[PATCH v14 54/74] memfd: Convert memfd_wait_for_pins to XArray

[Matthew Wilcox]

Simplify the locking by taking the spinlock while we walk the tree on
the assumption that many acquires and releases of the lock will be worse
than holding the lock while we process an entire batch of pages.

Signed-off-by: Matthew Wilcox <willy@xxxxxxxxxxxxx>
Reviewed-by: Mike Kravetz <mike.kravetz@xxxxxxxxxx>
---
 mm/memfd.c | 61 ++++++++++++++++++++++--------------------------------
 1 file changed, 25 insertions(+), 36 deletions(-)

diff --git a/mm/memfd.c b/mm/memfd.c
index 27069518e3c5..e7d6be725b7a 100644
--- a/mm/memfd.c
+++ b/mm/memfd.c
@@ -21,7 +21,7 @@
 #include <uapi/linux/memfd.h>
 
 /*
- * We need a tag: a new tag would expand every radix_tree_node by 8 bytes,
+ * We need a tag: a new tag would expand every xa_node by 8 bytes,
  * so reuse a tag which we firmly believe is never set or cleared on tmpfs
  * or hugetlbfs because they are memory only filesystems.
  */
@@ -72,9 +72,7 @@ static void memfd_tag_pins(struct address_space *mapping)
  */
 static int memfd_wait_for_pins(struct address_space *mapping)
 {
-	struct radix_tree_iter iter;
-	void __rcu **slot;
-	pgoff_t start;
+	XA_STATE(xas, &mapping->i_pages, 0);
 	struct page *page;
 	int error, scan;
 
@@ -82,7 +80,9 @@ static int memfd_wait_for_pins(struct address_space *mapping)
 
 	error = 0;
 	for (scan = 0; scan <= LAST_SCAN; scan++) {
-		if (!radix_tree_tagged(&mapping->i_pages, MEMFD_TAG_PINNED))
+		unsigned int tagged = 0;
+
+		if (!xas_tagged(&xas, MEMFD_TAG_PINNED))
 			break;
 
 		if (!scan)
@@ -90,45 +90,34 @@ static int memfd_wait_for_pins(struct address_space *mapping)
 		else if (schedule_timeout_killable((HZ << scan) / 200))
 			scan = LAST_SCAN;
 
-		start = 0;
-		rcu_read_lock();
-		radix_tree_for_each_tagged(slot, &mapping->i_pages, &iter,
-					   start, MEMFD_TAG_PINNED) {
-
-			page = radix_tree_deref_slot(slot);
-			if (radix_tree_exception(page)) {
-				if (radix_tree_deref_retry(page)) {
-					slot = radix_tree_iter_retry(&iter);
-					continue;
-				}
-
-				page = NULL;
-			}
-
-			if (page &&
-			    page_count(page) - page_mapcount(page) != 1) {
-				if (scan < LAST_SCAN)
-					goto continue_resched;
-
+		xas_set(&xas, 0);
+		xas_lock_irq(&xas);
+		xas_for_each_tagged(&xas, page, ULONG_MAX, MEMFD_TAG_PINNED) {
+			bool clear = true;
+			if (xa_is_value(page))
+				continue;
+			if (page_count(page) - page_mapcount(page) != 1) {
 				/*
 				 * On the last scan, we clean up all those tags
 				 * we inserted; but make a note that we still
 				 * found pages pinned.
 				 */
-				error = -EBUSY;
+				if (scan == LAST_SCAN)
+					error = -EBUSY;
+				else
+					clear = false;
 			}
+			if (clear)
+				xas_clear_tag(&xas, MEMFD_TAG_PINNED);
+			if (++tagged % XA_CHECK_SCHED)
+				continue;
 
-			xa_lock_irq(&mapping->i_pages);
-			radix_tree_tag_clear(&mapping->i_pages,
-					     iter.index, MEMFD_TAG_PINNED);
-			xa_unlock_irq(&mapping->i_pages);
-continue_resched:
-			if (need_resched()) {
-				slot = radix_tree_iter_resume(slot, &iter);
-				cond_resched_rcu();
-			}
+			xas_pause(&xas);
+			xas_unlock_irq(&xas);
+			cond_resched();
+			xas_lock_irq(&xas);
 		}
-		rcu_read_unlock();
+		xas_unlock_irq(&xas);
 	}
 
 	return error;
-- 
2.17.1