page_pool: Refactor page_pool to enable fragmenting after allocation

This change is meant to permit a driver to perform "fragmenting" of the
page from within the driver instead of the current model which requires
pre-partitioning the page. The main motivation behind this is to support
use cases where the page will be split up by the driver after DMA instead
of before.

With this change it becomes possible to start using page pool to replace
some of the existing use cases where multiple references were being used
for a single page, but the number needed was unknown as the size could be
dynamic.

For example, with this code it would be possible to do something like
the following to handle allocation:
  page = page_pool_alloc_pages();
  if (!page)
    return NULL;
  page_pool_fragment_page(page, DRIVER_PAGECNT_BIAS_MAX);
  rx_buf->page = page;
  rx_buf->pagecnt_bias = DRIVER_PAGECNT_BIAS_MAX;

Then we would process a received buffer by handling it with:
  rx_buf->pagecnt_bias--;

Once the page has been fully consumed we could then flush the remaining
instances with:
  if (page_pool_defrag_page(page, rx_buf->pagecnt_bias))
    continue;
  page_pool_put_defragged_page(pool, page -1, !!budget);

The general idea is that we want to have the ability to allocate a page
with excess fragment count and then trim off the unneeded fragments.

Signed-off-by: Alexander Duyck <alexanderduyck@fb.com>
Reviewed-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

include/net/page_pool.h

@@ -201,21 +201,67 @@ static inline void page_pool_put_page_bulk(struct page_pool *pool, void **data,
 }
 #endif
 
-void page_pool_put_page(struct page_pool *pool, struct page *page,
+void page_pool_put_defragged_page(struct page_pool *pool, struct page *page,
-			unsigned int dma_sync_size, bool allow_direct);
+				  unsigned int dma_sync_size,
+				  bool allow_direct);
 
-/* Same as above but will try to sync the entire area pool->max_len */
+static inline void page_pool_fragment_page(struct page *page, long nr)
-static inline void page_pool_put_full_page(struct page_pool *pool,
+{
-					   struct page *page, bool allow_direct)
+	atomic_long_set(&page->pp_frag_count, nr);
+}
+
+static inline long page_pool_defrag_page(struct page *page, long nr)
+{
+	long ret;
+
+	/* If nr == pp_frag_count then we have cleared all remaining
+	 * references to the page. No need to actually overwrite it, instead
+	 * we can leave this to be overwritten by the calling function.
+	 *
+	 * The main advantage to doing this is that an atomic_read is
+	 * generally a much cheaper operation than an atomic update,
+	 * especially when dealing with a page that may be partitioned
+	 * into only 2 or 3 pieces.
+	 */
+	if (atomic_long_read(&page->pp_frag_count) == nr)
+		return 0;
+
+	ret = atomic_long_sub_return(nr, &page->pp_frag_count);
+	WARN_ON(ret < 0);
+	return ret;
+}
+
+static inline bool page_pool_is_last_frag(struct page_pool *pool,
+					  struct page *page)
+{
+	/* If fragments aren't enabled or count is 0 we were the last user */
+	return !(pool->p.flags & PP_FLAG_PAGE_FRAG) ||
+	       (page_pool_defrag_page(page, 1) == 0);
+}
+
+static inline void page_pool_put_page(struct page_pool *pool,
+				      struct page *page,
+				      unsigned int dma_sync_size,
+				      bool allow_direct)
 {
 	/* When page_pool isn't compiled-in, net/core/xdp.c doesn't
 	 * allow registering MEM_TYPE_PAGE_POOL, but shield linker.
 	 */
 #ifdef CONFIG_PAGE_POOL
-	page_pool_put_page(pool, page, -1, allow_direct);
+	if (!page_pool_is_last_frag(pool, page))
+		return;
+
+	page_pool_put_defragged_page(pool, page, dma_sync_size, allow_direct);
 #endif
 }
 
+/* Same as above but will try to sync the entire area pool->max_len */
+static inline void page_pool_put_full_page(struct page_pool *pool,
+					   struct page *page, bool allow_direct)
+{
+	page_pool_put_page(pool, page, -1, allow_direct);
+}
+
 /* Same as above but the caller must guarantee safe context. e.g NAPI */
 static inline void page_pool_recycle_direct(struct page_pool *pool,
 					    struct page *page)
@@ -243,30 +289,6 @@ static inline void page_pool_set_dma_addr(struct page *page, dma_addr_t addr)
 		page->dma_addr_upper = upper_32_bits(addr);
 }
 
-static inline void page_pool_set_frag_count(struct page *page, long nr)
-{
-	atomic_long_set(&page->pp_frag_count, nr);
-}
-
-static inline long page_pool_atomic_sub_frag_count_return(struct page *page,
-							  long nr)
-{
-	long ret;
-
-	/* As suggested by Alexander, atomic_long_read() may cover up the
-	 * reference count errors, so avoid calling atomic_long_read() in
-	 * the cases of freeing or draining the page_frags, where we would
-	 * not expect it to match or that are slowpath anyway.
-	 */
-	if (__builtin_constant_p(nr) &&
-	    atomic_long_read(&page->pp_frag_count) == nr)
-		return 0;
-
-	ret = atomic_long_sub_return(nr, &page->pp_frag_count);
-	WARN_ON(ret < 0);
-	return ret;
-}
-
 static inline bool is_page_pool_compiled_in(void)
 {
 #ifdef CONFIG_PAGE_POOL

net/core/page_pool.c

@@ -423,11 +423,6 @@ static __always_inline struct page *
 __page_pool_put_page(struct page_pool *pool, struct page *page,
 		     unsigned int dma_sync_size, bool allow_direct)
 {
-	/* It is not the last user for the page frag case */
-	if (pool->p.flags & PP_FLAG_PAGE_FRAG &&
-	    page_pool_atomic_sub_frag_count_return(page, 1))
-		return NULL;
-
 	/* This allocator is optimized for the XDP mode that uses
 	 * one-frame-per-page, but have fallbacks that act like the
 	 * regular page allocator APIs.
@@ -471,8 +466,8 @@ __page_pool_put_page(struct page_pool *pool, struct page *page,
 	return NULL;
 }
 
-void page_pool_put_page(struct page_pool *pool, struct page *page,
+void page_pool_put_defragged_page(struct page_pool *pool, struct page *page,
-			unsigned int dma_sync_size, bool allow_direct)
+				  unsigned int dma_sync_size, bool allow_direct)
 {
 	page = __page_pool_put_page(pool, page, dma_sync_size, allow_direct);
 	if (page && !page_pool_recycle_in_ring(pool, page)) {
@@ -480,7 +475,7 @@ void page_pool_put_page(struct page_pool *pool, struct page *page,
 		page_pool_return_page(pool, page);
 	}
 }
-EXPORT_SYMBOL(page_pool_put_page);
+EXPORT_SYMBOL(page_pool_put_defragged_page);
 
 /* Caller must not use data area after call, as this function overwrites it */
 void page_pool_put_page_bulk(struct page_pool *pool, void **data,
@@ -491,6 +486,10 @@ void page_pool_put_page_bulk(struct page_pool *pool, void **data,
 	for (i = 0; i < count; i++) {
 		struct page *page = virt_to_head_page(data[i]);
 
+		/* It is not the last user for the page frag case */
+		if (!page_pool_is_last_frag(pool, page))
+			continue;
+
 		page = __page_pool_put_page(pool, page, -1, false);
 		/* Approved for bulk recycling in ptr_ring cache */
 		if (page)
@@ -526,8 +525,7 @@ static struct page *page_pool_drain_frag(struct page_pool *pool,
 	long drain_count = BIAS_MAX - pool->frag_users;
 
 	/* Some user is still using the page frag */
-	if (likely(page_pool_atomic_sub_frag_count_return(page,
+	if (likely(page_pool_defrag_page(page, drain_count)))
-							  drain_count)))
 		return NULL;
 
 	if (page_ref_count(page) == 1 && !page_is_pfmemalloc(page)) {
@@ -548,8 +546,7 @@ static void page_pool_free_frag(struct page_pool *pool)
 
 	pool->frag_page = NULL;
 
-	if (!page ||
+	if (!page || page_pool_defrag_page(page, drain_count))
-	    page_pool_atomic_sub_frag_count_return(page, drain_count))
 		return;
 
 	page_pool_return_page(pool, page);
@@ -588,7 +585,7 @@ frag_reset:
 		pool->frag_users = 1;
 		*offset = 0;
 		pool->frag_offset = size;
-		page_pool_set_frag_count(page, BIAS_MAX);
+		page_pool_fragment_page(page, BIAS_MAX);
 		return page;
 	}