Add special task handler to improve single threaded case.

CMakeLists.txt

@@ -7,7 +7,10 @@ find_package(Threads REQUIRED)
 add_executable(Sorter framework/runner.cpp
   src/container.cpp src/container.hpp
   src/sorter.cpp src/sorter.hpp
-  src/thread_pool.cpp src/thread_pool.hpp)
+  src/thread_pool.cpp src/thread_pool.hpp
+  src/single_task_handler.cpp src/single_task_handler.hpp
+  src/task_handler.hpp
+  )
 target_link_libraries(Sorter PUBLIC Threads::Threads)
 
 target_compile_features(Sorter PRIVATE cxx_std_20)
@@ -53,6 +56,9 @@ add_custom_command(
     ${CMAKE_CURRENT_LIST_DIR}/src/sorter.hpp
     ${CMAKE_CURRENT_LIST_DIR}/src/thread_pool.cpp
     ${CMAKE_CURRENT_LIST_DIR}/src/thread_pool.hpp
+    ${CMAKE_CURRENT_LIST_DIR}/src/src/single_task_handler.cpp
+    ${CMAKE_CURRENT_LIST_DIR}/src/src/single_task_handler.hpp
+    ${CMAKE_CURRENT_LIST_DIR}/src/src/task_handler.hpp
     ${plot_file}
     ${CMAKE_CURRENT_LIST_DIR}/description.md
   COMMENT "Creating submission"
@@ -76,6 +82,9 @@ set(dist_file_list
   src/sorter.hpp
   src/thread_pool.cpp
   src/thread_pool.hpp
+  src/single_task_handler.cpp 
+  src/single_task_handler.hpp
+  src/task_handler.hpp
 )
 
 set(framework_dist_file "ae-sorting.zip")

result.txt

@@ -1,30 +1,30 @@
-RESULT name=sort n=100 t=1 iterations=18 durationNanoseconds=57856 totalDurationNanoseconds=1041410 constructorNanoseconds=513 totalConstructorNanoseconds=9250
+RESULT name=sort n=100 t=1 iterations=95 durationNanoseconds=10561 totalDurationNanoseconds=1003361 constructorNanoseconds=460 totalConstructorNanoseconds=43740
-RESULT name=sort n=1000 t=1 iterations=8 durationNanoseconds=126883 totalDurationNanoseconds=1015070 constructorNanoseconds=1475 totalConstructorNanoseconds=11800
+RESULT name=sort n=1000 t=1 iterations=22 durationNanoseconds=46203 totalDurationNanoseconds=1016480 constructorNanoseconds=764 totalConstructorNanoseconds=16820
-RESULT name=sort n=10001 t=1 iterations=2 durationNanoseconds=615565 totalDurationNanoseconds=1231131 constructorNanoseconds=30890 totalConstructorNanoseconds=61780
+RESULT name=sort n=10001 t=1 iterations=3 durationNanoseconds=459633 totalDurationNanoseconds=1378901 constructorNanoseconds=18656 totalConstructorNanoseconds=55970
-RESULT name=sort n=100000 t=1 iterations=1 durationNanoseconds=9488107 totalDurationNanoseconds=9488107 constructorNanoseconds=524961 totalConstructorNanoseconds=524961
+RESULT name=sort n=100000 t=1 iterations=1 durationNanoseconds=8231415 totalDurationNanoseconds=8231415 constructorNanoseconds=514381 totalConstructorNanoseconds=514381
-RESULT name=sort n=10000000 t=1 iterations=1 durationNanoseconds=1416091993 totalDurationNanoseconds=1416091993 constructorNanoseconds=50440746 totalConstructorNanoseconds=50440746
+RESULT name=sort n=10000000 t=1 iterations=1 durationNanoseconds=744634713 totalDurationNanoseconds=744634713 constructorNanoseconds=51303216 totalConstructorNanoseconds=51303216
-RESULT name=sort n=100 t=2 iterations=16 durationNanoseconds=63258 totalDurationNanoseconds=1012141 constructorNanoseconds=375 totalConstructorNanoseconds=6000
+RESULT name=sort n=100 t=2 iterations=18 durationNanoseconds=58551 totalDurationNanoseconds=1053931 constructorNanoseconds=361 totalConstructorNanoseconds=6510
-RESULT name=sort n=1000 t=2 iterations=7 durationNanoseconds=154110 totalDurationNanoseconds=1078770 constructorNanoseconds=2008 totalConstructorNanoseconds=14060
+RESULT name=sort n=1000 t=2 iterations=9 durationNanoseconds=120015 totalDurationNanoseconds=1080141 constructorNanoseconds=1345 totalConstructorNanoseconds=12110
-RESULT name=sort n=10001 t=2 iterations=3 durationNanoseconds=451387 totalDurationNanoseconds=1354161 constructorNanoseconds=19620 totalConstructorNanoseconds=58860
+RESULT name=sort n=10001 t=2 iterations=3 durationNanoseconds=495777 totalDurationNanoseconds=1487331 constructorNanoseconds=19473 totalConstructorNanoseconds=58420
-RESULT name=sort n=100000 t=2 iterations=1 durationNanoseconds=6236655 totalDurationNanoseconds=6236655 constructorNanoseconds=514650 totalConstructorNanoseconds=514650
+RESULT name=sort n=100000 t=2 iterations=1 durationNanoseconds=7827986 totalDurationNanoseconds=7827986 constructorNanoseconds=508510 totalConstructorNanoseconds=508510
-RESULT name=sort n=10000000 t=2 iterations=1 durationNanoseconds=1380325518 totalDurationNanoseconds=1380325518 constructorNanoseconds=50373886 totalConstructorNanoseconds=50373886
+RESULT name=sort n=10000000 t=2 iterations=1 durationNanoseconds=1296871738 totalDurationNanoseconds=1296871738 constructorNanoseconds=49085315 totalConstructorNanoseconds=49085315
-RESULT name=sort n=100 t=4 iterations=9 durationNanoseconds=118743 totalDurationNanoseconds=1068691 constructorNanoseconds=436 totalConstructorNanoseconds=3930
+RESULT name=sort n=100 t=4 iterations=9 durationNanoseconds=118596 totalDurationNanoseconds=1067371 constructorNanoseconds=455 totalConstructorNanoseconds=4100
-RESULT name=sort n=1000 t=4 iterations=4 durationNanoseconds=272115 totalDurationNanoseconds=1088461 constructorNanoseconds=2415 totalConstructorNanoseconds=9660
+RESULT name=sort n=1000 t=4 iterations=5 durationNanoseconds=220166 totalDurationNanoseconds=1100831 constructorNanoseconds=2204 totalConstructorNanoseconds=11020
-RESULT name=sort n=10001 t=4 iterations=2 durationNanoseconds=569255 totalDurationNanoseconds=1138510 constructorNanoseconds=29920 totalConstructorNanoseconds=59840
+RESULT name=sort n=10001 t=4 iterations=2 durationNanoseconds=502925 totalDurationNanoseconds=1005851 constructorNanoseconds=29945 totalConstructorNanoseconds=59890
-RESULT name=sort n=100000 t=4 iterations=1 durationNanoseconds=6598125 totalDurationNanoseconds=6598125 constructorNanoseconds=507180 totalConstructorNanoseconds=507180
+RESULT name=sort n=100000 t=4 iterations=1 durationNanoseconds=5620814 totalDurationNanoseconds=5620814 constructorNanoseconds=528570 totalConstructorNanoseconds=528570
-RESULT name=sort n=10000000 t=4 iterations=1 durationNanoseconds=1300242690 totalDurationNanoseconds=1300242690 constructorNanoseconds=50475097 totalConstructorNanoseconds=50475097
+RESULT name=sort n=10000000 t=4 iterations=1 durationNanoseconds=1198704968 totalDurationNanoseconds=1198704968 constructorNanoseconds=50132516 totalConstructorNanoseconds=50132516
-RESULT name=sort n=100 t=8 iterations=3 durationNanoseconds=347863 totalDurationNanoseconds=1043591 constructorNanoseconds=2706 totalConstructorNanoseconds=8120
+RESULT name=sort n=100 t=8 iterations=5 durationNanoseconds=222358 totalDurationNanoseconds=1111791 constructorNanoseconds=1116 totalConstructorNanoseconds=5580
-RESULT name=sort n=1000 t=8 iterations=2 durationNanoseconds=610620 totalDurationNanoseconds=1221241 constructorNanoseconds=10400 totalConstructorNanoseconds=20800
+RESULT name=sort n=1000 t=8 iterations=3 durationNanoseconds=457883 totalDurationNanoseconds=1373650 constructorNanoseconds=3950 totalConstructorNanoseconds=11851
-RESULT name=sort n=10001 t=8 iterations=2 durationNanoseconds=706495 totalDurationNanoseconds=1412991 constructorNanoseconds=29600 totalConstructorNanoseconds=59200
+RESULT name=sort n=10001 t=8 iterations=2 durationNanoseconds=705680 totalDurationNanoseconds=1411361 constructorNanoseconds=30140 totalConstructorNanoseconds=60280
-RESULT name=sort n=100000 t=8 iterations=1 durationNanoseconds=7387085 totalDurationNanoseconds=7387085 constructorNanoseconds=557391 totalConstructorNanoseconds=557391
+RESULT name=sort n=100000 t=8 iterations=1 durationNanoseconds=6234564 totalDurationNanoseconds=6234564 constructorNanoseconds=530141 totalConstructorNanoseconds=530141
-RESULT name=sort n=10000000 t=8 iterations=1 durationNanoseconds=1261560682 totalDurationNanoseconds=1261560682 constructorNanoseconds=49470756 totalConstructorNanoseconds=49470756
+RESULT name=sort n=10000000 t=8 iterations=1 durationNanoseconds=1219668902 totalDurationNanoseconds=1219668902 constructorNanoseconds=50095186 totalConstructorNanoseconds=50095186
-RESULT name=sort n=100 t=12 iterations=3 durationNanoseconds=432037 totalDurationNanoseconds=1296111 constructorNanoseconds=1170 totalConstructorNanoseconds=3510
+RESULT name=sort n=100 t=12 iterations=3 durationNanoseconds=336683 totalDurationNanoseconds=1010050 constructorNanoseconds=1553 totalConstructorNanoseconds=4660
-RESULT name=sort n=1000 t=12 iterations=1 durationNanoseconds=1092461 totalDurationNanoseconds=1092461 constructorNanoseconds=12880 totalConstructorNanoseconds=12880
+RESULT name=sort n=1000 t=12 iterations=2 durationNanoseconds=709845 totalDurationNanoseconds=1419691 constructorNanoseconds=3685 totalConstructorNanoseconds=7370
-RESULT name=sort n=10001 t=12 iterations=1 durationNanoseconds=1019941 totalDurationNanoseconds=1019941 constructorNanoseconds=54540 totalConstructorNanoseconds=54540
+RESULT name=sort n=10001 t=12 iterations=1 durationNanoseconds=1008971 totalDurationNanoseconds=1008971 constructorNanoseconds=45810 totalConstructorNanoseconds=45810
-RESULT name=sort n=100000 t=12 iterations=1 durationNanoseconds=7159465 totalDurationNanoseconds=7159465 constructorNanoseconds=536730 totalConstructorNanoseconds=536730
+RESULT name=sort n=100000 t=12 iterations=1 durationNanoseconds=4991544 totalDurationNanoseconds=4991544 constructorNanoseconds=494530 totalConstructorNanoseconds=494530
-RESULT name=sort n=10000000 t=12 iterations=1 durationNanoseconds=1503813105 totalDurationNanoseconds=1503813105 constructorNanoseconds=50150056 totalConstructorNanoseconds=50150056
+RESULT name=sort n=10000000 t=12 iterations=1 durationNanoseconds=1295272477 totalDurationNanoseconds=1295272477 constructorNanoseconds=50175095 totalConstructorNanoseconds=50175095
-RESULT name=sort n=100 t=16 iterations=3 durationNanoseconds=432706 totalDurationNanoseconds=1298120 constructorNanoseconds=3833 totalConstructorNanoseconds=11500
+RESULT name=sort n=100 t=16 iterations=2 durationNanoseconds=668925 totalDurationNanoseconds=1337851 constructorNanoseconds=3290 totalConstructorNanoseconds=6580
-RESULT name=sort n=1000 t=16 iterations=2 durationNanoseconds=784875 totalDurationNanoseconds=1569751 constructorNanoseconds=5285 totalConstructorNanoseconds=10570
+RESULT name=sort n=1000 t=16 iterations=2 durationNanoseconds=828375 totalDurationNanoseconds=1656751 constructorNanoseconds=5660 totalConstructorNanoseconds=11320
-RESULT name=sort n=10001 t=16 iterations=1 durationNanoseconds=1953311 totalDurationNanoseconds=1953311 constructorNanoseconds=59420 totalConstructorNanoseconds=59420
+RESULT name=sort n=10001 t=16 iterations=1 durationNanoseconds=1319491 totalDurationNanoseconds=1319491 constructorNanoseconds=65160 totalConstructorNanoseconds=65160
-RESULT name=sort n=100000 t=16 iterations=1 durationNanoseconds=6820104 totalDurationNanoseconds=6820104 constructorNanoseconds=524961 totalConstructorNanoseconds=524961
+RESULT name=sort n=100000 t=16 iterations=1 durationNanoseconds=7145805 totalDurationNanoseconds=7145805 constructorNanoseconds=529550 totalConstructorNanoseconds=529550
-RESULT name=sort n=10000000 t=16 iterations=1 durationNanoseconds=1352546567 totalDurationNanoseconds=1352546567 constructorNanoseconds=52707158 totalConstructorNanoseconds=52707158
+RESULT name=sort n=10000000 t=16 iterations=1 durationNanoseconds=1483766932 totalDurationNanoseconds=1483766932 constructorNanoseconds=52444497 totalConstructorNanoseconds=52444497

result.withouttaskhandler.txt

@@ -0,0 +1,30 @@
+RESULT name=sort n=100 t=1 iterations=18 durationNanoseconds=57856 totalDurationNanoseconds=1041410 constructorNanoseconds=513 totalConstructorNanoseconds=9250
+RESULT name=sort n=1000 t=1 iterations=8 durationNanoseconds=126883 totalDurationNanoseconds=1015070 constructorNanoseconds=1475 totalConstructorNanoseconds=11800
+RESULT name=sort n=10001 t=1 iterations=2 durationNanoseconds=615565 totalDurationNanoseconds=1231131 constructorNanoseconds=30890 totalConstructorNanoseconds=61780
+RESULT name=sort n=100000 t=1 iterations=1 durationNanoseconds=9488107 totalDurationNanoseconds=9488107 constructorNanoseconds=524961 totalConstructorNanoseconds=524961
+RESULT name=sort n=10000000 t=1 iterations=1 durationNanoseconds=1416091993 totalDurationNanoseconds=1416091993 constructorNanoseconds=50440746 totalConstructorNanoseconds=50440746
+RESULT name=sort n=100 t=2 iterations=16 durationNanoseconds=63258 totalDurationNanoseconds=1012141 constructorNanoseconds=375 totalConstructorNanoseconds=6000
+RESULT name=sort n=1000 t=2 iterations=7 durationNanoseconds=154110 totalDurationNanoseconds=1078770 constructorNanoseconds=2008 totalConstructorNanoseconds=14060
+RESULT name=sort n=10001 t=2 iterations=3 durationNanoseconds=451387 totalDurationNanoseconds=1354161 constructorNanoseconds=19620 totalConstructorNanoseconds=58860
+RESULT name=sort n=100000 t=2 iterations=1 durationNanoseconds=6236655 totalDurationNanoseconds=6236655 constructorNanoseconds=514650 totalConstructorNanoseconds=514650
+RESULT name=sort n=10000000 t=2 iterations=1 durationNanoseconds=1380325518 totalDurationNanoseconds=1380325518 constructorNanoseconds=50373886 totalConstructorNanoseconds=50373886
+RESULT name=sort n=100 t=4 iterations=9 durationNanoseconds=118743 totalDurationNanoseconds=1068691 constructorNanoseconds=436 totalConstructorNanoseconds=3930
+RESULT name=sort n=1000 t=4 iterations=4 durationNanoseconds=272115 totalDurationNanoseconds=1088461 constructorNanoseconds=2415 totalConstructorNanoseconds=9660
+RESULT name=sort n=10001 t=4 iterations=2 durationNanoseconds=569255 totalDurationNanoseconds=1138510 constructorNanoseconds=29920 totalConstructorNanoseconds=59840
+RESULT name=sort n=100000 t=4 iterations=1 durationNanoseconds=6598125 totalDurationNanoseconds=6598125 constructorNanoseconds=507180 totalConstructorNanoseconds=507180
+RESULT name=sort n=10000000 t=4 iterations=1 durationNanoseconds=1300242690 totalDurationNanoseconds=1300242690 constructorNanoseconds=50475097 totalConstructorNanoseconds=50475097
+RESULT name=sort n=100 t=8 iterations=3 durationNanoseconds=347863 totalDurationNanoseconds=1043591 constructorNanoseconds=2706 totalConstructorNanoseconds=8120
+RESULT name=sort n=1000 t=8 iterations=2 durationNanoseconds=610620 totalDurationNanoseconds=1221241 constructorNanoseconds=10400 totalConstructorNanoseconds=20800
+RESULT name=sort n=10001 t=8 iterations=2 durationNanoseconds=706495 totalDurationNanoseconds=1412991 constructorNanoseconds=29600 totalConstructorNanoseconds=59200
+RESULT name=sort n=100000 t=8 iterations=1 durationNanoseconds=7387085 totalDurationNanoseconds=7387085 constructorNanoseconds=557391 totalConstructorNanoseconds=557391
+RESULT name=sort n=10000000 t=8 iterations=1 durationNanoseconds=1261560682 totalDurationNanoseconds=1261560682 constructorNanoseconds=49470756 totalConstructorNanoseconds=49470756
+RESULT name=sort n=100 t=12 iterations=3 durationNanoseconds=432037 totalDurationNanoseconds=1296111 constructorNanoseconds=1170 totalConstructorNanoseconds=3510
+RESULT name=sort n=1000 t=12 iterations=1 durationNanoseconds=1092461 totalDurationNanoseconds=1092461 constructorNanoseconds=12880 totalConstructorNanoseconds=12880
+RESULT name=sort n=10001 t=12 iterations=1 durationNanoseconds=1019941 totalDurationNanoseconds=1019941 constructorNanoseconds=54540 totalConstructorNanoseconds=54540
+RESULT name=sort n=100000 t=12 iterations=1 durationNanoseconds=7159465 totalDurationNanoseconds=7159465 constructorNanoseconds=536730 totalConstructorNanoseconds=536730
+RESULT name=sort n=10000000 t=12 iterations=1 durationNanoseconds=1503813105 totalDurationNanoseconds=1503813105 constructorNanoseconds=50150056 totalConstructorNanoseconds=50150056
+RESULT name=sort n=100 t=16 iterations=3 durationNanoseconds=432706 totalDurationNanoseconds=1298120 constructorNanoseconds=3833 totalConstructorNanoseconds=11500
+RESULT name=sort n=1000 t=16 iterations=2 durationNanoseconds=784875 totalDurationNanoseconds=1569751 constructorNanoseconds=5285 totalConstructorNanoseconds=10570
+RESULT name=sort n=10001 t=16 iterations=1 durationNanoseconds=1953311 totalDurationNanoseconds=1953311 constructorNanoseconds=59420 totalConstructorNanoseconds=59420
+RESULT name=sort n=100000 t=16 iterations=1 durationNanoseconds=6820104 totalDurationNanoseconds=6820104 constructorNanoseconds=524961 totalConstructorNanoseconds=524961
+RESULT name=sort n=10000000 t=16 iterations=1 durationNanoseconds=1352546567 totalDurationNanoseconds=1352546567 constructorNanoseconds=52707158 totalConstructorNanoseconds=52707158

src/single_task_handler.cpp

@@ -0,0 +1,11 @@
+#include "single_task_handler.hpp"
+
+void SingleTaskHandler::add(std::function<void()> task) {
+    task();
+}
+uint32_t SingleTaskHandler::size() {
+    return 0;
+}
+bool SingleTaskHandler::isWorking() {
+    return false;
+}

src/single_task_handler.hpp

@@ -0,0 +1,8 @@
+#pragma once
+#include "task_handler.hpp"
+
+class SingleTaskHandler: TaskHandler {
+    void add(std::function<void()> task);
+    uint32_t size();
+    bool isWorking();
+};

src/sorter.cpp

@@ -1,5 +1,8 @@
 #include "sorter.hpp"
 
+#include "single_task_handler.hpp"
+#include "thread_pool.hpp"
+
 #include <algorithm>
 #include <iterator>
 #include <cassert>
@@ -13,7 +16,11 @@ namespace ae {
 
 sorter::sorter(uint32_t num = 1) {
   sorter::num_threads = num;
-  sorter::pool = new ThreadPool(num);
+  if (num > 1) {
+    sorter::handler = (TaskHandler*) new ThreadPool(num);
+  } else {
+    sorter::handler = (TaskHandler*) new SingleTaskHandler();
+  }
 }
 
 void sorter::sort(container& data) {
@@ -32,7 +39,7 @@ void sorter::sort(container& data) {
     std::sort(copy.begin(), copy.end());
   #endif
   sorter::msd_inplace_radix_sort(data.placeholder_[0], 0, [&](auto span) {sorter::robin_hood_sort(span);});
-  while (sorter::pool->size() > 0 || sorter::pool->isWorking()) {};
+  while (sorter::handler->size() > 0 || sorter::handler->isWorking()) {};
   #if DEBUG
     for (int i = 0; i < copy.size(); i++) {
       if (copy[i] != data.placeholder_[0][i])
@@ -175,24 +182,20 @@ void sorter::msd_inplace_radix_sort(
 
   // sort each bucket recursively
   for (auto i = 0; i < sorter::RADIX_BUCKETS; i++) {
-    if (sorter::pool != nullptr) {
+    #if DEBUG
+      std::cerr << "Putting in task with depth " << passes << " of bucket " << i << std::endl;
+    #endif
+    auto start = buckets_start[i];
+    auto end = buckets_end[i];
+    sorter::handler->add([start, end, &bucket_sort, passes, this, i](){
       #if DEBUG
-        std::cerr << "Putting in task with depth " << passes << " of bucket " << i << std::endl;
+        std::cerr << "Starting task with depth " << passes << " of bucket " << i << std::endl;
       #endif
-      auto start = buckets_start[i];
+      sorter::msd_inplace_radix_sort(std::span<container::element_type> (start, end), passes + 1, bucket_sort);
-      auto end = buckets_end[i];
+      #if DEBUG
-      sorter::pool->add([start, end, &bucket_sort, passes, this, i](){
+        std::cerr << "Finishing task with depth " << passes << " of bucket " << i << std::endl;
-        #if DEBUG
+      #endif
-          std::cerr << "Starting task with depth " << passes << " of bucket " << i << std::endl;
+    });
-        #endif
-        sorter::msd_inplace_radix_sort(std::span<container::element_type> (start, end), passes + 1, bucket_sort);
-        #if DEBUG
-          std::cerr << "Finishing task with depth " << passes << " of bucket " << i << std::endl;
-        #endif
-      });
-    } else {
-      sorter::msd_inplace_radix_sort(std::span<container::element_type> (buckets_start[i], buckets_end[i]), passes + 1, bucket_sort);
-    }
   }
 }
 

src/sorter.hpp

@@ -1,7 +1,7 @@
 #pragma once
 
 #include "container.hpp"
-#include "thread_pool.hpp"
+#include "task_handler.hpp"
 #include "functional"
 #include "math.h"
 #include <mutex>
@@ -40,7 +40,7 @@ class sorter {
   const uint32_t RADIX_BUCKETS = std::pow(2, 4);
   const uint32_t RADIX_ITERATIONS = 8;
   uint32_t num_threads;
-  ThreadPool* pool = nullptr;
+  TaskHandler* handler = nullptr;
 
   void robin_hood_sort(std::span<container::element_type> range);
 };

src/task_handler.hpp

@@ -0,0 +1,11 @@
+#pragma once
+
+#include <functional>
+#include <cstdint>
+
+class TaskHandler {
+    public:
+        virtual void add(std::function<void()> task) = 0;
+        virtual uint32_t size() = 0;
+        virtual bool isWorking() = 0;
+};

src/thread_pool.hpp

@@ -1,5 +1,6 @@
 #pragma once
 
+#include "task_handler.hpp"
 #include <algorithm>
 #include <vector>
 #include <mutex>
@@ -11,7 +12,7 @@
 // This class was inspired by https://www.geeksforgeeks.org/cpp/thread-pool-in-cpp/ (access: 26/09/2025)
 // to more efficiently handle threads
 
-class ThreadPool {
+class ThreadPool: TaskHandler {
 private:
     std::vector<std::thread> threads;
     std::queue<std::function<void()>> tasks;