this patch implements feature which allows ext4 fs uses (e.g. Lustre) store data in ext4 dirent. data is stored in ext4 dirent after file-name, this space is accounted in de->rec_len. flag EXT4_DIRENT_LUFID added to d_type if extra data is present. Index: b/fs/ext4/dir.c =================================================================== --- a/fs/ext4/dir.c +++ b/fs/ext4/dir.c @@ -53,11 +53,18 @@ const struct file_operations ext4_dir_op static unsigned char get_dtype(struct super_block *sb, int filetype) { + int fl_index = filetype & EXT4_FT_MASK; + if (!EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_FILETYPE) || - (filetype >= EXT4_FT_MAX)) + (fl_index >= EXT4_FT_MAX)) return DT_UNKNOWN; - return (ext4_filetype_table[filetype]); + if (!EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_DIRDATA)) + return (ext4_filetype_table[fl_index]); + + return (ext4_filetype_table[fl_index]) | + (filetype & EXT4_DIRENT_LUFID); + } @@ -69,11 +76,11 @@ int ext4_check_dir_entry (const char * f const char * error_msg = NULL; const int rlen = ext4_rec_len_from_disk(de->rec_len); - if (rlen < EXT4_DIR_REC_LEN(1)) + if (rlen < __EXT4_DIR_REC_LEN(1)) error_msg = "rec_len is smaller than minimal"; else if (rlen % 4 != 0) error_msg = "rec_len % 4 != 0"; - else if (rlen < EXT4_DIR_REC_LEN(de->name_len)) + else if (rlen < EXT4_DIR_REC_LEN(de)) error_msg = "rec_len is too small for name_len"; else if (((char *) de - bh->b_data) + rlen > dir->i_sb->s_blocksize) error_msg = "directory entry across blocks"; @@ -175,7 +182,7 @@ revalidate: * failure will be detected in the * dirent test below. */ if (ext4_rec_len_from_disk(de->rec_len) - < EXT4_DIR_REC_LEN(1)) + < __EXT4_DIR_REC_LEN(1)) break; i += ext4_rec_len_from_disk(de->rec_len); } @@ -209,7 +216,6 @@ revalidate: * during the copy operation. */ u64 version = filp->f_version; - error = filldir(dirent, de->name, de->name_len, filp->f_pos, @@ -335,12 +341,17 @@ int ext4_htree_store_dirent(struct file struct fname * fname, *new_fn; struct dir_private_info *info; int len; + int extra_data = 1; info = (struct dir_private_info *) dir_file->private_data; p = &info->root.rb_node; /* Create and allocate the fname structure */ - len = sizeof(struct fname) + dirent->name_len + 1; + if (dirent->file_type & EXT4_DIRENT_LUFID) + extra_data = ext4_get_dirent_data_len(dirent); + + len = sizeof(struct fname) + dirent->name_len + extra_data; + new_fn = kzalloc(len, GFP_KERNEL); if (!new_fn) return -ENOMEM; @@ -349,7 +360,7 @@ int ext4_htree_store_dirent(struct file new_fn->inode = le32_to_cpu(dirent->inode); new_fn->name_len = dirent->name_len; new_fn->file_type = dirent->file_type; - memcpy(new_fn->name, dirent->name, dirent->name_len); + memcpy(new_fn->name, dirent->name, dirent->name_len + extra_data); new_fn->name[dirent->name_len] = 0; while (*p) { Index: b/fs/ext4/ext4.h =================================================================== --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -771,6 +771,7 @@ static inline int ext4_valid_inum(struct #define EXT4_FEATURE_INCOMPAT_64BIT 0x0080 #define EXT4_FEATURE_INCOMPAT_MMP 0x0100 #define EXT4_FEATURE_INCOMPAT_FLEX_BG 0x0200 +#define EXT4_FEATURE_INCOMPAT_DIRDATA 0x1000 #define EXT4_FEATURE_COMPAT_SUPP EXT2_FEATURE_COMPAT_EXT_ATTR #define EXT4_FEATURE_INCOMPAT_SUPP (EXT4_FEATURE_INCOMPAT_FILETYPE| \ @@ -779,7 +780,9 @@ static inline int ext4_valid_inum(struct EXT4_FEATURE_INCOMPAT_EXTENTS| \ EXT4_FEATURE_INCOMPAT_64BIT| \ EXT4_FEATURE_INCOMPAT_FLEX_BG| \ - EXT4_FEATURE_INCOMPAT_MMP) + EXT4_FEATURE_INCOMPAT_MMP| \ + EXT4_FEATURE_INCOMPAT_DIRDATA) + #define EXT4_FEATURE_RO_COMPAT_SUPP (EXT4_FEATURE_RO_COMPAT_SPARSE_SUPER| \ EXT4_FEATURE_RO_COMPAT_LARGE_FILE| \ EXT4_FEATURE_RO_COMPAT_GDT_CSUM| \ @@ -847,6 +850,44 @@ struct ext4_dir_entry_2 { #define EXT4_FT_SYMLINK 7 #define EXT4_FT_MAX 8 +#define EXT4_FT_MASK 0xf + + +#if EXT4_FT_MAX > EXT4_FT_MASK +#error "conflicting EXT4_FT_MAX and EXT4_FT_MASK" +#endif + +/* + * d_type has 4 unused bits, so it can hold four types data. these different + * type of data (e.g. lustre data, high 32 bits of 64-bit inode number) can be + * stored, in flag order, after file-name in ext4 dirent. +*/ +/* + * this flag is added to d_type if ext4 dirent has extra data after + * filename. this data length is variable and length is stored in first byte + * of data. data start after filename NUL byte. + * This is used by Lustre FS. + */ +#define EXT4_DIRENT_LUFID 0x10 + +#define EXT4_LUFID_MAGIC 0xAD200907UL +struct ext4_dentry_param { + __u32 edp_magic; /* EXT4_LUFID_MAGIC */ + char edp_len; /* size of edp_data in bytes */ + char edp_data[0]; /* packed array of data */ +} __attribute__((packed)); + +static inline unsigned char *ext4_dentry_get_data(struct super_block *sb, + struct ext4_dentry_param* p) + +{ + if (!EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_DIRDATA)) + return NULL; + if (p && p->edp_magic == EXT4_LUFID_MAGIC) + return &p->edp_len; + else + return NULL; +} /* * EXT4_DIR_PAD defines the directory entries boundaries @@ -855,8 +896,11 @@ struct ext4_dir_entry_2 { */ #define EXT4_DIR_PAD 4 #define EXT4_DIR_ROUND (EXT4_DIR_PAD - 1) -#define EXT4_DIR_REC_LEN(name_len) (((name_len) + 8 + EXT4_DIR_ROUND) & \ +#define __EXT4_DIR_REC_LEN(name_len) (((name_len) + 8 + EXT4_DIR_ROUND) & \ ~EXT4_DIR_ROUND) +#define EXT4_DIR_REC_LEN(de) (__EXT4_DIR_REC_LEN(de->name_len +\ + ext4_get_dirent_data_len(de))) + #define EXT4_MAX_REC_LEN ((1<<16)-1) static inline unsigned ext4_rec_len_from_disk(__le16 dlen) @@ -1155,7 +1199,7 @@ extern struct buffer_head * ext4_find_en struct ext4_dir_entry_2 ** res_dir); extern int ext4_add_dot_dotdot(handle_t *handle, struct inode *dir, - struct inode *inode); + struct inode *inode, const void *, const void *); extern int ext4_orphan_add(handle_t *, struct inode *); extern int ext4_orphan_del(handle_t *, struct inode *); extern int ext4_htree_fill_tree(struct file *dir_file, __u32 start_hash, @@ -1345,7 +1389,28 @@ static inline int ext4_is_group_locked(s return spin_is_locked(ext4_group_lock_ptr(sb, group)); } - +/* + * Compute the total directory entry data length. + * This includes the filename and an implicit NUL terminator (always present), + * and optional extensions. Each extension has a bit set in the high 4 bits of + * de->file_type, and the extension length is the first byte in each entry. + */ + +static inline int ext4_get_dirent_data_len(struct ext4_dir_entry_2 *de) +{ + char *len = de->name + de->name_len + 1 /* NUL terminator */; + int dlen = 0; + __u8 extra_data_flags = (de->file_type & ~EXT4_FT_MASK) >> 4; + + while (extra_data_flags) { + if (extra_data_flags & 1) { + dlen += *len + (dlen == 0); + len += *len; + } + extra_data_flags >>= 1; + } + return dlen; +} #endif /* __KERNEL__ */ Index: b/fs/ext4/namei.c =================================================================== --- a/fs/ext4/namei.c +++ b/fs/ext4/namei.c @@ -171,7 +171,8 @@ static unsigned dx_get_count (struct dx_ static unsigned dx_get_limit (struct dx_entry *entries); static void dx_set_count (struct dx_entry *entries, unsigned value); static void dx_set_limit (struct dx_entry *entries, unsigned value); -static unsigned dx_root_limit (struct inode *dir, unsigned infosize); +static inline unsigned dx_root_limit(__u32 blocksize, + struct ext4_dir_entry_2 *dot_de, unsigned infosize); static unsigned dx_node_limit (struct inode *dir); static struct dx_frame *dx_probe(struct dentry *dentry, struct inode *dir, @@ -212,11 +213,12 @@ ext4_next_entry(struct ext4_dir_entry_2 */ struct dx_root_info * dx_get_dx_info(struct ext4_dir_entry_2 *de) { - /* get dotdot first */ - de = (struct ext4_dir_entry_2 *)((char *)de + EXT4_DIR_REC_LEN(1)); + BUG_ON(de->name_len != 1); + /* get dotdot first */ + de = (struct ext4_dir_entry_2 *)((char *)de + EXT4_DIR_REC_LEN(de)); - /* dx root info is after dotdot entry */ - de = (struct ext4_dir_entry_2 *)((char *)de + EXT4_DIR_REC_LEN(2)); + /* dx root info is after dotdot entry */ + de = (struct ext4_dir_entry_2 *)((char *)de + EXT4_DIR_REC_LEN(de)); return (struct dx_root_info *) de; } @@ -261,16 +263,23 @@ static inline void dx_set_limit (struct ((struct dx_countlimit *) entries)->limit = cpu_to_le16(value); } -static inline unsigned dx_root_limit (struct inode *dir, unsigned infosize) +static inline unsigned dx_root_limit(__u32 blocksize, + struct ext4_dir_entry_2 *dot_de, unsigned infosize) { - unsigned entry_space = dir->i_sb->s_blocksize - EXT4_DIR_REC_LEN(1) - - EXT4_DIR_REC_LEN(2) - infosize; + struct ext4_dir_entry_2 *dotdot_de; + unsigned entry_space; + + BUG_ON(dot_de->name_len != 1); + dotdot_de = ext4_next_entry(dot_de); + entry_space = blocksize - EXT4_DIR_REC_LEN(dot_de) - + EXT4_DIR_REC_LEN(dotdot_de) - infosize; + return entry_space / sizeof(struct dx_entry); } static inline unsigned dx_node_limit (struct inode *dir) { - unsigned entry_space = dir->i_sb->s_blocksize - EXT4_DIR_REC_LEN(0); + unsigned entry_space = dir->i_sb->s_blocksize - __EXT4_DIR_REC_LEN(0); return entry_space / sizeof(struct dx_entry); } @@ -317,7 +326,7 @@ static struct stats dx_show_leaf(struct printk(":%x.%u ", h.hash, ((char *) de - base)); } - space += EXT4_DIR_REC_LEN(de->name_len); + space += EXT4_DIR_REC_LEN(de); names++; } de = ext4_next_entry(de); @@ -421,7 +430,8 @@ dx_probe(struct dentry *dentry, struct i entries = (struct dx_entry *) (((char *)info) + info->info_length); - if (dx_get_limit(entries) != dx_root_limit(dir, + if (dx_get_limit(entries) != dx_root_limit(dir->i_sb->s_blocksize, + (struct ext4_dir_entry_2*)bh->b_data, info->info_length)) { ext4_warning(dir->i_sb, __func__, "dx entry: limit != root limit"); @@ -611,7 +621,7 @@ static int htree_dirblock_to_tree(struct de = (struct ext4_dir_entry_2 *) bh->b_data; top = (struct ext4_dir_entry_2 *) ((char *) de + dir->i_sb->s_blocksize - - EXT4_DIR_REC_LEN(0)); + __EXT4_DIR_REC_LEN(0)); for (; de < top; de = ext4_next_entry(de)) { if (!ext4_check_dir_entry("htree_dirblock_to_tree", dir, de, bh, (block<i_sb)) @@ -1020,7 +1030,7 @@ static struct buffer_head * ext4_dx_find goto errout; de = (struct ext4_dir_entry_2 *) bh->b_data; top = (struct ext4_dir_entry_2 *) ((char *) de + sb->s_blocksize - - EXT4_DIR_REC_LEN(0)); + __EXT4_DIR_REC_LEN(0)); for (; de < top; de = ext4_next_entry(de)) { int off = (block << EXT4_BLOCK_SIZE_BITS(sb)) + ((char *) de - bh->b_data); @@ -1187,7 +1197,7 @@ dx_move_dirents(char *from, char *to, st while (count--) { struct ext4_dir_entry_2 *de = (struct ext4_dir_entry_2 *) (from + map->offs); - rec_len = EXT4_DIR_REC_LEN(de->name_len); + rec_len = EXT4_DIR_REC_LEN(de); memcpy (to, de, rec_len); ((struct ext4_dir_entry_2 *) to)->rec_len = ext4_rec_len_to_disk(rec_len); @@ -1211,7 +1221,7 @@ static struct ext4_dir_entry_2* dx_pack_ while ((char*)de < base + size) { next = ext4_next_entry(de); if (de->inode && de->name_len) { - rec_len = EXT4_DIR_REC_LEN(de->name_len); + rec_len = EXT4_DIR_REC_LEN(de); if (de > to) memmove(to, de, rec_len); to->rec_len = ext4_rec_len_to_disk(rec_len); @@ -1341,10 +1351,17 @@ static int add_dirent_to_buf(handle_t *h int namelen = dentry->d_name.len; unsigned long offset = 0; unsigned short reclen; - int nlen, rlen, err; + int nlen, rlen, err, dlen = 0; + unsigned char *data; char *top; - reclen = EXT4_DIR_REC_LEN(namelen); + data = ext4_dentry_get_data(inode->i_sb, + (struct ext4_dentry_param *) + dentry->d_fsdata); + if (data) + dlen = (*data) + 1; + + reclen = __EXT4_DIR_REC_LEN(namelen + dlen); if (!de) { de = (struct ext4_dir_entry_2 *)bh->b_data; top = bh->b_data + dir->i_sb->s_blocksize - reclen; @@ -1358,7 +1375,7 @@ static int add_dirent_to_buf(handle_t *h brelse (bh); return -EEXIST; } - nlen = EXT4_DIR_REC_LEN(de->name_len); + nlen = EXT4_DIR_REC_LEN(de); rlen = ext4_rec_len_from_disk(de->rec_len); if ((de->inode? rlen - nlen: rlen) >= reclen) break; @@ -1377,7 +1394,7 @@ static int add_dirent_to_buf(handle_t *h } /* By now the buffer is marked for journaling */ - nlen = EXT4_DIR_REC_LEN(de->name_len); + nlen = EXT4_DIR_REC_LEN(de); rlen = ext4_rec_len_from_disk(de->rec_len); if (de->inode) { struct ext4_dir_entry_2 *de1 = (struct ext4_dir_entry_2 *)((char *)de + nlen); @@ -1393,6 +1410,12 @@ static int add_dirent_to_buf(handle_t *h de->inode = 0; de->name_len = namelen; memcpy (de->name, name, namelen); + if (data) { + de->name[namelen] = 0; + memcpy(&de->name[namelen + 1], data, *(char *) data); + de->file_type |= EXT4_DIRENT_LUFID; + } + /* * XXX shouldn't update any times until successful * completion of syscall, but too many callers depend @@ -1482,7 +1505,8 @@ static int make_indexed_dir(handle_t *ha dx_set_block(entries, 1); dx_set_count(entries, 1); - dx_set_limit(entries, dx_root_limit(dir, sizeof(*dx_info))); + dx_set_limit(entries, dx_root_limit(dir->i_sb->s_blocksize, + dot_de, sizeof(*dx_info))); /* Initialize as for dx_probe */ hinfo.hash_version = dx_info->hash_version; @@ -2067,7 +2125,7 @@ static int empty_dir (struct inode * ino int err = 0; sb = inode->i_sb; - if (inode->i_size < EXT4_DIR_REC_LEN(1) + EXT4_DIR_REC_LEN(2) || + if (inode->i_size < __EXT4_DIR_REC_LEN(1) + __EXT4_DIR_REC_LEN(2) || !(bh = ext4_bread (NULL, inode, 0, 0, &err))) { if (err) ext4_error(inode->i_sb, __func__,