X-Git-Url: http://demsky.eecs.uci.edu/git/?a=blobdiff_plain;f=ms-queue%2Ftestcase3.c;h=628574ec9fdba4fbf653a5e57f912eed537f26d0;hb=357cc2d31eaf9df96c2fc676fa00f40a54e83019;hp=f6243f999abf5e807e545130132ad21445c1aa71;hpb=e4f5d19cd0d94393277381f8aab95126f201fb26;p=model-checker-benchmarks.git

diff --git a/ms-queue/testcase3.c b/ms-queue/testcase3.c
index f6243f9..628574e 100644
--- a/ms-queue/testcase3.c
+++ b/ms-queue/testcase3.c
@@ -25,28 +25,33 @@ int get_thread_num()
 bool succ1, succ2;
 atomic_int x[3];
 int idx1, idx2;
+unsigned int reclaimNode1, reclaimNode2;
+
+static int procs = 2;
+
+
+/** This testcase can infer w2->release & w4->acquire.
+	The initial node that Head and Tail points to is 1, so when T3 enqueue with
+	node 2, and dequeue(get node 1), and enqueue node 1 again, the second time
+	it enqueues node 1 it actually first initialize node1->next. At the same
+	time in T2, it reads that node1->next (because it reads the old Tail at the
+	very beginning), then loads the Tail agian (w4), it can actully reads an old
+	value. And this is a bug because if node 1 is again dequeued, then for T2 to
+	update node1->next, it can potentially contaminate the memory...
+*/
 
-static int procs = 4;
 static void main_task(void *param)
 {
 	unsigned int val;
 	int pid = *((int *)param);
 	if (pid % 4 == 0) {
-		atomic_store_explicit(&x[0], 1, memory_order_relaxed);
-		enqueue(queue, 0);
-	
-		succ1 = dequeue(queue, &idx1);
-		if (succ1) {
-			atomic_load_explicit(&x[idx1], memory_order_relaxed);
-		}
+		enqueue(queue, 0, 3);
 	} else if (pid % 4 == 1) {
-		atomic_store_explicit(&x[1], 1, memory_order_relaxed);
-		enqueue(queue, 1);
-
-		succ2 = dequeue(queue, &idx2);
-		if (succ2) {
-			atomic_load_explicit(&x[idx2], memory_order_relaxed);
-		}
+		enqueue(queue, 1, 2);
+		succ1 = dequeue(queue, &idx1, &reclaimNode1);
+		enqueue(queue, 1, 1);
+		//succ1 = dequeue(queue, &idx1, &reclaimNode1);
+		//enqueue(queue, 1, false, 2);
 	} else if (pid % 4 == 2) {
 
 	} else if (pid % 4 == 3) {