Change both stack and queue to same offset counters.

This allows the stack and deque class to use the same offset, making it work better with each other.

app/benchmark_fft/main.cpp

@@ -80,7 +80,7 @@ int main() {
   pls::internal::helpers::run_mini_benchmark([&] {
   pls::internal::helpers::run_mini_benchmark([&] {
     complex_vector input = initial_input;
     complex_vector input = initial_input;
     fft(input.begin(), input.size());
     fft(input.begin(), input.size());
-  }, 7, 30000);
+  }, 7, 1000);
 
 
   PROFILE_SAVE("test_profile.prof")
   PROFILE_SAVE("test_profile.prof")
 }
 }

lib/pls/include/pls/internal/base/alignment.h

@@ -21,7 +21,6 @@ void *allocate_aligned(size_t size);
 
 
 system_details::pointer_t next_alignment(system_details::pointer_t size);
 system_details::pointer_t next_alignment(system_details::pointer_t size);
 system_details::pointer_t previous_alignment(system_details::pointer_t size);
 system_details::pointer_t previous_alignment(system_details::pointer_t size);
-char *next_alignment(char *pointer);
 
 
 }
 }
 
 

lib/pls/include/pls/internal/data_structures/aligned_stack.h

@@ -27,16 +27,10 @@ using base::system_details::pointer_t;
  * stack.pop<T>(); // Remove the top object of type T
  * stack.pop<T>(); // Remove the top object of type T
  */
  */
 class aligned_stack {
 class aligned_stack {
-  // Keep bounds of our memory block
-  pointer_t memory_start_;
-  pointer_t memory_end_;
-
-  // Current head will always be aligned to cache lines
-  pointer_t head_;
  public:
  public:
-  typedef pointer_t state;
+  typedef size_t stack_offset;
 
 
-  aligned_stack() : memory_start_{0}, memory_end_{0}, head_{0} {};
+  aligned_stack() : aligned_memory_start_{0}, aligned_memory_end_{0}, max_offset_{0}, current_offset_{0} {};
   aligned_stack(pointer_t memory_region, std::size_t size);
   aligned_stack(pointer_t memory_region, std::size_t size);
   aligned_stack(char *memory_region, std::size_t size);
   aligned_stack(char *memory_region, std::size_t size);
 
 
@@ -48,8 +42,18 @@ class aligned_stack {
   template<typename T>
   template<typename T>
   T pop();
   T pop();
 
 
-  state save_state() const { return head_; }
-  void reset_state(state new_state) { head_ = new_state; }
+  void *memory_at_offset(stack_offset offset) const;
+
+  stack_offset save_offset() const { return current_offset_; }
+  void reset_offset(stack_offset new_offset) { current_offset_ = new_offset; }
+
+ private:
+  // Keep bounds of our memory block
+  pointer_t aligned_memory_start_;
+  pointer_t aligned_memory_end_;
+
+  stack_offset max_offset_;
+  stack_offset current_offset_;
 };
 };
 
 
 }
 }

lib/pls/include/pls/internal/data_structures/aligned_stack_impl.h

@@ -21,8 +21,10 @@ void *aligned_stack::push_bytes() {
 
 
 template<typename T>
 template<typename T>
 T aligned_stack::pop() {
 T aligned_stack::pop() {
-  head_ = head_ - base::alignment::next_alignment(sizeof(T));
-  return *reinterpret_cast<T *>(head_);
+  auto num_cache_lines = base::alignment::next_alignment(sizeof(T)) / base::system_details::CACHE_LINE_SIZE;
+  current_offset_ -= num_cache_lines;
+
+  return *reinterpret_cast<T *>(memory_at_offset(current_offset_));
 }
 }
 
 
 }
 }

lib/pls/include/pls/internal/data_structures/locking_deque.h

@@ -48,7 +48,7 @@ class locking_deque {
   base::spin_lock lock_;
   base::spin_lock lock_;
 
 
  public:
  public:
-  using state = aligned_stack::state;
+  using state = aligned_stack::stack_offset;
 
 
   explicit locking_deque(aligned_stack *stack)
   explicit locking_deque(aligned_stack *stack)
       : stack_{stack}, head_{nullptr}, tail_{nullptr}, lock_{} {}
       : stack_{stack}, head_{nullptr}, tail_{nullptr}, lock_{} {}

lib/pls/include/pls/internal/data_structures/locking_deque_impl.h

@@ -75,11 +75,11 @@ Item *locking_deque<Item>::pop_head() {
 
 
 template<typename Item>
 template<typename Item>
 void locking_deque<Item>::release_memory_until(state state) {
 void locking_deque<Item>::release_memory_until(state state) {
-  stack_->reset_state(state);
+  stack_->reset_offset(state);
 }
 }
 template<typename Item>
 template<typename Item>
 typename locking_deque<Item>::state locking_deque<Item>::save_state() {
 typename locking_deque<Item>::state locking_deque<Item>::save_state() {
-  return stack_->save_state();
+  return stack_->save_offset();
 }
 }
 
 
 }
 }

lib/pls/include/pls/internal/data_structures/work_stealing_deque.h

@@ -17,7 +17,7 @@ using base::system_details::pointer_t;
 
 
 // Integer split into two halfs, can be used in CAS operations
 // Integer split into two halfs, can be used in CAS operations
 using data_structures::stamped_integer;
 using data_structures::stamped_integer;
-using offset_t = stamped_integer::member_t;
+using deque_offset = stamped_integer::member_t;
 
 
 // Single Item in the deque
 // Single Item in the deque
 class work_stealing_deque_item {
 class work_stealing_deque_item {
@@ -29,8 +29,8 @@ class work_stealing_deque_item {
   // Pointer to the actual data
   // Pointer to the actual data
   std::atomic<pointer_t> data_;
   std::atomic<pointer_t> data_;
   // Index (relative to stack base) to the next and previous element
   // Index (relative to stack base) to the next and previous element
-  std::atomic<offset_t> next_item_;
-  offset_t previous_item_;
+  std::atomic<deque_offset> next_item_;
+  deque_offset previous_item_;
 
 
  public:
  public:
   work_stealing_deque_item() : data_{0}, next_item_{}, previous_item_{} {}
   work_stealing_deque_item() : data_{0}, next_item_{}, previous_item_{} {}
@@ -45,54 +45,48 @@ class work_stealing_deque_item {
     data_ = reinterpret_cast<pointer_t >(data);
     data_ = reinterpret_cast<pointer_t >(data);
   }
   }
 
 
-  offset_t next_item() const { return next_item_.load(); }
-  void set_next_item(offset_t next_item) { next_item_ = next_item; }
+  deque_offset next_item() const { return next_item_.load(); }
+  void set_next_item(deque_offset next_item) { next_item_ = next_item; }
 
 
-  offset_t previous_item() const { return previous_item_; }
-  void set_previous_item(offset_t previous_item) { previous_item_ = previous_item; }
+  deque_offset previous_item() const { return previous_item_; }
+  void set_previous_item(deque_offset previous_item) { previous_item_ = previous_item; }
 };
 };
-static_assert(sizeof(work_stealing_deque_item) < base::system_details::CACHE_LINE_SIZE,
-              "Work stealing deque relies on memory layout and requires cache lines to be longer than one 'work_stealing_deque_item' instance!");
 
 
-template<typename Item>
+template<typename Task>
 class work_stealing_deque {
 class work_stealing_deque {
   // Deque 'takes over' stack and handles memory management while in use.
   // Deque 'takes over' stack and handles memory management while in use.
   // At any point in time the deque can stop using more memory and the stack can be used by other entities.
   // At any point in time the deque can stop using more memory and the stack can be used by other entities.
   aligned_stack *stack_;
   aligned_stack *stack_;
-  pointer_t base_pointer_;
 
 
   std::atomic<stamped_integer> head_;
   std::atomic<stamped_integer> head_;
-  std::atomic<offset_t> tail_;
-  offset_t previous_tail_;
+  std::atomic<deque_offset> tail_;
+  deque_offset previous_tail_;
 
 
- public:
-  using state = aligned_stack::state;
+  Task* last_pushed_task_;
 
 
+ public:
   explicit work_stealing_deque(aligned_stack *stack) : stack_{stack},
   explicit work_stealing_deque(aligned_stack *stack) : stack_{stack},
-                                                       base_pointer_{0},
                                                        head_{stamped_integer{0, 0}},
                                                        head_{stamped_integer{0, 0}},
                                                        tail_{0},
                                                        tail_{0},
-                                                       previous_tail_{0} {
-    reset_base_pointer();
-  }
+                                                       previous_tail_{0},
+                                                       last_pushed_task_{0} {}
 
 
   template<typename T, typename ...ARGS>
   template<typename T, typename ...ARGS>
-  T *push_tail(ARGS &&... args);
-  template<typename T, typename Function, typename ...ARGS>
-  T *push_tail_cb(const Function &after_creation, ARGS &&... args);
-  Item *pop_tail();
-  Item *pop_head();
+  T *push_task(ARGS &&... args);
+  template<typename T, typename ...ARGS>
+  T *push_object(ARGS &&... args);
+  void *push_bytes(size_t size);
+  void publish_last_task();
 
 
-  void release_memory_until(state state);
-  state save_state();
+  Task *pop_local_task();
+  Task *pop_external_task();
 
 
- private:
-  void reset_base_pointer();
-  work_stealing_deque_item *item_at(offset_t offset);
-  offset_t current_stack_offset();
+  void reset_offset(deque_offset offset);
+  deque_offset save_offset();
 
 
-  template<typename T, typename ...ARGS>
-  std::pair<work_stealing_deque_item, T> *allocate_item(ARGS &&... args);
+ private:
+  work_stealing_deque_item *item_at(deque_offset offset);
+  deque_offset current_stack_offset();
 };
 };
 
 
 }
 }

lib/pls/include/pls/internal/data_structures/work_stealing_deque_impl.h

@@ -9,26 +9,23 @@ namespace pls {
 namespace internal {
 namespace internal {
 namespace data_structures {
 namespace data_structures {
 
 
-template<typename Item>
-void work_stealing_deque<Item>::reset_base_pointer() {
-  base_pointer_ = reinterpret_cast<pointer_t >(stack_->save_state()); // Keep the base of our region in the stack
+template<typename Task>
+work_stealing_deque_item *work_stealing_deque<Task>::item_at(deque_offset offset) {
+  return reinterpret_cast<work_stealing_deque_item *>(stack_->memory_at_offset(offset));
 }
 }
 
 
-template<typename Item>
-work_stealing_deque_item *work_stealing_deque<Item>::item_at(offset_t offset) {
-  return reinterpret_cast<work_stealing_deque_item *>(base_pointer_
-      + (base::system_details::CACHE_LINE_SIZE * offset));
+template<typename Task>
+deque_offset work_stealing_deque<Task>::current_stack_offset() {
+  return stack_->save_offset();
 }
 }
 
 
-template<typename Item>
-offset_t work_stealing_deque<Item>::current_stack_offset() {
-  return (stack_->save_state() - base_pointer_) / base::system_details::CACHE_LINE_SIZE;
-}
-
-template<typename Item>
+template<typename Task>
 template<typename T, typename ...ARGS>
 template<typename T, typename ...ARGS>
-std::pair<work_stealing_deque_item, T> *work_stealing_deque<Item>::allocate_item(ARGS &&... args) {
-  // 'Union' type to push both on stack
+T *work_stealing_deque<Task>::push_task(ARGS &&... args) {
+  static_assert(std::is_same<Task, T>::value || std::is_base_of<Task, T>::value,
+                "Must only push types of <Task> onto work_stealing_deque<Task>");
+
+  // 'Union' type to push both the task and the deque entry as one part onto the stack
   using pair_t = std::pair<work_stealing_deque_item, T>;
   using pair_t = std::pair<work_stealing_deque_item, T>;
   // Allocate space on stack
   // Allocate space on stack
   auto new_pair = reinterpret_cast<pair_t *>(stack_->push_bytes<pair_t>());
   auto new_pair = reinterpret_cast<pair_t *>(stack_->push_bytes<pair_t>());
@@ -36,43 +33,45 @@ std::pair<work_stealing_deque_item, T> *work_stealing_deque<Item>::allocate_item
   new((void *) &(new_pair->first)) work_stealing_deque_item();
   new((void *) &(new_pair->first)) work_stealing_deque_item();
   new((void *) &(new_pair->second)) T(std::forward<ARGS>(args)...);
   new((void *) &(new_pair->second)) T(std::forward<ARGS>(args)...);
 
 
-  return new_pair;
+  // Keep reference for later publishing
+  last_pushed_task_ = &new_pair->second;
+
+  // Item is not publicly visible until it is published
+  return &(new_pair->second);
 }
 }
 
 
-template<typename Item>
+template<typename Task>
 template<typename T, typename ...ARGS>
 template<typename T, typename ...ARGS>
-T *work_stealing_deque<Item>::push_tail(ARGS &&... args) {
-  return push_tail_cb<T>([](T *) {}, std::forward<ARGS>(args)...);
+T *work_stealing_deque<Task>::push_object(ARGS &&... args) {
+  // Simply add data to the stack, do not publish it in any way
+  return stack_->push<T>(std::forward(args)...);
 }
 }
 
 
-template<typename Item>
-template<typename T, typename Function, typename ...ARGS>
-T *work_stealing_deque<Item>::push_tail_cb(const Function &after_creation, ARGS &&... args) {
-  static_assert(std::is_same<Item, T>::value || std::is_base_of<Item, T>::value,
-                "Must only push types of <Item> onto work_stealing_deque<Item>");
-
-  offset_t local_tail = tail_;
+template<typename Task>
+void *work_stealing_deque<Task>::push_bytes(size_t size) {
+  // Simply add data to the stack, do not publish it in any way
+  return stack_->push_bytes(size);
+}
 
 
-  auto new_pair = allocate_item<T>(std::forward<ARGS>(args)...);
-  after_creation(&(new_pair->second)); // callback for time after creation but before being visible to others
+template<typename Task>
+void work_stealing_deque<Task>::publish_last_task() {
+  deque_offset local_tail = tail_;
 
 
-  // Prepare current tail to point to correct next items
+  // Prepare current tail to point to correct next task
   auto tail_deque_item = item_at(local_tail);
   auto tail_deque_item = item_at(local_tail);
-  tail_deque_item->set_data(&(new_pair->second));
+  tail_deque_item->set_data(last_pushed_task_);
   tail_deque_item->set_next_item(current_stack_offset());
   tail_deque_item->set_next_item(current_stack_offset());
   tail_deque_item->set_previous_item(previous_tail_);
   tail_deque_item->set_previous_item(previous_tail_);
   previous_tail_ = local_tail;
   previous_tail_ = local_tail;
 
 
-  // Linearization point, item appears after this write
-  offset_t new_tail = current_stack_offset();
+  // Linearization point, task appears after this write
+  deque_offset new_tail = current_stack_offset();
   tail_ = new_tail;
   tail_ = new_tail;
-
-  return &(new_pair->second);
 }
 }
 
 
-template<typename Item>
-Item *work_stealing_deque<Item>::pop_tail() {
-  offset_t local_tail = tail_;
+template<typename Task>
+Task *work_stealing_deque<Task>::pop_local_task() {
+  deque_offset local_tail = tail_;
   stamped_integer local_head = head_;
   stamped_integer local_head = head_;
 
 
   if (local_tail <= local_head.value) {
   if (local_tail <= local_head.value) {
@@ -80,7 +79,7 @@ Item *work_stealing_deque<Item>::pop_tail() {
   }
   }
 
 
   work_stealing_deque_item *previous_tail_item = item_at(previous_tail_);
   work_stealing_deque_item *previous_tail_item = item_at(previous_tail_);
-  offset_t new_tail = previous_tail_;
+  deque_offset new_tail = previous_tail_;
   previous_tail_ = previous_tail_item->previous_item();
   previous_tail_ = previous_tail_item->previous_item();
 
 
   // Publish our wish to set the tail back
   // Publish our wish to set the tail back
@@ -89,14 +88,14 @@ Item *work_stealing_deque<Item>::pop_tail() {
   local_head = head_; // Linearization point, outside knows list is empty
   local_head = head_; // Linearization point, outside knows list is empty
 
 
   if (local_head.value < new_tail) {
   if (local_head.value < new_tail) {
-    return previous_tail_item->data<Item>(); // Success, enough distance to other threads
+    return previous_tail_item->data<Task>(); // Success, enough distance to other threads
   }
   }
 
 
   if (local_head.value == new_tail) {
   if (local_head.value == new_tail) {
     stamped_integer new_head = stamped_integer{local_head.stamp + 1, new_tail};
     stamped_integer new_head = stamped_integer{local_head.stamp + 1, new_tail};
     // Try competing with consumers by updating the head's stamp value
     // Try competing with consumers by updating the head's stamp value
     if (head_.compare_exchange_strong(local_head, new_head)) {
     if (head_.compare_exchange_strong(local_head, new_head)) {
-      return previous_tail_item->data<Item>(); // SUCCESS, we won the competition with other threads
+      return previous_tail_item->data<Task>(); // SUCCESS, we won the competition with other threads
     }
     }
   }
   }
 
 
@@ -108,10 +107,10 @@ Item *work_stealing_deque<Item>::pop_tail() {
   return nullptr; // EMPTY, we lost the competition with other threads
   return nullptr; // EMPTY, we lost the competition with other threads
 }
 }
 
 
-template<typename Item>
-Item *work_stealing_deque<Item>::pop_head() {
+template<typename Task>
+Task *work_stealing_deque<Task>::pop_external_task() {
   stamped_integer local_head = head_;
   stamped_integer local_head = head_;
-  offset_t local_tail = tail_;
+  deque_offset local_tail = tail_;
 
 
   if (local_tail <= local_head.value) {
   if (local_tail <= local_head.value) {
     return nullptr; // EMPTY
     return nullptr; // EMPTY
@@ -121,8 +120,8 @@ Item *work_stealing_deque<Item>::pop_head() {
   // there has to be a competition over the tail -> the stamp increased and our next
   // there has to be a competition over the tail -> the stamp increased and our next
   // operation will fail anyways!
   // operation will fail anyways!
   work_stealing_deque_item *head_deque_item = item_at(local_head.value);
   work_stealing_deque_item *head_deque_item = item_at(local_head.value);
-  offset_t next_item_offset = head_deque_item->next_item();
-  Item *head_data_item = head_deque_item->data<Item>();
+  deque_offset next_item_offset = head_deque_item->next_item();
+  Task *head_data_item = head_deque_item->data<Task>();
 
 
   // We try to set the head to this new position.
   // We try to set the head to this new position.
   // Possible outcomes:
   // Possible outcomes:
@@ -137,26 +136,23 @@ Item *work_stealing_deque<Item>::pop_head() {
   return nullptr; // EMPTY, we lost the competition
   return nullptr; // EMPTY, we lost the competition
 }
 }
 
 
-template<typename Item>
-void work_stealing_deque<Item>::release_memory_until(state state) {
-  unsigned long item_offset = (state - base_pointer_) / base::system_details::CACHE_LINE_SIZE;
+template<typename Task>
+void work_stealing_deque<Task>::reset_offset(deque_offset offset) {
+  stack_->reset_offset(offset);
 
 
   stamped_integer local_head = head_;
   stamped_integer local_head = head_;
-  offset_t local_tail = tail_;
-
-  stack_->reset_state(state);
-
-  if (item_offset < local_tail) {
-    tail_ = item_offset;
+  deque_offset local_tail = tail_;
+  if (offset < local_tail) {
+    tail_ = offset;
     if (local_head.value >= local_tail) {
     if (local_head.value >= local_tail) {
-      head_ = stamped_integer{local_head.stamp + 1, item_offset};
+      head_ = stamped_integer{local_head.stamp + 1, offset};
     }
     }
   }
   }
 }
 }
 
 
-template<typename Item>
-typename work_stealing_deque<Item>::state work_stealing_deque<Item>::save_state() {
-  return stack_->save_state();
+template<typename Task>
+deque_offset work_stealing_deque<Task>::save_offset() {
+  return current_stack_offset();
 }
 }
 
 
 }
 }

lib/pls/include/pls/internal/helpers/mini_benchmark.h

@@ -21,7 +21,7 @@ void run_mini_benchmark(const Function &lambda,
   using namespace std;
   using namespace std;
   using namespace pls::internal::scheduling;
   using namespace pls::internal::scheduling;
 
 
-  malloc_scheduler_memory scheduler_memory{max_threads, 2u << 14};
+  malloc_scheduler_memory scheduler_memory{max_threads, 2u << 17u};
   for (unsigned int num_threads = 1; num_threads <= max_threads; num_threads++) {
   for (unsigned int num_threads = 1; num_threads <= max_threads; num_threads++) {
     scheduler local_scheduler{&scheduler_memory, num_threads};
     scheduler local_scheduler{&scheduler_memory, num_threads};
 
 

lib/pls/include/pls/internal/scheduling/scheduler_impl.h

@@ -28,7 +28,7 @@ void scheduler::perform_work(Function work_section) {
 
 
     // Do work (see if we can remove this duplicated code)
     // Do work (see if we can remove this duplicated code)
     root_task.parent_ = nullptr;
     root_task.parent_ = nullptr;
-    root_task.deque_state_ = my_state->deque_.save_state();
+    root_task.deque_offset_ = my_state->deque_.save_offset();
     root_task.execute();
     root_task.execute();
     work_section_done_ = true;
     work_section_done_ = true;
 
 

lib/pls/include/pls/internal/scheduling/task.h

@@ -24,7 +24,7 @@ class task {
   task *parent_;
   task *parent_;
 
 
   // Stack Management (reset stack pointer after wait_for_all() calls)
   // Stack Management (reset stack pointer after wait_for_all() calls)
-  data_structures::deque<task>::state deque_state_;
+  data_structures::deque_offset deque_offset_;
 
 
  protected:
  protected:
   /*
   /*
@@ -66,20 +66,36 @@ void task::spawn_child(ARGS &&... args) {
   ref_count_++;
   ref_count_++;
 
 
   // Push on our deque
   // Push on our deque
-  thread_state::get()->deque_.push_tail_cb<T>([this](T *item) {
-    // Assign forced values (for stack and parent management)
-    item->parent_ = this;
-    item->finished_construction_ = true;
-    item->deque_state_ = thread_state::get()->deque_.save_state();
-  }, std::forward<ARGS>(args)...);
+  auto item = thread_state::get()->deque_.push_task<T>(std::forward<ARGS>(args)...);
+
+  // Assign forced values (for stack and parent management)
+  item->parent_ = this;
+  item->finished_construction_ = true;
+  item->deque_offset_ = thread_state::get()->deque_.save_offset();
+
+  // Make new task visible to others
+  thread_state::get()->deque_.publish_last_task();
 }
 }
 
 
 template<typename T, typename ...ARGS>
 template<typename T, typename ...ARGS>
 void task::spawn_child_and_wait(ARGS &&... args) {
 void task::spawn_child_and_wait(ARGS &&... args) {
   static_assert(std::is_base_of<task, typename std::remove_reference<T>::type>::value, "Only pass task subclasses!");
   static_assert(std::is_base_of<task, typename std::remove_reference<T>::type>::value, "Only pass task subclasses!");
 
 
-  // TODO: See if we can inline this (avoid counters/deque) while maintaining memory management
   spawn_child<T>(std::forward<ARGS>(args)...);
   spawn_child<T>(std::forward<ARGS>(args)...);
+  // TODO: Check why 'direct spawn' (even when pushing it onto the tas queue) seems to be slower
+  //       (Also check if it even is slower or if it only appears so on our laptop)
+//  // Push on our deque
+//  auto task = thread_state::get()->deque_.push_task<T>(std::forward<ARGS>(args)...);
+//
+//  // Assign forced values (for stack and parent management)
+//  task->parent_ = nullptr; // ...do not assign this to a parent => it will not notify our reference counter
+//  task->finished_construction_ = true;
+//  task->deque_offset_ = thread_state::get()->deque_.save_offset();
+//
+//  // Execute it
+//  task->execute();
+
+  // Wait for the rest of the tasks
   wait_for_all();
   wait_for_all();
 }
 }
 
 

lib/pls/src/internal/data_structures/aligned_stack.cpp

@@ -6,21 +6,28 @@ namespace internal {
 namespace data_structures {
 namespace data_structures {
 
 
 aligned_stack::aligned_stack(pointer_t memory_region, const std::size_t size) :
 aligned_stack::aligned_stack(pointer_t memory_region, const std::size_t size) :
-    memory_start_{memory_region},
-    memory_end_{memory_region + size},
-    head_{base::alignment::next_alignment(memory_start_)} {}
+    aligned_memory_start_{base::alignment::next_alignment(memory_region)},
+    aligned_memory_end_{base::alignment::previous_alignment(memory_region + size)},
+    max_offset_{(aligned_memory_end_ - aligned_memory_start_) / base::system_details::CACHE_LINE_SIZE},
+    current_offset_{0} {}
 
 
 aligned_stack::aligned_stack(char *memory_region, const std::size_t size) :
 aligned_stack::aligned_stack(char *memory_region, const std::size_t size) :
-    memory_start_{(pointer_t) memory_region},
-    memory_end_{(pointer_t) memory_region + size},
-    head_{base::alignment::next_alignment(memory_start_)} {}
+    aligned_stack((pointer_t) memory_region, size) {}
+
+void *aligned_stack::memory_at_offset(stack_offset offset) const {
+  const auto byte_offset = offset * base::system_details::CACHE_LINE_SIZE;
+  return reinterpret_cast<void *>(aligned_memory_start_ + byte_offset);
+}
 
 
 void *aligned_stack::push_bytes(size_t size) {
 void *aligned_stack::push_bytes(size_t size) {
-  void *result = reinterpret_cast<void *>(head_);
+  size_t round_up_size = base::alignment::next_alignment(size);
+  size_t num_cache_lines = round_up_size / base::system_details::CACHE_LINE_SIZE;
+
+  void *result = memory_at_offset(current_offset_);
 
 
   // Move head to next aligned position after new object
   // Move head to next aligned position after new object
-  head_ = base::alignment::next_alignment(head_ + size);
-  if (head_ >= memory_end_) {
+  current_offset_ += num_cache_lines;
+  if (current_offset_ > max_offset_) {
     PLS_ERROR("Tried to allocate object on alligned_stack without sufficient memory!");
     PLS_ERROR("Tried to allocate object on alligned_stack without sufficient memory!");
   }
   }
 
 

lib/pls/src/internal/scheduling/scheduler.cpp

@@ -53,7 +53,7 @@ void scheduler::worker_routine() {
       // Main Thread
       // Main Thread
       auto root_task = scheduler->main_thread_root_task_;
       auto root_task = scheduler->main_thread_root_task_;
       root_task->parent_ = nullptr;
       root_task->parent_ = nullptr;
-      root_task->deque_state_ = my_state->deque_.save_state();
+      root_task->deque_offset_ = my_state->deque_.save_offset();
 
 
       root_task->execute();
       root_task->execute();
       scheduler->work_section_done_ = true;
       scheduler->work_section_done_ = true;
@@ -90,7 +90,7 @@ void scheduler::terminate() {
 
 
 task *scheduler::get_local_task() {
 task *scheduler::get_local_task() {
   PROFILE_STEALING("Get Local Task")
   PROFILE_STEALING("Get Local Task")
-  return thread_state::get()->deque_.pop_tail();
+  return thread_state::get()->deque_.pop_local_task();
 }
 }
 
 
 task *scheduler::steal_task() {
 task *scheduler::steal_task() {
@@ -112,7 +112,7 @@ task *scheduler::steal_task() {
 
 
     auto target_state = thread_state_for(target);
     auto target_state = thread_state_for(target);
     // TODO: See if we should re-try popping if it failed due to contention
     // TODO: See if we should re-try popping if it failed due to contention
-    auto result = target_state->deque_.pop_head();
+    auto result = target_state->deque_.pop_external_task();
     if (result != nullptr) {
     if (result != nullptr) {
       return result;
       return result;
     }
     }
@@ -137,7 +137,7 @@ bool scheduler::try_execute_local() {
 bool scheduler::try_execute_stolen() {
 bool scheduler::try_execute_stolen() {
   task *stolen_task = steal_task();
   task *stolen_task = steal_task();
   if (stolen_task != nullptr) {
   if (stolen_task != nullptr) {
-    stolen_task->deque_state_ = thread_state::get()->deque_.save_state();
+    stolen_task->deque_offset_ = thread_state::get()->deque_.save_offset();
     stolen_task->execute();
     stolen_task->execute();
     return true;
     return true;
   }
   }

lib/pls/src/internal/scheduling/task.cpp

@@ -12,13 +12,13 @@ task::task() :
     finished_construction_{false},
     finished_construction_{false},
     ref_count_{0},
     ref_count_{0},
     parent_{nullptr},
     parent_{nullptr},
-    deque_state_{0} {}
+    deque_offset_{0} {}
 
 
 void *task::allocate_memory(long size) {
 void *task::allocate_memory(long size) {
   if (finished_construction_) {
   if (finished_construction_) {
     PLS_ERROR("Must not allocate dynamic task memory after it's construction.")
     PLS_ERROR("Must not allocate dynamic task memory after it's construction.")
   }
   }
-  return thread_state::get()->task_stack_->push_bytes(size);
+  return thread_state::get()->deque_.push_bytes(size);
 }
 }
 
 
 void task::execute() {
 void task::execute() {
@@ -45,7 +45,7 @@ void task::wait_for_all() {
       scheduler->try_execute_stolen();
       scheduler->try_execute_stolen();
     }
     }
   }
   }
-  thread_state::get()->deque_.release_memory_until(deque_state_);
+  thread_state::get()->deque_.reset_offset(deque_offset_);
 }
 }
 
 
 }
 }

test/data_structures_test.cpp

@@ -85,77 +85,98 @@ TEST_CASE("work_stealing_deque functions correctly", "[internal/data_structures/
     int one = 1, two = 2, three = 3, four = 4;
     int one = 1, two = 2, three = 3, four = 4;
 
 
     SECTION("add and remove items form the tail") {
     SECTION("add and remove items form the tail") {
-      deque.push_tail<int>(one);
-      deque.push_tail<int>(two);
-      deque.push_tail<int>(three);
-
-      REQUIRE(*deque.pop_tail() == three);
-      REQUIRE(*deque.pop_tail() == two);
-      REQUIRE(*deque.pop_tail() == one);
+      deque.push_task<int>(one);
+      deque.publish_last_task();
+      deque.push_task<int>(two);
+      deque.publish_last_task();
+      deque.push_task<int>(three);
+      deque.publish_last_task();
+
+      REQUIRE(*deque.pop_local_task() == three);
+      REQUIRE(*deque.pop_local_task() == two);
+      REQUIRE(*deque.pop_local_task() == one);
     }
     }
 
 
     SECTION("handles getting empty by popping the tail correctly") {
     SECTION("handles getting empty by popping the tail correctly") {
-      deque.push_tail<int>(one);
-      REQUIRE(*deque.pop_tail() == one);
+      deque.push_task<int>(one);
+      deque.publish_last_task();
+      REQUIRE(*deque.pop_local_task() == one);
 
 
-      deque.push_tail<int>(two);
-      REQUIRE(*deque.pop_tail() == two);
+      deque.push_task<int>(two);
+      deque.publish_last_task();
+      REQUIRE(*deque.pop_local_task() == two);
     }
     }
 
 
     SECTION("remove items form the head") {
     SECTION("remove items form the head") {
-      deque.push_tail<int>(one);
-      deque.push_tail<int>(two);
-      deque.push_tail<int>(three);
-
-      REQUIRE(*deque.pop_head() == one);
-      REQUIRE(*deque.pop_head() == two);
-      REQUIRE(*deque.pop_head() == three);
+      deque.push_task<int>(one);
+      deque.publish_last_task();
+      deque.push_task<int>(two);
+      deque.publish_last_task();
+      deque.push_task<int>(three);
+      deque.publish_last_task();
+
+      REQUIRE(*deque.pop_external_task() == one);
+      REQUIRE(*deque.pop_external_task() == two);
+      REQUIRE(*deque.pop_external_task() == three);
     }
     }
 
 
     SECTION("handles getting empty by popping the head correctly") {
     SECTION("handles getting empty by popping the head correctly") {
-      deque.push_tail<int>(one);
-      REQUIRE(*deque.pop_head() == one);
+      deque.push_task<int>(one);
+      deque.publish_last_task();
+      REQUIRE(*deque.pop_external_task() == one);
 
 
-      deque.push_tail<int>(two);
-      REQUIRE(*deque.pop_head() == two);
+      deque.push_task<int>(two);
+      deque.publish_last_task();
+      REQUIRE(*deque.pop_external_task() == two);
     }
     }
 
 
     SECTION("handles getting empty by popping the head and tail correctly") {
     SECTION("handles getting empty by popping the head and tail correctly") {
-      deque.push_tail<int>(one);
-      REQUIRE(*deque.pop_tail() == one);
+      deque.push_task<int>(one);
+      deque.publish_last_task();
+      REQUIRE(*deque.pop_local_task() == one);
 
 
-      deque.push_tail<int>(two);
-      REQUIRE(*deque.pop_head() == two);
+      deque.push_task<int>(two);
+      deque.publish_last_task();
+      REQUIRE(*deque.pop_external_task() == two);
 
 
-      deque.push_tail<int>(three);
-      REQUIRE(*deque.pop_tail() == three);
+      deque.push_task<int>(three);
+      deque.publish_last_task();
+      REQUIRE(*deque.pop_local_task() == three);
     }
     }
 
 
     SECTION("handles jumps bigger 1 correctly") {
     SECTION("handles jumps bigger 1 correctly") {
-      deque.push_tail<int>(one);
-      deque.push_tail<int>(two);
-      REQUIRE(*deque.pop_tail() == two);
-
-      deque.push_tail<int>(three);
-      deque.push_tail<int>(four);
-      REQUIRE(*deque.pop_head() == one);
-      REQUIRE(*deque.pop_head() == three);
-      REQUIRE(*deque.pop_head() == four);
+      deque.push_task<int>(one);
+      deque.publish_last_task();
+      deque.push_task<int>(two);
+      deque.publish_last_task();
+      REQUIRE(*deque.pop_local_task() == two);
+
+      deque.push_task<int>(three);
+      deque.publish_last_task();
+      deque.push_task<int>(four);
+      deque.publish_last_task();
+      REQUIRE(*deque.pop_external_task() == one);
+      REQUIRE(*deque.pop_external_task() == three);
+      REQUIRE(*deque.pop_external_task() == four);
     }
     }
 
 
     SECTION("handles stack reset 1 correctly when emptied by tail") {
     SECTION("handles stack reset 1 correctly when emptied by tail") {
-      deque.push_tail<int>(one);
-      auto state = deque.save_state();
-      deque.push_tail<int>(two);
-      REQUIRE(*deque.pop_tail() == two);
-
-      deque.release_memory_until(state);
-      REQUIRE(*deque.pop_tail() == one);
-
-      deque.push_tail<int>(three);
-      deque.push_tail<int>(four);
-      REQUIRE(*deque.pop_head() == three);
-      REQUIRE(*deque.pop_tail() == four);
+      deque.push_task<int>(one);
+      deque.publish_last_task();
+      auto state = deque.save_offset();
+      deque.push_task<int>(two);
+      deque.publish_last_task();
+      REQUIRE(*deque.pop_local_task() == two);
+
+      deque.reset_offset(state);
+      REQUIRE(*deque.pop_local_task() == one);
+
+      deque.push_task<int>(three);
+      deque.publish_last_task();
+      deque.push_task<int>(four);
+      deque.publish_last_task();
+      REQUIRE(*deque.pop_external_task() == three);
+      REQUIRE(*deque.pop_local_task() == four);
     }
     }
   }
   }
 }
 }