Merge tag 'f2fs-for-3.18' of git://git.kernel.org/pub/scm/linux/kernel/git/jaegeuk/f2fs

Pull f2fs updates from Jaegeuk Kim:
 "This patch-set introduces a couple of new features such as large
  sector size, FITRIM, and atomic/volatile writes.

  Several patches enhance power-off recovery and checkpoint routines.

  The fsck.f2fs starts to support fixing corrupted partitions with
  recovery hints provided by this patch-set.

  Summary:
   - retain some recovery information for fsck.f2fs
   - enhance checkpoint speed
   - enhance flush command management
   - bug fix for lseek
   - tune in-place-update policies
   - enhance roll-forward speed
   - revisit all the roll-forward and fsync rules
   - support larget sector size
   - support FITRIM
   - support atomic and volatile writes

  And several clean-ups and bug fixes are included"

* tag 'f2fs-for-3.18' of git://git.kernel.org/pub/scm/linux/kernel/git/jaegeuk/f2fs: (42 commits)
  f2fs: support volatile operations for transient data
  f2fs: support atomic writes
  f2fs: remove unused return value
  f2fs: clean up f2fs_ioctl functions
  f2fs: potential shift wrapping buf in f2fs_trim_fs()
  f2fs: call f2fs_unlock_op after error was handled
  f2fs: check the use of macros on block counts and addresses
  f2fs: refactor flush_nat_entries to remove costly reorganizing ops
  f2fs: introduce FITRIM in f2fs_ioctl
  f2fs: introduce cp_control structure
  f2fs: use more free segments until SSR is activated
  f2fs: change the ipu_policy option to enable combinations
  f2fs: fix to search whole dirty segmap when get_victim
  f2fs: fix to clean previous mount option when remount_fs
  f2fs: skip punching hole in special condition
  f2fs: support large sector size
  f2fs: fix to truncate blocks past EOF in ->setattr
  f2fs: update i_size when __allocate_data_block
  f2fs: use MAX_BIO_BLOCKS(sbi)
  f2fs: remove redundant operation during roll-forward recovery
  ...

1 files changed, 98 insertions, 62 deletions

diff --git a/fs/f2fs/segment.h b/fs/f2fs/segment.h
index ff483257283b..2495bec1c621 100644
--- a/fs/f2fs/segment.h
+++ b/fs/f2fs/segment.h
@@ -45,16 +45,26 @@
 	 (secno == CURSEG_I(sbi, CURSEG_COLD_NODE)->segno /		\
 	  sbi->segs_per_sec))	\
 
-#define START_BLOCK(sbi, segno)						\
-	(SM_I(sbi)->seg0_blkaddr +					\
+#define MAIN_BLKADDR(sbi)	(SM_I(sbi)->main_blkaddr)
+#define SEG0_BLKADDR(sbi)	(SM_I(sbi)->seg0_blkaddr)
+
+#define MAIN_SEGS(sbi)	(SM_I(sbi)->main_segments)
+#define MAIN_SECS(sbi)	(sbi->total_sections)
+
+#define TOTAL_SEGS(sbi)	(SM_I(sbi)->segment_count)
+#define TOTAL_BLKS(sbi)	(TOTAL_SEGS(sbi) << sbi->log_blocks_per_seg)
+
+#define MAX_BLKADDR(sbi)	(SEG0_BLKADDR(sbi) + TOTAL_BLKS(sbi))
+#define SEGMENT_SIZE(sbi)	(1ULL << (sbi->log_blocksize +		\
+					sbi->log_blocks_per_seg))
+
+#define START_BLOCK(sbi, segno)	(SEG0_BLKADDR(sbi) +			\
 	 (GET_R2L_SEGNO(FREE_I(sbi), segno) << sbi->log_blocks_per_seg))
+
 #define NEXT_FREE_BLKADDR(sbi, curseg)					\
 	(START_BLOCK(sbi, curseg->segno) + curseg->next_blkoff)
 
-#define MAIN_BASE_BLOCK(sbi)	(SM_I(sbi)->main_blkaddr)
-
-#define GET_SEGOFF_FROM_SEG0(sbi, blk_addr)				\
-	((blk_addr) - SM_I(sbi)->seg0_blkaddr)
+#define GET_SEGOFF_FROM_SEG0(sbi, blk_addr)	((blk_addr) - SEG0_BLKADDR(sbi))
 #define GET_SEGNO_FROM_SEG0(sbi, blk_addr)				\
 	(GET_SEGOFF_FROM_SEG0(sbi, blk_addr) >> sbi->log_blocks_per_seg)
 #define GET_BLKOFF_FROM_SEG0(sbi, blk_addr)				\
@@ -77,23 +87,21 @@
 
 #define SIT_ENTRY_OFFSET(sit_i, segno)					\
 	(segno % sit_i->sents_per_block)
-#define SIT_BLOCK_OFFSET(sit_i, segno)					\
+#define SIT_BLOCK_OFFSET(segno)					\
 	(segno / SIT_ENTRY_PER_BLOCK)
-#define	START_SEGNO(sit_i, segno)		\
-	(SIT_BLOCK_OFFSET(sit_i, segno) * SIT_ENTRY_PER_BLOCK)
+#define	START_SEGNO(segno)		\
+	(SIT_BLOCK_OFFSET(segno) * SIT_ENTRY_PER_BLOCK)
 #define SIT_BLK_CNT(sbi)			\
-	((TOTAL_SEGS(sbi) + SIT_ENTRY_PER_BLOCK - 1) / SIT_ENTRY_PER_BLOCK)
+	((MAIN_SEGS(sbi) + SIT_ENTRY_PER_BLOCK - 1) / SIT_ENTRY_PER_BLOCK)
 #define f2fs_bitmap_size(nr)			\
 	(BITS_TO_LONGS(nr) * sizeof(unsigned long))
-#define TOTAL_SEGS(sbi)	(SM_I(sbi)->main_segments)
-#define TOTAL_SECS(sbi)	(sbi->total_sections)
 
-#define SECTOR_FROM_BLOCK(sbi, blk_addr)				\
-	(((sector_t)blk_addr) << (sbi)->log_sectors_per_block)
-#define SECTOR_TO_BLOCK(sbi, sectors)					\
-	(sectors >> (sbi)->log_sectors_per_block)
-#define MAX_BIO_BLOCKS(max_hw_blocks)					\
-	(min((int)max_hw_blocks, BIO_MAX_PAGES))
+#define SECTOR_FROM_BLOCK(blk_addr)					\
+	(((sector_t)blk_addr) << F2FS_LOG_SECTORS_PER_BLOCK)
+#define SECTOR_TO_BLOCK(sectors)					\
+	(sectors >> F2FS_LOG_SECTORS_PER_BLOCK)
+#define MAX_BIO_BLOCKS(sbi)						\
+	((int)min((int)max_hw_blocks(sbi), BIO_MAX_PAGES))
 
 /*
  * indicate a block allocation direction: RIGHT and LEFT.
@@ -167,6 +175,11 @@ struct segment_allocation {
 	void (*allocate_segment)(struct f2fs_sb_info *, int, bool);
 };
 
+struct inmem_pages {
+	struct list_head list;
+	struct page *page;
+};
+
 struct sit_info {
 	const struct segment_allocation *s_ops;
 
@@ -237,6 +250,12 @@ struct curseg_info {
 	unsigned int next_segno;		/* preallocated segment */
 };
 
+struct sit_entry_set {
+	struct list_head set_list;	/* link with all sit sets */
+	unsigned int start_segno;	/* start segno of sits in set */
+	unsigned int entry_cnt;		/* the # of sit entries in set */
+};
+
 /*
  * inline functions
  */
@@ -316,7 +335,7 @@ static inline void __set_free(struct f2fs_sb_info *sbi, unsigned int segno)
 	clear_bit(segno, free_i->free_segmap);
 	free_i->free_segments++;
 
-	next = find_next_bit(free_i->free_segmap, TOTAL_SEGS(sbi), start_segno);
+	next = find_next_bit(free_i->free_segmap, MAIN_SEGS(sbi), start_segno);
 	if (next >= start_segno + sbi->segs_per_sec) {
 		clear_bit(secno, free_i->free_secmap);
 		free_i->free_sections++;
@@ -430,8 +449,10 @@ static inline int reserved_sections(struct f2fs_sb_info *sbi)
 
 static inline bool need_SSR(struct f2fs_sb_info *sbi)
 {
-	return (prefree_segments(sbi) / sbi->segs_per_sec)
-			+ free_sections(sbi) < overprovision_sections(sbi);
+	int node_secs = get_blocktype_secs(sbi, F2FS_DIRTY_NODES);
+	int dent_secs = get_blocktype_secs(sbi, F2FS_DIRTY_DENTS);
+	return free_sections(sbi) <= (node_secs + 2 * dent_secs +
+						reserved_sections(sbi) + 1);
 }
 
 static inline bool has_not_enough_free_secs(struct f2fs_sb_info *sbi, int freed)
@@ -466,48 +487,47 @@ static inline int utilization(struct f2fs_sb_info *sbi)
  * F2FS_IPU_UTIL - if FS utilization is over threashold,
  * F2FS_IPU_SSR_UTIL - if SSR mode is activated and FS utilization is over
  *                     threashold,
+ * F2FS_IPU_FSYNC - activated in fsync path only for high performance flash
+ *                     storages. IPU will be triggered only if the # of dirty
+ *                     pages over min_fsync_blocks.
  * F2FS_IPUT_DISABLE - disable IPU. (=default option)
  */
 #define DEF_MIN_IPU_UTIL	70
+#define DEF_MIN_FSYNC_BLOCKS	8
 
 enum {
 	F2FS_IPU_FORCE,
 	F2FS_IPU_SSR,
 	F2FS_IPU_UTIL,
 	F2FS_IPU_SSR_UTIL,
-	F2FS_IPU_DISABLE,
+	F2FS_IPU_FSYNC,
 };
 
 static inline bool need_inplace_update(struct inode *inode)
 {
-	struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb);
+	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
+	unsigned int policy = SM_I(sbi)->ipu_policy;
 
 	/* IPU can be done only for the user data */
-	if (S_ISDIR(inode->i_mode))
+	if (S_ISDIR(inode->i_mode) || f2fs_is_atomic_file(inode))
 		return false;
 
-	/* this is only set during fdatasync */
-	if (is_inode_flag_set(F2FS_I(inode), FI_NEED_IPU))
+	if (policy & (0x1 << F2FS_IPU_FORCE))
+		return true;
+	if (policy & (0x1 << F2FS_IPU_SSR) && need_SSR(sbi))
+		return true;
+	if (policy & (0x1 << F2FS_IPU_UTIL) &&
+			utilization(sbi) > SM_I(sbi)->min_ipu_util)
+		return true;
+	if (policy & (0x1 << F2FS_IPU_SSR_UTIL) && need_SSR(sbi) &&
+			utilization(sbi) > SM_I(sbi)->min_ipu_util)
 		return true;
 
-	switch (SM_I(sbi)->ipu_policy) {
-	case F2FS_IPU_FORCE:
+	/* this is only set during fdatasync */
+	if (policy & (0x1 << F2FS_IPU_FSYNC) &&
+			is_inode_flag_set(F2FS_I(inode), FI_NEED_IPU))
 		return true;
-	case F2FS_IPU_SSR:
-		if (need_SSR(sbi))
-			return true;
-		break;
-	case F2FS_IPU_UTIL:
-		if (utilization(sbi) > SM_I(sbi)->min_ipu_util)
-			return true;
-		break;
-	case F2FS_IPU_SSR_UTIL:
-		if (need_SSR(sbi) && utilization(sbi) > SM_I(sbi)->min_ipu_util)
-			return true;
-		break;
-	case F2FS_IPU_DISABLE:
-		break;
-	}
+
 	return false;
 }
 
@@ -534,18 +554,13 @@ static inline unsigned short curseg_blkoff(struct f2fs_sb_info *sbi, int type)
 #ifdef CONFIG_F2FS_CHECK_FS
 static inline void check_seg_range(struct f2fs_sb_info *sbi, unsigned int segno)
 {
-	unsigned int end_segno = SM_I(sbi)->segment_count - 1;
-	BUG_ON(segno > end_segno);
+	BUG_ON(segno > TOTAL_SEGS(sbi) - 1);
 }
 
 static inline void verify_block_addr(struct f2fs_sb_info *sbi, block_t blk_addr)
 {
-	struct f2fs_sm_info *sm_info = SM_I(sbi);
-	block_t total_blks = sm_info->segment_count << sbi->log_blocks_per_seg;
-	block_t start_addr = sm_info->seg0_blkaddr;
-	block_t end_addr = start_addr + total_blks - 1;
-	BUG_ON(blk_addr < start_addr);
-	BUG_ON(blk_addr > end_addr);
+	BUG_ON(blk_addr < SEG0_BLKADDR(sbi));
+	BUG_ON(blk_addr >= MAX_BLKADDR(sbi));
 }
 
 /*
@@ -554,8 +569,6 @@ static inline void verify_block_addr(struct f2fs_sb_info *sbi, block_t blk_addr)
 static inline void check_block_count(struct f2fs_sb_info *sbi,
 		int segno, struct f2fs_sit_entry *raw_sit)
 {
-	struct f2fs_sm_info *sm_info = SM_I(sbi);
-	unsigned int end_segno = sm_info->segment_count - 1;
 	bool is_valid  = test_bit_le(0, raw_sit->valid_map) ? true : false;
 	int valid_blocks = 0;
 	int cur_pos = 0, next_pos;
@@ -564,7 +577,7 @@ static inline void check_block_count(struct f2fs_sb_info *sbi,
 	BUG_ON(GET_SIT_VBLOCKS(raw_sit) > sbi->blocks_per_seg);
 
 	/* check boundary of a given segment number */
-	BUG_ON(segno > end_segno);
+	BUG_ON(segno > TOTAL_SEGS(sbi) - 1);
 
 	/* check bitmap with valid block count */
 	do {
@@ -583,16 +596,39 @@ static inline void check_block_count(struct f2fs_sb_info *sbi,
 	BUG_ON(GET_SIT_VBLOCKS(raw_sit) != valid_blocks);
 }
 #else
-#define check_seg_range(sbi, segno)
-#define verify_block_addr(sbi, blk_addr)
-#define check_block_count(sbi, segno, raw_sit)
+static inline void check_seg_range(struct f2fs_sb_info *sbi, unsigned int segno)
+{
+	if (segno > TOTAL_SEGS(sbi) - 1)
+		sbi->need_fsck = true;
+}
+
+static inline void verify_block_addr(struct f2fs_sb_info *sbi, block_t blk_addr)
+{
+	if (blk_addr < SEG0_BLKADDR(sbi) || blk_addr >= MAX_BLKADDR(sbi))
+		sbi->need_fsck = true;
+}
+
+/*
+ * Summary block is always treated as an invalid block
+ */
+static inline void check_block_count(struct f2fs_sb_info *sbi,
+		int segno, struct f2fs_sit_entry *raw_sit)
+{
+	/* check segment usage */
+	if (GET_SIT_VBLOCKS(raw_sit) > sbi->blocks_per_seg)
+		sbi->need_fsck = true;
+
+	/* check boundary of a given segment number */
+	if (segno > TOTAL_SEGS(sbi) - 1)
+		sbi->need_fsck = true;
+}
 #endif
 
 static inline pgoff_t current_sit_addr(struct f2fs_sb_info *sbi,
 						unsigned int start)
 {
 	struct sit_info *sit_i = SIT_I(sbi);
-	unsigned int offset = SIT_BLOCK_OFFSET(sit_i, start);
+	unsigned int offset = SIT_BLOCK_OFFSET(start);
 	block_t blk_addr = sit_i->sit_base_addr + offset;
 
 	check_seg_range(sbi, start);
@@ -619,7 +655,7 @@ static inline pgoff_t next_sit_addr(struct f2fs_sb_info *sbi,
 
 static inline void set_to_next_sit(struct sit_info *sit_i, unsigned int start)
 {
-	unsigned int block_off = SIT_BLOCK_OFFSET(sit_i, start);
+	unsigned int block_off = SIT_BLOCK_OFFSET(start);
 
 	if (f2fs_test_bit(block_off, sit_i->sit_bitmap))
 		f2fs_clear_bit(block_off, sit_i->sit_bitmap);
@@ -666,7 +702,7 @@ static inline unsigned int max_hw_blocks(struct f2fs_sb_info *sbi)
 {
 	struct block_device *bdev = sbi->sb->s_bdev;
 	struct request_queue *q = bdev_get_queue(bdev);
-	return SECTOR_TO_BLOCK(sbi, queue_max_sectors(q));
+	return SECTOR_TO_BLOCK(queue_max_sectors(q));
 }
 
 /*
@@ -683,7 +719,7 @@ static inline int nr_pages_to_skip(struct f2fs_sb_info *sbi, int type)
 	else if (type == NODE)
 		return 3 * sbi->blocks_per_seg;
 	else if (type == META)
-		return MAX_BIO_BLOCKS(max_hw_blocks(sbi));
+		return MAX_BIO_BLOCKS(sbi);
 	else
 		return 0;
 }
@@ -706,7 +742,7 @@ static inline long nr_pages_to_write(struct f2fs_sb_info *sbi, int type,
 	else if (type == NODE)
 		desired = 3 * max_hw_blocks(sbi);
 	else
-		desired = MAX_BIO_BLOCKS(max_hw_blocks(sbi));
+		desired = MAX_BIO_BLOCKS(sbi);
 
 	wbc->nr_to_write = desired;
 	return desired - nr_to_write;