f2fs: swap: support migrating swapfile in aligned write mode

This patch supports to migrate swapfile in aligned write mode during
swapon in order to keep swapfile being aligned to section as much as
possible, then pinned swapfile will locates fully filled section which
may not affected by GC.

However, for the case that swapfile's size is not aligned to section
size, it will still leave last extent in file's tail as unaligned due
to its size is smaller than section size, like case #2.

case #1
xfs_io -f /mnt/f2fs/file -c "pwrite 0 4M" -c "fsync"

Before swapon:
 EXT: FILE-OFFSET      BLOCK-RANGE      TOTAL FLAGS
   0: [0..3047]:       1123352..1126399  3048 0x1000
   1: [3048..7143]:    237568..241663    4096 0x1000
   2: [7144..8191]:    245760..246807    1048 0x1001
After swapon:
 EXT: FILE-OFFSET      BLOCK-RANGE      TOTAL FLAGS
   0: [0..8191]:       249856..258047    8192 0x1001
Kmsg:
F2FS-fs (zram0): Swapfile (2) is not align to section:
1) creat(), 2) ioctl(F2FS_IOC_SET_PIN_FILE), 3) fallocate(2097152 * n)

case #2
xfs_io -f /mnt/f2fs/file -c "pwrite 0 3M" -c "fsync"

Before swapon:
 EXT: FILE-OFFSET      BLOCK-RANGE      TOTAL FLAGS
   0: [0..3047]:       246808..249855    3048 0x1000
   1: [3048..6143]:    237568..240663    3096 0x1001
After swapon:
 EXT: FILE-OFFSET      BLOCK-RANGE      TOTAL FLAGS
   0: [0..4095]:       258048..262143    4096 0x1000
   1: [4096..6143]:    238616..240663    2048 0x1001
Kmsg:
F2FS-fs (zram0): Swapfile: last extent is not aligned to section
F2FS-fs (zram0): Swapfile (2) is not align to section:
1) creat(), 2) ioctl(F2FS_IOC_SET_PIN_FILE), 3) fallocate(2097152 * n)

Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>

1 files changed, 94 insertions, 14 deletions

diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c
index ee04e6b63218..3a01a1b50104 100644
--- a/fs/f2fs/data.c
+++ b/fs/f2fs/data.c
@@ -2472,6 +2472,10 @@ static inline bool check_inplace_update_policy(struct inode *inode,
 
 bool f2fs_should_update_inplace(struct inode *inode, struct f2fs_io_info *fio)
 {
+	/* swap file is migrating in aligned write mode */
+	if (is_inode_flag_set(inode, FI_ALIGNED_WRITE))
+		return false;
+
 	if (f2fs_is_pinned_file(inode))
 		return true;
 
@@ -2494,6 +2498,11 @@ bool f2fs_should_update_outplace(struct inode *inode, struct f2fs_io_info *fio)
 		return true;
 	if (f2fs_is_atomic_file(inode))
 		return true;
+
+	/* swap file is migrating in aligned write mode */
+	if (is_inode_flag_set(inode, FI_ALIGNED_WRITE))
+		return true;
+
 	if (fio) {
 		if (page_private_gcing(fio->page))
 			return true;
@@ -3830,6 +3839,65 @@ int f2fs_migrate_page(struct address_space *mapping,
 #endif
 
 #ifdef CONFIG_SWAP
+static int f2fs_migrate_blocks(struct inode *inode, block_t start_blk,
+							unsigned int blkcnt)
+{
+	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
+	unsigned int blkofs;
+	unsigned int blk_per_sec = BLKS_PER_SEC(sbi);
+	unsigned int secidx = start_blk / blk_per_sec;
+	unsigned int end_sec = secidx + blkcnt / blk_per_sec;
+	int ret = 0;
+
+	down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
+	down_write(&F2FS_I(inode)->i_mmap_sem);
+
+	set_inode_flag(inode, FI_ALIGNED_WRITE);
+
+	for (; secidx < end_sec; secidx++) {
+		down_write(&sbi->pin_sem);
+
+		f2fs_lock_op(sbi);
+		f2fs_allocate_new_section(sbi, CURSEG_COLD_DATA_PINNED, false);
+		f2fs_unlock_op(sbi);
+
+		set_inode_flag(inode, FI_DO_DEFRAG);
+
+		for (blkofs = 0; blkofs < blk_per_sec; blkofs++) {
+			struct page *page;
+			unsigned int blkidx = secidx * blk_per_sec + blkofs;
+
+			page = f2fs_get_lock_data_page(inode, blkidx, true);
+			if (IS_ERR(page)) {
+				up_write(&sbi->pin_sem);
+				ret = PTR_ERR(page);
+				goto done;
+			}
+
+			set_page_dirty(page);
+			f2fs_put_page(page, 1);
+		}
+
+		clear_inode_flag(inode, FI_DO_DEFRAG);
+
+		ret = filemap_fdatawrite(inode->i_mapping);
+
+		up_write(&sbi->pin_sem);
+
+		if (ret)
+			break;
+	}
+
+done:
+	clear_inode_flag(inode, FI_DO_DEFRAG);
+	clear_inode_flag(inode, FI_ALIGNED_WRITE);
+
+	up_write(&F2FS_I(inode)->i_mmap_sem);
+	up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
+
+	return ret;
+}
+
 static int check_swap_activate(struct swap_info_struct *sis,
 				struct file *swap_file, sector_t *span)
 {
@@ -3843,7 +3911,8 @@ static int check_swap_activate(struct swap_info_struct *sis,
 	sector_t highest_pblock = 0;
 	int nr_extents = 0;
 	unsigned long nr_pblocks;
-	unsigned int blocks_per_sec = BLKS_PER_SEC(sbi);
+	unsigned int blks_per_sec = BLKS_PER_SEC(sbi);
+	unsigned int sec_blks_mask = BLKS_PER_SEC(sbi) - 1;
 	unsigned int not_aligned = 0;
 	int ret = 0;
 
@@ -3856,7 +3925,7 @@ static int check_swap_activate(struct swap_info_struct *sis,
 
 	while (cur_lblock < last_lblock && cur_lblock < sis->max) {
 		struct f2fs_map_blocks map;
-
+retry:
 		cond_resched();
 
 		memset(&map, 0, sizeof(map));
@@ -3881,16 +3950,28 @@ static int check_swap_activate(struct swap_info_struct *sis,
 		pblock = map.m_pblk;
 		nr_pblocks = map.m_len;
 
-		if ((pblock - SM_I(sbi)->main_blkaddr) & (blocks_per_sec - 1) ||
-				nr_pblocks & (blocks_per_sec - 1)) {
-			if (f2fs_is_pinned_file(inode)) {
-				f2fs_err(sbi, "Swapfile does not align to section");
-				ret = -EINVAL;
-				goto out;
-			}
+		if ((pblock - SM_I(sbi)->main_blkaddr) & sec_blks_mask ||
+				nr_pblocks & sec_blks_mask) {
 			not_aligned++;
-		}
 
+			nr_pblocks = roundup(nr_pblocks, blks_per_sec);
+			if (cur_lblock + nr_pblocks > sis->max)
+				nr_pblocks -= blks_per_sec;
+
+			if (!nr_pblocks) {
+				/* this extent is last one */
+				nr_pblocks = map.m_len;
+				f2fs_warn(sbi, "Swapfile: last extent is not aligned to section");
+				goto next;
+			}
+
+			ret = f2fs_migrate_blocks(inode, cur_lblock,
+							nr_pblocks);
+			if (ret)
+				goto out;
+			goto retry;
+		}
+next:
 		if (cur_lblock + nr_pblocks >= sis->max)
 			nr_pblocks = sis->max - cur_lblock;
 
@@ -3917,11 +3998,10 @@ static int check_swap_activate(struct swap_info_struct *sis,
 	sis->max = cur_lblock;
 	sis->pages = cur_lblock - 1;
 	sis->highest_bit = cur_lblock - 1;
-
-	if (not_aligned)
-		f2fs_warn(sbi, "Swapfile (%u) is not align to section: 1) creat(), 2) ioctl(F2FS_IOC_SET_PIN_FILE), 3) fallocate()",
-			  not_aligned);
 out:
+	if (not_aligned)
+		f2fs_warn(sbi, "Swapfile (%u) is not align to section: 1) creat(), 2) ioctl(F2FS_IOC_SET_PIN_FILE), 3) fallocate(%u * N)",
+			  not_aligned, blks_per_sec * F2FS_BLKSIZE);
 	return ret;
 }