cuda: advanced: exercises: fixups

labs/cuda/advanced/exercises/inserts/Makefile

 EXEC := $(notdir $(CURDIR))
 
+LDFLAGS := -lstdc++
+
 include ../../../Makefile

labs/cuda/advanced/exercises/inserts/inserts.cu

@@ -34,8 +34,9 @@ __global__ void kernel_write(unsigned long long int *data,
   // TODO
 }
 
-// do not modify validateDB and main
-//
+////////////////////////////////////////////////////////////////////////////////
+// DO NOT modify below
+
 void validateDB(unsigned long long int *database, int dbSize, int expNZElem) {
   bool isValid = true;
   int numNZElem = 0;
@@ -109,4 +110,4 @@ int main(void) {
   cudaFree(data);
   cudaFree(database);
   return 0;
-}
\ No newline at end of file
+}

labs/cuda/advanced/exercises/peculiar-sums/Makefile

 EXEC := $(notdir $(CURDIR))
 
+LDFLAGS := -lstdc++
+
 include ../../../Makefile

labs/cuda/advanced/exercises/peculiar-sums/REAMDE.md

+# Sume particulare
+
+Fiind dat un vector, `data`, cu valori mai mici decât 10, generați un nou
+vector `result` după următoarea formulă.
+
+$$result(i) = \sum_{j=0}^{data(i)} data(j)$$

labs/cuda/advanced/exercises/peculiar-sums/peculiar-sums.cu

-#include <stdio.h>
-#include <stdlib.h>
-
+#include <cstdlib>
 #include <iostream>
 
 #define NUM_ELEM 128
@@ -11,71 +9,72 @@ using namespace std;
 // workers will compute sum on first N elements
 __global__ void worker() {
   // TODO, compute sum and store in result
+}
 
-  // TODO
-  // master will launch threads to compute sum on first N elements
-  __global__ void master() {
-    // TODO, schedule worker threads
-  }
+// TODO
+// master will launch threads to compute sum on first N elements
+__global__ void master() {
+  // TODO, schedule worker threads
+}
 
-  void generateData(int *data, int num) {
-    srand(time(0));
+void generateData(int *data, int num) {
+  srand(time(0));
 
-    for (int i = 0; i < num; i++) {
-      data[i] = rand() % 8 + 2;
-    }
+  for (int i = 0; i < num; i++) {
+    data[i] = rand() % 8 + 2;
   }
+}
 
-  void print(int *data, int num) {
-    for (int i = 0; i < num; i++) {
-      cout << data[i] << " ";
-    }
-    cout << endl;
+void print(int *data, int num) {
+  for (int i = 0; i < num; i++) {
+    cout << data[i] << " ";
   }
-
-  // check
-  // each element result[i] should be sum of first data[i] elements of data[i]
-  bool checkResult(int *data, int num, int *result) {
-    for (int i = 0; i < num; i++) {
-      int sum = 0;
-      for (int j = 0; j < data[i] && j < num; j++) {
-        sum += data[j];
-      }
-
-      if (result[i] != sum) {
-        cout << "Error at " << i << ", requested sum of first " << data[i]
-             << " elem, got " << result[i] << endl;
-        return false;
-      }
+  cout << endl;
+}
+
+// check
+// each element result[i] should be sum of first data[i] elements of data
+bool checkResult(int *data, int num, int *result) {
+  for (int i = 0; i < num; i++) {
+    int sum = 0;
+    for (int j = 0; j < data[i] && j < num; j++) {
+      sum += data[j];
     }
 
-    return true;
+    if (result[i] != sum) {
+      cout << "Error at " << i << ", requested sum of first " << data[i]
+           << " elem, got " << result[i] << endl;
+      return false;
+    }
   }
 
-  int main(int argc, char *argv[]) {
-    int *data = NULL;
-    cudaMallocManaged(&data, NUM_ELEM * sizeof(int));
-
-    int *result = NULL;
-    cudaMallocManaged(&result, NUM_ELEM * sizeof(int));
+  return true;
+}
 
-    generateData(data, NUM_ELEM);
+int main(int argc, char *argv[]) {
+  int *data = NULL;
+  cudaMallocManaged(&data, NUM_ELEM * sizeof(int));
 
-    // TODO schedule master threads and pass data/result/num
-    master<<<1, 1>>>();
-    cudaDeviceSynchronize();
+  int *result = NULL;
+  cudaMallocManaged(&result, NUM_ELEM * sizeof(int));
 
-    print(data, NUM_ELEM);
-    print(result, NUM_ELEM);
+  generateData(data, NUM_ELEM);
 
-    if (checkResult(data, NUM_ELEM, result)) {
-      cout << "Result OK" << endl;
-    } else {
-      cout << "Result ERR" << endl;
-    }
+  // TODO schedule master threads and pass data/result/num
+  master<<<1, 1>>>();
+  cudaDeviceSynchronize();
 
-    cudaFree(data);
-    cudaFree(result);
+  print(data, NUM_ELEM);
+  print(result, NUM_ELEM);
 
-    return 0;
+  if (checkResult(data, NUM_ELEM, result)) {
+    cout << "Result OK" << endl;
+  } else {
+    cout << "Result ERR" << endl;
   }
+
+  cudaFree(data);
+  cudaFree(result);
+
+  return 0;
+}

labs/cuda/advanced/exercises/stats/full-atomics/Makefile

 EXEC := $(notdir $(CURDIR))
 
+LDFLAGS := -lstdc++
+
 include ../../../../Makefile

labs/cuda/advanced/exercises/stats/no-atomics/Makefile

 EXEC := $(notdir $(CURDIR))
 
+LDFLAGS := -lstdc++
+
 include ../../../../Makefile

labs/cuda/advanced/exercises/stats/partial-atomics/Makefile

 EXEC := $(notdir $(CURDIR))
 
+LDFLAGS := -lstdc++
+
 include ../../../../Makefile