1 files changed, 63 insertions, 57 deletions

diff --git a/kernel/events/ring_buffer.c b/kernel/events/ring_buffer.c
index 192b8abc6330..180509132d4b 100644
--- a/kernel/events/ring_buffer.c
+++ b/kernel/events/ring_buffer.c
@@ -22,7 +22,11 @@ static void perf_output_wakeup(struct perf_output_handle *handle)
 	atomic_set(&handle->rb->poll, EPOLLIN);
 
 	handle->event->pending_wakeup = 1;
-	irq_work_queue(&handle->event->pending);
+
+	if (*perf_event_fasync(handle->event) && !handle->event->pending_kill)
+		handle->event->pending_kill = POLL_IN;
+
+	irq_work_queue(&handle->event->pending_irq);
 }
 
 /*
@@ -147,6 +151,7 @@ ring_buffer_has_space(unsigned long head, unsigned long tail,
 
 static __always_inline int
 __perf_output_begin(struct perf_output_handle *handle,
+		    struct perf_sample_data *data,
 		    struct perf_event *event, unsigned int size,
 		    bool backward)
 {
@@ -171,8 +176,10 @@ __perf_output_begin(struct perf_output_handle *handle,
 		goto out;
 
 	if (unlikely(rb->paused)) {
-		if (rb->nr_pages)
+		if (rb->nr_pages) {
 			local_inc(&rb->lost);
+			atomic64_inc(&event->lost_samples);
+		}
 		goto out;
 	}
 
@@ -188,9 +195,10 @@ __perf_output_begin(struct perf_output_handle *handle,
 
 	perf_output_get_handle(handle);
 
+	offset = local_read(&rb->head);
 	do {
+		head = offset;
 		tail = READ_ONCE(rb->user_page->data_tail);
-		offset = head = local_read(&rb->head);
 		if (!rb->overwrite) {
 			if (unlikely(!ring_buffer_has_space(head, tail,
 							    perf_data_size(rb),
@@ -214,7 +222,7 @@ __perf_output_begin(struct perf_output_handle *handle,
 			head += size;
 		else
 			head -= size;
-	} while (local_cmpxchg(&rb->head, offset, head) != offset);
+	} while (!local_try_cmpxchg(&rb->head, &offset, head));
 
 	if (backward) {
 		offset = head;
@@ -237,24 +245,23 @@ __perf_output_begin(struct perf_output_handle *handle,
 	handle->size = (1UL << page_shift) - offset;
 
 	if (unlikely(have_lost)) {
-		struct perf_sample_data sample_data;
-
 		lost_event.header.size = sizeof(lost_event);
 		lost_event.header.type = PERF_RECORD_LOST;
 		lost_event.header.misc = 0;
 		lost_event.id          = event->id;
 		lost_event.lost        = local_xchg(&rb->lost, 0);
 
-		perf_event_header__init_id(&lost_event.header,
-					   &sample_data, event);
+		/* XXX mostly redundant; @data is already fully initializes */
+		perf_event_header__init_id(&lost_event.header, data, event);
 		perf_output_put(handle, lost_event);
-		perf_event__output_id_sample(event, handle, &sample_data);
+		perf_event__output_id_sample(event, handle, data);
 	}
 
 	return 0;
 
 fail:
 	local_inc(&rb->lost);
+	atomic64_inc(&event->lost_samples);
 	perf_output_put_handle(handle);
 out:
 	rcu_read_unlock();
@@ -263,22 +270,25 @@ out:
 }
 
 int perf_output_begin_forward(struct perf_output_handle *handle,
-			     struct perf_event *event, unsigned int size)
+			      struct perf_sample_data *data,
+			      struct perf_event *event, unsigned int size)
 {
-	return __perf_output_begin(handle, event, size, false);
+	return __perf_output_begin(handle, data, event, size, false);
 }
 
 int perf_output_begin_backward(struct perf_output_handle *handle,
+			       struct perf_sample_data *data,
 			       struct perf_event *event, unsigned int size)
 {
-	return __perf_output_begin(handle, event, size, true);
+	return __perf_output_begin(handle, data, event, size, true);
 }
 
 int perf_output_begin(struct perf_output_handle *handle,
+		      struct perf_sample_data *data,
 		      struct perf_event *event, unsigned int size)
 {
 
-	return __perf_output_begin(handle, event, size,
+	return __perf_output_begin(handle, data, event, size,
 				   unlikely(is_write_backward(event)));
 }
 
@@ -327,6 +337,8 @@ ring_buffer_init(struct perf_buffer *rb, long watermark, int flags)
 	 */
 	if (!rb->nr_pages)
 		rb->paused = 1;
+
+	mutex_init(&rb->aux_mutex);
 }
 
 void perf_aux_output_flag(struct perf_output_handle *handle, u64 flags)
@@ -604,8 +616,8 @@ static struct page *rb_alloc_aux_page(int node, int order)
 {
 	struct page *page;
 
-	if (order > MAX_ORDER)
-		order = MAX_ORDER;
+	if (order > MAX_PAGE_ORDER)
+		order = MAX_PAGE_ORDER;
 
 	do {
 		page = alloc_pages_node(node, PERF_AUX_GFP, order);
@@ -631,7 +643,6 @@ static void rb_free_aux_page(struct perf_buffer *rb, int idx)
 	struct page *page = virt_to_page(rb->aux_pages[idx]);
 
 	ClearPagePrivate(page);
-	page->mapping = NULL;
 	__free_page(page);
 }
 
@@ -672,23 +683,39 @@ int rb_alloc_aux(struct perf_buffer *rb, struct perf_event *event,
 	if (!has_aux(event))
 		return -EOPNOTSUPP;
 
-	/*
-	 * We need to start with the max_order that fits in nr_pages,
-	 * not the other way around, hence ilog2() and not get_order.
-	 */
-	max_order = ilog2(nr_pages);
+	if (nr_pages <= 0)
+		return -EINVAL;
 
-	/*
-	 * PMU requests more than one contiguous chunks of memory
-	 * for SW double buffering
-	 */
 	if (!overwrite) {
-		if (!max_order)
-			return -EINVAL;
+		/*
+		 * Watermark defaults to half the buffer, and so does the
+		 * max_order, to aid PMU drivers in double buffering.
+		 */
+		if (!watermark)
+			watermark = min_t(unsigned long,
+					  U32_MAX,
+					  (unsigned long)nr_pages << (PAGE_SHIFT - 1));
 
-		max_order--;
+		/*
+		 * Use aux_watermark as the basis for chunking to
+		 * help PMU drivers honor the watermark.
+		 */
+		max_order = get_order(watermark);
+	} else {
+		/*
+		 * We need to start with the max_order that fits in nr_pages,
+		 * not the other way around, hence ilog2() and not get_order.
+		 */
+		max_order = ilog2(nr_pages);
+		watermark = 0;
 	}
 
+	/*
+	 * kcalloc_node() is unable to allocate buffer if the size is larger
+	 * than: PAGE_SIZE << MAX_PAGE_ORDER; directly bail out in this case.
+	 */
+	if (get_order((unsigned long)nr_pages * sizeof(void *)) > MAX_PAGE_ORDER)
+		return -ENOMEM;
 	rb->aux_pages = kcalloc_node(nr_pages, sizeof(void *), GFP_KERNEL,
 				     node);
 	if (!rb->aux_pages)
@@ -741,9 +768,6 @@ int rb_alloc_aux(struct perf_buffer *rb, struct perf_event *event,
 	rb->aux_overwrite = overwrite;
 	rb->aux_watermark = watermark;
 
-	if (!rb->aux_watermark && !rb->aux_overwrite)
-		rb->aux_watermark = nr_pages << (PAGE_SHIFT - 1);
-
 out:
 	if (!ret)
 		rb->aux_pgoff = pgoff;
@@ -794,7 +818,6 @@ static void perf_mmap_free_page(void *addr)
 {
 	struct page *page = virt_to_page(addr);
 
-	page->mapping = NULL;
 	__free_page(page);
 }
 
@@ -802,15 +825,16 @@ struct perf_buffer *rb_alloc(int nr_pages, long watermark, int cpu, int flags)
 {
 	struct perf_buffer *rb;
 	unsigned long size;
-	int i;
+	int i, node;
 
 	size = sizeof(struct perf_buffer);
 	size += nr_pages * sizeof(void *);
 
-	if (order_base_2(size) >= PAGE_SHIFT+MAX_ORDER)
+	if (order_base_2(size) > PAGE_SHIFT+MAX_PAGE_ORDER)
 		goto fail;
 
-	rb = kzalloc(size, GFP_KERNEL);
+	node = (cpu == -1) ? cpu : cpu_to_node(cpu);
+	rb = kzalloc_node(size, GFP_KERNEL, node);
 	if (!rb)
 		goto fail;
 
@@ -854,11 +878,6 @@ void rb_free(struct perf_buffer *rb)
 }
 
 #else
-static int data_page_nr(struct perf_buffer *rb)
-{
-	return rb->nr_pages << page_order(rb);
-}
-
 static struct page *
 __perf_mmap_to_page(struct perf_buffer *rb, unsigned long pgoff)
 {
@@ -869,28 +888,13 @@ __perf_mmap_to_page(struct perf_buffer *rb, unsigned long pgoff)
 	return vmalloc_to_page((void *)rb->user_page + pgoff * PAGE_SIZE);
 }
 
-static void perf_mmap_unmark_page(void *addr)
-{
-	struct page *page = vmalloc_to_page(addr);
-
-	page->mapping = NULL;
-}
-
 static void rb_free_work(struct work_struct *work)
 {
 	struct perf_buffer *rb;
-	void *base;
-	int i, nr;
 
 	rb = container_of(work, struct perf_buffer, work);
-	nr = data_page_nr(rb);
-
-	base = rb->user_page;
-	/* The '<=' counts in the user page. */
-	for (i = 0; i <= nr; i++)
-		perf_mmap_unmark_page(base + (i * PAGE_SIZE));
 
-	vfree(base);
+	vfree(rb->user_page);
 	kfree(rb);
 }
 
@@ -904,11 +908,13 @@ struct perf_buffer *rb_alloc(int nr_pages, long watermark, int cpu, int flags)
 	struct perf_buffer *rb;
 	unsigned long size;
 	void *all_buf;
+	int node;
 
 	size = sizeof(struct perf_buffer);
 	size += sizeof(void *);
 
-	rb = kzalloc(size, GFP_KERNEL);
+	node = (cpu == -1) ? cpu : cpu_to_node(cpu);
+	rb = kzalloc_node(size, GFP_KERNEL, node);
 	if (!rb)
 		goto fail;