diff --git a/fs/aio.c b/fs/aio.c
index dedeea0..0e23dfa 100644
--- a/fs/aio.c
+++ b/fs/aio.c
@@ -94,7 +94,13 @@
 	struct work_struct	rcu_work;
 
 	struct {
-		atomic_t	reqs_active;
+		/*
+		 * This counts the number of available slots in the ringbuffer,
+		 * so we avoid overflowing it: it's decremented (if positive)
+		 * when allocating a kiocb and incremented when the resulting
+		 * io_event is pulled off the ringbuffer.
+		 */
+		atomic_t	reqs_available;
 	} ____cacheline_aligned_in_smp;
 
 	struct {
@@ -404,19 +410,20 @@
 	head = ring->head;
 	kunmap_atomic(ring);
 
-	while (atomic_read(&ctx->reqs_active) > 0) {
+	while (atomic_read(&ctx->reqs_available) < ctx->nr_events - 1) {
 		wait_event(ctx->wait,
-				head != ctx->tail ||
-				atomic_read(&ctx->reqs_active) <= 0);
+			   (head != ctx->tail) ||
+			   (atomic_read(&ctx->reqs_available) >=
+			    ctx->nr_events - 1));
 
 		avail = (head <= ctx->tail ? ctx->tail : ctx->nr_events) - head;
 
-		atomic_sub(avail, &ctx->reqs_active);
+		atomic_add(avail, &ctx->reqs_available);
 		head += avail;
 		head %= ctx->nr_events;
 	}
 
-	WARN_ON(atomic_read(&ctx->reqs_active) < 0);
+	WARN_ON(atomic_read(&ctx->reqs_available) > ctx->nr_events - 1);
 
 	aio_free_ring(ctx);
 
@@ -475,6 +482,8 @@
 	if (aio_setup_ring(ctx) < 0)
 		goto out_freectx;
 
+	atomic_set(&ctx->reqs_available, ctx->nr_events - 1);
+
 	/* limit the number of system wide aios */
 	spin_lock(&aio_nr_lock);
 	if (aio_nr + nr_events > aio_max_nr ||
@@ -586,7 +595,7 @@
 				"exit_aio:ioctx still alive: %d %d %d\n",
 				atomic_read(&ctx->users),
 				atomic_read(&ctx->dead),
-				atomic_read(&ctx->reqs_active));
+				atomic_read(&ctx->reqs_available));
 		/*
 		 * We don't need to bother with munmap() here -
 		 * exit_mmap(mm) is coming and it'll unmap everything.
@@ -615,12 +624,9 @@
 {
 	struct kiocb *req;
 
-	if (atomic_read(&ctx->reqs_active) >= ctx->nr_events)
+	if (atomic_dec_if_positive(&ctx->reqs_available) <= 0)
 		return NULL;
 
-	if (atomic_inc_return(&ctx->reqs_active) > ctx->nr_events - 1)
-		goto out_put;
-
 	req = kmem_cache_alloc(kiocb_cachep, GFP_KERNEL|__GFP_ZERO);
 	if (unlikely(!req))
 		goto out_put;
@@ -630,7 +636,7 @@
 
 	return req;
 out_put:
-	atomic_dec(&ctx->reqs_active);
+	atomic_inc(&ctx->reqs_available);
 	return NULL;
 }
 
@@ -701,7 +707,7 @@
 
 	/*
 	 * Take rcu_read_lock() in case the kioctx is being destroyed, as we
-	 * need to issue a wakeup after decrementing reqs_active.
+	 * need to issue a wakeup after incrementing reqs_available.
 	 */
 	rcu_read_lock();
 
@@ -719,7 +725,7 @@
 	 */
 	if (unlikely(xchg(&iocb->ki_cancel,
 			  KIOCB_CANCELLED) == KIOCB_CANCELLED)) {
-		atomic_dec(&ctx->reqs_active);
+		atomic_inc(&ctx->reqs_available);
 		/* Still need the wake_up in case free_ioctx is waiting */
 		goto put_rq;
 	}
@@ -857,7 +863,7 @@
 
 	pr_debug("%li  h%u t%u\n", ret, head, ctx->tail);
 
-	atomic_sub(ret, &ctx->reqs_active);
+	atomic_add(ret, &ctx->reqs_available);
 out:
 	mutex_unlock(&ctx->ring_lock);
 
@@ -1241,7 +1247,7 @@
 	aio_put_req(req);	/* drop extra ref to req */
 	return 0;
 out_put_req:
-	atomic_dec(&ctx->reqs_active);
+	atomic_inc(&ctx->reqs_available);
 	aio_put_req(req);	/* drop extra ref to req */
 	aio_put_req(req);	/* drop i/o ref to req */
 	return ret;