remove .sh in favor of .bat. Do some testing and verify that running w/ 15 threads does indeed use the whole cpu (via btop)

.gitignore

@@ -1,3 +1,5 @@
+*.idb
+*.pdb
 **obj**
 **bin**
 .cache**

activate.bat

@@ -1 +1 @@
-"C:\Program Files\Microsoft Visual Studio\2022\Community\VC\Auxiliary\Build\vcvars64.bat" && bash
+"C:\Program Files\Microsoft Visual Studio\2022\Community\VC\Auxiliary\Build\vcvars64.bat"

build.bat

@@ -0,0 +1,10 @@
+@ECHO off
+SET flags=-Od -ZI -MTd
+SET config=Debug
+IF "%1" == "release" (
+    SET flags=-O2 -MT
+    SET config=Release
+)
+SET srcs=src\*
+mkdir bin obj
+cl %srcs% -I inc %flags% -std:c++20 -MP -Fo:obj\\ -Fe:bin\\

build.sh

@@ -1,13 +0,0 @@
-(
-    cd bin
-    rm main.*
-    srcs=../src/*
-    if [ $# -eq 1 ] && [ "$1" == "release" ]
-    then
-	flags="-O2"
-    else
-	flags="-Od -ZI"
-    fi
-    echo $flags
-    cl $srcs -I ../inc/ $flags -std:c++20 -Fe 
-)

debug.rad

@@ -1,14 +1,31 @@
 // raddbg 0.9.21 project file
 
+recent_file: path: "C:/Program Files (x86)/Windows Kits/10/include/10.0.26100.0/ucrt/stdio.h"
+recent_file: path: "d:/os/obj/amd64fre/minkernel/crts/ucrt/src/appcrt/stdio/xmt/objfre/amd64/minkernel/crts/ucrt/src/appcrt/stdio/output.cpp"
+recent_file: path: "C:/Program Files/Microsoft Visual Studio/2022/Community/VC/Tools/MSVC/14.44.35207/include/xutility"
+recent_file: path: "C:/Users/sethh/Documents/repos/Petri/inc/genetic.h"
+recent_file: path: "C:/Program Files/Microsoft Visual Studio/2022/Community/VC/Tools/MSVC/14.44.35207/include/algorithm"
+recent_file: path: "C:/Users/sethh/Documents/repos/Petri/inc/sync.h"
 recent_file: path: "inc/genetic.h"
 recent_file: path: "inc/sync.h"
+recent_file: path: "d:/os/obj/amd64fre/minkernel/crts/ucrt/src/appcrt/startup/mt/objfre/amd64/minkernel/crts/ucrt/src/appcrt/startup/abort.cpp"
+recent_file: path: "src/main.cpp"
+recent_file: path: "d:/os/obj/amd64fre/minkernel/crts/ucrt/src/appcrt/startup/mt/objfre/amd64/minkernel/crts/ucrt/src/appcrt/startup/assert.cpp"
+recent_file: path: "../../../../../Program Files/Microsoft Visual Studio/2022/Community/VC/Tools/MSVC/14.44.35207/include/vector"
+recent_file: path: "d:/os/obj/amd64fre/minkernel/crts/ucrt/src/appcrt/misc/mt/objfre/amd64/minkernel/crts/ucrt/src/appcrt/misc/invalid_parameter.cpp"
+recent_file: path: "../../../../../Program Files/Microsoft Visual Studio/2022/Community/VC/Tools/MSVC/14.44.35207/include/xmemory"
 recent_file: path: "../../../../../Program Files/Microsoft Visual Studio/2022/Community/VC/Tools/MSVC/14.42.34433/include/algorithm"
 recent_file: path: "../../../../../Program Files/Microsoft Visual Studio/2022/Community/VC/Tools/MSVC/14.42.34433/include/xutility"
-recent_file: path: "src/main.cpp"
 target:
 {
   executable: "bin/main.exe"
   working_directory: bin
   label: main
   enabled: 1
+  arguments: 1
+}
+breakpoint:
+{
+  source_location: "inc/genetic.h:292:1"
+  hit_count: 0
 }

export.bat

@@ -0,0 +1,22 @@
+@ECHO off
+SETLOCAL ENABLEDELAYEDEXPANSION
+
+ECHO [ > compile_commands.json
+
+FOR /r "src\" %%F IN (*.cpp) DO (
+
+SET "file=%%F"
+SET "file=!file:\=/!"
+SET "directory=%~dp0"
+SET "directory=!directory:\=/!"
+
+ECHO { >> compile_commands.json
+ECHO "directory": "!directory!", >> compile_commands.json
+ECHO "command": "cl !file! -I inc %flags% -std:c++20 -MP -Fo:obj\\ ", >> compile_commands.json
+
+ECHO "file": "!file!" >> compile_commands.json
+ECHO }, >> compile_commands.json
+
+)
+
+ECHO ] >> compile_commands.json

export_compdb.sh

@@ -1,5 +0,0 @@
-srcs=src/*
-
-echo [ > compile_commands.json
-find src -iname "*.cpp" -exec sh -c 'echo { \"directory\": \"$(cygpath -m $(pwd))\", \"command\": \"cl "$(cygpath -m {})" -I inc -Od -std:c++20 -Fo\", \"file\": \"$(cygpath -m {})\" }, >> compile_commands.json' \;
-echo ] >> compile_commands.json

inc/genetic.h

@@ -1,32 +1,67 @@
 #pragma once 
 
 #include <algorithm>
+#include <cfloat>
+#include <cstdarg>
+#include <cstdio>
 #include <cstdlib>
-#include <vector>
 
+#include "util.h"
 #include "sync.h"
 #include "rand.h"
 
 using namespace sync;
-using namespace std;
 
 namespace genetic {
 
-template <class T> struct Array;
 template <class T> struct Stats;
 template <class T> struct Strategy;
 struct CellTracker;
 
-template <class T> Stats<T> run(Strategy<T>);
+
+const char *global_stat_format_str = "GLOBAL, Progress %.1f%%, Top: %.5e, Overhead Per: %.4f%%, Gen: %.4f, Overhead: %.4f, Cross: %.4f (s), Mutate: %.4f (s), Fitness: %.4f (s), Sorting: %.4f (s)\n";
+const char *thread_stat_format_str = "%d, Progress %d/%d, Top: %.5e, Overhead Per: %.4f%%, Gen: %.4f, Overhead: %.4f, Cross: %.4f (s), Mutate: %.4f (s), Fitness: %.4f (s), Sorting: %.4f (s)\n";
+
+static int stat_str_len = 2*max(strlen(thread_stat_format_str), strlen(global_stat_format_str));
+static char *stat_str = (char*)malloc(stat_str_len);
+static char *filename = (char*)malloc(64);
+static int n_threads = 0;
+
+void log(const char *format_str, ...) {
+    va_list list;
+    va_start(list, format_str);
+    vsprintf_s(stat_str, 2*max(strlen(thread_stat_format_str), strlen(global_stat_format_str)), format_str, list);
+
+    printf("%s", stat_str);
+
+    FILE *f;
+    sprintf(filename, "logs/logs-%d.txt", n_threads);
+    fopen_s(&f, filename, "a");
+    fwrite(stat_str, sizeof(char), strlen(stat_str), f);
+    fclose(f);
+}
+
+template <class T> T run(Strategy<T>);
 
 template <class T> struct Strategy {
     // Number of worker threads that will be evaluating cell fitness
     int num_threads;
 
-    int batch_size; // Number of cells a worker thread tries to work on in a row
+    // Period of print statements (in seconds)
-                    // before accessing/locking the work queue again.
+    float stats_print_period_s;
-    int num_cells;  // Size of the population pool
+
-    int num_generations; // Number of times (epochs) to run the algorithm
+    // Size of the population pool per sim thread
+    int num_cells_per_thread;
+
+    // Number of times (epochs) to run the algorithm
+    int num_generations; 
+
+    // Each thread will integrate the best globally performing cell
+    bool share_breakthroughs;
+
+    // How many generations to explore before resyncing with the global best
+    int share_breakthrough_gen_period;
+
     bool test_all; // Sets whether or not every cell's fitness is evaluated every
                    // generation
     float test_chance; // Chance to test any given cell's fitness. Relevant only
@@ -59,10 +94,16 @@ template <class T> struct Strategy {
 };
 
 template<class T> struct Stats {
-    std::vector<T> best_cell;
+    DynArray<T> best_cells;
-    std::vector<float> best_cell_fitness;
+    DynArray<float> best_cell_fitness;
-    TimeSpan setup_time;
+    int gen;
-    TimeSpan run_time;
+    bool done;
+    DynArray<TimeSpan> gen_time;
+    DynArray<TimeSpan> crossover_time;
+    DynArray<TimeSpan> mutate_time;
+    DynArray<TimeSpan> fitness_time;
+    DynArray<TimeSpan> sorting_time;
+    Mutex m;
 };
 
 struct CellTracker {
@@ -70,301 +111,262 @@ struct CellTracker {
     int cellid;
 };
 
-template <class T> struct Array {
-    T *data;
-    int len;
-
-    T &operator[](int i) { return data[i]; }
-};
-
-template <class T> Array<T> make_array(int len) {
-    return {
-    .data = (T*)malloc(sizeof(T)*len),
-    .len = len
-     };
-}
-
-template<class T>
-struct MutateJob {
-    T* cell;
-};
-
-template<class T>
-struct CrossoverJob {
-    Array<T*> parents;
-    Array<T*> children;
-};
-
-template<class T>
-struct FitnessJob {
-    T* cell;
-    CellTracker* track;
-};
-
-enum class JobType {
-    MUTATE,
-    CROSSOVER,
-    FITNESS
-};
-
-template<class T>
-union Job {
-    MutateJob<T> m;
-    CrossoverJob<T> c;
-    FitnessJob<T> f;
-};
-
-// Yes. I am aware of variant
-// For some reason I like this better
-template<class T>
-struct TaggedJob {
-    Job<T> data;
-    JobType type;
-};
-
-template<class T>
-struct WorkQueue {
-    Array<TaggedJob<T>> jobs;
-    int read_i, write_i, batch_size;
-    bool done_writing, work_complete, stop; // These catch some edge conditions
-    Mutex m;
-    ConditionVar done;
-    ConditionVar jobs_ready;
-};
-
 template<class T>
 struct WorkerThreadArgs {
-    WorkQueue<T> &q;
+    Strategy<T> strat;
-    Strategy<T> &s;
+    Array<T> cells;
+    Array<CellTracker> trackers;
+    Stats<T> *stats;
+
+    Mutex m;
+    float *best_global_score;
+    T* best_global_cell;
 };
 
-template<class T> WorkQueue<T> make_work_queue(int len, int batch_size);
+template<class T> T* _cellp(Array<T> cells, CellTracker tracker) { return &cells[tracker.cellid]; }
-template<class T> bool tryget_job_batch(WorkQueue<T> &q, int len, Array<TaggedJob<T>>* out_batch, bool* out_batch_is_end);
 
-template<class T>
+template <class T> DWORD worker(LPVOID args) {
-DWORD worker(LPVOID args);
+    // Unpack everything...
+    WorkerThreadArgs<T>* worker_args = static_cast<WorkerThreadArgs<T>*>(args); 
+    Strategy<T> strat = worker_args->strat;
+    Array<T> cells = worker_args->cells;
+    Array<CellTracker> trackers = worker_args->trackers;
+    Stats<T> &stats = *worker_args->stats;
+    float* best_global_score = worker_args->best_global_score;
+    T* best_global_cell = worker_args->best_global_cell;
+    Mutex best_m = worker_args->m;
 
-template <class T> Stats<T> run(Strategy<T> strat) {
+    // Prepare crossover operations as these will be the same every time except
-    Stats<T> stats;
+    // for the exact cell pointers
-
-    // ************* SETUP **************
-    TimeSpan start_setup = now();
-
-    // Create cells
-    Array<T> cells = make_array<T>(strat.num_cells);
-    for (int i = 0; i < cells.len; i++) cells[i] = strat.make_default_cell();
-
-    // Create cell trackers
-    Array<CellTracker> trackers = make_array<CellTracker>(strat.num_cells);
-    for (int i = 0; i < trackers.len; i++) trackers[i] = { .score=0, .cellid=i };
-
-    // Create work queue
-    // Worst case size is every cell mutated, crossed, and evaluated...? Not quite, but 3x should be upper bound
-    WorkQueue<T> q = make_work_queue<T>(3*strat.num_cells, strat.batch_size);
-    WorkerThreadArgs<T> args = {q, strat};
-
-    // Create worker threads
-    Thread *threads = (Thread*)malloc(sizeof(Thread*)*strat.num_threads);
-    for (int i = 0; i < strat.num_threads; i++) {
-	threads[i] = make_thread(worker<T>, &args);
-    }
-
-    stats.setup_time = now() - start_setup;
-
-    // *********** ALGORITHM ************
-    TimeSpan start_algo = now();
-    for (int gen = 0; gen < strat.num_generations; gen++) {
-	// Reset work queue
-	lock(q.m);
-	q.read_i = 0;
-	q.write_i = 0;
-	q.work_complete = false;
-	q.done_writing = false;
-	unlock(q.m);
-
-	// 1. mutate
-	for (int i = 0; i < trackers.len; i++) {
-	    if (abs(norm_rand(strat.rand_seed)) < strat.mutation_chance) {
-		MutateJob<T> mj = {&cells[trackers[i].cellid]};
-		TaggedJob<T> job;
-		job.data.m = mj;
-		job.type=JobType::MUTATE;
-		q.jobs[q.write_i++] = job;
-	    }
-	}
-	wake_all(q.jobs_ready); // There are available jobs for the worker threads!
-
-	// 2. crossover
-	if (strat.enable_crossover) {
     int npar = strat.crossover_parent_num;
     int nchild = strat.crossover_children_num;
+    Array<T*> parents = make_array<T*>(npar);
+    Array<T*> children = make_array<T*>(nchild);
 
+    bool gt = strat.higher_fitness_is_better; // Writing strat.higher... is annoying
+
+    TimeSpan start, diff, gen_start;
+    while(stats.gen < strat.num_generations) {
+	gen_start = now();
+
+	// 0. Share/Integrate global breakthrough
+	if (strat.share_breakthroughs && (stats.gen + get_affinity()) % strat.share_breakthrough_gen_period) {
+	    lock(best_m);
+	    if (better(gt, front(trackers).score, *best_global_score) != *best_global_score) {
+		// Share
+		*best_global_cell = *_cellp(cells, trackers[0]);
+		*best_global_score = trackers[0].score;
+		
+	    } else {
+		// Integrate
+		*_cellp(cells, trackers[0]) = *best_global_cell;
+		trackers[0].score = *best_global_score;
+	    }
+	    unlock(best_m);
+	}
+
+	// 1. crossover
+	start = now();
+	if (strat.enable_crossover) {
 	    int parent_end = npar;
 	    int child_begin = trackers.len-nchild;
 	    while (parent_end <= child_begin) {
 
-		// TODO: Variable size arrays please. This is rediculous.
-		Array<T*> parents = make_array<T*>(npar);
-		Array<T*> children = make_array<T*>(nchild);
 		// Get pointers to all the parent cells
 		for (int i = parent_end-npar; i < parent_end; i++) {
-		    parents[i - (parent_end-npar)] = &cells[trackers[i].cellid];
+		    T* cell = _cellp(cells, trackers[i]);
+		    assert(cell != NULL);
+		    parents[i - (parent_end-npar)] = cell;
 		}
 
 		// Get pointers to all the child cells (these will be overwritten)
 		for (int i = child_begin; i < child_begin+nchild; i++) {
-		    children[i-child_begin] = &cells[trackers[i].cellid];
+		    T* cell = _cellp(cells, trackers[i]);
+		    assert(cell != NULL);
+		    children[i-child_begin] = cell;
 		}
-		CrossoverJob<T> cj = {parents, children};
+		strat.crossover(parents, children);
-		TaggedJob<T> job;
-		job.data.c=cj;
-		job.type=JobType::CROSSOVER;
-		q.jobs[q.write_i++] = job;
 		parent_end += strat.crossover_parent_stride;
 		child_begin -= nchild;
 	    }
-	    wake_all(q.jobs_ready); // There are available jobs for the worker threads!
 	}
+	lock(stats.m);
+	append(stats.crossover_time, now() - start);
+	unlock(stats.m);
+
+
+	// 2. mutate
+	start = now();
+	for (int i = 0; i < trackers.len; i++) {
+	    if (abs(norm_rand(strat.rand_seed)) < strat.mutation_chance) {
+		strat.mutate(cells[trackers[i].cellid]);
+	    }
+	}
+	lock(stats.m);
+	append(stats.mutate_time, now() - start);
+	unlock(stats.m);
 
 	// 3. evaluate
+	start = now();
 	if (strat.test_all) {
 	    for (int i = 0; i < trackers.len; i++) {
-		FitnessJob<T> fj = {&cells[trackers[i].cellid], &trackers[i]};
+		trackers[i].score = strat.fitness(cells[trackers[i].cellid]);
-		TaggedJob<T> job;
-		job.data.f=fj;
-		job.type=JobType::FITNESS;
-		if (i == trackers.len-1) lock(q.m);
-		q.jobs[q.write_i++] = job;
-		if (i == trackers.len-1) { q.done_writing = true; unlock(q.m); }
 	    }
 	} else {
-	    lock(q.m);
 	    for (int i = 0; i < trackers.len; i++) {
 		if (abs(norm_rand(strat.rand_seed)) < strat.test_chance) {
-		    FitnessJob<T> fj = {&cells[trackers[i].cellid], &trackers[i]};
+		    trackers[i].score = strat.fitness(cells[trackers[i].cellid]);
-		    TaggedJob<T> job;
-		    job.data.f=fj;
-		    job.type=JobType::FITNESS;
-		    q.jobs[q.write_i++] = job;
 		}
 	    }
-	    q.done_writing = true;
-	    unlock(q.m);
 	}
-	wake_all(q.jobs_ready);
+	lock(stats.m);
+	append(stats.fitness_time, now() - start);
+	unlock(stats.m);
 
-	// Wait until the work is finished
+	auto comp = [strat](CellTracker &a, CellTracker &b){
-	lock(q.m);
+	    return strat.higher_fitness_is_better ? (a.score > b.score) : (a.score < b.score);
-	if (!q.work_complete) 
+	};
-	    wait(q.done, q.m, infinite_ts);
-	unlock(q.m);
 
 	// 4. sort
-	std::sort(&trackers[0], &trackers[trackers.len-1], [strat](CellTracker &a, CellTracker &b){ return strat.higher_fitness_is_better ? a.score > b.score : a.score < b.score; });
+	start = now();
+	std::sort(&trackers[0], &trackers[trackers.len-1], comp);
+	lock(stats.m);
+	append(stats.sorting_time, now() - start);
 
-	printf("Gen: %d, Best Score: %f\n", gen, trackers[0].score);
+	append(stats.best_cells, cells[trackers[0].cellid]);
-	stats.best_cell.push_back(cells[trackers[0].cellid]);
+	append(stats.best_cell_fitness, trackers[0].score);
-	stats.best_cell_fitness.push_back(trackers[0].score);
+	append(stats.gen_time, now() - gen_start);
+	stats.gen++;
+	unlock(stats.m);
+    }
+    stats.done = true;
+    return 0;
 }
 
-    q.stop = true;
+template <class T> T run(Strategy<T> strat) {
-    wake_all(q.jobs_ready);
+    Array<Stats<T>> stats = make_array<Stats<T>>(strat.num_threads);
-    // TODO: join all threads
+    Array<Thread> threads = make_array<Thread>(strat.num_threads);
+    Array<WorkerThreadArgs<T>> args = make_array<WorkerThreadArgs<T>>(strat.num_threads);
 
-    // TODO: There's some data freeing that should really be done here
+    float best_global_score = strat.higher_fitness_is_better ? FLT_MIN : FLT_MAX;
-    stats.run_time = now() - start_algo;
+    T best_global_cell;
-    return stats;
-}
 
-template<class T> WorkQueue<T> make_work_queue(int len, int batch_size) {
+    allow_all_processors();
-    return {
+    set_affinity(0);
-	.jobs=make_array<TaggedJob<T>>(len),
+
-	.read_i=0,
+    for (int i = 0; i < strat.num_threads; i++) {
-	.write_i=0,
+	stats[i] = {
-	.batch_size=batch_size,
+	    .best_cells=make_dynarray<T>(strat.num_generations),
-	.done_writing=false,
+	    .best_cell_fitness=make_dynarray<float>(strat.num_generations),
-	.work_complete=false,
+	    .gen_time=make_dynarray<TimeSpan>(strat.num_generations),
-	.m=make_mutex(),
+	    .crossover_time=make_dynarray<TimeSpan>(strat.num_generations),
-	.done=make_condition_var(),
+	    .mutate_time=make_dynarray<TimeSpan>(strat.num_generations),
-	.jobs_ready=make_condition_var()
+	    .fitness_time=make_dynarray<TimeSpan>(strat.num_generations),
+	    .sorting_time=make_dynarray<TimeSpan>(strat.num_generations),
+	    .m=make_mutex()
 	};
+	Array<T> cells = make_array<T>(strat.num_threads*strat.num_cells_per_thread);
+	Array<CellTracker> trackers = make_array<CellTracker>(strat.num_cells_per_thread);
+	for (int i = 0; i < strat.num_cells_per_thread; i++) {
+	    cells[i] = strat.make_default_cell();
+	    trackers[i] = {0, i};
 	}
 
-template<class T> bool tryget_job_batch(WorkQueue<T> &q, Array<TaggedJob<T>>* out_batch, bool* out_batch_is_end) {
+	args[i].strat=strat;
-    lock(q.m);
+	args[i].cells=cells;
-    if (q.stop) {
+	args[i].trackers=trackers;
-	unlock(q.m);
+	args[i].stats=&stats[i];
-	return false;
+	args[i].best_global_score=&best_global_score;
+	args[i].best_global_cell=&best_global_cell;
+	args[i].m = make_mutex();
+
+	threads[i] = make_thread(worker<T>, &args[i], i+1);
     }
 
-    // Keep waiting till jobs are available
-    while (q.read_i >= q.write_i) {
-	wait(q.jobs_ready, q.m, infinite_ts); 
 
-	if (q.stop) {
+    // We are the stats thread
-	    unlock(q.m);
+    bool complete = false;
-	    return false;
+    while (!complete) {
+	sleep(from_s(strat.stats_print_period_s));
+
+	log("**********************\n");
+	float g_avg_gen_time = 0;
+	float g_avg_crossover_time = 0;
+	float g_avg_mutate_time = 0;
+	float g_avg_fitness_time = 0;
+	float g_avg_sorting_time = 0;
+	float g_avg_overhead_time = 0;
+	float g_progress_per = 0;
+	float g_best_fitness = strat.higher_fitness_is_better ? FLT_MIN : FLT_MAX;
+
+	complete = true;
+
+
+	for (int i = 0; i < stats.len; i++) {
+	    lock(stats[i].m);
+	    complete &= stats[i].done;
+
+	    int end = stats[i].gen_time.end-1;
+
+	    float gen_time = to_s(stats[i].gen_time[end]);
+	    float crossover_time = to_s(stats[i].crossover_time[end]);
+	    float mutate_time = to_s(stats[i].mutate_time[end]);
+	    float fitness_time = to_s(stats[i].fitness_time[end]);
+	    float sorting_time = to_s(stats[i].sorting_time[end]);
+	    float progress_per = static_cast<float>(stats[i].gen) / static_cast<float>(strat.num_generations) * 100;
+	    float best_score = back(stats[i].best_cell_fitness);
+
+	    float overhead = max(0, gen_time - (crossover_time + mutate_time + fitness_time + sorting_time));
+
+	    float overhead_per = overhead / gen_time * 100;
+
+	    g_avg_gen_time += gen_time;
+	    g_avg_crossover_time += crossover_time;
+	    g_avg_mutate_time += mutate_time;
+	    g_avg_fitness_time += fitness_time;
+	    g_avg_sorting_time += sorting_time;
+	    g_progress_per += progress_per;
+	    g_best_fitness = better(strat.higher_fitness_is_better, best_score, g_best_fitness);
+
+	    g_avg_overhead_time += overhead;
+
+	    log(thread_stat_format_str, i, stats[i].gen, strat.num_generations, best_score, overhead_per, gen_time, overhead, crossover_time, mutate_time, fitness_time, sorting_time);
+
+	    unlock(stats[i].m);
+	}
+
+	g_avg_gen_time       /= stats.len;
+	g_avg_crossover_time /= stats.len;
+	g_avg_mutate_time    /= stats.len;
+	g_avg_fitness_time   /= stats.len;
+	g_avg_sorting_time   /= stats.len;
+	g_progress_per       /= stats.len;
+
+	g_avg_overhead_time  /= stats.len;
+
+	float g_avg_overhead_per = g_avg_overhead_time / g_avg_gen_time * 100;
+
+	log(global_stat_format_str, g_progress_per, g_best_fitness, g_avg_overhead_per, g_avg_gen_time, g_avg_overhead_time, g_avg_crossover_time, g_avg_mutate_time, g_avg_fitness_time, g_avg_sorting_time);
+	
+	if (complete) break;
+    }
+
+    for (int i = 0; i < threads.len; i++) {
+	join(threads[i]);
+    }
+
+    T best_cell;
+    // TODO: bad
+    float best_score = strat.higher_fitness_is_better ? FLT_MIN : FLT_MAX;
+    for (int i = 0; i < stats.len; i++) {
+	float score = back(stats[i].best_cell_fitness);
+	if (strat.higher_fitness_is_better ? score > best_score : score < best_score) {
+	    best_cell = back(stats[i].best_cells);
+	    best_score = score;
 	}
     }
 
-    // Yay! Let's grab some jobs to do
+    return best_cell;
-
-    // If the batch we're about to grab moves read_i to write_i and the producer
-    // is done writing, we should let our callee know it's handling this gen's last 
-    // batch know that way it sets work_complete and signals done.
-    *out_batch_is_end = q.done_writing && q.read_i + q.batch_size >= q.write_i;
-
-    out_batch->data = &q.jobs[q.read_i];
-    out_batch->len = min(q.batch_size, q.write_i - q.read_i);
-    q.read_i += q.batch_size;
-    unlock(q.m);
-    return true;
-}
-
-template<class T>
-void work_batch(Array<TaggedJob<T>> batch, Strategy<T> &s) {
-    for (int i = 0; i < batch.len; i++) {
-	switch (batch[i].type) {
-	    case JobType::MUTATE: {
-		MutateJob<T> mj = batch[i].data.m;
-		s.mutate(*mj.cell);
-	    } break;
-	    case JobType::CROSSOVER: {
-		CrossoverJob<T> cj = batch[i].data.c;
-		s.crossover(cj.parents, cj.children);
-	    } break;
-	    case JobType::FITNESS: {	
-		FitnessJob<T> fj = batch[i].data.f;
-		fj.track->score = s.fitness(*fj.cell);
-	    } break;
-	    default: {
-		assert(false);
-	    }
-	}
-    }
-}
-
-template<class T>
-DWORD worker(LPVOID args) {
-    WorkerThreadArgs<T>* wa = static_cast<WorkerThreadArgs<T>*>(args);
-    WorkQueue<T> &q = wa->q;
-    Strategy<T> &s = wa->s;
-
-    // These are written by tryget_job_batch
-    bool batch_is_end;
-    Array<TaggedJob<T>> batch;
-
-    while (tryget_job_batch(q, &batch, &batch_is_end)) {
-	work_batch(batch, s);
-	if (batch_is_end) {
-	    lock(q.m);
-	    q.work_complete = true;
-	    wake_one(q.done);
-	    unlock(q.m);
-	}
-    }
-
-    return NULL;
 }
 
 } // namespace genetic

inc/sync.h

@@ -1,7 +1,11 @@
 #pragma once 
 
+#include <cassert>
+#include <cstdint>
+#include <cstdio>
+
 #ifdef _WIN32
-#include "windows.h"
+#include <windows.h>
 #endif
 
 namespace sync {
@@ -17,6 +21,14 @@ typedef LPVOID ThreadArg;
 
 const TimeSpan infinite_ts = { .QuadPart = LLONG_MAX };
 
+int get_num_cores() {
+    SYSTEM_INFO sysinfo;
+    GetSystemInfo(&sysinfo);
+    return sysinfo.dwNumberOfProcessors;
+}
+
+const int num_cores = get_num_cores();
+
 LARGE_INTEGER _init_freq() {
     LARGE_INTEGER freq;
     QueryPerformanceFrequency(&freq);
@@ -27,7 +39,13 @@ static LARGE_INTEGER freq = _init_freq();
 #endif
 
 Thread make_thread(ThreadFunc t, ThreadArg a);
+Thread make_thread(ThreadFunc t, ThreadArg a, int core_affinity);
 void join(Thread t);
+void sleep(TimeSpan ts);
+void allow_all_processors();
+void set_affinity(Thread &t, int core);
+void set_affinity(int core);
+int get_affinity();
 
 Mutex make_mutex();
 void lock(Mutex &m);
@@ -52,23 +70,106 @@ TimeSpan from_min(double minutes);
 TimeSpan from_hours(double hours);
 TimeSpan now();
 TimeSpan operator-(const TimeSpan &a, const TimeSpan &b);
+TimeSpan operator+(const TimeSpan &a, const TimeSpan &b);
+TimeSpan operator*(const TimeSpan &a, const TimeSpan &b);
+TimeSpan operator/(const TimeSpan &a, const TimeSpan &b);
 
-double to_ms(TimeSpan &sp);
+double to_ms(TimeSpan &ts);
-double to_s(TimeSpan &sp);
+double to_s(TimeSpan &ts);
-double to_min(TimeSpan &sp);
+double to_min(TimeSpan &ts);
-double to_hours(TimeSpan &sp);
+double to_hours(TimeSpan &ts);
 
 #ifdef _WIN32
 
+uint64_t bitmask (unsigned short n) {
+  if (n == 64) return -((uint64_t)1);
+  return (((uint64_t) 1) << n) - 1;
+}
+
+const int tab64[64] = {
+    63,  0, 58,  1, 59, 47, 53,  2,
+    60, 39, 48, 27, 54, 33, 42,  3,
+    61, 51, 37, 40, 49, 18, 28, 20,
+    55, 30, 34, 11, 43, 14, 22,  4,
+    62, 57, 46, 52, 38, 26, 32, 41,
+    50, 36, 17, 19, 29, 10, 13, 21,
+    56, 45, 25, 31, 35, 16,  9, 12,
+    44, 24, 15,  8, 23,  7,  6,  5};
+
+int log2_64 (uint64_t value)
+{
+    value |= value >> 1;
+    value |= value >> 2;
+    value |= value >> 4;
+    value |= value >> 8;
+    value |= value >> 16;
+    value |= value >> 32;
+    return tab64[((uint64_t)((value - (value >> 1))*0x07EDD5E59A4E28C2)) >> 58];
+}
+
 Thread make_thread(ThreadFunc f, ThreadArg a) {
     DWORD tid;
     return CreateThread(NULL, 0, f, a, 0, &tid);
 }
 
+struct DummyThreadArgs {
+    int core_affinity;
+    ThreadFunc f;
+    ThreadArg a;
+};
+
+DWORD _dummy_thread(LPVOID a) {
+    DummyThreadArgs *wrap = static_cast<DummyThreadArgs*>(a);
+    set_affinity(wrap->core_affinity);
+    return wrap->f(wrap->a);
+}
+
+Thread make_thread(ThreadFunc f, ThreadArg a, int core_affinity) {
+    DWORD tid;
+    DummyThreadArgs *args = (DummyThreadArgs*)malloc(sizeof(DummyThreadArgs));
+    *args = {
+	.core_affinity=core_affinity,
+	.f=f,
+	.a=a
+    };
+    return CreateThread(NULL, 0, _dummy_thread, args, 0, &tid);
+}
+
 void join(Thread t) {
     WaitForSingleObject(t, INFINITE);
 }
 
+void sleep(TimeSpan ts) {
+    Sleep(static_cast<DWORD>(to_ms(ts)));
+}
+
+void allow_all_processors() {
+    Thread t = GetCurrentThread();
+    DWORD affinity = bitmask(num_cores);
+    SetProcessAffinityMask(t, affinity);
+}
+
+void set_affinity(Thread &t, int core) {
+    DWORD mask = 1 << (core % num_cores);
+    DWORD old = SetThreadAffinityMask(t, mask);
+    DWORD confirm = SetThreadAffinityMask(t, mask);
+    assert(old && GetLastError() != ERROR_INVALID_PARAMETER && mask == confirm);
+}
+
+void set_affinity(int core) {
+    Thread cur = GetCurrentThread();
+    set_affinity(cur, core);
+}
+
+int get_affinity() {
+    Thread t = GetCurrentThread();
+    DWORD mask = 1;
+    DWORD affinity = SetThreadAffinityMask(t, (DWORD_PTR)mask);
+    DWORD check = SetThreadAffinityMask(t, (DWORD_PTR)affinity);
+    assert(check == mask);
+    return log2_64(affinity);
+}
+
 Mutex make_mutex() {
     Mutex m;
     InitializeCriticalSection(&m);
@@ -135,25 +236,25 @@ void dispose(Semaphore &s) {
 
 TimeSpan from_ms(double milliseconds) {
     TimeSpan ts;
-    ts.QuadPart = static_cast<int64_t>(milliseconds/1000.0)*freq.QuadPart;
+    ts.QuadPart = static_cast<LONGLONG>(milliseconds/1000.0)*freq.QuadPart;
     return ts;
 }
 
 TimeSpan from_s(double seconds) {
     TimeSpan ts;
-    ts.QuadPart = static_cast<int64_t>(seconds)*freq.QuadPart;
+    ts.QuadPart = static_cast<LONGLONG>(seconds)*freq.QuadPart;
     return ts;
 }
 
 TimeSpan from_min(double minutes) {
     TimeSpan ts;
-    ts.QuadPart = static_cast<int64_t>(minutes*60.0)*freq.QuadPart;
+    ts.QuadPart = static_cast<LONGLONG>(minutes*60.0)*freq.QuadPart;
     return ts;
 }
 
 TimeSpan from_hours(double hours) {
     TimeSpan ts;
-    ts.QuadPart = static_cast<int64_t>(hours*60.0*60.0)*freq.QuadPart;
+    ts.QuadPart = static_cast<LONGLONG>(hours*60.0*60.0)*freq.QuadPart;
     return ts;
 }
 
@@ -163,26 +264,44 @@ TimeSpan now() {
     return ts;
 }
 
-TimeSpan operator-(const TimeSpan &a, TimeSpan &b) {
+TimeSpan operator-(const TimeSpan &a, const TimeSpan &b) {
     TimeSpan ts;
     ts.QuadPart = a.QuadPart - b.QuadPart;
     return ts;
 }
 
-double to_ms(TimeSpan &sp) {
+TimeSpan operator+(const TimeSpan &a, const TimeSpan &b) {
-    return static_cast<double>(sp.QuadPart*1000)/static_cast<double>(freq.QuadPart);
+    TimeSpan ts;
+    ts.QuadPart = a.QuadPart + b.QuadPart;
+    return ts;
 }
 
-double to_s(TimeSpan &sp) {
+TimeSpan operator*(const TimeSpan &a, const TimeSpan &b) {
-    return static_cast<double>(sp.QuadPart)/static_cast<double>(freq.QuadPart);
+    TimeSpan ts;
+    ts.QuadPart = a.QuadPart * b.QuadPart;
+    return ts;
 }
 
-double to_min(TimeSpan &sp) {
+TimeSpan operator/(const TimeSpan &a, const TimeSpan &b) {
-    return static_cast<double>(sp.QuadPart)/static_cast<double>(freq.QuadPart*60);
+    TimeSpan ts;
+    ts.QuadPart = a.QuadPart / b.QuadPart;
+    return ts;
 }
 
-double to_hours(TimeSpan &sp) {
+double to_ms(TimeSpan &ts) {
-    return static_cast<double>(sp.QuadPart)/static_cast<double>(freq.QuadPart*60*60);
+    return static_cast<double>(ts.QuadPart*1000)/static_cast<double>(freq.QuadPart);
+}
+
+double to_s(TimeSpan &ts) {
+    return static_cast<double>(ts.QuadPart)/static_cast<double>(freq.QuadPart);
+}
+
+double to_min(TimeSpan &ts) {
+    return static_cast<double>(ts.QuadPart)/static_cast<double>(freq.QuadPart*60);
+}
+
+double to_hours(TimeSpan &ts) {
+    return static_cast<double>(ts.QuadPart)/static_cast<double>(freq.QuadPart*60*60);
 }
  
 #endif

inc/util.h

@@ -0,0 +1,56 @@
+#pragma once 
+
+#include <cstring>
+#define min(A, B) ((A < B) ? (A) : (B))
+#define max(A, B) ((A > B) ? (A) : (B))
+#define better(GT, A, B) (GT ? max((A), (B)) : min((A), (B)))
+
+template <class T> struct Array {
+    T *data;
+    int len;
+
+    T &operator[](int i) { return data[i]; }
+};
+
+
+template <class T> Array<T> make_array(int len) {
+    return {
+	.data=(T*)malloc(sizeof(T)*len),
+	.len=len
+     };
+}
+template <class T> T back(Array<T> &a) { return a.data[a.len-1]; }
+template <class T> T front(Array<T> &a) { return a.data[0]; }
+
+template <class T> struct DynArray {
+    T* _data;
+    int end;
+    int cap;
+
+    T &operator[](int i) { return _data[i]; }
+};
+
+template <class T> DynArray<T> make_dynarray(int cap) {
+    return {
+	._data=(T*)malloc(sizeof(T)*cap),
+	.end=0,
+	.cap=cap
+    };
+}
+
+template <class T> void resize(DynArray<T> &a, int new_cap) {
+    T* old = a._data;
+    a._data = (T*)malloc(sizeof(T)*new_cap);
+    memcpy(a._data, old, min(sizeof(T)*a.end, sizeof(T)*new_cap));
+    a.cap = new_cap;
+    free(old);
+}
+
+template <class T> void append(DynArray<T> &a, T el) {
+    if (a.end == a.cap) resize(a, min(1, a.cap*2));
+    a[a.end++] = el;
+}
+
+template <class T> T back(DynArray<T> &a) { return a._data[a.end-1]; }
+template <class T> T front(DynArray<T> &a) { return a._data[0]; }
+

src/main.cpp

@@ -3,6 +3,7 @@
 #include <cstdlib>
 #include "genetic.h"
 #include "rand.h"
+#include "sync.h"
 
 using namespace genetic;
 
@@ -40,9 +41,6 @@ void crossover(const Array<Array<float>*> parents, const Array<Array<float> *> o
     }
 }
 
-// norm_rand can go negative. fix in genetic.cpp
-// child stride doesn't make sense. Should always skip over child num
-
 float fitness(const Array<float> &cell) {
     float sum = 0;
     float product = 1;
@@ -54,12 +52,14 @@ float fitness(const Array<float> &cell) {
 }
 
 int main(int argc, char **argv) {
-    int num_gens = 2000;
+    int num_gens = 10000;
     Strategy<Array<float>> strat {
-        .num_threads = 15,
+        .num_threads = atoi(argv[1]),
-        .batch_size  = 1000,
+	.stats_print_period_s = 2,
-        .num_cells   = 100000,
+        .num_cells_per_thread = 100000,
         .num_generations = num_gens,
+	.share_breakthroughs=true,
+	.share_breakthrough_gen_period=10,
         .test_all = true,
         .test_chance = 0.0, // doesn't matter
         .enable_crossover = true,
@@ -75,22 +75,24 @@ int main(int argc, char **argv) {
         .crossover=crossover,
         .fitness=fitness
     };
+    n_threads = atoi(argv[1]);
 
-    auto res = run(strat);
+    log("Running w/ %d threads\n", atoi(argv[1]));
+    TimeSpan start = now();
+    auto best_cell = run(strat);
+    TimeSpan runtime = now() - start;
 
     float sum = 0;
     float product = 1;
-    printf("Winning cell: ");
+    log("Winning cell: ");
-    for (int i = 0; i < res.best_cell.back().len; i++) {
+    for (int i = 0; i < best_cell.len; i++) {
-	float val = res.best_cell.back()[i];
+	float val = best_cell[i];
 	sum += val;
 	product *= val;
-	printf("%f ", val);
+	log("%f ", val);
     }
-    printf("\n");
+    log("\n");
-    printf("Final Sum: %f\n", sum);
+    log("Final Sum: %f\n", sum);
-    printf("Final Product: %f\n", product);
+    log("Final Product: %f\n", product);
-    printf("Setup Time (s): %f\n", sync::to_s(res.setup_time));
+    log("Execution Time %d (min) %f (s)\n", static_cast<int>(sync::to_min(runtime)), fmod(to_s(runtime), 60) );
-    printf("Run Time (s): %f\n", sync::to_s(res.run_time));
-    printf("Average Gen Time (s): %f\n", sync::to_s(res.run_time)/num_gens);
 }