Flush cout at various points so we can more easily watch progress

controller.cc

@@ -409,6 +409,7 @@ static void computation_complete(void)
 	cout << "base^exp = " << base_exp << "\n";
 	cout << "target   = " << ctrlstate.target << "\n";
     }
+    cout.flush();
 
     ctrlstate.reset();
 }

parrhoasm.cu

@@ -538,6 +538,7 @@ cout << i << ": " << l_Z[i] << " != " << l_z[i + t * WORDS] << "\n";
 
     unsigned long long totnanos = totmicros * 1000;
     cout << totmicros << " us / " << nthreads << " = " << totmicros / nthreads << " us / " << (nmult*launchcount) << " = " << totnanos / ((unsigned long long)nthreads * nmult * launchcount) << " ns\n";
+    cout.flush();
     //cudaPrintfEnd();
     DPstreamEnd();
 

worker.cc

@@ -136,6 +136,7 @@ static void start_working(void)
     cerr << "Starting work\n";
 #endif
     cout << "Subproblem " << wrkctrlstate.current_problem->problemid << "\n";
+    cout.flush();
     wrkctrlstate.worker_thread_state = WT_RUNNING;
     if (pthread_create(&wrkctrlstate.worker_thread, NULL,
 			worker_thread_start, NULL)) {