destroy an event and its job
+ */
+static void event_destroy(event_t *event)
+{
+	event->job->destroy(event->job);
+	free(event);
+}
+
+typedef struct private_scheduler_t private_scheduler_t;
+
+/**
+ * Private data of a scheduler_t object.
+ */
+struct private_scheduler_t {
+
+	/**
+	 * Public part of a scheduler_t object.
+	 */
+	 scheduler_t public;
+
+	/**
+	 * Job which queues scheduled jobs to the processor.
+	 */
+	callback_job_t *job;
+
+	/**
+	 * The heap in which the events are stored.
+	 */
+	event_t **heap;
+
+	/**
+	 * The size of the heap.
+	 */
+	u_int heap_size;
+
+	/**
+	 * The number of scheduled events.
+	 */
+	u_int event_count;
+
+	/**
+	 * Exclusive access to list
+	 */
+	mutex_t *mutex;
+
+	/**
+	 * Condvar to wait for next job.
+	 */
+	condvar_t *condvar;
+};
+
+/**
+ * Comparse two timevals, return >0 if a > b, <0 if a < b and =0 if equal
+ */
+static int timeval_cmp(timeval_t *a, timeval_t *b)
+{
+	if (a->tv_sec > b->tv_sec)
+	{
+		return 1;
+	}
+	if (a->tv_sec < b->tv_sec)
+	{
+		return -1;
+	}
+	if (a->tv_usec > b->tv_usec)
+	{
+		return 1;
+	}
+	if (a->tv_usec < b->tv_usec)
+	{
+		return -1;
+	}
+	return 0;
+}
+
+/**
+ * Returns the top event without removing it. Returns NULL if the heap is empty.
+ */
+static event_t *peek_event(private_scheduler_t *this)
+{
+	return this->event_count > 0 ? this->heap[1] : NULL;
+}
+
+/**
+ * Removes the top event from the heap and returns it. Returns NULL if the heap
+ * is empty.
+ */
+static event_t *remove_event(private_scheduler_t *this)
+{
+	event_t *event, *top;
+	if (!this->event_count)
+	{
+		return NULL;
+	}
+
+	/* store the value to return */
+	event = this->heap[1];
+	/* move the bottom event to the top */
+	top = this->heap[1] = this->heap[this->event_count];
+
+	if (--this->event_count > 1)
+	{
+		/* seep down the top event */
+		u_int position = 1;
+		while ((position << 1) <= this->event_count)
+		{
+			u_int child = position << 1;
+
+			if ((child + 1) <= this->event_count &&
+				timeval_cmp(&this->heap[child + 1]->time,
+							&this->heap[child]->time) < 0)
+			{
+				/* the "right" child is smaller */
+				child++;
+			}
+
+			if (timeval_cmp(&top->time, &this->heap[child]->time) <= 0)
+			{
+				/* the top event fires before the smaller of the two children,
+				 * stop */
+				break;
+			}
+
+			/* swap with the smaller child */
+			this->heap[position] = this->heap[child];
+			position = child;
+		}
+		this->heap[position] = top;
+	}
+	return event;
+}
+
+/**
+ * Get events from the queue and pass it to the processor
+ */
+static job_requeue_t schedule(private_scheduler_t * this)
+{
+	timeval_t now;
+	event_t *event;
+	bool timed = FALSE, oldstate;
+
+	this->mutex->lock(this->mutex);
+
+	time_monotonic(&now);
+
+	if ((event = peek_event(this)) != NULL)
+	{
+		if (timeval_cmp(&now, &event->time) >= 0)
+		{
+			remove_event(this);
+			this->mutex->unlock(this->mutex);
+			DBG2(DBG_JOB, "got event, queuing job for execution");
+			lib->processor->queue_job(lib->processor, event->job);
+			free(event);
+			return JOB_REQUEUE_DIRECT;
+		}
+		timersub(&event->time, &now, &now);
+		if (now.tv_sec)
+		{
+			DBG2(DBG_JOB, "next event in %ds %dms, waiting",
+				 now.tv_sec, now.tv_usec/1000);
+		}
+		else
+		{
+			DBG2(DBG_JOB, "next event in %dms, waiting", now.tv_usec/1000);
+		}
+		timed = TRUE;
+	}
+	thread_cleanup_push((thread_cleanup_t)this->mutex->unlock, this->mutex);
+	oldstate = thread_cancelability(TRUE);
+
+	if (timed)
+	{
+		this->condvar->timed_wait_abs(this->condvar, this->mutex, event->time);
+	}
+	else
+	{
+		DBG2(DBG_JOB, "no events, waiting");
+		this->condvar->wait(this->condvar, this->mutex);
+	}
+	thread_cancelability(oldstate);
+	thread_cleanup_pop(TRUE);
+	return JOB_REQUEUE_DIRECT;
+}
+
+/**
+ * Implements scheduler_t.get_job_load
+ */
+static u_int get_job_load(private_scheduler_t *this)
+{
+	int count;
+	this->mutex->lock(this->mutex);
+	count = this->event_count;
+	this->mutex->unlock(this->mutex);
+	return count;
+}
+
+/**
+ * Implements scheduler_t.schedule_job_tv.
+ */
+static void schedule_job_tv(private_scheduler_t *this, job_t *job, timeval_t tv)
+{
+	event_t *event;
+	u_int position;
+
+	event = malloc_thing(event_t);
+	event->job = job;
+	event->time = tv;
+
+	this->mutex->lock(this->mutex);
+
+	this->event_count++;
+	if (this->event_count > this->heap_size)
+	{
+		/* double the size of the heap */
+		this->heap_size <<= 1;
+		this->heap = (event_t**)realloc(this->heap,
+									(this->heap_size + 1) * sizeof(event_t*));
+	}
+	/* "put" the event to the bottom */
+	position = this->event_count;
+
+	/* then bubble it up */
+	while (position > 1 && timeval_cmp(&this->heap[position >> 1]->time,
+									   &event->time) > 0)
+	{
+		/* parent has to be fired after the new event, move up */
+		this->heap[position] = this->heap[position >> 1];
+		position >>= 1;
+	}
+	this->heap[position] = event;
+
+	this->condvar->signal(this->condvar);
+	this->mutex->unlock(this->mutex);
+}
+
+/**
+ * Implements scheduler_t.schedule_job.
+ */
+static void schedule_job(private_scheduler_t *this, job_t *job, u_int32_t s)
+{
+	timeval_t tv;
+
+	time_monotonic(&tv);
+	tv.tv_sec += s;
+
+	schedule_job_tv(this, job, tv);
+}
+
+/**
+ * Implements scheduler_t.schedule_job_ms.
+ */
+static void schedule_job_ms(private_scheduler_t *this, job_t *job, u_int32_t ms)
+{
+	timeval_t tv, add;
+
+	time_monotonic(&tv);
+	add.tv_sec = ms / 1000;
+	add.tv_usec = (ms % 1000) * 1000;
+
+	timeradd(&tv, &add, &tv);
+
+	schedule_job_tv(this, job, tv);
+}
+
+/**
+ * Implementation of scheduler_t.destroy.
+ */
+static void destroy(private_scheduler_t *this)
+{
+	event_t *event;
+	this->job->cancel(this->job);
+	this->condvar->destroy(this->condvar);
+	this->mutex->destroy(this->mutex);
+	while ((event = remove_event(this)) != NULL)
+	{
+		event_destroy(event);
+	}
+	free(this->heap);
+	free(this);
+}
+
+/*
+ * Described in header.
+ */
+scheduler_t * scheduler_create()
+{
+	private_scheduler_t *this = malloc_thing(private_scheduler_t);
+
+	this->public.get_job_load = (u_int (*) (scheduler_t *this)) get_job_load;
+	this->public.schedule_job = (void (*) (scheduler_t *this, job_t *job, u_int32_t s)) schedule_job;
+	this->public.schedule_job_ms = (void (*) (scheduler_t *this, job_t *job, u_int32_t ms)) schedule_job_ms;
+	this->public.schedule_job_tv = (void (*) (scheduler_t *this, job_t *job, timeval_t tv)) schedule_job_tv;
+	this->public.destroy = (void(*)(scheduler_t*)) destroy;
+
+	/* Note: the root of the heap is at index 1 */
+	this->event_count = 0;
+	this->heap_size = HEAP_SIZE_DEFAULT;
+	this->heap = (event_t**)calloc(this->heap_size + 1, sizeof(event_t*));
+
+	this->mutex = mutex_create(MUTEX_TYPE_DEFAULT);
+	this->condvar = condvar_create(CONDVAR_TYPE_DEFAULT);
+
+	this->job = callback_job_create((callback_job_cb_t)schedule, this, NULL, NULL);
+	lib->processor->queue_job(lib->processor, (job_t*)this->job);
+
+	return &this->public;
+}
+
diff --git a/src/libstrongswan/processing/scheduler.h b/src/libstrongswan/processing/scheduler.h
new file mode 100644
index 000000000..f2c72550f
--- /dev/null
+++ b/src/libstrongswan/processing/scheduler.h
@@ -0,0 +1,130 @@
+/*
+ * Copyright (C) 2009 Tobias Brunner
+ * Copyright (C) 2005-2007 Martin Willi
+ * Copyright (C) 2005 Jan Hutter
+ * Hochschule fuer Technik Rapperswil
+ *
+ * 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.  See <http://www.fsf.org/copyleft/gpl.txt>.
+ *
+ * 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.
+ */
+
+/**
+ * @defgroup scheduler scheduler
+ * @{ @ingroup processing
+ */
+
+#ifndef SCHEDULER_H_
+#define SCHEDULER_H_
+
+typedef struct scheduler_t scheduler_t;
+
+#include <library.h>
+#include <processing/jobs/job.h>
+
+/**
+ * The scheduler queues timed events which are then passed to the processor.
+ *
+ * The scheduler is implemented as a heap. A heap is a special kind of tree-
+ * based data structure that satisfies the following property: if B is a child
+ * node of A, then key(A) >= (or <=) key(B). So either the element with the
+ * greatest (max-heap) or the smallest (min-heap) key is the root of the heap.
+ * We use a min-heap whith the key being the absolute unix time at which an
+ * event is scheduled. So the root is always the event that will fire next.
+ *
+ * An earlier implementation of the scheduler used a sorted linked list to store
+ * the events. That had the advantage that removing the next event was extremely
+ * fast, also, adding an event scheduled before or after all other events was
+ * equally fast (all in O(1)). The problem was, though, that adding an event
+ * in-between got slower, as the number of events grew larger (O(n)).
+ * For each connection there could be several events: IKE-rekey, NAT-keepalive,
+ * retransmissions, expire (half-open), and others. So a gateway that probably
+ * has to handle thousands of concurrent connnections has to be able to queue a
+ * large number of events as fast as possible. Locking makes this even worse, to
+ * provide thread-safety, no events can be processed, while an event is queued,
+ * so making the insertion fast is even more important.
+ *
+ * That's the advantage of the heap. Adding an element to the heap can be
+ * achieved in O(log n) - on the other hand, removing the root node also
+ * requires O(log n) operations. Consider 10000 queued events. Inserting a new
+ * event in the list implementation required up to 10000 comparisons. In the
+ * heap implementation, the worst case is about 13.3 comparisons. That's a
+ * drastic improvement.
+ *
+ * The implementation itself uses a binary tree mapped to a one-based array to
+ * store the elements. This reduces storage overhead and simplifies navigation:
+ * the children of the node at position n are at position 2n and 2n+1 (likewise
+ * the parent node of the node at position n is at position [n/2]). Thus,
+ * navigating up and down the tree is reduced to simple index computations.
+ *
+ * Adding an element to the heap works as follows: The heap is always filled
+ * from left to right, until a row is full, then the next row is filled. Mapped
+ * to an array this gets as simple as putting the new element to the first free
+ * position. In a one-based array that position equals the number of elements
+ * currently stored in the heap. Then the heap property has to be restored, i.e.
+ * the new element has to be "bubbled up" the tree until the parent node's key
+ * is smaller or the element got the new root of the tree.
+ *
+ * Removing the next event from the heap works similarly. The event itself is
+ * the root node and stored at position 1 of the array. After removing it, the
+ * root has to be replaced and the heap property has to be restored. This is
+ * done by moving the bottom element (last row, rightmost element) to the root
+ * and then "seep it down" by swapping it with child nodes until none of the
+ * children has a smaller key or it is again a leaf node.
+ */
+struct scheduler_t {
+
+	/**
+	 * Adds a event to the queue, using a relative time offset in s.
+	 *
+	 * @param job			job to schedule
+	 * @param time			relative time to schedule job, in s
+	 */
+	void (*schedule_job) (scheduler_t *this, job_t *job, u_int32_t s);
+
+	/**
+	 * Adds a event to the queue, using a relative time offset in ms.
+	 *
+	 * @param job			job to schedule
+	 * @param time			relative time to schedule job, in ms
+	 */
+	void (*schedule_job_ms) (scheduler_t *this, job_t *job, u_int32_t ms);
+
+	/**
+	 * Adds a event to the queue, using an absolut time.
+	 *
+	 * The passed timeval should be calculated based on the time_monotonic()
+	 * function.
+	 *
+	 * @param job			job to schedule
+	 * @param time			absolut time to schedule job
+	 */
+	void (*schedule_job_tv) (scheduler_t *this, job_t *job, timeval_t tv);
+
+	/**
+	 * Returns number of jobs scheduled.
+	 *
+	 * @return				number of scheduled jobs
+	 */
+	u_int (*get_job_load) (scheduler_t *this);
+
+	/**
+	 * Destroys a scheduler object.
+	 */
+	void (*destroy) (scheduler_t *this);
+};
+
+/**
+ * Create a scheduler.
+ *
+ * @return		scheduler_t object
+ */
+scheduler_t *scheduler_create(void);
+
+#endif /** SCHEDULER_H_ @}*/