Differences

This shows you the differences between two versions of the page.

--- c:pthreads:parallel-increment-vector [2024/01/16 21:51] – created odefta
+++ c:pthreads:parallel-increment-vector [2025/01/02 18:22] (current) – external edit 127.0.0.1
@@ Line 8: / Line 8: @@
 #include <pthread.h>
-#define NUM_THREADS 4  // Numărul de thread-uri
+#define NUM_THREADS 4  // Number of threads to use
-#define VECTOR_SIZE 100  // Dimensiunea vectorului
+#define VECTOR_SIZE 100  // Size of the array
-// Structura pentru a stoca informațiile pentru fiecare thread
+// Structure to pass information to threads
 typedef struct {
-    int start;  // Indexul de început pentru acest thread
+    int start;  // Starting index for this thread
-    int end;    // Indexul de sfârșit pentru acest thread
+    int end;    // Ending index for this thread
-    int *vector;  // Pointer către vector
+    int *vector;  // Pointer to the array
 } ThreadData;
-// Funcția executată de fiecare thread
+// Function that each thread will execute
 void *increment_vector(void *arg) {
     ThreadData *data = (ThreadData *)arg;
+    // Increment the array elements in the specified range
     for (int i = data->start; i < data->end; i++) {
         data->vector[i]++;
@@ Line 35: / Line 36: @@
     int segment_size = VECTOR_SIZE / NUM_THREADS;
-    // Inițializează vectorul
+    // Initialize the array
     for (int i = 0; i < VECTOR_SIZE; i++) {
         vector[i] = i;
     }
-    // Creează thread-urile
+    // Create threads and assign them segments of the array
     for (int i = 0; i < NUM_THREADS; i++) {
         thread_data[i].start = i * segment_size;
@@ Line 49: / Line 50: @@
     }
-    // Așteaptă ca thread-urile să termine
+    // Wait for all threads to complete
     for (int i = 0; i < NUM_THREADS; i++) {
         pthread_join(threads[i], NULL);
     }
-    // Afișează vectorul incrementat
+    // Display the incremented array
     for (int i = 0; i < VECTOR_SIZE; i++) {
         printf("%d\n", vector[i]);
     }
     return 0;
 }
 </code>
+The process involves splitting the array into nearly equal segments and assigning each segment to a different thread for incrementing. \\
+**Define an Array and Thread Information**: First, define a 100-element array and decide on the number of threads. The array will be split into segments, with each segment processed by a different thread.
+**Thread Function**: Write a function that each thread will execute. This function should accept a range of indices to process (start and end indices) and increment the elements within this range.
+**Create and Initialize Threads**: Create threads and pass them their respective array segments to process. Ensure the segments cover the entire array and are as evenly distributed as possible.
+** Wait for Threads to Complete**: After starting all threads, use pthread_join to wait for all threads to finish their execution.
+**Finalize**: Optionally, after all threads have completed, you can perform any final processing or output the updated array.
+This method efficiently utilizes multiple cores (if available) to perform the increment operation in parallel, potentially reducing the total processing time compared to a single-threaded approach. It's important in such parallel operations to carefully manage thread creation and synchronization to avoid issues like race conditions.