Message ID | 20240913212810.912171-1-daeho43@gmail.com (mailing list archive) |
---|---|
State | New |
Headers | show |
Series | [f2fs-dev] f2fs: introduce device aliasing file | expand |
On 2024/9/14 5:28, Daeho Jeong wrote: > From: Daeho Jeong <daehojeong@google.com> > > 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 extrent, 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> > --- > fs/f2fs/data.c | 5 +++++ > fs/f2fs/extent_cache.c | 10 ++++++++++ > fs/f2fs/f2fs.h | 5 +++++ > fs/f2fs/file.c | 36 ++++++++++++++++++++++++++++++++---- > fs/f2fs/inode.c | 10 ++++++++-- > fs/f2fs/sysfs.c | 2 ++ > 6 files changed, 62 insertions(+), 6 deletions(-) > > diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c > index 6457e5bca9c9..9ce92093ba1e 100644 > --- a/fs/f2fs/data.c > +++ b/fs/f2fs/data.c > @@ -3423,6 +3423,11 @@ static int prepare_write_begin(struct f2fs_sb_info *sbi, > > 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 fd1fc06359ee..03883963b991 100644 > --- a/fs/f2fs/extent_cache.c > +++ b/fs/f2fs/extent_cache.c > @@ -401,6 +401,11 @@ 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 +468,11 @@ 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 ac19c61f0c3e..59179b9b3a83 100644 > --- a/fs/f2fs/f2fs.h > +++ b/fs/f2fs/f2fs.h > @@ -208,6 +208,7 @@ 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) > @@ -3001,6 +3002,7 @@ 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) > > @@ -3016,6 +3018,8 @@ 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)) > @@ -4478,6 +4482,7 @@ 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 168f08507004..0f4af6b303ff 100644 > --- a/fs/f2fs/file.c > +++ b/fs/f2fs/file.c > @@ -727,6 +727,11 @@ 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)) > @@ -741,6 +746,21 @@ 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); > @@ -776,7 +796,7 @@ int f2fs_do_truncate_blocks(struct inode *inode, u64 from, bool lock) > /* 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; > } > @@ -994,7 +1014,8 @@ 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, > @@ -1855,7 +1876,7 @@ 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 */ > @@ -3264,6 +3285,9 @@ 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); > @@ -3294,6 +3318,9 @@ 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; > @@ -4711,7 +4738,8 @@ 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); > diff --git a/fs/f2fs/inode.c b/fs/f2fs/inode.c > index aef57172014f..f118e955ba88 100644 > --- a/fs/f2fs/inode.c > +++ b/fs/f2fs/inode.c > @@ -367,6 +367,12 @@ static bool sanity_check_inode(struct inode *inode, struct page *node_page) > return false; > } > > + if ((fi->i_flags & F2FS_DEVICE_ALIAS_FL) && !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; > + } Do we need to do sanity check device_alias feature flag w/ sb.devs[].path format? and related inode? > + > return true; > } > > @@ -818,8 +824,6 @@ 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); For hardlink inode, it missed to call f2fs_destroy_extent_tree()? Thanks, > - > if (inode->i_nlink || is_bad_inode(inode)) > goto no_delete; > > @@ -874,6 +878,8 @@ void f2fs_evict_inode(struct inode *inode) > goto retry; > } > > + f2fs_destroy_extent_tree(inode); > + > if (err) { > f2fs_update_inode_page(inode); > if (dquot_initialize_needed(inode)) > diff --git a/fs/f2fs/sysfs.c b/fs/f2fs/sysfs.c > index fee7ee45ceaa..bf64f4cc3522 100644 > --- a/fs/f2fs/sysfs.c > +++ b/fs/f2fs/sysfs.c > @@ -1281,6 +1281,7 @@ 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), > @@ -1297,6 +1298,7 @@ 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);
diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c index 6457e5bca9c9..9ce92093ba1e 100644 --- a/fs/f2fs/data.c +++ b/fs/f2fs/data.c @@ -3423,6 +3423,11 @@ static int prepare_write_begin(struct f2fs_sb_info *sbi, 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 fd1fc06359ee..03883963b991 100644 --- a/fs/f2fs/extent_cache.c +++ b/fs/f2fs/extent_cache.c @@ -401,6 +401,11 @@ 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 +468,11 @@ 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 ac19c61f0c3e..59179b9b3a83 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -208,6 +208,7 @@ 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) @@ -3001,6 +3002,7 @@ 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) @@ -3016,6 +3018,8 @@ 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)) @@ -4478,6 +4482,7 @@ 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 168f08507004..0f4af6b303ff 100644 --- a/fs/f2fs/file.c +++ b/fs/f2fs/file.c @@ -727,6 +727,11 @@ 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)) @@ -741,6 +746,21 @@ 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); @@ -776,7 +796,7 @@ int f2fs_do_truncate_blocks(struct inode *inode, u64 from, bool lock) /* 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; } @@ -994,7 +1014,8 @@ 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, @@ -1855,7 +1876,7 @@ 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 */ @@ -3264,6 +3285,9 @@ 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); @@ -3294,6 +3318,9 @@ 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; @@ -4711,7 +4738,8 @@ 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); diff --git a/fs/f2fs/inode.c b/fs/f2fs/inode.c index aef57172014f..f118e955ba88 100644 --- a/fs/f2fs/inode.c +++ b/fs/f2fs/inode.c @@ -367,6 +367,12 @@ static bool sanity_check_inode(struct inode *inode, struct page *node_page) return false; } + if ((fi->i_flags & F2FS_DEVICE_ALIAS_FL) && !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; + } + return true; } @@ -818,8 +824,6 @@ 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 (inode->i_nlink || is_bad_inode(inode)) goto no_delete; @@ -874,6 +878,8 @@ void f2fs_evict_inode(struct inode *inode) goto retry; } + f2fs_destroy_extent_tree(inode); + if (err) { f2fs_update_inode_page(inode); if (dquot_initialize_needed(inode)) diff --git a/fs/f2fs/sysfs.c b/fs/f2fs/sysfs.c index fee7ee45ceaa..bf64f4cc3522 100644 --- a/fs/f2fs/sysfs.c +++ b/fs/f2fs/sysfs.c @@ -1281,6 +1281,7 @@ 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), @@ -1297,6 +1298,7 @@ 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);