f2fs: introduce device aliasing file

F2FS should understand how the device aliasing file works and support
deleting the file after use. A device aliasing file can be created by
mkfs.f2fs tool and it can map the whole device with an extent, not
using node blocks. The file space should be pinned and normally used for
read-only usages.

Signed-off-by: Daeho Jeong <daehojeong@google.com>
Signed-off-by: Chao Yu <chao@kernel.org>
Reviewed-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>

7 files changed, 119 insertions, 6 deletions

diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c
index 94f7b084f601..90fa8ab85194 100644
--- a/fs/f2fs/data.c
+++ b/fs/f2fs/data.c
@@ -3441,6 +3441,11 @@ restart:
 
 	if (!f2fs_lookup_read_extent_cache_block(inode, index,
 						 &dn.data_blkaddr)) {
+		if (IS_DEVICE_ALIASING(inode)) {
+			err = -ENODATA;
+			goto out;
+		}
+
 		if (locked) {
 			err = f2fs_reserve_block(&dn, index);
 			goto out;
diff --git a/fs/f2fs/extent_cache.c b/fs/f2fs/extent_cache.c
index 62ac440d9416..019c1f7b7fa5 100644
--- a/fs/f2fs/extent_cache.c
+++ b/fs/f2fs/extent_cache.c
@@ -24,6 +24,7 @@ bool sanity_check_extent_cache(struct inode *inode, struct page *ipage)
 	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
 	struct f2fs_extent *i_ext = &F2FS_INODE(ipage)->i_ext;
 	struct extent_info ei;
+	int devi;
 
 	get_read_extent_info(&ei, i_ext);
 
@@ -38,7 +39,36 @@ bool sanity_check_extent_cache(struct inode *inode, struct page *ipage)
 			  ei.blk, ei.fofs, ei.len);
 		return false;
 	}
-	return true;
+
+	if (!IS_DEVICE_ALIASING(inode))
+		return true;
+
+	for (devi = 0; devi < sbi->s_ndevs; devi++) {
+		if (FDEV(devi).start_blk != ei.blk ||
+				FDEV(devi).end_blk != ei.blk + ei.len - 1)
+			continue;
+
+		if (devi == 0) {
+			f2fs_warn(sbi,
+			    "%s: inode (ino=%lx) is an alias of meta device",
+			    __func__, inode->i_ino);
+			return false;
+		}
+
+		if (bdev_is_zoned(FDEV(devi).bdev)) {
+			f2fs_warn(sbi,
+			    "%s: device alias inode (ino=%lx)'s extent info "
+			    "[%u, %u, %u] maps to zoned block device",
+			    __func__, inode->i_ino, ei.blk, ei.fofs, ei.len);
+			return false;
+		}
+		return true;
+	}
+
+	f2fs_warn(sbi, "%s: device alias inode (ino=%lx)'s extent info "
+			"[%u, %u, %u] is inconsistent w/ any devices",
+			__func__, inode->i_ino, ei.blk, ei.fofs, ei.len);
+	return false;
 }
 
 static void __set_extent_info(struct extent_info *ei,
@@ -76,6 +106,9 @@ static bool __init_may_extent_tree(struct inode *inode, enum extent_type type)
 
 static bool __may_extent_tree(struct inode *inode, enum extent_type type)
 {
+	if (IS_DEVICE_ALIASING(inode) && type == EX_READ)
+		return true;
+
 	/*
 	 * for recovered files during mount do not create extents
 	 * if shrinker is not registered.
@@ -401,6 +434,11 @@ void f2fs_init_read_extent_tree(struct inode *inode, struct page *ipage)
 	if (atomic_read(&et->node_cnt) || !ei.len)
 		goto skip;
 
+	if (IS_DEVICE_ALIASING(inode)) {
+		et->largest = ei;
+		goto skip;
+	}
+
 	en = __attach_extent_node(sbi, et, &ei, NULL,
 				&et->root.rb_root.rb_node, true);
 	if (en) {
@@ -463,6 +501,11 @@ static bool __lookup_extent_tree(struct inode *inode, pgoff_t pgofs,
 		goto out;
 	}
 
+	if (IS_DEVICE_ALIASING(inode)) {
+		ret = false;
+		goto out;
+	}
+
 	en = __lookup_extent_node(&et->root, et->cached_en, pgofs);
 	if (!en)
 		goto out;
diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h
index dd47dbf6d9e6..f3ef4dc50992 100644
--- a/fs/f2fs/f2fs.h
+++ b/fs/f2fs/f2fs.h
@@ -213,6 +213,7 @@ struct f2fs_mount_info {
 #define F2FS_FEATURE_CASEFOLD			0x00001000
 #define F2FS_FEATURE_COMPRESSION		0x00002000
 #define F2FS_FEATURE_RO				0x00004000
+#define F2FS_FEATURE_DEVICE_ALIAS		0x00008000
 
 #define __F2FS_HAS_FEATURE(raw_super, mask)				\
 	((raw_super->feature & cpu_to_le32(mask)) != 0)
@@ -3046,6 +3047,7 @@ static inline void f2fs_change_bit(unsigned int nr, char *addr)
 #define F2FS_DIRSYNC_FL			0x00010000 /* dirsync behaviour (directories only) */
 #define F2FS_PROJINHERIT_FL		0x20000000 /* Create with parents projid */
 #define F2FS_CASEFOLD_FL		0x40000000 /* Casefolded file */
+#define F2FS_DEVICE_ALIAS_FL		0x80000000 /* File for aliasing a device */
 
 #define F2FS_QUOTA_DEFAULT_FL		(F2FS_NOATIME_FL | F2FS_IMMUTABLE_FL)
 
@@ -3061,6 +3063,8 @@ static inline void f2fs_change_bit(unsigned int nr, char *addr)
 /* Flags that are appropriate for non-directories/regular files. */
 #define F2FS_OTHER_FLMASK	(F2FS_NODUMP_FL | F2FS_NOATIME_FL)
 
+#define IS_DEVICE_ALIASING(inode)	(F2FS_I(inode)->i_flags & F2FS_DEVICE_ALIAS_FL)
+
 static inline __u32 f2fs_mask_flags(umode_t mode, __u32 flags)
 {
 	if (S_ISDIR(mode))
@@ -4526,6 +4530,7 @@ F2FS_FEATURE_FUNCS(sb_chksum, SB_CHKSUM);
 F2FS_FEATURE_FUNCS(casefold, CASEFOLD);
 F2FS_FEATURE_FUNCS(compression, COMPRESSION);
 F2FS_FEATURE_FUNCS(readonly, RO);
+F2FS_FEATURE_FUNCS(device_alias, DEVICE_ALIAS);
 
 #ifdef CONFIG_BLK_DEV_ZONED
 static inline bool f2fs_blkz_is_seq(struct f2fs_sb_info *sbi, int devi,
diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c
index 348ef73bf8dd..75a8b22da664 100644
--- a/fs/f2fs/file.c
+++ b/fs/f2fs/file.c
@@ -725,6 +725,11 @@ int f2fs_do_truncate_blocks(struct inode *inode, u64 from, bool lock)
 
 	trace_f2fs_truncate_blocks_enter(inode, from);
 
+	if (IS_DEVICE_ALIASING(inode) && from) {
+		err = -EINVAL;
+		goto out_err;
+	}
+
 	free_from = (pgoff_t)F2FS_BLK_ALIGN(from);
 
 	if (free_from >= max_file_blocks(inode))
@@ -739,6 +744,21 @@ int f2fs_do_truncate_blocks(struct inode *inode, u64 from, bool lock)
 		goto out;
 	}
 
+	if (IS_DEVICE_ALIASING(inode)) {
+		struct extent_tree *et = F2FS_I(inode)->extent_tree[EX_READ];
+		struct extent_info ei = et->largest;
+		unsigned int i;
+
+		for (i = 0; i < ei.len; i++)
+			f2fs_invalidate_blocks(sbi, ei.blk + i);
+
+		dec_valid_block_count(sbi, inode, ei.len);
+		f2fs_update_time(sbi, REQ_TIME);
+
+		f2fs_put_page(ipage, 1);
+		goto out;
+	}
+
 	if (f2fs_has_inline_data(inode)) {
 		f2fs_truncate_inline_inode(inode, ipage, from);
 		f2fs_put_page(ipage, 1);
@@ -774,7 +794,7 @@ free_partial:
 	/* lastly zero out the first data page */
 	if (!err)
 		err = truncate_partial_data_page(inode, from, truncate_page);
-
+out_err:
 	trace_f2fs_truncate_blocks_exit(inode, err);
 	return err;
 }
@@ -992,7 +1012,8 @@ int f2fs_setattr(struct mnt_idmap *idmap, struct dentry *dentry,
 		return -EPERM;
 
 	if ((attr->ia_valid & ATTR_SIZE)) {
-		if (!f2fs_is_compress_backend_ready(inode))
+		if (!f2fs_is_compress_backend_ready(inode) ||
+				IS_DEVICE_ALIASING(inode))
 			return -EOPNOTSUPP;
 		if (is_inode_flag_set(inode, FI_COMPRESS_RELEASED) &&
 			!IS_ALIGNED(attr->ia_size,
@@ -1861,7 +1882,7 @@ static long f2fs_fallocate(struct file *file, int mode,
 		return -EIO;
 	if (!f2fs_is_checkpoint_ready(F2FS_I_SB(inode)))
 		return -ENOSPC;
-	if (!f2fs_is_compress_backend_ready(inode))
+	if (!f2fs_is_compress_backend_ready(inode) || IS_DEVICE_ALIASING(inode))
 		return -EOPNOTSUPP;
 
 	/* f2fs only support ->fallocate for regular file */
@@ -3297,6 +3318,9 @@ int f2fs_pin_file_control(struct inode *inode, bool inc)
 	struct f2fs_inode_info *fi = F2FS_I(inode);
 	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
 
+	if (IS_DEVICE_ALIASING(inode))
+		return -EINVAL;
+
 	if (fi->i_gc_failures >= sbi->gc_pin_file_threshold) {
 		f2fs_warn(sbi, "%s: Enable GC = ino %lx after %x GC trials",
 			  __func__, inode->i_ino, fi->i_gc_failures);
@@ -3327,6 +3351,9 @@ static int f2fs_ioc_set_pin_file(struct file *filp, unsigned long arg)
 	if (f2fs_readonly(sbi->sb))
 		return -EROFS;
 
+	if (!pin && IS_DEVICE_ALIASING(inode))
+		return -EOPNOTSUPP;
+
 	ret = mnt_want_write_file(filp);
 	if (ret)
 		return ret;
@@ -3392,6 +3419,12 @@ static int f2fs_ioc_get_pin_file(struct file *filp, unsigned long arg)
 	return put_user(pin, (u32 __user *)arg);
 }
 
+static int f2fs_ioc_get_dev_alias_file(struct file *filp, unsigned long arg)
+{
+	return put_user(IS_DEVICE_ALIASING(file_inode(filp)) ? 1 : 0,
+			(u32 __user *)arg);
+}
+
 int f2fs_precache_extents(struct inode *inode)
 {
 	struct f2fs_inode_info *fi = F2FS_I(inode);
@@ -4491,6 +4524,8 @@ static long __f2fs_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
 		return f2fs_ioc_decompress_file(filp);
 	case F2FS_IOC_COMPRESS_FILE:
 		return f2fs_ioc_compress_file(filp);
+	case F2FS_IOC_GET_DEV_ALIAS_FILE:
+		return f2fs_ioc_get_dev_alias_file(filp, arg);
 	default:
 		return -ENOTTY;
 	}
@@ -4766,7 +4801,8 @@ static int f2fs_preallocate_blocks(struct kiocb *iocb, struct iov_iter *iter,
 	else
 		return 0;
 
-	map.m_may_create = true;
+	if (!IS_DEVICE_ALIASING(inode))
+		map.m_may_create = true;
 	if (dio) {
 		map.m_seg_type = f2fs_rw_hint_to_seg_type(sbi,
 						inode->i_write_hint);
@@ -5203,6 +5239,7 @@ long f2fs_compat_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
 	case F2FS_IOC_SET_COMPRESS_OPTION:
 	case F2FS_IOC_DECOMPRESS_FILE:
 	case F2FS_IOC_COMPRESS_FILE:
+	case F2FS_IOC_GET_DEV_ALIAS_FILE:
 		break;
 	default:
 		return -ENOIOCTLCMD;
diff --git a/fs/f2fs/inode.c b/fs/f2fs/inode.c
index 10780e37fc7b..282fd320bdb3 100644
--- a/fs/f2fs/inode.c
+++ b/fs/f2fs/inode.c
@@ -372,6 +372,19 @@ static bool sanity_check_inode(struct inode *inode, struct page *node_page)
 		return false;
 	}
 
+	if (IS_DEVICE_ALIASING(inode)) {
+		if (!f2fs_sb_has_device_alias(sbi)) {
+			f2fs_warn(sbi, "%s: inode (ino=%lx) has device alias flag, but the feature is off",
+				  __func__, inode->i_ino);
+			return false;
+		}
+		if (!f2fs_is_pinned_file(inode)) {
+			f2fs_warn(sbi, "%s: inode (ino=%lx) has device alias flag, but is not pinned",
+				  __func__, inode->i_ino);
+			return false;
+		}
+	}
+
 	return true;
 }
 
@@ -825,7 +838,8 @@ void f2fs_evict_inode(struct inode *inode)
 	f2fs_bug_on(sbi, get_dirty_pages(inode));
 	f2fs_remove_dirty_inode(inode);
 
-	f2fs_destroy_extent_tree(inode);
+	if (!IS_DEVICE_ALIASING(inode))
+		f2fs_destroy_extent_tree(inode);
 
 	if (inode->i_nlink || is_bad_inode(inode))
 		goto no_delete;
@@ -881,6 +895,9 @@ retry:
 		goto retry;
 	}
 
+	if (IS_DEVICE_ALIASING(inode))
+		f2fs_destroy_extent_tree(inode);
+
 	if (err) {
 		f2fs_update_inode_page(inode);
 		if (dquot_initialize_needed(inode))
diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c
index 8d4ecb2e855e..aa14c8fce7d9 100644
--- a/fs/f2fs/super.c
+++ b/fs/f2fs/super.c
@@ -834,6 +834,10 @@ static int parse_options(struct super_block *sb, char *options, bool is_remount)
 			set_opt(sbi, READ_EXTENT_CACHE);
 			break;
 		case Opt_noextent_cache:
+			if (F2FS_HAS_FEATURE(sbi, F2FS_FEATURE_DEVICE_ALIAS)) {
+				f2fs_err(sbi, "device aliasing requires extent cache");
+				return -EINVAL;
+			}
 			clear_opt(sbi, READ_EXTENT_CACHE);
 			break;
 		case Opt_noinline_data:
diff --git a/fs/f2fs/sysfs.c b/fs/f2fs/sysfs.c
index c56e8c873935..e51304bc65ea 100644
--- a/fs/f2fs/sysfs.c
+++ b/fs/f2fs/sysfs.c
@@ -1313,6 +1313,7 @@ F2FS_SB_FEATURE_RO_ATTR(sb_checksum, SB_CHKSUM);
 F2FS_SB_FEATURE_RO_ATTR(casefold, CASEFOLD);
 F2FS_SB_FEATURE_RO_ATTR(compression, COMPRESSION);
 F2FS_SB_FEATURE_RO_ATTR(readonly, RO);
+F2FS_SB_FEATURE_RO_ATTR(device_alias, DEVICE_ALIAS);
 
 static struct attribute *f2fs_sb_feat_attrs[] = {
 	ATTR_LIST(sb_encryption),
@@ -1329,6 +1330,7 @@ static struct attribute *f2fs_sb_feat_attrs[] = {
 	ATTR_LIST(sb_casefold),
 	ATTR_LIST(sb_compression),
 	ATTR_LIST(sb_readonly),
+	ATTR_LIST(sb_device_alias),
 	NULL,
 };
 ATTRIBUTE_GROUPS(f2fs_sb_feat);