Merge tag 'for-f2fs-3.15' of git://git.kernel.org/pub/scm/linux/kernel/git/jaegeuk/f2fs

Pull f2fs updates from Jaegeuk Kim:
 "This patch-set includes the following major enhancement patches.
   - introduce large directory support
   - introduce f2fs_issue_flush to merge redundant flush commands
   - merge write IOs as much as possible aligned to the segment
   - add sysfs entries to tune the f2fs configuration
   - use radix_tree for the free_nid_list to reduce in-memory operations
   - remove costly bit operations in f2fs_find_entry
   - enhance the readahead flow for CP/NAT/SIT/SSA blocks

  The other bug fixes are as follows:
   - recover xattr node blocks correctly after sudden-power-cut
   - fix to calculate the maximum number of node ids
   - enhance to handle many error cases

  And, there are a bunch of cleanups"

* tag 'for-f2fs-3.15' of git://git.kernel.org/pub/scm/linux/kernel/git/jaegeuk/f2fs: (62 commits)
  f2fs: fix wrong statistics of inline data
  f2fs: check the acl's validity before setting
  f2fs: introduce f2fs_issue_flush to avoid redundant flush issue
  f2fs: fix to cover io->bio with io_rwsem
  f2fs: fix error path when fail to read inline data
  f2fs: use list_for_each_entry{_safe} for simplyfying code
  f2fs: avoid free slab cache under spinlock
  f2fs: avoid unneeded lookup when xattr name length is too long
  f2fs: avoid unnecessary bio submit when wait page writeback
  f2fs: return -EIO when node id is not matched
  f2fs: avoid RECLAIM_FS-ON-W warning
  f2fs: skip unnecessary node writes during fsync
  f2fs: introduce fi->i_sem to protect fi's info
  f2fs: change reclaim rate in percentage
  f2fs: add missing documentation for dir_level
  f2fs: remove unnecessary threshold
  f2fs: throttle the memory footprint with a sysfs entry
  f2fs: avoid to drop nat entries due to the negative nr_shrink
  f2fs: call f2fs_wait_on_page_writeback instead of native function
  f2fs: introduce nr_pages_to_write for segment alignment
  ...

1 files changed, 17 insertions, 8 deletions

diff --git a/fs/f2fs/node.h b/fs/f2fs/node.h
index c4c79885c993..5decc1a375f0 100644
--- a/fs/f2fs/node.h
+++ b/fs/f2fs/node.h
@@ -17,14 +17,11 @@
 /* # of pages to perform readahead before building free nids */
 #define FREE_NID_PAGES 4
 
-/* maximum # of free node ids to produce during build_free_nids */
-#define MAX_FREE_NIDS (NAT_ENTRY_PER_BLOCK * FREE_NID_PAGES)
-
 /* maximum readahead size for node during getting data blocks */
 #define MAX_RA_NODE		128
 
-/* maximum cached nat entries to manage memory footprint */
-#define NM_WOUT_THRESHOLD	(64 * NAT_ENTRY_PER_BLOCK)
+/* control the memory footprint threshold (10MB per 1GB ram) */
+#define DEF_RAM_THRESHOLD	10
 
 /* vector size for gang look-up from nat cache that consists of radix tree */
 #define NATVEC_SIZE	64
@@ -45,6 +42,7 @@ struct node_info {
 struct nat_entry {
 	struct list_head list;	/* for clean or dirty nat list */
 	bool checkpointed;	/* whether it is checkpointed or not */
+	bool fsync_done;	/* whether the latest node has fsync mark */
 	struct node_info ni;	/* in-memory node information */
 };
 
@@ -58,9 +56,15 @@ struct nat_entry {
 #define nat_set_version(nat, v)		(nat->ni.version = v)
 
 #define __set_nat_cache_dirty(nm_i, ne)					\
-	list_move_tail(&ne->list, &nm_i->dirty_nat_entries);
+	do {								\
+		ne->checkpointed = false;				\
+		list_move_tail(&ne->list, &nm_i->dirty_nat_entries);	\
+	} while (0);
 #define __clear_nat_cache_dirty(nm_i, ne)				\
-	list_move_tail(&ne->list, &nm_i->nat_entries);
+	do {								\
+		ne->checkpointed = true;				\
+		list_move_tail(&ne->list, &nm_i->nat_entries);		\
+	} while (0);
 #define inc_node_version(version)	(++version)
 
 static inline void node_info_from_raw_nat(struct node_info *ni,
@@ -71,6 +75,11 @@ static inline void node_info_from_raw_nat(struct node_info *ni,
 	ni->version = raw_ne->version;
 }
 
+enum nid_type {
+	FREE_NIDS,	/* indicates the free nid list */
+	NAT_ENTRIES	/* indicates the cached nat entry */
+};
+
 /*
  * For free nid mangement
  */
@@ -236,7 +245,7 @@ static inline bool IS_DNODE(struct page *node_page)
 {
 	unsigned int ofs = ofs_of_node(node_page);
 
-	if (ofs == XATTR_NODE_OFFSET)
+	if (f2fs_has_xattr_block(ofs))
 		return false;
 
 	if (ofs == 3 || ofs == 4 + NIDS_PER_BLOCK ||