diff --git a/cs677/pa2/Makefile b/cs677/pa2/Makefile
index e792a6d..ed7875b 100644
--- a/cs677/pa2/Makefile
+++ b/cs677/pa2/Makefile
@@ -11,7 +11,7 @@ test: all
 	./sequential e1.*
 	./threaded e1.*
 	./sequential e2.*
-	./sequential e2.*
+	./threaded e2.*
 
 clean:
 	-rm -f $(PROGS) *~ *.o
diff --git a/cs677/pa2/threaded.cc b/cs677/pa2/threaded.cc
index 3f47e51..f530c88 100644
--- a/cs677/pa2/threaded.cc
+++ b/cs677/pa2/threaded.cc
@@ -143,7 +143,7 @@ void * calcSimMatrixThread(void * arg)
     int s_size = s->size();
     int t_size = t->size();
     int last_step = s_size + t_size;
-    int (*F)[s_size][t_size] = (int (*) [s_size][t_size]) matrix;
+    int (*F)[s_size+1][t_size+1] = (int (*) [s_size+1][t_size+1]) matrix;
     int max_x = 0, max_y = 0, max_val = 0;
     int first_task_id, num_tasks;
     taskAllocate(t_size+1, num_threads, id, &first_task_id, &num_tasks);
@@ -153,13 +153,32 @@ void * calcSimMatrixThread(void * arg)
     for (int i = 0, idx = first_task_id; i < num_tasks; i++, idx++)
         (*F)[idx][0] = 0;               /* set first column to 0's */
     pthread_barrier_wait(&barrier);
-    for (int step = 2; i <= last_step; i++)
+    for (int step = 2; step <= last_step; step++)
     {
-        /* TODO: figure out how many tasks there are in this step */
-        /* TODO: call taskAllocate() to get mine */
-        /* TODO: loop over tasks assigned by taskAllocate() */
+        int first_i = step - 1;
+        int first_j = 1;
+        int last_i = 1;
+        int last_j = step - 1;
+        if (first_i > s_size)
+        {
+            first_j += (first_i - s_size);
+            first_i = s_size;
+        }
+        if (last_j > t_size)
+        {
+            last_i += (last_j - t_size);
+            last_j = t_size;
+        }
+        num_tasks = (last_j - first_j) + 1;
+        taskAllocate(num_tasks, num_threads, id, &first_task_id, &num_tasks);
+        for (int i = first_i - first_task_id,
+                 j = first_j + first_task_id,
+                 task_id = 0;
+             task_id < num_tasks;
+             i--,
+             j++,
+             task_id++)
         {
-            /* TODO: figure out i, j from step and task */
             /* Compute the value for the matrix */
             (*F)[i][j] =
                 max(
@@ -194,6 +213,17 @@ void * calcSimMatrixThread(void * arg)
         pthread_barrier_wait(&barrier);
     }
 
+#if 0
+    cout << "Matrix: " << s_size+1 << " x " << t_size+1 << endl;
+    for (int i = 0; i <= s_size; i++)
+    {
+        for (int j = 0; j <= t_size; j++)
+        {
+            printf("%2d ", (*F)[i][j]);
+        }
+        printf("\n");
+    }
+#endif
     cout << "Maximum value is " << max_val << " at position ("
         << max_x << ", " << max_y << ")" << endl;