/*
 * Copyright 2018 Google, Inc.
 *
 * 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 copyright holders 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 THE COPYRIGHT
 * OWNER OR CONTRIBUTORS 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.
 */

#ifndef __SYSTEMC_CORE_SCHEDULER_HH__
#define __SYSTEMC_CORE_SCHEDULER_HH__

#include <atomic>
#include <functional>
#include <list>
#include <map>
#include <mutex>
#include <set>
#include <vector>

#include "base/logging.hh"
#include "sim/eventq.hh"
#include "systemc/core/channel.hh"
#include "systemc/core/list.hh"
#include "systemc/core/process.hh"
#include "systemc/core/sched_event.hh"

class Fiber;

namespace sc_gem5
{

class TraceFile;

typedef NodeList<Process> ProcessList;
typedef NodeList<Channel> ChannelList;

/*
 * The scheduler supports three different mechanisms, the initialization phase,
 * delta cycles, and timed notifications.
 *
 * INITIALIZATION PHASE
 *
 * The initialization phase has three parts:
 * 1. Run requested channel updates.
 * 2. Make processes which need to initialize runnable (methods and threads
 *    which didn't have dont_initialize called on them).
 * 3. Process delta notifications.
 *
 * First, the Kernel SimObject calls the update() method during its startup()
 * callback which handles the requested channel updates. The Kernel also
 * schedules an event to be run at time 0 with a slightly elevated priority
 * so that it happens before any "normal" event.
 *
 * When that t0 event happens, it calls the schedulers prepareForInit method
 * which performs step 2 above. That indirectly causes the scheduler's
 * readyEvent to be scheduled with slightly lowered priority, ensuring it
 * happens after any "normal" event.
 *
 * Because delta notifications are scheduled at the standard priority, all
 * of those events will happen next, performing step 3 above. Once they finish,
 * if the readyEvent was scheduled above, there shouldn't be any higher
 * priority events in front of it. When it runs, it will start the first
 * evaluate phase of the first delta cycle.
 *
 * DELTA CYCLE
 *
 * A delta cycle has three phases within it.
 * 1. The evaluate phase where runnable processes are allowed to run.
 * 2. The update phase where requested channel updates hapen.
 * 3. The delta notification phase where delta notifications happen.
 *
 * The readyEvent runs all three steps of the delta cycle. It first goes
 * through the list of runnable processes and executes them until the set is
 * empty, and then immediately runs the update phase. Since these are all part
 * of the same event, there's no chance for other events to intervene and
 * break the required order above.
 *
 * During the update phase above, the spec forbids any action which would make
 * a process runnable. That means that once the update phase finishes, the set
 * of runnable processes will be empty. There may, however, have been some
 * delta notifications/timeouts which will have been scheduled during either
 * the evaluate or update phase above. Those will have been accumulated in the
 * scheduler, and are now all executed.
 *
 * If any processes became runnable during the delta notification phase, the
 * readyEvent will have been scheduled and will be waiting and ready to run
 * again, effectively starting the next delta cycle.
 *
 * TIMED NOTIFICATION PHASE
 *
 * If no processes became runnable, the event queue will continue to process
 * events until it comes across an event which represents all the timed
 * notifications which are supposed to happen at a particular time. The object
 * which tracks them will execute all those notifications, and then destroy
 * itself. If the readyEvent is now ready to run, the next delta cycle will
 * start.
 *
 * PAUSE/STOP
 *
 * To inject a pause from sc_pause which should happen after the current delta
 * cycle's delta notification phase, an event is scheduled with a lower than
 * normal priority, but higher than the readyEvent. That ensures that any
 * delta notifications which are scheduled with normal priority will happen
 * first, since those are part of the current delta cycle. Then the pause
 * event will happen before the next readyEvent which would start the next
 * delta cycle. All of these events are scheduled for the current time, and so
 * would happen before any timed notifications went off.
 *
 * To inject a stop from sc_stop, the delta cycles should stop before even the
 * delta notifications have happened, but after the evaluate and update phases.
 * For that, a stop event with slightly higher than normal priority will be
 * scheduled so that it happens before any of the delta notification events
 * which are at normal priority.
 *
 * MAX RUN TIME
 *
 * When sc_start is called, it's possible to pass in a maximum time the
 * simulation should run to, at which point sc_pause is implicitly called. The
 * simulation is supposed to run up to the latest timed notification phase
 * which is less than or equal to the maximum time. In other words it should
 * run timed notifications at the maximum time, but not the subsequent evaluate
 * phase. That's implemented by scheduling an event at the max time with a
 * priority which is lower than all the others except the ready event. Timed
 * notifications will happen before it fires, but it will override any ready
 * event and prevent the evaluate phase from starting.
 */

class Scheduler
{
  public:
    typedef std::list<ScEvent *> ScEvents;

    class TimeSlot : public gem5::Event
    {
      public:
        TimeSlot(Scheduler* scheduler) : gem5::Event(Default_Pri, AutoDelete),
                                         parent_scheduler(scheduler) {}
        // Event::when() is only set after it's scheduled to an event queue.
        // However, TimeSlot won't be scheduled before init is done. We need
        // to keep the real 'targeted_when' information before scheduled.
        gem5::Tick targeted_when;
        Scheduler* parent_scheduler;
        ScEvents events;
        void process() override;

      protected:
        void
        releaseImpl() override
        {
            if (!scheduled())
                parent_scheduler->releaseTimeSlot(this);
        }

    };

    typedef std::list<TimeSlot *> TimeSlots;

    Scheduler();
    ~Scheduler();

    void clear();

    const std::string name() const { return "systemc_scheduler"; }

    uint64_t numCycles() { return _numCycles; }
    Process *current() { return _current; }

    void initPhase();

    // Register a process with the scheduler.
    void reg(Process *p);

    // Run the next process, if there is one.
    void yield();

    // Put a process on the ready list.
    void ready(Process *p);

    // Mark a process as ready if init is finished, or put it on the list of
    // processes to be initialized.
    void resume(Process *p);

    // Remove a process from the ready/init list if it was on one of them, and
    // return if it was.
    bool suspend(Process *p);

    // Schedule an update for a given channel.
    void requestUpdate(Channel *c);
    // Same as above, but may be called from a different thread.
    void asyncRequestUpdate(Channel *c);

    // Run the given process immediately, preempting whatever may be running.
    void
    runNow(Process *p)
    {
        // This function may put a process on the wrong list, ie a thread
        // the method list. That's fine since that's just a performance
        // optimization, and the important thing here is how the processes are
        // ordered.

        // If a process is running, schedule it/us to run again.
        if (_current)
            readyListMethods.pushFirst(_current);
        // Schedule p to run first.
        readyListMethods.pushFirst(p);
        yield();
    }

    // Run this process at the next opportunity.
    void
    runNext(Process *p)
    {
        // Like above, it's ok if this isn't a method. Putting it on this list
        // just gives it priority.
        readyListMethods.pushFirst(p);
        if (!inEvaluate())
            scheduleReadyEvent();
    }

    // Set an event queue for scheduling events.
    void setEventQueue(gem5::EventQueue *_eq) { eq = _eq; }

    // Get the current time according to gem5.
    gem5::Tick getCurTick() { return eq ? eq->getCurTick() : 0; }

    gem5::Tick
    delayed(const ::sc_core::sc_time &delay)
    {
        return getCurTick() + delay.value();
    }

    // For scheduling delayed/timed notifications/timeouts.
    void
    schedule(ScEvent *event, const ::sc_core::sc_time &delay)
    {
        gem5::Tick tick = delayed(delay);
        if (tick < getCurTick())
            tick = getCurTick();

        // Delta notification/timeout.
        if (delay.value() == 0) {
            event->schedule(deltas, tick);
            if (!inEvaluate() && !inUpdate())
                scheduleReadyEvent();
            return;
        }

        // Timed notification/timeout.
        auto it = timeSlots.begin();
        while (it != timeSlots.end() && (*it)->targeted_when < tick)
            it++;
        if (it == timeSlots.end() || (*it)->targeted_when != tick) {
            it = timeSlots.emplace(it, acquireTimeSlot(tick));
            schedule(*it, tick);
        }
        event->schedule((*it)->events, tick);
    }

    // For descheduling delayed/timed notifications/timeouts.
    void
    deschedule(ScEvent *event)
    {
        using namespace gem5;

        ScEvents *on = event->scheduledOn();

        if (on == &deltas) {
            event->deschedule();
            return;
        }

        // Timed notification/timeout.
        auto tsit = timeSlots.begin();
        while (tsit != timeSlots.end() &&
               (*tsit)->targeted_when < event->when())
            tsit++;

        panic_if(tsit == timeSlots.end() ||
                 (*tsit)->targeted_when != event->when(),
                "Descheduling event at time with no events.");
        TimeSlot *ts = *tsit;
        ScEvents &events = ts->events;
        assert(on == &events);
        event->deschedule();

        // If no more events are happening at this time slot, get rid of it.
        if (events.empty()) {
            deschedule(ts);
            timeSlots.erase(tsit);
        }
    }

    void
    completeTimeSlot(TimeSlot *ts)
    {
        assert(ts == timeSlots.front());
        timeSlots.erase(timeSlots.begin());
        if (!runToTime && starved())
            scheduleStarvationEvent();
        scheduleTimeAdvancesEvent();
    }

    // Pending activity ignores gem5 activity, much like how a systemc
    // simulation wouldn't know about asynchronous external events (socket IO
    // for instance) that might happen before time advances in a pure
    // systemc simulation. Also the spec lists what specific types of pending
    // activity needs to be counted, which obviously doesn't include gem5
    // events.

    // Return whether there's pending systemc activity at this time.
    bool
    pendingCurr()
    {
        return !readyListMethods.empty() || !readyListThreads.empty() ||
            !updateList.empty() || !deltas.empty();
    }

    // Return whether there are pending timed notifications or timeouts.
    bool
    pendingFuture()
    {
        return !timeSlots.empty();
    }

    // Return how many ticks there are until the first pending event, if any.
    gem5::Tick
    timeToPending()
    {
        if (pendingCurr())
            return 0;
        if (pendingFuture())
            return timeSlots.front()->targeted_when - getCurTick();
        return gem5::MaxTick - getCurTick();
    }

    // Run scheduled channel updates.
    void runUpdate();

    // Run delta events.
    void runDelta();

    void start(gem5::Tick max_tick, bool run_to_time);
    void oneCycle();

    void schedulePause();
    void scheduleStop(bool finish_delta);

    enum Status
    {
        StatusOther = 0,
        StatusEvaluate,
        StatusUpdate,
        StatusDelta,
        StatusTiming,
        StatusPaused,
        StatusStopped
    };

    bool elaborationDone() { return _elaborationDone; }
    void elaborationDone(bool b) { _elaborationDone = b; }

    bool paused() { return status() == StatusPaused; }
    bool stopped() { return status() == StatusStopped; }
    bool inEvaluate() { return status() == StatusEvaluate; }
    bool inUpdate() { return status() == StatusUpdate; }
    bool inDelta() { return status() == StatusDelta; }
    bool inTiming() { return status() == StatusTiming; }

    uint64_t changeStamp() { return _changeStamp; }
    void stepChangeStamp() { _changeStamp++; }

    // Throw upwards, either to sc_main or to the report handler if sc_main
    // isn't running.
    void throwUp();

    Status status() { return _status; }
    void status(Status s) { _status = s; }

    void registerTraceFile(TraceFile *tf) { traceFiles.insert(tf); }
    void unregisterTraceFile(TraceFile *tf) { traceFiles.erase(tf); }

    TimeSlot*
    acquireTimeSlot(gem5::Tick tick)
    {
        TimeSlot *ts = nullptr;
        if (!freeTimeSlots.empty()) {
            ts = freeTimeSlots.top();
            freeTimeSlots.pop();
        } else {
            ts = new TimeSlot(this);
        }
        ts->targeted_when = tick;
        ts->events.clear();
        return ts;
    }

    void
    releaseTimeSlot(TimeSlot *ts)
    {
        freeTimeSlots.push(ts);
    }

  private:
    typedef const gem5::EventBase::Priority Priority;
    static Priority DefaultPriority = gem5::EventBase::Default_Pri;

    static Priority StopPriority = DefaultPriority - 1;
    static Priority PausePriority = DefaultPriority + 1;
    static Priority MaxTickPriority = DefaultPriority + 2;
    static Priority ReadyPriority = DefaultPriority + 3;
    static Priority StarvationPriority = ReadyPriority;
    static Priority TimeAdvancesPriority = gem5::EventBase::Maximum_Pri;

    gem5::EventQueue *eq;

    // For gem5 style events.
    void
    schedule(gem5::Event *event, gem5::Tick tick)
    {
        if (initDone)
            eq->schedule(event, tick);
        else
            eventsToSchedule[event] = tick;
    }

    void schedule(gem5::Event *event) { schedule(event, getCurTick()); }

    void
    deschedule(gem5::Event *event)
    {
        if (initDone)
            eq->deschedule(event);
        else
            eventsToSchedule.erase(event);
    }

    ScEvents deltas;
    TimeSlots timeSlots;
    std::stack<TimeSlot*> freeTimeSlots;

    Process *
    getNextReady()
    {
        Process *p = readyListMethods.getNext();
        return p ? p : readyListThreads.getNext();
    }

    void runReady();
    gem5::EventWrapper<Scheduler, &Scheduler::runReady> readyEvent;
    void scheduleReadyEvent();

    void pause();
    void stop();
    gem5::EventWrapper<Scheduler, &Scheduler::pause> pauseEvent;
    gem5::EventWrapper<Scheduler, &Scheduler::stop> stopEvent;

    const ::sc_core::sc_report *_throwUp;

    bool
    starved()
    {
        return (readyListMethods.empty() && readyListThreads.empty() &&
                updateList.empty() && deltas.empty() &&
                (timeSlots.empty() ||
                 timeSlots.front()->targeted_when > maxTick) &&
                initList.empty());
    }
    gem5::EventWrapper<Scheduler, &Scheduler::pause> starvationEvent;
    void scheduleStarvationEvent();

    bool _elaborationDone;
    bool _started;
    bool _stopNow;

    Status _status;

    gem5::Tick maxTick;
    gem5::Tick lastReadyTick;
    void
    maxTickFunc()
    {
        if (lastReadyTick != getCurTick())
            _changeStamp++;
        pause();
    }
    gem5::EventWrapper<Scheduler, &Scheduler::maxTickFunc> maxTickEvent;

    void timeAdvances() { trace(false); }
    gem5::EventWrapper<Scheduler, &Scheduler::timeAdvances> timeAdvancesEvent;
    void
    scheduleTimeAdvancesEvent()
    {
        if (!traceFiles.empty() && !timeAdvancesEvent.scheduled())
            schedule(&timeAdvancesEvent);
    }

    uint64_t _numCycles;
    uint64_t _changeStamp;

    Process *_current;

    bool initDone;
    bool runToTime;
    bool runOnce;

    ProcessList initList;

    ProcessList readyListMethods;
    ProcessList readyListThreads;

    ChannelList updateList;

    ChannelList asyncUpdateList;
    std::mutex asyncListMutex;
    std::atomic<bool> hasAsyncUpdate;

    std::map<gem5::Event *, gem5::Tick> eventsToSchedule;

    std::set<TraceFile *> traceFiles;

    void trace(bool delta);
};

extern Scheduler scheduler;

// A proxy function to avoid having to expose the scheduler in header files.
Process *getCurrentProcess();

inline void
Scheduler::TimeSlot::process()
{
    scheduler.stepChangeStamp();
    scheduler.status(StatusTiming);

    try {
        while (!events.empty())
            events.front()->run();
    } catch (...) {
        if (events.empty())
            scheduler.completeTimeSlot(this);
        else
            scheduler.schedule(this);
        scheduler.throwUp();
    }

    scheduler.status(StatusOther);
    scheduler.completeTimeSlot(this);
}

const ::sc_core::sc_report reportifyException();

} // namespace sc_gem5

#endif // __SYSTEMC_CORE_SCHEDULER_H__