ring-buffer: Return reader page back into existing ring buffer

When reading the ring buffer for consuming, it is optimized for splice,
where a page is taken out of the ring buffer (zero copy) and sent to the
reading consumer. When the read is finished with the page, it calls
ring_buffer_free_read_page(), which simply frees the page. The next time the
reader needs to get a page from the ring buffer, it must call
ring_buffer_alloc_read_page() which allocates and initializes a reader page
for the ring buffer to be swapped into the ring buffer for a new filled page
for the reader.

The problem is that there's no reason to actually free the page when it is
passed back to the ring buffer. It can hold it off and reuse it for the next
iteration. This completely removes the interaction with the page_alloc
mechanism.

Using the trace-cmd utility to record all events (causing trace-cmd to
require reading lots of pages from the ring buffer, and calling
ring_buffer_alloc/free_read_page() several times), and also assigning a
stack trace trigger to the mm_page_alloc event, we can see how many times
the ring_buffer_alloc_read_page() needed to allocate a page for the ring
buffer.

Before this change:

  # trace-cmd record -e all -e mem_page_alloc -R stacktrace sleep 1
  # trace-cmd report |grep ring_buffer_alloc_read_page | wc -l
  9968

After this change:

  # trace-cmd record -e all -e mem_page_alloc -R stacktrace sleep 1
  # trace-cmd report |grep ring_buffer_alloc_read_page | wc -l
  4

Signed-off-by: Steven Rostedt (VMware) <rostedt@goodmis.org>

1 files changed, 37 insertions, 3 deletions

diff --git a/kernel/trace/ring_buffer.c b/kernel/trace/ring_buffer.c
index 96fc3c043ad6..01b4ee5326cf 100644
--- a/kernel/trace/ring_buffer.c
+++ b/kernel/trace/ring_buffer.c
@@ -438,6 +438,7 @@ struct ring_buffer_per_cpu {
         raw_spinlock_t                  reader_lock;    /* serialize readers */
         arch_spinlock_t                 lock;
         struct lock_class_key           lock_key;
+        struct buffer_data_page         *free_page;
         unsigned long                   nr_pages;
         unsigned int                    current_context;
         struct list_head                *pages;
@@ -4377,9 +4378,25 @@ EXPORT_SYMBOL_GPL(ring_buffer_swap_cpu);
  */
 void *ring_buffer_alloc_read_page(struct ring_buffer *buffer, int cpu)
 {
-        struct buffer_data_page *bpage;
+        struct ring_buffer_per_cpu *cpu_buffer = buffer->buffers[cpu];
+        struct buffer_data_page *bpage = NULL;
+        unsigned long flags;
         struct page *page;
 
+        local_irq_save(flags);
+        arch_spin_lock(&cpu_buffer->lock);
+
+        if (cpu_buffer->free_page) {
+                bpage = cpu_buffer->free_page;
+                cpu_buffer->free_page = NULL;
+        }
+
+        arch_spin_unlock(&cpu_buffer->lock);
+        local_irq_restore(flags);
+
+        if (bpage)
+                goto out;
+
         page = alloc_pages_node(cpu_to_node(cpu),
                                 GFP_KERNEL | __GFP_NORETRY, 0);
         if (!page)
@@ -4387,6 +4404,7 @@ void *ring_buffer_alloc_read_page(struct ring_buffer *buffer, int cpu)
 
         bpage = page_address(page);
 
+ out:
         rb_init_page(bpage);
 
         return bpage;
@@ -4396,13 +4414,29 @@ EXPORT_SYMBOL_GPL(ring_buffer_alloc_read_page);
 /**
  * ring_buffer_free_read_page - free an allocated read page
  * @buffer: the buffer the page was allocate for
+ * @cpu: the cpu buffer the page came from
  * @data: the page to free
  *
  * Free a page allocated from ring_buffer_alloc_read_page.
  */
-void ring_buffer_free_read_page(struct ring_buffer *buffer, void *data)
+void ring_buffer_free_read_page(struct ring_buffer *buffer, int cpu, void *data)
 {
-        free_page((unsigned long)data);
+        struct ring_buffer_per_cpu *cpu_buffer = buffer->buffers[cpu];
+        struct buffer_data_page *bpage = data;
+        unsigned long flags;
+
+        local_irq_save(flags);
+        arch_spin_lock(&cpu_buffer->lock);
+
+        if (!cpu_buffer->free_page) {
+                cpu_buffer->free_page = bpage;
+                bpage = NULL;
+        }
+
+        arch_spin_unlock(&cpu_buffer->lock);
+        local_irq_restore(flags);
+
+        free_page((unsigned long)bpage);
 }
 EXPORT_SYMBOL_GPL(ring_buffer_free_read_page);