scheduler_impl.h

#ifndef PLS_SCHEDULER_IMPL_H
#define PLS_SCHEDULER_IMPL_H

#include <utility>

#include "context_switcher/context_switcher.h"
#include "context_switcher/continuation.h"

#include "pls/internal/scheduling/task_manager.h"
#include "pls/internal/scheduling/base_task.h"
#include "base_task.h"

#include "pls/internal/profiling/dag_node.h"

namespace pls::internal::scheduling {

template<typename ALLOC>
scheduler::scheduler(unsigned int num_threads,
                     size_t computation_depth,
                     size_t stack_size,
                     bool reuse_thread,
                     size_t serial_stack_size,
                     ALLOC &&stack_allocator) :
    num_threads_{num_threads},
    reuse_thread_{reuse_thread},
    sync_barrier_{num_threads + 1 - reuse_thread},
    worker_threads_{},
    thread_states_{},
    main_thread_starter_function_{nullptr},
    work_section_done_{false},
    terminated_{false},
    stack_allocator_{std::make_shared<ALLOC>(std::forward<ALLOC>(stack_allocator))},
    serial_stack_size_{serial_stack_size}
#if PLS_PROFILING_ENABLED
, profiler_{num_threads}
#endif
{

  worker_threads_.reserve(num_threads);
  task_managers_.reserve(num_threads);
  thread_states_.reserve(num_threads);
  std::atomic<unsigned> num_spawned{0};
  for (unsigned int i = 0; i < num_threads_; i++) {
    auto &this_task_manager =
        task_managers_.emplace_back(std::make_unique<task_manager>(i,
                                                                   computation_depth,
                                                                   stack_size,
                                                                   stack_allocator_));
    auto &this_thread_state = thread_states_.emplace_back(std::make_unique<thread_state>(*this,
                                                                                         i,
                                                                                         *this_task_manager,
                                                                                         stack_allocator_,
                                                                                         serial_stack_size));

    if (reuse_thread && i == 0) {
      worker_threads_.emplace_back();
      num_spawned++;
      continue; // Skip over first/main thread when re-using the users thread, as this one will replace the first one.
    }

    auto *this_thread_state_pointer = this_thread_state.get();

    worker_threads_.emplace_back([this_thread_state_pointer, &num_spawned] {
      thread_state::set(this_thread_state_pointer);
      num_spawned++;
      work_thread_main_loop();
    });
  }

  while (num_spawned < num_threads) {
    std::this_thread::yield();
  }
}

class scheduler::init_function {
 public:
  virtual void run() = 0;
};
template<typename F>
class scheduler::init_function_impl : public init_function {
 public:
  explicit init_function_impl(F &function) : function_{function} {}
  void run() override {
    base_task *root_task = thread_state::get().get_active_task();
    root_task->attached_resources_.store(nullptr, std::memory_order_relaxed);

#if PLS_PROFILING_ENABLED
    thread_state::get().get_scheduler().profiler_.task_start_running(thread_state::get().get_thread_id(),
                                                                     root_task->profiling_node_);
    thread_state::get().get_scheduler().profiler_.task_prepare_stack_measure(thread_state::get().get_thread_id(),
                                                                             root_task->stack_memory_,
                                                                             root_task->stack_size_);
#endif

    root_task->run_as_task([root_task, this](auto cont) {
      root_task->is_synchronized_ = true;
      thread_state::get().main_continuation() = std::move(cont);
      function_();

      thread_state::get().get_scheduler().work_section_done_.store(true);
      PLS_ASSERT(thread_state::get().main_continuation().valid(), "Must return valid continuation from main task.");

#if PLS_SLEEP_WORKERS_ON_EMPTY
      thread_state::get().get_scheduler().empty_queue_counter_.store(-1000);
      thread_state::get().get_scheduler().empty_queue_decrease_counter_and_wake();
#endif

#if PLS_PROFILING_ENABLED
      thread_state::get().get_scheduler().profiler_.task_finish_stack_measure(thread_state::get().get_thread_id(),
                                                                              root_task->stack_memory_,
                                                                              root_task->stack_size_,
                                                                              root_task->profiling_node_);
      thread_state::get().get_scheduler().profiler_.task_stop_running(thread_state::get().get_thread_id(),
                                                                      root_task->profiling_node_);
#endif

      return std::move(thread_state::get().main_continuation());
    });
  }
 private:
  F &function_;
};

template<typename Function>
void scheduler::perform_work(Function work_section) {
  // Prepare main root task
  init_function_impl<Function> starter_function{work_section};
  main_thread_starter_function_ = &starter_function;

#if PLS_PROFILING_ENABLED
  auto *root_task = thread_state_for(0).get_active_task();
  auto *root_node = profiler_.start_profiler_run();
  root_task->profiling_node_ = root_node;
#endif
#if PLS_SLEEP_WORKERS_ON_EMPTY
  empty_queue_counter_.store(0);
  for (auto &thread_state : thread_states_) {
    thread_state->get_queue_empty_flag().store({EMPTY_QUEUE_STATE::QUEUE_NON_EMPTY});
  }
#endif

  work_section_done_ = false;
  if (reuse_thread_) {
    auto &my_state = thread_state_for(0);
    thread_state::set(&my_state); // Make THIS THREAD become the main worker

    sync_barrier_.wait(); // Trigger threads to wake up
    work_thread_work_section(); // Simply also perform the work section on the main loop
    sync_barrier_.wait(); // Wait for threads to finish
  } else {
    // Simply trigger the others to do the work, this thread will sleep/wait for the time being
    sync_barrier_.wait(); // Trigger threads to wake up
    sync_barrier_.wait(); // Wait for threads to finish
  }
#if PLS_PROFILING_ENABLED
  profiler_.stop_profiler_run();
#endif
}

template<typename Function>
void scheduler::spawn_internal(Function &&lambda) {
  if (thread_state::is_scheduler_active()) {
    thread_state &spawning_state = thread_state::get();

    base_task *last_task = spawning_state.get_active_task();
    base_task *spawned_task = last_task->next_;

    // We are at the end of our allocated tasks, go over to serial execution.
    if (spawned_task == nullptr) {
      scheduler::serial_internal(std::forward<Function>(lambda));
      return;
    }
    // We are in a serial section.
    // For now we simply stay serial. Later versions could add nested parallel sections.
    if (last_task->is_serial_section_) {
      lambda();
      return;
    }

    // Carry over the resources allocated by parents
    auto *attached_resources = last_task->attached_resources_.load(std::memory_order_relaxed);
    spawned_task->attached_resources_.store(attached_resources, std::memory_order_relaxed);

#if PLS_PROFILING_ENABLED
    spawning_state.get_scheduler().profiler_.task_prepare_stack_measure(spawning_state.get_thread_id(),
                                                                        spawned_task->stack_memory_,
                                                                        spawned_task->stack_size_);
    auto *child_dag_node = spawning_state.get_scheduler().profiler_.task_spawn_child(spawning_state.get_thread_id(),
                                                                                     last_task->profiling_node_);
    spawned_task->profiling_node_ = child_dag_node;
#endif

    auto continuation = spawned_task->run_as_task([last_task, spawned_task, lambda, &spawning_state](auto cont) {
      // allow stealing threads to continue the last task.
      last_task->continuation_ = std::move(cont);

      // we are now executing the new task, allow others to steal the last task continuation.
      spawned_task->is_synchronized_ = true;
      spawning_state.set_active_task(spawned_task);

      spawning_state.get_task_manager().push_local_task(last_task);
#if PLS_SLEEP_WORKERS_ON_EMPTY
      data_structures::stamped_integer
          queue_empty_flag = spawning_state.get_queue_empty_flag().load(std::memory_order_relaxed);
      switch (queue_empty_flag.value_) {
        case EMPTY_QUEUE_STATE::QUEUE_NON_EMPTY: {
          // The queue was not found empty, ignore it.
          break;
        }
        case EMPTY_QUEUE_STATE::QUEUE_MAYBE_EMPTY: {
          // Someone tries to mark us empty and might be re-stealing right now.
          data_structures::stamped_integer
              queue_non_empty_flag{queue_empty_flag.stamp_++, EMPTY_QUEUE_STATE::QUEUE_NON_EMPTY};
          auto actual_empty_flag =
              spawning_state.get_queue_empty_flag().exchange(queue_non_empty_flag, std::memory_order_acq_rel);
          if (actual_empty_flag.value_ == EMPTY_QUEUE_STATE::QUEUE_EMPTY) {
            spawning_state.get_scheduler().empty_queue_decrease_counter_and_wake();
          }
          break;
        }
        case EMPTY_QUEUE_STATE::QUEUE_EMPTY: {
          // Someone already marked the queue empty, we must revert its action on the central queue.
          data_structures::stamped_integer
              queue_non_empty_flag{queue_empty_flag.stamp_++, EMPTY_QUEUE_STATE::QUEUE_NON_EMPTY};
          spawning_state.get_queue_empty_flag().store(queue_non_empty_flag, std::memory_order_release);
          spawning_state.get_scheduler().empty_queue_decrease_counter_and_wake();
          break;
        }
      }
#endif

      // execute the lambda itself, which could lead to a different thread returning.
#if PLS_PROFILING_ENABLED
      spawning_state.get_scheduler().profiler_.task_stop_running(spawning_state.get_thread_id(),
                                                                 last_task->profiling_node_);
      spawning_state.get_scheduler().profiler_.task_start_running(spawning_state.get_thread_id(),
                                                                  spawned_task->profiling_node_);
#endif
      lambda();

      thread_state &syncing_state = thread_state::get();
      PLS_ASSERT(syncing_state.get_active_task() == spawned_task,
                 "Task manager must always point its active task onto whats executing.");

      // try to pop a task of the syncing task manager.
      // possible outcomes:
      // - this is a different task manager, it must have an empty deque and fail
      // - this is the same task manager and someone stole last tasks, thus this will fail
      // - this is the same task manager and no one stole the last task, this this will succeed
      base_task *popped_task = syncing_state.get_task_manager().pop_local_task();
      if (popped_task) {
        // Fast path, simply continue execution where we left of before spawn.
        PLS_ASSERT(popped_task == last_task,
                   "Fast path, nothing can have changed until here.");
        PLS_ASSERT(&spawning_state == &syncing_state,
                   "Fast path, we must only return if the task has not been stolen/moved to other thread.");
        PLS_ASSERT(last_task->continuation_.valid(),
                   "Fast path, no one can have continued working on the last task.");

        syncing_state.set_active_task(last_task);
#if PLS_PROFILING_ENABLED
        syncing_state.get_scheduler().profiler_.task_finish_stack_measure(syncing_state.get_thread_id(),
                                                                          spawned_task->stack_memory_,
                                                                          spawned_task->stack_size_,
                                                                          spawned_task->profiling_node_);
        syncing_state.get_scheduler().profiler_.task_stop_running(syncing_state.get_thread_id(),
                                                                  spawned_task->profiling_node_);
#endif
        return std::move(last_task->continuation_);
      } else {
        // Slow path, the last task was stolen. This path is common to sync() events.
#if PLS_PROFILING_ENABLED
        syncing_state.get_scheduler().profiler_.task_finish_stack_measure(syncing_state.get_thread_id(),
                                                                          spawned_task->stack_memory_,
                                                                          spawned_task->stack_size_,
                                                                          spawned_task->profiling_node_);
        syncing_state.get_scheduler().profiler_.task_stop_running(syncing_state.get_thread_id(),
                                                                  spawned_task->profiling_node_);
#endif
        auto continuation = slow_return(syncing_state, false);
        return continuation;
      }
    });

#if PLS_PROFILING_ENABLED
    thread_state::get().get_scheduler().profiler_.task_start_running(thread_state::get().get_thread_id(),
                                                                     last_task->profiling_node_);
#endif

    if (continuation.valid()) {
      // We jumped in here from the main loop, keep track!
      thread_state::get().main_continuation() = std::move(continuation);
    }
  } else {
    // Scheduler not active...
    lambda();
  }
}

template<typename Function>
void scheduler::spawn_and_sync_internal(Function &&lambda) {
  thread_state &spawning_state = thread_state::get();

  base_task *last_task = spawning_state.get_active_task();
  base_task *spawned_task = last_task->next_;

  // We are at the end of our allocated tasks, go over to serial execution.
  if (spawned_task == nullptr) {
    scheduler::serial_internal(std::forward<Function>(lambda));
    return;
  }
  // We are in a serial section.
  // For now we simply stay serial. Later versions could add nested parallel sections.
  if (last_task->is_serial_section_) {
    lambda();
    return;
  }

  // Carry over the resources allocated by parents
  auto *attached_resources = last_task->attached_resources_.load(std::memory_order_relaxed);
  spawned_task->attached_resources_.store(attached_resources, std::memory_order_relaxed);

#if PLS_PROFILING_ENABLED
  spawning_state.get_scheduler().profiler_.task_prepare_stack_measure(spawning_state.get_thread_id(),
                                                                      spawned_task->stack_memory_,
                                                                      spawned_task->stack_size_);
  auto *child_dag_node = spawning_state.get_scheduler().profiler_.task_spawn_child(spawning_state.get_thread_id(),
                                                                                   last_task->profiling_node_);
  spawned_task->profiling_node_ = child_dag_node;
#endif

  auto continuation = spawned_task->run_as_task([last_task, spawned_task, lambda, &spawning_state](auto cont) {
    // allow stealing threads to continue the last task.
    last_task->continuation_ = std::move(cont);

    // we are now executing the new task, allow others to steal the last task continuation.
    spawned_task->is_synchronized_ = true;
    spawning_state.set_active_task(spawned_task);

    // execute the lambda itself, which could lead to a different thread returning.
#if PLS_PROFILING_ENABLED
    spawning_state.get_scheduler().profiler_.task_finish_stack_measure(spawning_state.get_thread_id(),
                                                                       last_task->stack_memory_,
                                                                       last_task->stack_size_,
                                                                       last_task->profiling_node_);
    spawning_state.get_scheduler().profiler_.task_stop_running(spawning_state.get_thread_id(),
                                                               last_task->profiling_node_);
    auto *next_dag_node =
        spawning_state.get_scheduler().profiler_.task_sync(spawning_state.get_thread_id(), last_task->profiling_node_);
    last_task->profiling_node_ = next_dag_node;
    spawning_state.get_scheduler().profiler_.task_start_running(spawning_state.get_thread_id(),
                                                                spawned_task->profiling_node_);
#endif
    lambda();

    thread_state &syncing_state = thread_state::get();
    PLS_ASSERT(syncing_state.get_active_task() == spawned_task,
               "Task manager must always point its active task onto whats executing.");

    if (&syncing_state == &spawning_state && last_task->is_synchronized_) {
      // Fast path, simply continue execution where we left of before spawn.
      PLS_ASSERT(last_task->continuation_.valid(),
                 "Fast path, no one can have continued working on the last task.");

      syncing_state.set_active_task(last_task);
#if PLS_PROFILING_ENABLED
      syncing_state.get_scheduler().profiler_.task_finish_stack_measure(syncing_state.get_thread_id(),
                                                                        spawned_task->stack_memory_,
                                                                        spawned_task->stack_size_,
                                                                        spawned_task->profiling_node_);
      syncing_state.get_scheduler().profiler_.task_stop_running(syncing_state.get_thread_id(),
                                                                spawned_task->profiling_node_);
#endif
      return std::move(last_task->continuation_);
    } else {
      // Slow path, the last task was stolen. This path is common to sync() events.
#if PLS_PROFILING_ENABLED
      syncing_state.get_scheduler().profiler_.task_finish_stack_measure(syncing_state.get_thread_id(),
                                                                        spawned_task->stack_memory_,
                                                                        spawned_task->stack_size_,
                                                                        spawned_task->profiling_node_);
      syncing_state.get_scheduler().profiler_.task_stop_running(syncing_state.get_thread_id(),
                                                                spawned_task->profiling_node_);
#endif
      auto continuation = slow_return(syncing_state, false);
      return continuation;
    }
  });

#if PLS_PROFILING_ENABLED
  thread_state::get().get_scheduler().profiler_.task_start_running(thread_state::get().get_thread_id(),
                                                                   last_task->profiling_node_);
#endif

  PLS_ASSERT(!continuation.valid(), "We must not jump to a not-published (synced) spawn.");
}

template<typename Function>
void scheduler::serial_internal(Function &&lambda) {
  thread_state &spawning_state = thread_state::get();
  base_task *active_task = spawning_state.get_active_task();

  if (active_task->is_serial_section_) {
    lambda();
  } else {
    active_task->is_serial_section_ = true;
    context_switcher::enter_context(spawning_state.get_serial_call_stack(),
                                    spawning_state.get_serial_call_stack_size(),
                                    [&](auto cont) {
                                      lambda();
                                      return cont;
                                    });
    active_task->is_serial_section_ = false;
  }
}

}

#endif //PLS_SCHEDULER_IMPL_H